Diego e Machuca Skepticism From Antiquity To The Present 1

SKEPTICISM
Also available from Bloomsbury
The Bloomsbury Companion to Analytic Philosophy,

edited by Barry Dainton and Howard Robinson
The Bloomsbury Companion to Epistemology,

edited by Andrew Cullison
A Critical Introduction to Skepticism, Allan Hazlett

SKEPTICISM
FROM ANTIQUITY TO
THE PRESENT
Edited by Diego E. Machuca and Baron Reed
Bloomsbury Academic
An imprint of Bloomsbury Publishing Plc
LON DON • OX F O R D • N E W YO R K • N E W D E L H I • SY DN EY
Bloomsbury Academic
An imprint of Bloomsbury Publishing Plc
50 Bedford Square 1385 Broadway

London New York
WC1B 3DP NY 10018
UK USA
www.bloomsbury.com
BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc
First published 2018
© Diego Machuca and Baron Reed, 2018
Diego Machuca and Baron Reed have asserted their right under the Copyright, Designs and
Patents Act, 1988, to be identified as the Editors of this work.
All rights reserved. No part of this publication may be reproduced or transmitted in any form or
by any means, electronic or mechanical, including photocopying, recording, or any information
storage or retrieval system, without prior permission in writing from the publishers.
No responsibility for loss caused to any individual or organization acting on or refraining

from action as a result of the material in this publication can be accepted by Bloomsbury
or the editors.
British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.
ISBN: HB: 978-1-4725-0771-6

ePDF: 978-1-4725-1436-3
ePub: 978-1-4725-1149-2
Library of Congress Cataloging-in-Publication Data

Names: Machuca, Diego E., editor.
Title: Skepticism: from antiquity to the present / [edited by] Diego Machuca and Baron Reed.
Description: New York : Bloomsbury, 2018. | Includes bibliographical references and index.
Identifiers: LCCN 2017030623 (print) | LCCN 2017043043 (ebook) | ISBN
9781472511492 (ePub) | ISBN 9781472514363 (ePDF) | ISBN 9781472507716 (hardback)
Subjects: LCSH: Skepticism–History.
Classification: LCC B837 (ebook) | LCC B837 .S53 2018 (print) | DDC 149/.7309–dc23
LC record available at https://lccn.loc.gov/2017030623
Cover design: Clare Turner

Cover image © Bridgeman Images
Typeset by Deanta Global Publishing Services, Chennai, India

CONTENTS
L ist of C ontributors ix
L ist of A bbreviAtions xi
General Introduction xiii

Baron Reed and Diego E. Machuca
Part I: Ancient Skepticism 1
Introduction Diego E. Machuca 3
1 The Cyrenaics and Skepticism 14

Richard Bett
2 Pyrrho and Timon 24

Casey Perin
3 Arcesilaus 36
Anna Maria Ioppolo
4 Carneades 51
Harald Thorsrud
5 Aenesidemus 67
Luca Castagnoli
6 Philo of Larissa 81
Harold Tarrant
7 Cicero 93
J. P. F. Wynne
8 Menodotus and Medical Empiricism 102

James Allen
9 Middle Platonism and Skepticism: Plutarch and Favorinus 114

Carlos Lévy
vi CONTENTS
10 Sextus Empiricus 125

Tad Brennan and Cliff Roberts
11 Skepticism in Classical Indian Philosophy 145

Matthew R. Dasti
Part II: Medieval and Renaissance Skepticism 163
Introduction Diego E. Machuca 165
12 Augustine and Skepticism 175

Stéphane Marchand
13 John of Salisbury’s Skepticism 186

David Bloch
14 Skepticism in Classical Islam: The Case of Ghazali 196

Paul L. Heck
15 Nicholas of Autrecourt and John Buridan on Skepticism 209

Christophe Grellard
16 Divine Deception 222

Henrik Lagerlund
17 Michel de Montaigne 232

Gianni Paganini
18 Pierre Charron 247

Gianni Paganini
19 Francisco Sanchez: A Renaissance Pyrrhonist against

Aristotelian Dogmatism 260
Damian Caluori
Part III: Modern Skepticism 271
Introduction Baron Reed 273
20 François de La Mothe Le Vayer 283

Sylvia Giocanti
21 Gassendi on Skepticism 295

Antonia LoLordo
22 Descartes and the Force of Skepticism 306

David Cunning
CONTENTS vii
23 Varieties of Modern Academic Skepticism:

Pierre-Daniel Huet and Simon Foucher 320
Michael W. Hickson
24 Spinoza on Skepticism 342

Alison Peterman
25 Pierre Bayle 355

John Christian Laursen
26 Berkeley and Skepticism 369

Margaret Atherton
27 A Pyrrhonian Interpretation of Hume on Assent 380

Donald L. M. Baxter
28 Thomas Reid’s Engagement with Skeptics and Skepticism 395

René van Woudenberg
29 Johann Georg Hamann 407

Gwen Griffith-Dickson
30 Kant and External World Skepticism 419

Georges Dicker
31 Hegel: Philosophy as a Kind of Skepticism 430

Dietmar H. Heidemann
32 Friedrich Nietzsche 442

Andreas Urs Sommer
33 Russell’s Logical Construction of the External World 454

Peter J. Graham
34 Moore and Mooreanism 467

Annalisa Coliva
35 Wittgenstein and Skepticism: Illusory Doubts 481

Michael Williams
36 Richard Popkin on the History of Skepticism 506

José R. Maia Neto
Part IV: Contemporary Skepticism 521
Introduction Baron Reed 523

viii CONTENTS
37 Regress Arguments and Skepticism 535

Richard Fumerton
38 The Problem of the Criterion 550

Andrew D. Cling
39 Neo-Pyrrhonism 565
Markus Lammenranta
40 Disagreement and Skepticism 581

Bryan Frances
41 Skepticism and the Internalism-Externalism Debate 592

Matthias Steup
42 Skepticism and Fallibilism 609

Stephen Hetherington
43 Skepticism and Contextualism 620

Michael Blome-Tillmann
44 External World Skepticism 634

Ram Neta
45 Disjunctivism and Skepticism 652

Duncan Pritchard and Chris Ranalli
46 Inductive Skepticism: A Functional Investigation 668

Ruth Weintraub
47 Skepticism about A Priori Knowledge 685

Otávio Bueno
48 Skepticism about Other Minds 700

Anil Gomes
49 Moral Skepticism 714

Richard Joyce
50 Religious Skepticism 727

J. L. Schellenberg
n Ame i ndex 739

s ubjeCt i ndex 742
LIST OF CONTRIBUTORS
James Allen, University of Toronto, Canada

Margaret Atherton, University of Wisconsin-Milwaukee, United States
Donald L. M. Baxter, University of Connecticut, United States
Richard Bett, Johns Hopkins University, United States
David Bloch, University of Copenhagen, Denmark
Michael Blome-Tillmann, McGill University, Canada
Tad Brennan, Cornell University, United States
Otávio Bueno, University of Miami, United States
Damian Caluori, Trinity University, United States
Luca Castagnoli, University of Oxford, United Kingdom
Andrew Cling, University of Alabama in Huntsville, United States
Annalisa Coliva, University of California, Irvine, United States
David Cunning, University of Iowa, United States
Matthew Dasti, Bridgewater State University, United States
Georges Dicker, SUNY College at Brockport, United States
Bryan Frances, University of Tartu, Estonia
Richard Fumerton, University of Iowa, United States
Sylvia Giocanti, Université de Toulouse Le Mirail, France
Anil Gomes, University of Oxford, United Kingdom
Peter Graham, University of California, Riverside, United States
Christophe Grellard, EPHE, PSL Research University, France
Gwen Griffith-Dickson, King’s College London, United Kingdom
Paul Heck, Georgetown University, United States
Stephen Hetherington, University of New South Wales, Australia
Dietmar Heidemann, Université du Luxembourg, Luxembourg
Michael Hickson, Trent University, Canada
Anna Maria Ioppolo, Sapienza - Università di Roma, Italy
Richard Joyce, Victoria University of Wellington, New Zealand
Henrik Lagerlund, Stockholm University, Sweden
x LIST OF CONTRIBUTORS
Markus Lammenranta, University of Helsinki, Finland

John Christian Laursen, University of California, Riverside, United States
Carlos Lévy, Université Paris-Sorbonne, France
Antonia LoLordo, University of Virginia, United States
Diego E. Machuca, Consejo Nacional de Investigaciones Científicas y Técnicas,
Argentina
José Raimundo Maia Neto, Universidade Federal de Minas Gerais, Brazil
Stéphane Marchand, École Normale Supérieure de Lyon, France
Ram Neta, University of North Carolina at Chapel Hill, United States
Gianni Paganini, Università del Piemonte Orientale, Italy
Casey Perin, University of California, Irvine, United States
Alison Peterman, University of Rochester, United States
Duncan Pritchard, University of Edinburgh, United Kingdom
Baron Reed, Northwestern University, United States
Cliff Roberts, University of Victoria, Canada
J. L. Schellenberg, Mount Saint Vincent University, Canada
Matthias Steup, University of Colorado Boulder, United States
Harold Tarrant, The University of Newcastle, Australia
Harald Thorsrud, Agnes Scott College, United States
Andreas Urs Sommer, Albert-Ludwigs-Universität Freiburg, Germany
Ruth Weintraub, Tel Aviv University, Israel
Michael Williams, Johns Hopkins University, United States
René van Woudenberg, VU University, The Netherlands
J. P. F. Wynne, Northwestern University, United States
LIST OF ABBREVIATIONS
3D Berkeley, Three Dialogues between Hylas and Philonous

Acad. Cicero, Academica
Ad Att. Cicero, Epistulae ad Atticum
Adv. Col. Plutarch, Adversus Colotem
AM Sextus Empiricus, Adversus Mathematicos
Apol. Plato, Apology of Socrates
AT Œuvres de Descartes, 11 vols., edited by C. Adam & P.
Tannery. Paris: Vrin, 1996.
Bibl. Photius, Bibliotheca
C. Acad. Augustine, Contra Academicos
Chronic. Can. Eusebius, Chronicus Canon
Conf. Augustine, Confessiones
CSM The Philosophical Writings of Descartes, vols. 1 & 2, edited
by J. Cottingham, R. Stoothoff, and D. Murdoch. Cambridge:
Cambridge University Press, 1984/5.
CSMK The Philosophical Writings of Descartes, vol. 3, edited by
J. Cottingham, R. Stoothoff, D. Murdoch, and A. Kenny.
Cambridge: Cambridge University Press, 1991.
De Alex. fort. Plutarch, De Alexandri Magni Fortuna
Comp. med. gen. Galen, De compositione medicamentorum per genera
De E Plutarch, De E apud Delphos
De Or. Cicero, De Oratore
De Prim. Frig. Plutarch, De Primo Frigido
De rep. Cicero, De re publica
De sect. ingred. Galen, De sectis ingredientibus
De ut. cred. Augustine, De utilitate credendi
Div. Cicero, De Divinatione
DL Diogenes Laertius, Lives of Eminent Philosophers
EHU Hume, An Enquiry concerning Human Understanding
Fat. Cicero, De Fato
xii LIST OF ABBREVIATIONS
Fin. Cicero, De Finibus Bonorum et Malorum

Gorg. Plato, Gorgias
Met. Aristotle, Metaphysics
MM Galen, De methodo medendi
Mur. Cicero, Pro Murena
ND Cicero, De Natura Deorum
NA Aulas Gellius, Noctes Atticae
Off. Cicero, De Officiis
Opt. Doct. Galen, De Optima Doctrina
PH Sextus Empiricus, Pyrrhonian Outlines
PHK Berkeley, The Principles of Human Knowledge
Praep. evang. Eusebius, Praeparatio evangelica
Proem. Celsus, Proemium
Prot. Plato, Protagoras
QNS Sanchez, Quod nihil scitur
Rep. Plato, Republic
St. Rep. Plutarch, De Stoicorum Repugnantiis
Subfig. emp. Galen, Subfiguratio empirica
T Hume, A Treatise of Human Nature
Theaet. Plato, Theaetetus
Tusc. Cicero, Tusculanae Disputationes
Vit. Soph. Philostratus, Vitae Sophistarum
General Introduction
BARON REED AND DIEGO E. MACHUCA
Skeptical concerns arise in a variety of cultures. Most famously, the possibility that
our experiences of the world are merely a dream is explored by many different
thinkers, including by the Chinese sage, Zhuangzi:
Last night Zhuang Zhou dreamed he was a butterfly . . . and did not know about
[Zhuang] Zhou. When all of a sudden he awoke, he was Zhou with all his wits
about him. He does not know whether he is Zhou who dreams he is a butterfly
or a butterfly who dreams he is Zhou. (Quoted in Raphals 1994: 501)
By a Nahuatl tlamatinimi (poet-sage) in pre-Conquest Mexico:
Do we speak the truth here, oh Giver of Life?
We merely dream, we only rise from a dream.
All is like a dream . . .
No one speaks here of truth . . . . (León-Portilla 1963: 7)
And, in the history of Western philosophy, by Plato:
The question I imagine you have often heard asked—what evidence could be
appealed to, supposing we were asked at this very moment whether we are asleep
or awake, dreaming all that passes through our minds or talking to one another
in the waking state? (Plato, Theaetetus 158b)
By Cicero:
But you quite miss the point when you refute the false impressions of the
insane or dreamers by their own subsequent recollection. The question isn’t
what recollection dreamers or the insane have when they are awake or their fits
subside, but what kind of impression they had at the time. (Acad. II 90)
By al-Ghāzalī:
Do you not see how, when you are asleep, you believe things and imagine
circumstances, holding them to be stable and enduring, and, so long as you are
in that dream-condition, have no doubts about them? And is it not the case that
when you awake you know that all you have imagined and believed is unfounded
and ineffectual? Why are you confident that all your waking beliefs, whether
from sense or intellect, are genuine? They are true in respect of your present
state; but it is possible that a state will come upon you whose relation to your
waking consciousness is analogous to the relation of the latter to dreaming. In
xiv SKEPTICISM
comparison with this state your waking consciousness would be like dreaming!
(From al-Ghazālī’s Deliverance from Error, quoted in Watt 1953: 23.)
And, of course, by Descartes:
How often, asleep at night, am I convinced of just such familiar events—that I am
here in my dressing-gown, sitting by the fire—when in fact I am lying undressed
in bed! . . . As I think about this more carefully, I see plainly that there are never
any sure signs by means of which being awake can be distinguished from being
asleep. (Descartes 1984, CSM II 13; AT VII 19)
Although this curiosity about how and whether our minds are properly fitted to
reality is widely shared, perhaps even universal, it is also important not to lose sight
of the differences in how that concern has been shaped and answered. Skepticism
has been many things in its long history: a puzzle, a paradox, a challenge, an
argument, an attitude, a way of life. Adopting a broad perspective allows us to see
that it has also been a tradition, linking philosophers across the centuries who have
been impelled to ask, why do we believe what we believe? and do we know what
we take ourselves to know? These questions are as old as philosophy itself—and
perhaps essential to it. Understanding skepticism in its full variety is of particular
importance in the present philosophical age, as the predominant reactions to it have
been attempts at refuting or solving it (see, e.g., Klein 1981 and DeRose 1995).
The skeptical tradition is marked by a pair of dualities. The more obvious contrast
is between skepticism and anti-skepticism. Some philosophers, like the Pyrrhonists
and Montaigne, put forth skeptical arguments because they think skepticism is true,
or rational, or part of a good life, or leads us toward faith instead of knowledge,
or perhaps simply because they find it unavoidable. Others, like Descartes and G.
E. Moore, take up skeptical arguments because they think skepticism is a useful
challenge, or an obstacle to be overcome, or a deep and possibly dangerous
philosophical problem. And some philosophers, like Wittgenstein, have a more
ambiguous relationship to skepticism, both resisting it and articulating new forms
of it.
The second duality was submerged through most of the twentieth century, though
it is perhaps now becoming more visible. The contrast here is between radical and
mitigated skepticism. Where radical skeptics take most or all of our beliefs to be
without epistemic legitimacy, mitigated skeptics do allow for the possibility that
some beliefs are better than others, though most or all of them fall short of certainty.
Both forms of skepticism appear throughout the history of philosophy, with the
radical sort being championed by the Pyrrhonists in the ancient world and used
methodologically by Descartes in the early modern period. Mitigated skepticism is
associated with some of the later Academic skeptics, with Cicero, and with Charron
and Gassendi. More common, especially in the modern world, are philosophers like
Montaigne and Bayle who draw on both varieties as it suits their own needs. Hume
is the best-known skeptic of this sort, as he favors a “more mitigated skepticism or
academical philosophy,” which results when Pyrrhonism is “corrected by common
sense and reflection” (Hume 2000, XII 24).1
GENERAL INTRODUCTION xv
Like philosophical views of every sort, the various developments of skepticism—

including in present-day philosophy—must be understood in the context of wider
concerns. Some philosophers have been drawn to skepticism as a broadly ethical
way of life—for example, the Pyrrhonists, Cicero, and Charron—while others, like
Descartes, have used it in the service of scientific certainty or, like al-Ghazālī and
perhaps Bayle, have thought that skepticism clears the way for a purer religious
faith. The aims for which skeptical arguments can be mustered often determine the
force and value of those arguments. To borrow an image from Sextus Empiricus,
skepticism is like the color of a dove’s neck, changing as one’s perspective on it shifts
(PH I 120). If that makes it difficult to come to grips with skepticism, it also makes
the effort all the more rewarding.
The present collection aims to provide an up-to-date critical survey of both the entire
history of the various skeptical traditions and the systematic discussion of skeptical
problems and arguments in the current philosophical scene. Although its scope could
perhaps be expanded even further, the ground covered in the fifty chapters collected
here is much more extensive than anything found in other scholarly works on skepticism,
making it the first of its kind. The book is chronologically divided into four parts—
ancient, medieval and Renaissance, modern, and contemporary skepticism—each
preceded by an introductory chapter. Each chapter provides an authoritative overview
of its topic, advances ongoing debates, and indicates new directions of research.
We are grateful to Colleen Coalter, the philosophy senior commissioning
editor at Bloomsbury, who approached us with the project of putting together an
anthology on skepticism and later accepted the proposal to edit instead a book of
newly written chapters. She has been patient and supportive through a variety of
vicissitudes, including the tragically early death of Tony Brueckner, one of the very
first people to join the list of contributors. His passing was a grievous loss both to
this book and to the entire community of philosophers who work on skepticism. We
are also grateful to the authors, some of whom overcame injury and illness in the
writing of their chapters, and all of whom have greatly enriched our understanding
of skepticism as it has been, as it is, and as it might yet become.
NOTE
1. It is perhaps worth noting here that the editors disagree with each other over the most
fruitful way to think of skepticism. One favors a mitigated approach, while the other
prefers his Pyrrhonism uncorrected by common sense and reflection.
REFERENCES
Cicero. 2006. On Academic Scepticism, translated by C. Brittain. Indianapolis, IN: Hackett.
DeRose, Keith. 1995. “Solving the Skeptical Problem,” Philosophical Review 104: 1–52.
Descartes, René. 1984. The Philosophical Writings of Descartes, vol. II, translated by J.
Cottingham, R. Stoothoff, and D. Murdoch. Cambridge: Cambridge University Press.
Hume, David. 2000. An Enquiry Concerning Human Understanding, edited by T.
Beauchamp. Oxford: Oxford University Press.
xvi SKEPTICISM
Klein, Peter. 1981. Certainty: A Refutation of Scepticism. Minneapolis, MN: University of

Minnesota Press.
León-Portilla, Miguel. 1963. Aztec Thought and Culture, translated by J. E. Davis. Norman,
OK: University of Oklahoma Press.
Plato. 1961. The Collected Dialogues of Plato, edited by E. Hamilton and H. Cairns.
Princeton, NJ: Princeton University Press.
Raphals, Lisa. 1994. “Skeptical Strategies in the Zuangzhi and Theaetetus,” Philosophy
East and West 44: 501–526.
Sextus Empiricus. 2000. Outlines of Scepticism, edited by J. Annas and J. Barnes.
Cambridge: Cambridge University Press.
Watt, W. Montgomery. 1953. The Faith and Practice of Al-Ghazālī. London: George Allen
and Unwin.
PART ONE
Ancient Skepticism
Introduction
Ancient Skepticism
DIEGO E. MACHUCA
As noted in the General Introduction, we have aimed to offer a comprehensive

picture of both the history of the various skeptical traditions and the contemporary
philosophical discussions of skeptical views, problems, and arguments. Consequently,
this first part devoted to ancient skepticism includes chapters not only on the two
main Western traditions, namely Pyrrhonism and Academic skepticism, but also
on the skeptical aspects of two schools with which they bear close historical and
philosophical connections, namely Cyrenaicism and medical Empiricism. Likewise,
we have decided to include chapters not only on the chief representatives of
Pyrrhonism (Pyrrho, Timon, Aenesidemus, and Sextus) and the skeptical Academy
(Arcesilaus and Carneades), but also on less important figures (Menodotus, Philo,
Cicero, Plutarch, and Favorinus). Finally, in order not to focus exclusively on Western
thought, we have included a chapter devoted to the skeptical strands within classical
Indian philosophy, which exhibit remarkable richness and sophistication. Some
scholars have argued that there exist both historical and philosophical connections
between Pyrrhonism and classical Indian thought. While the evidence for the
historical link is both meager and problematic, we do find certain philosophical
similarities—particularly in the skeptical argumentative strategies used by members
of the two traditions—as anyone can realize by reading the chapter on Indian
skepticism.1 The reader will also find significant parallels with skeptical views and
arguments discussed in contemporary analytic epistemology. It should be observed
that the authors covered by that chapter (Nāgārjuna, Jayarāśi, and Śrīharṣa) range
from the second to the twelfth century CE, a timespan that falls within the classical
period of Indian philosophy. Although this span of time goes, in the periodization of
the history of Western philosophy, from the Imperial period to the Middle Ages, it
did not make much sense to split the chapter into two and place the second portion
in the part on medieval and Renaissance skepticism.
Until not long ago, most scholars of Greco-Roman philosophy undervalued
or neglected the skeptical traditions of the Hellenistic and Imperial periods,
the main reason being the widespread view that any form of skepticism is of no
intrinsic philosophical significance. The voluminous extant corpus of a Pyrrhonist
like Sextus Empiricus was read almost exclusively as a key source of information
about the views of non-Pyrrhonian thinkers and schools. The powerful skeptical
4 SKEPTICISM
argumentative armory expounded therein was either ignored or regarded as a

collection of sophisms. This tendency began to recede particularly in the early
1980s, and at present scholarship on ancient skepticism has reached a high degree
of development, rigor, and sophistication. Scholars have not only addressed
complex historical and exegetical issues, but also examined the consistency and
livability of the distinct strands of ancient skepticism as well as assessed the scope
and soundness of the skeptics’ arguments. This reevaluation of ancient skepticism
has no doubt been essential to a better understanding of Hellenistic and Imperial
philosophy and medicine. But it has also had a much broader effect inasmuch as the
form of Pyrrhonism described in Sextus’s writings has exerted a strong influence
on contemporary philosophy, primarily in the field of epistemology. As a result,
hardly anyone familiar, first- or second-hand, with ancient skepticism would today
question its philosophical import.
As already observed, ancient Western philosophy knew two major and closely
related skeptical traditions, the Pyrrhonian and the Academic, both of which
originated in the Hellenistic era and continued into the Imperial age. Even though
in pre-Hellenistic philosophy one can find skeptical inclinations, themes, and
arguments, it is only at the beginning of the Hellenistic age that the possibility of
knowledge and justified belief was called into question in a systematic way by means
of elaborate argumentative strategies.
Those approaching ancient skepticism for the first time should be aware that it
was not uniform, not only because there were different traditions but also because
there were significant transformations within each of them. This is also a cautionary
reminder to those already familiar with ancient skepticism. For there is sometimes
a tendency to neglect especially the differences between the members of a single
tradition, ascribing to earlier figures a stance or an argument that was arrived at only
at a later stage. Keeping in mind the heterogeneous character of ancient skepticism
not only prevents historical and interpretive mistakes, but also allows one to
appreciate its richness and to make sense of some of the discrepancies between
our sources or within a single source. This is not to say that it is easy or in all
cases possible to provide a precise account of the distinct skeptical traditions or the
various brands of skepticism within the same tradition, and of how they influenced
each other or diverged from each other. The main reason for such a difficulty is that
some key figures like Pyrrho, Arcesilaus, and Carneades wrote nothing, and that the
works of most of those who did are lost to us. Most of the time we must therefore
content ourselves with fragments, reports, and summaries that are usually meager,
sometimes of doubtful reliability, and on occasion hard to reconcile.2
Pyrrhonism is no doubt the most prominent and influential form of skepticism
in the history of Western philosophy. Not only was it an important philosophical
movement in the Hellenistic and Imperial periods, but, as we will see in the other
three parts of this book, it also had a profound impact on Renaissance and modern
philosophy and is a topic of discussion in contemporary epistemology. Pyrrhonism
was thus named after its eponymous figurehead, Pyrrho of Elis (360–270 BCE),
whose stance is known primarily through the poems and prose works of his leading
disciple Timon of Phlius (320–230 BCE), parts of which survive in fragments and
INTRODUCTION 5
second-hand reports. What we know of Pyrrho’s thought, however, differs in key

respects from our current conception of Greek Pyrrhonism, which derives from the
works of the later Pyrrhonists. The later ancient Pyrrhonian tradition, sometimes
called “neo-Pyrrhonism” by ancient philosophy scholars,3 began in the first century
BCE with Aenesidemus of Cnossos and culminated with Sextus Empiricus (probably
late second century CE) and his immediate successors. Succinctly put, by the
standards of the later representatives of the tradition, Pyrrho should be regarded
as a “Dogmatist,” since he made assertions, even if negative in kind, about what
things are like by nature. In describing Pyrrho’s stance in terms of the suspending of
judgment, the refraining from dogmatic determinations, and the following of that
which appears (DL IX 62, 106), Aenesidemus was actually projecting onto Pyrrho
his own skepticism. It could likewise be argued that the picture of Pyrrho painted
by Timon was molded by the latter’s own skeptical stance. Timon is close to the
later Pyrrhonists in refraining from affirming that things are as they appear and
in taking that which appears as a guide in practical matters (DL IX 105). For this
reason, it has even been claimed that the epistemological views ascribed to Pyrrho
by Timon are actually the latter’s, who should therefore be considered the first
Pyrrhonist.4 As for Sextus, he cautiously explains the connection between Pyrrho
and the skeptical philosophy in relative terms: the skeptical way of thought is called
“Pyrrhonian” because “Pyrrho appears to us to have attached himself to Skepticism
more tangibly and more conspicuously than his predecessors” (PH I 7). One should
therefore refrain from assuming that the type of skepticism expounded in great
detail in Sextus’s substantive extant writings, which are our chief source for ancient
Pyrrhonism, is nothing but a more developed version of Pyrrho’s stance. Nonetheless,
it is not arbitrary that Pyrrho was taken as a forerunner or an inspiration by the later
Pyrrhonists, for it can be explained by his having claimed that the state of tranquility
or undisturbedness (ἀταραξία), which the later Pyrrhonists took as their aim, is
attained by mistrusting our sensations and opinions as a guide to determining the
nature of things.5
The preceding remarks may give the misleading impression that the later
Pyrrhonian tradition was homogeneous. In fact, there appear to be considerable
differences between what we can reconstruct of Aenesidemus’s outlook and the
skeptical stance expounded in at least most of Sextus’s surviving works.6 For
Aenesidemus seems to have made both negative and relativized claims that would
have been viewed by Sextus as violating suspension of judgment in that they amount
to assertions about non-evident matters.7 It should also be noted that even within
Sextus’s own corpus there seem to coexist differing varieties of Pyrrhonism that
cannot always be reconciled and that can at least in part be accounted for by his
drawing on different sources to compose his works. One of the tensions between
distinct forms of skepticism that is detectable in his works concerns the sets of modes
by means of which the Pyrrhonist seeks to induce suspension of judgment. In book
one of PH, Sextus expounds the Ten Modes (PH I 36–163) and the Five Modes
(PH I 164–177). These two sets of modes differ in nature: whereas each of the Ten
Modes applies to a particular conflict of perceptual appearances or opinions,8 the
Five Modes are thematically neutral and of universal application, since they can
6 SKEPTICISM
be employed in relation to any assertion or doctrine.9 In this case, it is possible to

identify Sextus’s sources. First, at PH I 36 he ascribes the Ten Modes to “the older
Skeptics,” and at AM VII 345 he speaks of “the Ten Modes of Aenesidemus.” We
also know from Diogenes Laertius that Aenesidemus made use of the Ten Modes
(DL IX 87).10 Second, Sextus ascribes the Five Modes to “the more recent Skeptics”
(PH I 164), whereas Diogenes attributes them to Agrippa (DL IX 88), of whom we
know nothing, except that he probably lived in the first century CE.11 Now, both
the fact that the Aenesideman and the Agrippan modes differ in character and the
fact that they were proposed by Pyrrhonists belonging to different phases of the
Pyrrhonian tradition indicate that they correspond to differing types of skepticism.
It should be remarked, however, that although the Ten Modes and the Five Modes
differ in their scope and force, they are not incompatible and their joint presence in
Sextus’s work does not actually threaten the coherence of his Pyrrhonism.12
As regards the Academy, most scholars agree that its skeptical phase arose out
of both the epistemological debate between Academics13 and Stoics and the return
to Socrates’s dialectical style of philosophizing, but they also think it possible that
the figure of Pyrrho played a part in the origin of Academic skepticism. It should be
noted that, while the Pyrrhonists were in general reluctant to recognize forebears
or to acknowledge similarities between Pyrrhonism and other philosophies,
the Academics believed that their skepticism was the culmination of a gradual
development commencing with the Presocratics and that it was perfectly in keeping
with the tradition of Socrates and Plato.
The skeptical phase of the Academy ranges from Arcesilaus (316/5–241/0
BCE), through Carneades (214–129/8 BCE) and his student Clitomachus (187–
10 BCE), to Philo of Larissa (159/8–84/3 BCE) and Antiochus of Ascalon (130–69/8
BCE), to name only the main representatives. Arcesilaus adopted a radical form of
epistemological skepticism characterized by universal suspension of judgment.14 In
Carneades, skepticism was extended to other domains such as ethics and theology,
but it may also have been mitigated by the espousal of a kind of fallibilism. There
is, however, much dispute about what his stance was and even whether he adopted
any stance in propria persona.15 For his part, Clitomachus defended a strong form of
skepticism that he ascribed to Carneades, but after him Academic skepticism began
to soften. Philo first defended a radical skepticism but later held that it is possible
to have knowledge, albeit not of the type proposed by the Stoics.16 This decline
of Academic skepticism was followed by a total departure from it on the part of
Antiochus, who after espousing his teacher Philo’s skeptical outlook returned to
what he took to be the position of the Old Academy, embracing what most scholars
describe as a Stoicizing form of Platonism. With Antiochus begins the period known
as “Middle Platonism,” which extends to the early third century CE.17
The influence of Academic skepticism did not however cease, since it is still present
in Cicero (106–43 BCE), a student of both Philo and Antiochus, who considered
himself to belong to that philosophical tradition. In relation to skepticism, Cicero is
generally studied only as a major source of information on the skeptical Academy; his
own moderate skeptical outlook is not usually discussed in companions and general
introductions to ancient skepticism. For this reason, we decided to include a chapter
INTRODUCTION 7
entirely devoted to him.18 The presence of a mitigated form of Academic skepticism

is also found in Favorinus of Arles (80–160 CE), who claimed to be an Academic
skeptic, although he was also influenced by Pyrrhonism. In On the Best Method
of Teaching, which is our most important source for Favorinus’s philosophical
stance, Galen (129–210 CE) attacks the “younger” Academics, among whom was
Favorinus.19 Thus, even though the Academy as an institution ceased to exist in
the first century BCE, in the second century CE there was a group of Academics
who were skeptical and prominent enough to merit Galen’s attack. It should also
be noted that Plutarch (ca. 50–120 CE), Favorinus’s teacher, was sympathetic to
Academic skepticism, albeit of course not in its radical form. Not only did he seek
to defend Academic skepticism from the attacks of rival schools, but he also viewed
it as part of a unitary tradition that extended from Plato to himself. He thought
that the skeptical interpretation of Plato did justice to the aporetic aspect of the
latter’s dialogues, but did not take it to be incompatible with the fact that Plato
held doctrines.20
It is rather surprising that what may be called the “epistemological” position of
the Cyrenaic school (late fifth to mid-third centuries BCE) is not generally discussed
in presentations of Greek skepticism. It is true both that the Cyrenaics were not
interested in epistemological issues per se, but only insofar as these were relevant
to their ethical theory, and that they did not draw all the skeptical consequences of
their epistemology that they could have drawn. Nonetheless, their epistemological
stance seems to be more elaborate and subtle than those of the pre-Hellenistic
philosophers in whom one may detect skeptical elements, and it bears close relations
with both Academic and Pyrrhonian skepticism. The Cyrenaics claimed that we
have knowledge of our own “affections” (πάθη),21 but not of the properties of the
external objects that cause them because they are inapprehensible, and that we
cannot know the content of other people’s affections. They even coined curious
neologisms (“being whitened,” “being sweetened,” “being darkened”) so as to
restrict themselves to talking about the various ways they are affected and to avoid
any reference to the causes of the affections. Such neologisms are occasionally used
by Sextus. In short, the Cyrenaic epistemological stance should be taken into careful
consideration by anyone seriously interested in ancient skeptical thought.22
Another current in ancient skepticism was that of medical Empiricism, which
stretched from the third century BCE to the second century CE and was one of the
three main medical sects of the Hellenistic and Imperial ages, the other two being
the Rationalist or Dogmatic and the Methodical. Medical Empiricism did not in
fact constitute an independent skeptical tradition, since it was in intimate relation
with Pyrrhonism. Not only are there significant philosophical similarities between
the two outlooks (as the testimonies in Galen and Celsus attest), but we also know
that several Pyrrhonists were Empirical doctors, the first of whom being probably
Menodotus of Nicomedia, who flourished in the first part of the second century CE.
Sextus Empiricus himself was an Empiricist, as both the external evidence and his
sobriquet indicate, even though he explicitly distinguishes Pyrrhonism from medical
Empiricism and expresses more sympathy for medical Methodism (PH I 236–241).23
It has been suggested that medical Empiricism contributed to the revival of the
8 SKEPTICISM
Pyrrhonian philosophy in the first century BCE. The Empirical doctors also seem
to have been heavily influenced by the skeptical Academy, which is not in any case
strange given the close similarities between Pyrrhonism and Academic skepticism
and their mutual influence.24
I would like to conclude this brief introduction by making a few remarks about
the application of the words “skeptic” and “skepticism.” While Aenesidemus did not
probably use “skeptic” but only “Pyrrhonist,” Sextus prefers to employ “skeptic”
and “skepticism” (or “the skeptical way”) when talking about the Pyrrhonist and his
stance. Diogenes Laertius (third century CE) too uses “skeptic” and “Pyrrhonist” as
synonyms. As is widely known, the Greek σκεπτικός means “inquirer” and σκέψις
“inquiry.” It is commonly agreed that only in the early second century CE did those
words come to be generally used as a designation of the Pyrrhonist and his stance.
Now, the application of such terms in their technical sense to medical Empiricism
is plainly legitimate, given its intimate historical and philosophical connection with
Pyrrhonism. Their use in relation to Cyrenaicism, and to certain strands in Indian
philosophy, is grounded on the fact that in such positions one can identify points
of similarity with Pyrrhonism—and even with contemporary versions of skepticism.
As for the Academics, it is sometimes remarked that, unlike the Pyrrhonists,
they did not call themselves “skeptics,” nor did they describe their stance as
“skepticism.” Some of our sources, however, indicate that “skeptic” was also applied
to the Academics. For instance, in his Attic Nights (XI v 1), the second-century CE
antiquarian Aulus Gellius tells us that Pyrrhonists are referred to as “skeptics,” but
he later remarks that both Pyrrhonists and Academics are so designated (XI v 6).
Note in addition that, since Favorinus is likely the source of Gellius’s text, one might
infer that at least he and other Academics of the second century CE did actually call
themselves “skeptics”—which may be accounted for by their having been influenced
by Pyrrhonism. Be that as it may, the application of the label “skeptics” to figures such
as Arcesilaus, Carneades, Clitomachus, Philo, Cicero, and Favorinus is grounded on
the close similarities between their outlooks and the Pyrrhonian stance: for example,
the advocacy of suspension of judgment, the practice of arguing on both sides of a
question, and the dialectical style of argumentation. Such similarities explain, at least
in part, why in the second-century CE authors like Seneca, Epictetus, Galen, Lucian,
and even Favorinus tended to assimilate Academic skepticism to Pyrrhonism. There is
also a historical connection between the two traditions, since it is almost unanimously
accepted that Aenesidemus was a former member of the Academy whose desertion
seems to have been motivated by Philo’s abandonment of the rigorous skepticism of
the earlier Academics. It seems clear that the source of Aenesidemus’s argumentative
practice is to be found in the dialectical method of the skeptical Academy, and our
sources state that both Arcesilaus (PH I 232, cf. Acad. I 45) and Aenesidemus (DL IX
107, cf. PH I 30) conceived of suspension of judgment as their aim (τέλος). For such
reasons, some have suggested that Aenesidemus’s revival of Pyrrho’s outlook was
actually a return to radical Academic skepticism. In this connection, it must also be
borne in mind both that it is possible that Arcesilaus was influenced to some extent
by Pyrrho (PH I 234, DL IV 33) and that Aenesidemus viewed Pyrrho as a proponent
of suspension of judgment (DL IX 62).
INTRODUCTION 9
NOTES
1. On the historical and philosophical connections between Pyrrhonism and Buddhism,
besides Matthew Dasti’s chapter in this book, see e.g. Frenkian (1957), Flintoff
(1980), Bett (2000: 169–178), Kuzminski (2008), and Beckwith (2015).
2. For useful book-length general presentations of ancient skepticism, see Hankinson
(1998), Brochard (2002), Thorsrud (2009), and Bett (2010). For a shorter overview,
with extensive bibliographical references in the main European languages, see
Machuca (2011a,b,c).
3. This usage is common among French and Italian scholars. In English, the term “neo-
Pyrrhonism” is employed in contemporary analytic epistemology to refer to present-
day versions of Pyrrhonism or to the application of Pyrrhonian argumentative
strategies to current debates.
4. This is the view defended by Brunschwig (1994, 1999).
5. On the early Pyrrhonism of Pyrrho and Timon, besides Casey Perin’s chapter in this
book, see e.g. Ausland (1989), Brunschwig (1994), Conche (1994), Bett (2000),
Long (2006), Clayman (2009), and Svavarsson (2010).
6. Bett (1997) claims that in AM XI one detects traces of the Aenesideman variety
of skepticism.
7. See Woodruff (1988, 2010) and Bett (2000: ch. 4). On these and other aspects
of Aenesidemus’s philosophical stance, including his so-called Heracliteanism, see,
besides Luca Castagnoli’s chapter in this book, Polito (2004, 2014), Pérez-Jean
(2005), and Schofield (2007).
8. The Aenesideman modes are also expounded by Diogenes Laertius (DL IX 78–
88) and Philo of Alexandria (On Drunkenness 169–205). On these arguments, see
especially Annas and Barnes (1985); also Striker (1996) and Woodruff (2010).
9. The Agrippan modes are also expounded by Diogenes (DL IX 88–89). On these
modes, see especially Barnes (1990). It should be noted that those modes have
been the focus of much attention among contemporary epistemologists, who have
recognized the seriousness of the challenge posed by the so-called “Agrippa’s
trilemma” and proposed different ways to cope with it. On this issue, see Klein
(2008), Lammenranta (2008), and Machuca (2015).
10. Scholars have usually thought that Aenesidemus was not strictly the author of the
Ten Modes, but their compiler.
11. Sextus also expounds Two Modes of suspension of judgment (PH I 178–179), but
they actually work by the application of three of the Five Modes and most likely
derive from Agrippa.
12. See Machuca (2008: 51–57) for an overview of the varieties of skepticism found in
Sextus’s corpus and the tensions between them. For book-length studies on Sextus’s
Pyrrhonism, see especially Bailey (2002) and Perin (2010).
13. Following a usage found in some of our sources, I employ “Academics” to refer to
the Academic skeptics.
14. On the skepticism of Arcesilaus, in addition to Anna Maria Ioppolo’s chapter, see
Ioppolo (1986, 2000), Cooper (2004), and Vezzoli (2016).
15. On Carneades’s skepticism, besides Harald Thorsrud’s chapter in this book, see e.g.
Bett (1989, 1990), Thorsrud (2002, 2010), and Obdrzalek (2006).
10 SKEPTICISM
16. On Philo’s skepticism, besides Harold Tarrant’s chapter in the present book, see
Brittain (2001) and Glucker (2004).
17. On the philosophy of Antiochus and Middle Platonism in general, see, in addition
to Carlos Lévy’s chapter, Glucker (1978), Tarrant (1985), and Sedley (2012).
18. On Cicero’s skepticism, see also Thorsrud (2009: ch. 5), Woolf (2015), and
Nicgorski (2016).
19. On Favorinus’s skeptical views, besides Carlos Lévy’s chapter, see Ioppolo (1993,
1994, 2002) and Holford-Strevens (1997).
20. On Plutarch and skepticism, besides Lévy’s chapter, see De Lacy (1953) and
Bonazzi (2014).
21. It should be noted that a πάθος is that which happens to someone or something as
a result of being affected by an agent in the broad sense of this term. It refers to the
physical and/or psychological state or condition in which the affected person or thing
is. Although in ordinary English “affection” does not have this meaning anymore, it
has become in the specialist literature a technical term to translate πάθος because it
shows the connection that πάθος bears with the Greek verb πάσχειν (to affect).
22. On the skepticism of the Cyrenaic school, besides Richard Bett’s chapter, see Tsouna
(1998) and O’Keefe (2011, 2015).
23. On Sextus’s relationship with the medical schools, see Machuca (2008: 40–50).
24. On Menodotus and medical Empiricism more generally, besides James Allen’s
chapter in this book, see e.g. Edelstein (1967), Mudry (1982, 1990), Frede (1987,
1988, 1990), Hankinson (1987), and Perilli (2004).
REFERENCES
Annas, Julia and Jonathan Barnes. 1985. The Modes of Scepticism. Cambridge: Cambridge
University Press.
Ausland, Hayden. 1989. “On the Moral Origin of the Pyrrhonian Philosophy,” Elenchos
10: 359–434.
Bailey, Alan. 2002. Sextus Empiricus and Pyrrhonean Scepticism. Oxford: Oxford
University Press.
Beckwith, Christopher. 2015. Greek Buddha: Pyrrho’s Encounter with Early Buddhism in
Central Asia. Princeton: Princeton University Press.
Bett, Richard. 1989. “Carneades’ Pithanon: A Reappraisal of its Role and Status,” Oxford
Studies in Ancient Philosophy 7: 59–94.
Bett, Richard. 1990. “Carneades’ Distinction between Assent and Approval,” Monist 73:
3–20.
Bett, Richard. 1997. Sextus Empiricus: Against the Ethicists. English translation with an
introduction and commentary. Oxford: Oxford University Press.
Bett, Richard. 2000. Pyrrho, His Antecedents, and His Legacy. Oxford: Oxford University
Press.
Bett, Richard (ed.). 2010. The Cambridge Companion to Ancient Scepticism. Cambridge:
Cambridge University Press.
INTRODUCTION 11
Bonazzi, Mauro. 2014. “Plutarch and the Skeptics.” In M. Beck (ed.), A Companion to
Plutarch, 121–134. Malden and Oxford: Blackwell.
Brittain, Charles. 2001. Philo of Larissa: The Last of the Academic Sceptics. Oxford: Oxford
University Press.
Brochard, Victor. 2002. Les Sceptiques Grecs. 4th ed. Paris: Le livre de poche. (First
published in 1887.)
Brunschwig, Jacques. 1994. “Once again on Eusebius on Aristocles on Timon on Pyrrho.” In
his Papers in Hellenistic Philosophy, 190–211. Cambridge: Cambridge University Press.
Brunschwig, Jacques. 1999. “The Beginnings of Hellenistic Epistemology.” In K. Algra,
J. Barnes, J. Mansfeld, and M. Schofield (eds.), The Cambridge History of Hellenistic
Philosophy, 229–259. Cambridge: Cambridge University Press.
Clayman, Dee. 2009. Timon of Phlius: Pyrrhonism into Poetry. Berlin and New York:
Walter de Gruyter.
Conche, Marcel. 1994. Pyrrhon ou l’apparence. 2nd ed. Paris: Presses Universitaires de
France. (First published in 1973.)
Cooper, John. 2004. “Arcesilaus: Socratic and Sceptic.” In his Knowledge, Nature, and the
Good: Essays on Ancient Philosophy, 81–103. Princeton: Princeton University Press.
De Lacy, Phillip. 1953. “Plutarch and the Academic Sceptics,” The Classical Journal 49:
79–85.
Edelstein, Ludwig. 1967. “Empiricism and Scepticism in the Teaching of the Greek
Empiricist School.” In his Ancient Medicine: Selected Papers of Ludwig Edelstein, 195–
203, edited by O. Temkin and C. Temkin. Baltimore: Johns Hopkins University Press.
Flintoff, Everard. 1980. “Pyrrho and India,” Phronesis 25: 88–108.
Frede, Michael. 1987. “The Ancient Empiricists.” In his Essays in Ancient Philosophy, 234–
260. Oxford: Oxford University Press.
Frede, Michael. 1988. “The Empiricist Attitude Towards Reason and Theory.” In R. J.
Hankinson (ed.), Method, Metaphysics and Medicine: Studies in the Philosophy of
Ancient Medicine, special volume of Apeiron 21, 79–97. Edmonton: Academic Printing
and Publishing.
Frede, Michael. 1990. “An Empiricist View of Knowledge: Memorism.” In S. Everson (ed.),
Companions to Ancient Thought I: Epistemology, 225–250. Cambridge: Cambridge
University Press.
Frenkian, Aram. 1957. “Sextus Empiricus and Indian Logic,” The Philosophical Quarterly
(India) 30: 115–126.
Glucker, John. 1978. Antiochus and the Late Academy. Göttingen: Vandenhoeck &
Ruprecht.
Glucker, John. 2004. “The Philonian/Metrodorians: Problems of Method in Ancient
Philosophy,” Elenchos 25: 99–152.
Greco, John (ed.). 2008. The Oxford Handbook of Skepticism. New York: Oxford
University Press.
Hankinson, Robert. 1987. “Causes and Empiricism: A Problem in the Interpretation of
Later Greek Medical Method,” Phronesis 32: 329–348.
Hankinson, Robert. 1998. The Sceptics. 2nd ed. London and New York: Routledge.
(First published in 1995.)
12 SKEPTICISM
Holford-Strevens, Leofranc. 1997. “Favorinus: The Man of Paradoxes.” In J. Barnes and

M. Griffin (eds.), Philosophia Togata II: Plato and Aristotle at Rome, 188–217. Oxford:
Oxford University Press.
Ioppolo, Anna Maria. 1986. Opinione e scienza. Il dibattito tra Stoici e Accademici nel terzo
e secondo secolo a. C. Napoli: Bibliopolis.
Ioppolo, Anna Maria. 1993. “The Academic Position of Favorinus of Arelate,” Phronesis
38: 183–213.
Ioppolo, Anna Maria. 1994. “Accademici e pirroniani nel II secolo D.C.” In A. Alberti
(ed.), Realtà e ragione: Studi di filosofia antica, 85–103. Firenze: Olschki.
Ioppolo, Anna Maria. 2000. “Su alcune recenti interpretazioni dello scetticismo
dell’Academia. Plutarch. Adv. Col. 26, 1121F–1122F: una testimonianza su Arcesilao,”
Elenchos 21: 333–360.
Ioppolo, Anna Maria. 2002. “Gli accademici neōteroi nel secondo secolo d.C.,” Méthexis
15: 45–70.
Klein, Peter. 2008. “Contemporary Responses to Agrippa’s Trilemma.” In Greco 2008,
484–503.
Kuzminski, Adrian. 2008. Pyrrhonism: How the Ancient Greeks Reinvented Buddhism.
Lanham: Lexington Books.
Lammenranta, Markus. 2008. “The Pyrrhonian Problematic.” In Greco 2008, 9–33.
Long, Anthony. 2006. “Timon of Phlius: Pyrrhonist and Satirist.” In his From Epicurus
to Epictetus: Studies in Hellenistic and Roman Philosophy, 70–95. Oxford: Oxford
University Press. (First published in 1973.)
Machuca, Diego. 2008. “Sextus Empiricus: His Outlook, Works, and Legacy,” Freiburger
Zeitschrift für Philosophie und Theologie 55: 28–63.
Machuca, Diego. 2011a. “Ancient Skepticism: Overview,” Philosophy Compass 6: 234–
245.
Machuca, Diego. 2011b. “Ancient Skepticism: Pyrrhonism,” Philosophy Compass 6: 246–
258.
Machuca, Diego. 2011c. “Ancient Skepticism: The Skeptical Academy,” Philosophy
Compass 6: 259–266.
Machuca, Diego. 2015. “Agrippan Pyrrhonism and the Challenge of Disagreement,”
Journal of Philosophical Research 40: 23–39.
Mudry, Philippe. 1982. La préface du De medicina de Celsus. Lausanne: Imprimerie des
Arts et Métiers.
Mudry, Philippe. 1990. “Le scepticisme des médecins empiriques dans le traité De
la médecine de Celse: modèles et modalités.” Le scepticisme antique: perspectives
historiques et systématiques, 85–96. Cahiers de la Revue de Théologie et de Philosophie
15. Genève: Lausanne & Neuchâtel.
Nicgorski, Walter. 2016. Cicero’s Skepticism and His Recovery of Political Philosophy.
Basingstoke: Palgrave.
Obdrzalek, Suzanne. 2006. “Living in Doubt: Carneades’ Pithanon Reconsidered,” Oxford
O’Keefe, Timothy. 2011. “The Cyrenaics vs. the Pyrrhonists on Knowledge of Appearances.”
In D. Machuca (ed.), New Essays on Ancient Pyrrhonism, 27–40. Leiden: Brill.
INTRODUCTION 13
O’Keefe, Timothy. 2015. “The Sources and Scope of Cyrenaic Scepticism.” In U. Zilioli
(ed.), From the Socratics to the Socratic Schools, 99–113. New York: Routledge.
Pérez-Jean, Brigitte. 2005. Dogmatisme et scepticisme: L’héraclitisme d’Énésidème. Lille:
Septentrion.
Perilli, Lorenzo. 2004. Menodoto di Nicomedia: Contributo a una storia galeniana della
medicina empirica. München and Leipzig: Saur.
Perin, Casey. 2010. The Demands of Reason: An Essay on Pyrrhonian Scepticism. Oxford:
Polito, Roberto. 2004. The Sceptical Road: Aenesidemus’ Appropriation of Heraclitus.
Leiden and Boston: Brill.
Polito, Roberto. 2014. Aenesidemus of Cnossus: Testimonia. Cambridge: Cambridge
University Press.
Schofield, Malcolm. 2007. “Aenesidemus: Pyrrhonist and ‘Heraclitean’.” In A. M. Ioppolo
and D. Sedley (eds.), Pyrrhonists, Patricians, Platonizers: Hellenistic Philosophy in the
Period 155–86 BC, 269–338. Napoli: Bibliopolis.
Sedley, David (ed.). 2012. The Philosophy of Antiochus. Cambridge: Cambridge University
Press.
Striker, Gisela. 1996. “The Ten Modes of Aenesidemus.” In her Essays on Hellenistic
Epistemology and Ethics, 116–134. Cambridge: Cambridge University Press.
(First published in 1983.)
Svavarsson, Svavar. 2010. “Pyrrho and Early Pyrrhonism.” In Bett 2010, 36–57.
Tarrant, Harold. 1985. Scepticism or Platonism? The Philosophy of the Fourth Academy.
Thorsrud, Harald. 2002. “Cicero on His Academic Predecessors: The Fallibilism of
Arcesilaus and Carneades,” Journal of the History of Philosophy 40: 1–18.
Thorsrud, Harald. 2009. Ancient Scepticism. Stocksfield: Acumen.
Thorsrud, Harald. 2010. “Arcesilaus and Carneades.” In Bett 2010, 58–80.
Tsouna, Voula. 1998. The Epistemology of the Cyrenaic School. Cambridge: Cambridge
University Press.
Vezzoli, Simone. 2016. Arcesilao di Pitane: l’origine del Platonismo neoaccademico. Analisi
e fonti. Turnhout: Brepols.
Woodruff, Paul. 1988. “Aporetic Pyrrhonism,” Oxford Studies in Ancient Philosophy 6:
139–168.
Woodruff, Paul. 2010. “The Pyrrhonian Modes.” In Bett 2010, 208–231.
Woolf, Raphael. 2015. Cicero: The Philosophy of a Roman Sceptic. New York: Routledge.
CHAPTER ONE
The Cyrenaics
and Skepticism
RICHARD BETT
1 SHOULD THE CYRENAICS COUNT AS SKEPTICS?

The Cyrenaics are regularly described as having a skeptical epistemology.1 But some
would say that the Cyrenaics do not belong in a book on skepticism. Prominent
among them would be the Pyrrhonist skeptic Sextus Empiricus. At the end of the
first book of his Outlines of Pyrrhonism, Sextus discusses several philosophies that
were thought to be equivalent or at least similar to skepticism. It is not clear who
thought so; were they skeptics themselves, or adherents of some other philosophical
outlook? But Sextus is opposed to the whole project, arguing in each case that there
are central differences between skepticism as he understands it and the philosophy
in question. And one of these philosophies is that of the Cyrenaics (PH 1.215).
Sextus points to two differences. First, the Cyrenaics posit pleasure as the end
or aim (τέλος) of life, whereas the skeptic’s end is ἀταραξία, freedom from worry;
referring to his earlier chapter on the skeptic’s end (PH 1.25–30), Sextus counters
that a firm conviction that pleasure is the end will (whether or not one actually
achieves it) lead to a great deal of trouble—that is, the opposite of ἀταραξία. The
earlier chapter did not actually mention the specific case of pleasure, but argued
that taking anything to be by nature good or bad, and thereby caring intensely
about its presence or absence, will lead to turmoil; the same point can be found
elsewhere (PH 3.235–238, AM 11.110–140). There are really two distinct issues
here: the identity of the ends posited by each school, and the ways in which they
are posited. First, pleasure and ἀταραξία are two different things, so the Cyrenaics
do not have the same end as the skeptics. But second, holding to one’s view of
the end (whatever one takes it to be) as a matter of firm conviction (in the present
passage, διαβεβαιούμενος) will lead to results that are actually contrary to the
skeptic’s end. The Cyrenaics are assumed to adopt their end in this committed way.
By contrast, when Sextus states the skeptics’ own end, he is careful to put it much
less decisively. He says: “Up to now we say that the skeptic’s end is freedom from
worry” (PH 1.25).2 Rather than being posited, as ends usually are, as the end for
human beings in general—as what humans as such should, or naturally do, strive
THE CYRENAICS AND SKEPTICISM 15
toward—it is explicitly restricted to the skeptics themselves; and it is put forward

tentatively, as a matter of their experience “up to now,” rather than definitively.
In both respects it conforms to the suspension of judgment (ἐποχή) that Sextus
regularly identifies as the centerpiece of his skeptical attitude.
The centrality of suspension of judgment for the Pyrrhonist skeptic is also the basis
of the second contrast Sextus draws between his school and the Cyrenaics. Here and
elsewhere, Sextus distinguishes between the way objects strike us and the objects
themselves, and says that concerning the latter, here referred to as “the external
underlying things” (τῶν ἐκτὸς ὑποκειμένων), the skeptics suspend judgment,3
whereas the Cyrenaics “declare that they have an inapprehensible nature.” The
Cyrenaics, then, assert that these objects are inherently unknowable, and that is
an intellectual commitment, just as much as if one were to claim that they could
be known; in modern scholarship, though Sextus does not use this terminology
himself, such a position is generally called “negative dogmatism,” and Sextus is
clear that this is different from skepticism. Now, the other side of this distinction—
what I referred to as “the way things strike us”—is what led, as Sextus reports
it, to the claim that the Cyrenaic position is the same as skepticism; the reason
given is that “it too [i.e., skepticism] says that only the ways we are affected [πάθη]
are apprehended,” just as the Cyrenaics did.4 Sextus does not necessarily endorse
this characterization of skepticism; it is presented simply as the support offered by
those who claim the identity. In fact, as I shall suggest in closing, he would be well
advised not to endorse it; even here, the Cyrenaic and the skeptical positions are not
identical. But his point in drawing this second contrast is that, even if we accept the
supposed point of similarity, the Cyrenaics differ importantly from the skeptics in
what they are prepared to say about the objects themselves.
So Sextus, a card-carrying skeptic, explicitly repudiates the idea that the
Cyrenaics are skeptics. Why, then, should we treat them as skeptics? The answer
is that skepticism has taken numerous different forms in the history of philosophy.
Although the Cyrenaic position may differ in crucial ways from skepticism as
understood in the ancient period—the hallmark of which, both in Sextus’s hands
and elsewhere, was suspension of judgment of one form or another—anyone
approaching the Cyrenaics with a background in reflections about skepticism in the
modern period will immediately see their philosophy as having a strong skeptical
component. For skepticism in the modern period has generally been regarded as
centering around the denial of the possibility of knowledge in some domain: about
the “external world,” about other minds, about induction, and so on. As we have
already noted, the Cyrenaics, if Sextus is right, hold something that clearly has at
least a family resemblance with external world skepticism; we shall also see hints of
a position resembling other minds skepticism. So to the extent that the Cyrenaics
qualify as skeptics, it is in an anachronistic fashion, because of the resemblance of
some of their views to views that would now be called “skeptical,” not because of
anything that would have been recognized as skepticism in the ancient world.
How close the Cyrenaic position is even to varieties of skepticism discussed today
is an interesting and subtle question; I shall suggest in what follows that we should
be careful not to try to assimilate them too closely. If I am right about this last point,
16 SKEPTICISM
then the Cyrenaics serve as an illuminating foil to skepticism in both the ancient
and the modern periods. In order to develop this line of thought, however, we
need to look more closely at what the Cyrenaic position actually was. I shall say no
more about their ideas on pleasure, nor will I address the very interesting questions
surrounding their picture of the relation between pleasure and happiness,5 except
to say that this side of their thought seems to reveal the same tendency to restrict
their concerns to what is immediately accessible to experience as we find in their
epistemological ideas. It is the latter on which I shall concentrate from now on.
2 THE VIEW IN MORE DETAIL

The passage of Sextus referred to in the previous section attributes to the Cyrenaics the
view that only the πάθη, and not the objects causing them, are or can be apprehended
(PH 1.215). We find the same attribution in another, considerably longer passage of
Sextus in which the Cyrenaic view is summarized (AM 7.191–200—on the present
point see especially 191, 195); this is a section of Sextus’s lengthy review, in the first
book of Against the Logicians, of positions, positive and negative, concerning the
criterion of truth. Several other authors also cite them as holding that only the πάθη
can be apprehended: the Anonymous Commentator on the Theaetetus (on 152b,
col. 65, 29–39), Aristocles of Messene in a passage on the Cyrenaics quoted by
Eusebius (Praep. evang. 19.1–2), Eusebius himself in introducing this passage (Praep.
evang. 18.31),6 and Diogenes Laertius in his life of Aristippus (DL 2.92). Despite the
close verbal agreement among all these passages, there is room for some doubt as to
whether the terminology of apprehension (κατάληψις) goes back to the Cyrenaics’
own original formulation of the position. As far as we can tell, this terminology
was first given an epistemological connotation by Zeno of Citium, the first Stoic
(Cicero, Acad. 2.145). Now, there is an unresolved debate (and the evidence may
be insufficient to settle this) over whether the Cyrenaics’ epistemology was part of
their position from its beginning with the older Aristippus, a disciple of Socrates, or
was added later—probably by his grandson, also named Aristippus.7 But although
the older Aristippus’s precise dates are uncertain, and those of his grandson even
more so, it is clear that both must have predated Zeno, whose birthdate is generally
placed in the later 330s BCE.8 Hence, if the report that Zeno invented κατάληψις
as a philosophical term is correct, the Cyrenaic view is likely to have been developed
too early to make use of it. By the time of the authors who are our sources, it was
no longer restricted to Stoicism, but was common philosophical parlance for any
kind of secure cognitive awareness; and this would explain why it became standard
in describing the Cyrenaic position, even if the Cyrenaics themselves did not use
this language.
However, this need not trouble us, since what is clearly the same position is
also reported to us without the language of apprehension.9 Plutarch tells us that
the Cyrenaics saw each πάθος as “having its own evidence [ἐνάργειαν] within
itself, which is not to be challenged” (Adv. Col. 1120E). In Hellenistic and later
philosophical contexts, ἐνάργεια regularly indicates something’s being a matter of
plain and incontrovertible experience, and hence it plays a role equivalent to that
of apprehension in the other passages: it specifies that there is no room for doubt
or dispute about how these πάθη come across to us.10 By contrast, Plutarch says
that they “thought that the assurance [πίστιν] coming from them [i.e., from the
πάθη] was not sufficient for firm declarations about the objects”—that is, the objects
causing those πάθη (1120D); examples are whether honey is sweet, an olive-shoot
bitter, hail cold, wine warm, the sun bright, or night dark (1120E). Similarly, Cicero
tells us that the Cyrenaics held that “they do not know what has which color or
which sound, but only sense that they are affected in a certain way” (Acad. 2.76.).11
The outlines of the view, then, are clear enough: we can have a certain kind of
secure grasp of something about the πάθη themselves, but the objects that caused
them are not amenable to any such grasp. It is now time to get clearer about what
this amounts to.
First of all, what is a πάθος? Most generally, it is something that happens to
someone or something, or that someone or something undergoes. Most often, but
certainly not always, it is human beings who are described as having πάθη, and this
is clearly true for the Cyrenaics; it is our own πάθη of which we are said to have
apprehension. In human contexts the word is frequently translated “emotions,”
“passions,” or “feelings”; these can often be appropriate, but in the present case
they are too narrow. From the texts already adduced, it is clear that the πάθη
include cases of ordinary sense-perception, with no particular affective component;
this is suggested by the frequent talk of the external objects (the character of which
we cannot apprehend) versus our πάθη of them, and is confirmed by the examples
from Plutarch and Cicero in the previous paragraph. Indeed, these appear to be the
cases of most immediate interest for our purposes, although pleasure and pain are
also included by the Cyrenaics among the πάθη (Sextus, AM 7.199, DL 2.86, Cic.
Acad. 2.76). Above, in an attempt to capture this diversity, I translated πάθη as
“ways we are affected”; it is not clear that any single English word will do.12 From
now on I will simply use the Greek word.
What, then, is apprehended on the Cyrenaic view, and what cannot be
apprehended? On the side of the objects, it seems clear enough that we cannot
apprehend the objects’ real or intrinsic character; we can know how we are affected
by these objects, but not how they are in themselves.13 And it may seem that the
answer on the side of the subject having the πάθη is equally clear: one apprehends,
or has infallible access to, one’s own subjective experience of these objects. This may
also seem to be supported by Sextus’s talk of the πάθη as “apparent” (φαινόμενα)
to us (AM 7.194, 197).14 But the other sources do not use this language, and it may
well be that Sextus is translating the Cyrenaics’ position into language with which
he himself and his assumed audience are familiar (so Tsouna 1998: chapter 4.V).
Sextus also reports the Cyrenaic view using a different form of language; the
same form of words also appears in Plutarch’s report, and Plutarch explicitly claims
to be giving the view as the Cyrenaics themselves present it—by contrast with the
Epicurean Colotes, whom he criticizes for misrepresenting it (Adv. Col. 1120D–E).
According to these sources, in characterizing the πάθη the Cyrenaics said that we
are “sweetened,” “whitened,” “reddened,” “chilled,” “heated,” etc. (AM 7.192–
193, Adv. Col. 1120E); in what appears to be interchangeable terminology, Sextus
18 SKEPTICISM
also has them saying that we are “affected [or “activated,” κινεῖται] yellowly,”
“redly,” etc. (AM 7.192, 198). Now, to say that we are “sweetened” or “whitened”
does not seem the most natural language to use if one wanted to talk purely about
experiencing something as sweet or white. Rather, it sounds as if something in us is
being said to turn sweet or white.15 And this recalls another, roughly contemporary
account of sense perception—at least as some scholars have read it. On one
interpretation of Aristotle’s account of sense perception,16 the sense organs actually
acquire, when one perceives, the sensory qualities belonging to the objects of sense-
perception; for example, the interior of one’s eye turns red when one sees a red
tomato. Whether or not this is right about Aristotle, it seems a natural reading of the
Cyrenaic terminology—thinking now, of course, about the perceiver’s side alone.
For someone to be “reddened” is for something in that person to become red, the
sense organ for vision being the obvious candidate; and the same would be true,
mutatis mutandis, for “sweetened,” “chilled,” and the rest. The other language in
Sextus (“affected redly,” etc.) does not point so clearly in this direction, but seems
at least compatible with the reading just suggested.
This reading might be challenged on the ground that the peculiar verbal forms
reported in our sources extend considerably beyond the kind of basic sensory
properties considered so far. Plutarch’s report includes the claim that the Cyrenaics
also spoke of being “walled,” “horsed,” or “manned” (Adv. Col. 1120D). Now
it would be hard to understand these other than as shorthand for “experience a
perception as of a wall/horse/man”; that perceiving subjects in any sense become
walls, horses, or men during these perceptions would surely be absurd. However,
there is good reason not to accept this account of what the Cyrenaics said. Plutarch
presents these terms as coming from Colotes’s portrait of the Cyrenaic view, which
he says was offered in a spirit of mockery (κωμῳδῶν, 1120D) and which, as noted
above, he explicitly rejects, contrasting it with his own account of what the Cyrenaics
actually said, where only sensory properties are included. No other evidence includes
these outlandish terms, which one would expect to have excited broader ridicule if
they had really been part of the view.17 Sextus’s summary does say that “the person
who presses on his eye is activated as if by two things,” and goes on to use the verb
“doubled” to express this (AM 7.192–193); here is another purported Cyrenaic
coinage outside the range of sensory properties, and given Plutarch’s evidence we
may find this suspect as well. But, unlike in the case of “walled,” etc., it is not hard to
understand “doubled” as referring to an actual change in the sense organ; pressing
on your eye might well be thought to divide the material inside into two. Indeed, the
reference to pressing on the eye may even support the general reading of these verbs
as referring to physiological changes rather than to subjective experiences.
If this is indeed what the Cyrenaics mean, we need not deny that they are
interested in the experience of tasting something as sweet or seeing it as white; in
speaking of the πάθη as apprehensible (or whatever equivalent terms they actually
used), they clearly have in mind our grasp of something of which we are subjectively
aware.18 Rather, the point is that they do not have a sharp distinction, such as is
taken for granted in Descartes and much early modern philosophy, between mental
and physical occurrences. One tastes something as sweet, and that thing turns one’s
tongue sweet. Are these the very same event described in two different ways? There
is no indication in the sources that the Cyrenaics have any concern to distinguish
them. And if not, then our πάθη, the ways we are affected by objects, are not purely
mental events; they are also effects on our bodies. By the same token, they are not
purely subjective, but also things happening in the world, describable from a third-
person perspective.19 And this is one reason why one should not try to assimilate the
Cyrenaics too closely to the typical figure of “the skeptic” in modern philosophy.
While it is not false that the Cyrenaics deny that we can know about the external
world, “external” in this case means “external to our bodies,” not “external to our
subjective awareness.”20
Another reason why it is hazardous to assimilate the Cyrenaics to the modern
skeptic is that they do not seem to question the existence of the external world (in
any sense of “external”). Instead, the question is consistently whether we know what
the objects causing our πάθη are like. Their answer is that we do not know this; but
that there are objects out there, to which our πάθη are somehow a response, is taken
for granted. For example, Sextus tells us that, on the Cyrenaic view, “that we are
whitened and we are sweetened, it is possible to say . . . irrefutably; but that the thing
productive of the effect is white or sweet, it is not possible to assert” (AM 7.191).
And Aristocles reports that “they said that, when being burned or cut, they knew that
they were affected in some way; but whether the thing burning them was fire or the
thing cutting them was iron, they were not able to say” (in Eusebius, Praep. evang.
14.19.1). Sextus’s report once seems to show them inching toward a question about
existence; he says that “the external thing productive of the effect is perhaps (τάχα)
a being, but it is not apparent to us” (AM 7.194). But there is no indication that they
ever followed through on that suggestion and explicitly raised the question.21 One
might say that this is simply a failure to see where the logic of their position was
leading them. But if, as I have suggested, the πάθη are physiological events as much
as experiences, it is not perhaps surprising; a view according to which there were
human bodies, but maybe no world for them to inhabit, would be a strange view
indeed. So if it is true that “external” means something different for the Cyrenaics
from what it means for a Cartesian, this is not unconnected with their focus on the
nature of external objects rather than their existence.
In addition to doubts about our knowledge of the world around us, Sextus’s
report includes doubts about our knowledge of the πάθη of others, which seems
to anticipate what is today called “skepticism about other minds.” No other source
mentions this, but it is hard to see why Sextus would simply invent it in the course
of a description of what the Cyrenaics say. It appears to be introduced as part of the
argument for external objects being inapprehensible; the idea seems to be that we
cannot tell what objects are like because we cannot compare our πάθη with those of
others and confirm that they are qualitatively the same (AM 7.196–197). One may
wonder why, even if we could do this, it would give us knowledge of the way things
are; might not all humans see things distortedly?22 But I am not sure the argument
has to deny this; the idea may be that, since we cannot even tell whether we see
things the same way as others—which, even if we could, might not be enough—
the prospects for getting to a secure grasp of the nature of those things are truly
20 SKEPTICISM
hopeless. In any case, Sextus says that according to the Cyrenaics, “each person
grasps the πάθος that is his own,” but that since we do not grasp each other’s πάθη,
“it is rash to say that what appears a certain way to me also appears that way to the
next person.” If our own πάθη are all we can be sure about, it is indeed not hard to
see how the πάθη of others might be judged just as elusive as the character of the
objects that produce them. But again, just as the existence of the external world does
not seem to be an issue for the Cyrenaics, nor is there any hint of solipsism. Nothing
in this passage suggests any doubt that there are other minds; what goes on in them
is the question. Again, part of the reason for this may be that the πάθη are, as I put
it earlier, not purely mental,23 but the connection is not as obvious as in the case of
the external world.
So the Cyrenaics interestingly anticipate what in modern times has been called
skepticism about the external world, and also, if Sextus is to be believed, skepticism
about other minds. There are also interesting and important differences. But certainly
they are much closer to these modern forms of skepticism than are any other ancient
thinkers—including, to return to my opening point, the ancient thinkers who called
themselves skeptics. Finally, we can now see why Sextus ought to distinguish himself
from the Cyrenaics not only, as he says, on the inapprehensibility of external things,
but also on the apprehensibility of the πάθη, which was the point on which those
who wished to assimilate the two schools were said to rely (PH 1.215). Sextus does
not question the way things appear to him; indeed, he says that this is not a topic
for discussion (PH 1.22). But this is not the same as asserting that a whole class
of items, the πάθη, can be infallibly known about—which is what the claim of
apprehensibility amounts to; that would be a theoretical claim.24 The Cyrenaics
have an epistemological doctrine of which apprehensibility of the πάθη and the
inapprehensibility of their objects are interconnected parts; both should equally be
anathema to Sextus. True skeptics in the ancient sense are not philosophers, in any
usual sense of the term; the Cyrenaics are.
NOTES
1. See, e.g., Matson (2006) and Taylor (2012), both in standard reference works.
2. Translations are my own.
3. The sentence also contains the difficult phrase ὅσον ἐπὶ τῷ λόγῳ: “as far as the
argument goes,” “as far as reason is concerned.” However precisely one translates it,
the point seems to be that the skeptic makes no attempt to theorize about the nature
of these “external” objects. See Brunschwig (1994: 250).
4. I shall return to the question of what the πάθη are and how the word should be
translated; the somewhat cumbersome translation “ways we are affected” is intended
as a place-holder until that point.
5. On this, see especially DL 2.87–88, and for two partially differing interpretations, see
O’Keefe (2002) and Tsouna-McKirahan (2002).
6. On this division of the text between Eusebius and Aristocles, see Chiesara (2001:
fragment F5 and Introduction, section II, especially xxvi–xxviii).
7. In this chapter, for reasons of space, I do not attempt to address the history of the
school or possible differences among individual members. Our sources in any case
consistently report Cyrenaic epistemology as adopted by the school in general (unlike
their ethics, where we hear of a number of individual differences—see especially
DL 2.93–100); the only question is whether it was part of their philosophy from
the start.
8. The evidence on the older Aristippus’s chronology is conflicting and hard to
interpret; see Giannantoni (1990: Note 13, vol. 4, 135–140), and for the ancient
sources, Giannantoni (1990: vol. 2, section IV-A). There are no precise chronological
indications for the younger Aristippus. But someone who was an adult in Socrates’s
lifetime must have been born no later than the 420s, which almost certainly puts the
younger Aristippus at least a generation earlier than Zeno.
9. Here (as on almost every other point in this section) I am in agreement with Tsouna
(1998); see chapter 4.I on the present issue. This book is an essential resource for
anyone wishing to pursue this subject further.
10. Hence ἀπερίσπαστος can confidently be translated (as above) “not to be challenged,”
not merely “not challenged.” Adjectives ending in –τος routinely have this
descriptive/prescriptive ambiguity, but the occurrence of ἐνάργεια shows that the
prescriptive translation is not misleading. (Incidentally, ἐνάργεια is an emendation.
But the mss. reading ἐνέργεια, “activity” makes no sense, and the correction is an
easy one.)
11. In the same section, Cicero says that “they deny that anything external can be
perceived, and that they perceive only those things that they sense with internal
touch.” But “perceive” (percipi) in Cicero’s Academica is used as equivalent to
the Greek καταλαμβάνειν, “apprehend,” even though his actual translation of
καταλαμβάνειν is comprehendere (2.145–146); since my present interest is in
evidence independent of the notion of apprehension, I leave these words aside.
12. The translation “affections” is still sometimes used, presumably intended as
shorthand for “ways we are affected”; but in contemporary English the word
“affections” has connotations that are hopelessly misleading in the current context.
In the past I have used “effects,” but this fails to bring out the fact that a πάθος is
something that happens to a subject (be that subject animate or inanimate).
13. Zilioli (2012) has argued that this is because, according to the Cyrenaics, there
is no way things are in themselves; things are indeterminate in their nature. On
this reading the Cyrenaics would resemble Pyrrho, on some interpretations of his
thought. I have argued against this in Bett (2015); see also O’Keefe (2013).
14. See, however, Everson (1991), which argues that even the category of the apparent,
in the hands of the Pyrrhonists, should not be construed in subjectivist terms. For
some well-founded doubts about this, see Fine (2003), especially section 6.
15. Diego Machuca commented that the point may simply be to put the focus on the
ways in which one is affected, and to avoid any reference to the causes. I quite
agree that this is the intention. But what are the ways in which one is affected? This
peculiar form of language suggests that the Cyrenaics mean to identify something
other than pure subjective experience. The account that I go on to sketch seems to
me the most natural way to make sense of this idiosyncratic terminology; given the
state of our evidence, it would be rash to claim anything stronger.
22 SKEPTICISM
16. E.g., Ross (1923: 137), Sorabji (1974).

17. Here I follow Kechagia (2011: chapter 8.2). Tsouna (1998: chapter 6.III) also
rejects Colotes’s picture.
18. My account is thus not inconsistent with that of Fine (2003: section 5) Fine’s focus
is on the subjective aspect of the Cyrenaics’ πάθη; her concern is to rebut the idea,
stemming from Burnyeat (1982), that ancient Greek philosophy does not recognize
a category of the subjective that is anything like that which we find in philosophy
since Descartes. I think that she is successful in this aim. But Fine does not deny
that these πάθη are also physical states, though she is not particularly interested in
their character.
19. Of course, whether these events could ever be reliably observed from a third-person
perspective is put in doubt by the other side of the Cyrenaic position, that the
objects causing the πάθη are not apprehensible. That you are “reddened” may be
apprehensible to you, and this may have an objective aspect (your eye’s turning
red) as well as a subjective one (something’s looking red to you). But your being
reddened is not apprehensible to anyone else, because to everyone else your eye’s
turning red is not a πάθος, but an external event.
20. That the conception of one’s body as “external” originates with Descartes is not a
new idea; it was first made prominent in Burnyeat (1982).
21. Another passage of Sextus (AM 6.53) claims that the Cyrenaics thought that only
the πάθη exist, which sounds like an outright denial, not just a questioning,
of the existence of the external world. But this is in a polemical context in
which some highly dubious claims are also made about Plato and Democritus;
it should almost certainly be discounted. See Tsouna (1998: 80–81) and Bett
(2013: 177–178).
22. Tsouna (1998: chapter 7) raises this objection and then explores alter-
native interpretations.
23. Tsouna (1998: 93–95) has some suggestions in that direction.
24. On this, see further O’Keefe (2011). Perin (2010) argues that Sextus takes us to
have noninferential knowledge of how things appear to us. (Fine (2003: 209) also
considers this possibility, but is uncertain.) But, first, this is an inference from other
things that he says; he certainly does not say anything that could be translated by
these words. What we think an ancient philosopher was committed to holding may
not always correspond to what that ancient philosopher actually held. Second,
even if we accept Perin’s view, it does not follow that Sextus would or should
be comfortable with the dogmatic notion of apprehension; one can claim in a
commonsense way to know things (even noninferentially, as in “I just know it”),
without having or wishing to have any precise commitments concerning what that
amounts to. Both Perin (2010: 160) and Fine (2003: 208) draw attention to the
fact that Sextus reports the view of those who would assimilate the Cyrenaics to
the skeptics with the words “it too says that only the πάθη are apprehended,” and
does not reject the implication that this is the skeptic’s view. However, as I noted
earlier, he also does not endorse this implication; and my claim here is that he is
well-advised not to do so.
REFERENCES
Bett, Richard. 2013. “A Sceptic Looks at Art (but not Very Closely): Sextus Empiricus on
Music,” International Journal for the Study of Skepticism 3: 155–181.
Bett, Richard. 2015. “Pyrrho and the Socratic Schools.” In U. Zilioli (ed.), From the
Socratics to the Socratic Schools, 149–167. London: Routledge.
Brunschwig, Jacques 1994. “The ὅσον ἐπὶ τῷ λόγῳ Formula in Sextus Empiricus.” In his,
Papers in Hellenistic Philosophy, 244–258. Cambridge: Cambridge University Press.
Burnyeat, Myles. 1982. “Idealism and Greek Philosophy: What Descartes Saw and Berkeley
Missed,” Philosophical Review 91: 3–40.
Chiesara, Maria Lorenza. 2001. Aristocles of Messene: Testimonia and Fragments, edited
with translation and commentary. Oxford and New York: Oxford University Press.
Everson, Stephen. 1991. “The Objective Appearance of Pyrrhonism.” In S. Everson (ed.),
Companions to Ancient Thought 2: Psychology, 121–147. Cambridge: Cambridge
University Press.
Fine, Gail. 2003. “Subjectivity, Ancient and Modern: The Cyrenaics, Sextus, and Descartes.”
In J. Miller and Brad Inwood (eds.), Hellenistic and Early Modern Philosophy, 192–231.
Giannantoni, Gabriele. 1990. Socratis et Socraticorum Reliquiae, 4 vols. Naples: Bibliopolis.
Kechagia, Eleni. 2011. Plutarch Against Colotes: A Lesson in the History of Philosophy.
Oxford: Oxford University Press.
Matson, Wallace. 2006. “Cyrenaics.” In D. M. Borchert (ed.), The Encyclopedia of
Philosophy, 2nd ed., vol. 2, 619–620. Farmington Hills, MI: Thomson Gale.
O’Keefe, Tim. 2002. “The Cyrenaics on Pleasure, Happiness, and Future-Concern,”
Phronesis 47: 395–416.
O’Keefe, Tim. 2011. “The Cyrenaics vs. the Pyrrhonists on Knowledge of Appearances.” In
D. Machuca (ed.), New Essays on Ancient Pyrrhonism, 27–40. Leiden and Boston: Brill.
O’Keefe, Tim. 2013. Review of Zilioli (2012), Notre Dame Philosophical Reviews
2013.04.14.
Perin, Casey. 2010. “Scepticism and Belief.” In R. Bett (ed.), The Cambridge Companion to
Ancient Scepticism, 145–64. Cambridge: Cambridge University Press.
Ross, David. 1923. Aristotle. London: Methuen.
Sorabji, Richard. 1974. “Body and Soul in Aristotle,” Philosophy 49: 63–89.
Taylor, C. C. W. 2012. “Cyrenaics.” In S. Hornblower and A. Spawforth (eds.), The Oxford
Classical Dictionary, 4th ed., 405. Oxford: Oxford University Press.
Tsouna, Voula. 1998. The Epistemology of the Cyrenaic School. Cambridge: Cambridge
University Press.
Tsouna-McKirahan, Voula. 2002. “Is there an Exception to Greek eudaimonism?” In M.
Canto-Sperber and P. Pellegrin (eds.), Le style de la pensée: receuil de texts en hommage
à Jacques Brunschwig, 464–489. Paris: Les Belles Lettres.
Zilioli, Ugo. 2012. The Cyrenaics. Durham: Acumen.
CHAPTER TWO
Pyrrho and Timon

CASEY PERIN
1 PYRRHO THE OBSCURE

We know remarkably little about the life of Pyrrho. Born sometime around 365
BCE in Elis on the Peloponnesos, he was, according to Antigonus of Carystus by
way of Diogenes Laertius, a failed painter turned philosopher. Several ancient
authors report that Pyrrho traveled to India as part of Alexander’s expedition.
There he is said by Diogenes Laertius to have encountered certain “naked wise men”
(γυμνοσοφισταί) and, as a result, to have developed certain philosophical views.
However, he did not record these views in writing as he appears to have written
nothing. He died sometime around 275 BCE.
It is arguable that we know even less about Pyrrho’s philosophical views than
we do about his life. There is a consensus among scholars that the most important
text for understanding Pyrrho’s philosophical position is a passage from the late-
first-century BCE or early-first-century CE Peripatetic philosopher Aristocles’s
On Philosophy. This and other passages from Aristocles’s treatise are preserved in
Eusebius’s (260–340 CE) Praeparatio evangelica. Aristocles reports a number of
claims he attributes to Pyrrho’s disciple and publicist Timon of Phlius (c. 325–c.
235 BCE). Timon, in turn, attributes at least one of these claims to Pyrrho. It is not
easy to tease out just what in the passage comes from Aristocles, what from Timon,
and what, if anything, from Pyrrho. Few scholars think Aristocles is quoting Timon
directly. Many, though, think his paraphrase of Timon is, at least for the most
part, accurate. And many of these scholars think the text by Timon that Aristocles
accurately paraphrases itself accurately reports Pyrrho’s views. Hence, these scholars
think we can read back through Aristocles via Timon to Pyrrho. Other scholars are
less optimistic and think that what we get in Aristocles’s passage is either mostly the
views of Timon or, worse, the views of a Timon badly distorted by Aristocles.1 I see
no way to settle these issues. However, unless we assume that what Aristocles tells
us about Timon, and what Timon tells us about Pyrrho, is accurate, then Pyrrho
as a philosopher is more or less lost to us. For this reason, I will proceed on the
assumption that the Aristocles passage tells us something about Pyrrho and Timon
if only we can make sense of what it is.2
PYRRHO AND TIMON 25
2 THE PASSAGE FROM ARISTOCLES

Here, in as neutral a translation as I think possible, is the Aristocles passage
in full. (For ease of exposition I have divided the passage into six sections [A]
through [F].)
A. It is necessary above all to consider our own knowledge; for if it is our
nature to know nothing, there is no need to inquire any further into things.
There were some among the ancients, too, who made this statement, whom
Aristotle has argued against. Pyrrho of Elis was also a powerful advocate of
such a position (ἴσχυσε μὲν τοιαῦτα λέγων καὶ Πύρρων ὁ Ἠλεῖος).
B. He himself has left nothing in writing; his pupil Timon, however, says that
the person who is to be happy must look to these three points: first, what
are things like by nature (ὁποῖα πέφυκε τὰ πρὰγματα)? Second, in what
way ought we to be disposed toward them (τίνα χρὴ τρόπον ἡμᾶς πρὸς
αὐτὰ διακεῖσθαι)? And, finally, what will be the result for those who are
so disposed?
C. He [Timon] says that he [Pyrrho] reveals that things are equally
indistinguishable/indifferent (ἀδιάφορα) and unmeasurable/unstable
(ἀστάθμητα) and indeterminable/indeterminate (ἀνεπίκριτα);
D. for this reason (διὰ τοῦτο) neither our sensations nor our opinions (μήτε
τὰς αἰσθήσεις ἡμῶν μήτε τὰς δόξας) tell the truth or lie (ἀληθεύειν
ἢ ψεύδεσθαι).
E. For this reason, then (διὰ τοῦτο οὖν), we should not trust them, but should
be without opinions (ἀδοξάστους) and without inclinations and without
wavering, saying about each single thing that it no more (οὐ μᾶλλον) is than
is not or both is and is not or neither is nor is not.
F. Timon says that the result for those who are so disposed will be first an
inability to say anything (ἀφασία), but then tranquility (ἀταραξία); and
Aenesidemus says pleasure. These, then, are the main points of what they
say. (Eusebius 14.18.1-5)3
The passage has the following structure. In [A] Aristocles attributes to Pyrrho the
view that in virtue of our nature we can’t know anything. However, Aristocles
himself notes that Pyrrho wrote nothing. Hence, when he attributes this view to
Pyrrho, Aristocles must by relying on what someone else has said or written about
Pyrrho. So in [B] Aristocles either quotes or paraphrases Timon. But the connection
between the quotation from or paraphrase of Timon and the view attributed to
Pyrrho in [A] is not obvious. Pyrrho is said to have held the view that we can’t know
anything. Timon is said to have made a claim about what is required for someone to
be happy. Specifically, Timon is reported as saying that to be happy a person must
consider, and presumably answer correctly, three questions. Then, in [C], we get the
only claim clearly and explicitly attributed to Pyrrho by Timon. And, apparently,
this claim constitutes an answer, and is understood by Timon to be Pyrrho’s answer,
26 SKEPTICISM
to the first of the three questions relevant to happiness. In [D] an inference is made,
either by Pyrrho or Timon or both, from the claim attributed to Pyrrho in [C]. A
second inference is made in [E] and generates the answer given by either Pyrrho or
Timon or both to the second of the three questions relevant to happiness. Finally,
in [F], Aristocles reports the answer Timon gives, and may implicitly attribute to
Pyrrho, to the third of the three questions relevant to happiness. Aristocles then
claims to have presented the “main points” of what “they” (Pyrrho and Timon? and
Aenesidemus?) say.
3 TWO READINGS OF PYRRHO

The claim attributed to Pyrrho by Timon in section [C] is supposed to be an answer
to the first of the three questions relevant to happiness. Hence, it is supposed to
be a claim about what things are like by nature. The initial, and perhaps most
fundamental, interpretative problem with the passage is how to understand the trio
of adjectives by which Pyrrho describes what things are like by nature. Here there
are two general interpretative options: the so-called subjective versus objective or,
better, the epistemic versus the metaphysical reading of the adjectives. According
to the former, Pyrrho’s fundamental claim is that things are indeterminable by us
(as, on this reading, ἀνεπίκριτα is taken to mean), that is, we can’t determine, and
hence can’t know, what things are like.4 More precisely, Pyrrho’s is the minimally
informative claim that
(α) Things by nature are such that we can’t know what they are like.
A problem arises if, as appears to be the case, (α) falls within its own scope. (α) is a
claim about what things are like by nature (otherwise it fails to answer the first of the
three questions relevant to happiness). But (α) says that we can’t know what things
are like. A fortiori we can’t know what things are like by nature. Hence, we can’t
know that things by nature are such that we can’t know what they are like. And if
that is so, we can’t know that (α) is true. So if (α) falls within its own scope, and if
Pyrrho asserts (α) as something he knows, Pyrrho’s assertion of (α) is self-refuting.
If (α) is something Pyrrho knows, it is false; if (α) is true, it is not something Pyrrho
knows. Pyrrho can avoid these problems if he thinks that somehow (α) does not fall
within its own scope. He would then think, for reasons we can only guess, that the
one thing we can know about what things are like is that they are by nature such that
we can’t know what they are like.
On the metaphysical reading, Pyrrho is making a more substantial claim about
what things are like by nature. This is the strange, almost paradoxical, claim that
(as Richard Bett formulates it) “things in themselves have no fixed and definite
character” (Bett 2000: 20).5 That is, things are “indifferent” in the sense that any
one thing lacks properties that would distinguish it from any other thing; things are
“unstable” in the sense that their natures change; and things are “indeterminate”
in the sense that that they have no determinate features or properties. On the
metaphysical reading, Pyrrho’s answer to the question “What are things like by
nature?” is that there is nothing at all that things are like by nature.
PYRRHO AND TIMON 27
4 ZELLER’S EMENDATION
Partisans of the metaphysical reading have argued, and many partisans of the
epistemic reading have agreed, that the epistemic reading of Pyrrho’s claim wrecks
the logic of Aristocles’s text as we have it. Either Pyrrho or Timon or both make an
inference from the claim attributed to Pyrro in section [C] to the claim in section
[D] that “neither our sensations nor our beliefs tell the truth or lie.” Most, though
not all, scholars take the latter claim to mean that our sensations and beliefs neither
constantly nor consistently tell the truth nor constantly nor consistently lie, that
is, they are unreliable. Suppose the claim in section [D] is understood in this way.
Suppose, further, that Pyrrho’s claim in section [C] is equivalent to or at least clearly
implies that we can’t know what things are like.6 In Aristocles’s text as we have it,
either Pyrrho or Timon or both would be making an inference from
(1) We can’t know what things are like.
to
(2) Our sensations and beliefs are unreliable.
Yet, as commentators regularly note, the inference ought to go the other way: at
best (1) is a consequence of (2), not (2) of (1).
Since this is so, partisans of the epistemic reading of Pyrrho’s claim have been
thought to have only one option, namely, to emend Aristocles’s text so that either
Pyrrho or Timon or both infer (1) from (2) rather than (2) from (1).7 Hence, partisans of
the epistemic reading have accepted an emendation first proposed by the nineteenth-
century German scholar Eduard Zeller. Sections [C] and [D] of Aristocles’s text as we
have it read (in the translation favored by the epistemic reading):
[C] He [Timon] says that he [Pyrrho] reveals that things are equally
indistinguishable and unmeasurable and indeterminable; [D] for this reason, (διὰ
τοῦτο) neither our sensations nor our opinions tell the truth or lie.
Zeller altered the italicized phrase so that the emended text reads:
[C] He [Timon] says that he [Pyrrho] reveals that things are equally indifferent
and unstable and indeterminate [D] on account of the fact that (διὰ τό neither
our sensations nor our opinions tell the truth or lie.
Zeller’s emendation reverses the direction of the inference made in sections [C] and
[D]. Now either Pyrrho or Timon or both make an inference from (what they take
to be the fact that) our sensations and beliefs are unreliable to the conclusion that
we can’t know what things are like. And, the partisans of the epistemic reading of
Pyrrho’s claim insist that inference is at least intelligible. Moreover, and importantly,
there are linguistic and paleographic considerations that independently support,
even if they do not strictly require, Zeller’s emendation.8
5 PYRRHO THE METAPHYSICIAN

Partisans of the metaphysical reading of Pyrrho’s claim resist Zeller’s emendation as
unnecessary. In their view, once Pyrrho’s claim and the claim made in section [D]
28 SKEPTICISM
are properly understood, the logic of Aristocles’s text as we have it is clear. On the
metaphysical reading, the claim Pyrrho makes in section [C] is that
(1) Things are indeterminate, that is, things do not have any
determinate property.
Now the inference from (1*) to
(2) Our sensations and beliefs are unreliable.
is not promising. But Bett argues that (2) does not capture the meaning of the claim
reported in section [D] of Aristocles’s text. That claim is not (2) but
(2) Our sensations and beliefs are neither true nor false.
And, Bett insists, the inference from (1*) to (2*) is perspicuous. According to him,
a sensation or belief is either true or false only if there is “some definite state of
affairs” that the sensation or belief “either correctly or incorrectly represents” (Bett
2000: 23). The import of (1*) is that there are no definite states of affairs at all. So
there is no state of affairs of, for example, lemons being yellow. Why? Lemons, like
any other thing, have no determinate properties. Hence, they have no determinate
color properties. Since this is so, the sensation or belief that lemons are yellow is not
true. Nor, Bett argues, is it false. For it is false that lemons are yellow only if there
is a state of affairs of lemons being not yellow but some other color. But, given their
lack of determinate color properties, there is no color lemons are.
Bett’s argument turns on two points, one linguistic and the other philosophical.
First, it isn’t clear that the claim made in section [D]—that “neither our sensations
nor our beliefs tell the truth or lie”—is best, or even plausibly, interpreted as (2*).
We might think that in section [D] Aristocles preserves in its original formulation
a claim made by Timon (and we might think this regardless of whether we think
Timon, in turn, preserves in its original formulation a claim made by Pyrrho). But
we might also think that what Aristocles gives us in section [D] is not unadulterated
Timon but rather his—that is, Aristocles’s—explication of a claim made by Timon.
If the words in [D] are not Timon’s but Aristocles’s, then the context in which
they occur might give us some clue about how to understand them. That context is
Aristocles’s treatise On Philosophy. Eusebius transcribes five sections of this treatise
in his Praeparatio evangelica. Tad Brennan has shown in convincing detail that in
this context the claim in section [D] is best understood as a claim about the general
reliability of our sensations and beliefs and so as (2) rather than (2*) (Brennan
1998: 426–429).
Second, Bett’s claim that the inference from (1*) to (2*) is perspicuous depends
on a particular view about the conditions under which a proposition is false. On his
view, a proposition of the form ‘x is F’ is false just in case some proposition of the
form ‘x is G’ (where being G is incompatible with being F) is true. That is, ‘x is F’
is false just in case some determinate state of affairs contrary to x’s being F obtains
(Bett 2000: 22–23 and n.19). Call this the strong conception of falsehood. Since, on
the metaphysical reading of Pyrrho’s claim that Bett favors, no determinate state
of affairs obtains, no proposition of the form ‘x is F’ is false. However, there are
PYRRHO AND TIMON 29
two problems with any version of the metaphysical reading of Pyrrho that relies
on the strong conception of falsehood. First, as several scholars have noted, this
conception of falsehood is neither mandatory nor standard.9 We might reject it in
favor of the weak conception of falsehood according to which ‘x is F’ is false if the
state of affairs specified by that proposition—x’s being F—fails to obtain. On the
weak conception of falsehood, that is, a proposition’s being false requires only that
a certain state of affairs does not obtain and not, as on the strong conception, that
some other state of affairs does obtain. Given the weak conception of falsehood,
the inference from (1*) to (2*) is invalid. If, as (1*) says or implies, no determinate
states of affairs obtain, then it follows that every proposition of the form ‘x is F’ is
false. Hence, every sensation or belief whose propositional content is of this form is
false. So Bett must show that there is reason to think that either Pyrrho or Timon,
or both, accepted the strong rather than the weak conception of falsehood.
Even if he can do so, however, there is a second and even more serious problem.
For it appears that the metaphysical reading can’t attribute the strong conception
of falsehood to Pyrrho or Timon, or both, without undermining itself.10 On this
reading they reject as false the claim that things are determinate where that just
is or implies the claim that there are no determinate states of affairs. But this they
can’t do if they accept the strong conception of falsehood. For on that conception
of falsehood, it is false that
(3) There are determinate states of affairs.
only if some determinate state of affairs contrary to (3) obtains. But if any
determinate state of affairs obtains, then (3) is true. So if Pyrrho or Timon, or both,
accept the strong conception of falsehood, they can’t reject (3) as false. Hence, on
pain of inconsistency, they must reject (1*) as false. If they do so, however, then the
metaphysical reading is false. Alternatively, if they accept (1*), then, again on pain
of inconsistency, they must reject the strong conception of falsehood.
There is another option open to those inclined to favor a metaphysical reading
of Pyrrho’s claim and to read the claim in section [D] as (2*) rather than (2). Again,
the claim attributed to Pyrrho in section [C] is, on the metaphysical reading, that
“things (πράγματα) are equally indifferent and unstable and indeterminate.” Jacques
Brunschwig has argued, ingeniously, that the inference from (1*) to (2*) is perfectly
acceptable if we make two assumptions. First, we assume that our sensations and
beliefs are themselves “things” and so fall within the scope of Pyrrho’s claim.
And, second, we assume that the way in which sensations and beliefs are “equally
indifferent and unstable and indeterminate” is that they are neither true nor false
(Brunschwig 1994: 199–202). It is not clear, though, that we have good reason to
make either of these assumptions, especially the first. Brunschwig seems to think
that the claim in section [D] must be part of the answer being given to the first of
the three questions relevant to happiness, and so must be a claim about what things
are like by nature. This is so only if our sensations and beliefs, and not just the
things in the world they are of or about, are things (πράγματα). But why think that
everything said in sections [C] through [F] of the passage must be part of an answer
to one or another of the three questions relevant to happiness? Why not think, as
30 SKEPTICISM
Bett does, that [D] is the conclusion of an inference made from the answer given in
[C] to the first of the three questions concerning happiness?
6 A LIFE WITHOUT BELIEFS

In section [E] we get the answer either Pyrrho or Timon, or both, give to the second
of the three questions relevant to happiness (“In what way ought we to be disposed
toward things?”). We are told that we should not trust our sensations or beliefs or
even have any beliefs at all. This claim is inferred from the claim made in section [D].
If that claim is construed as (2), the inference is easy: the claim that our sensations
and beliefs are unreliable is the basis for the recommendation not to trust them but
to suspend judgment. Bett has argued that if the claim in section [D] is construed
as (2*), the inference is no less easy. But here we should ask why the fact that our
sensations and beliefs are neither true nor false (the claim made in section [D] as
Bett reads it) is a reason not to trust them and to avoid having any beliefs at all.
According to Bett, we are subject to a misconception if we trust, and so accept as
true, a sensation or belief that purports to be true but in fact is neither true nor false.
Now I can think of only two possible conceptions of a misconception. It is, or
consists in, either (a) a belief that isn’t true or (b) a belief that is false. (a) is the more
capacious conception of misconception because the class of propositions that aren’t
true (including, as it does, those propositions that are neither true nor false as well
as those propositions that are false) is wider than the class of false propositions.11
Consider first (a). In this case Bett would be claiming that the reason to avoid having
beliefs that are neither true nor false is that by having them we are thereby subject
to a misconception. But, given (a), that misconception just is, or consists in, a belief
that is neither true nor false. So Bett’s claim would then be that the reason not to
have beliefs that are neither true nor false is that by having them we thereby have a
belief that is neither true nor false. But, that, obviously, is no reason at all. It does
not tell us what is wrong, in the first place, with having beliefs that are neither true
nor false. Consider, next, (b)—the conception of a misconception as consisting in
one or more false beliefs. What might be the false belief I acquire by trusting, and
so accepting as true, a sensation or belief? On Bett’s own reading of the claim made
in section [D] as (2*) rather than (2), it can’t be a false belief about the world. For
(2*) says that our beliefs about the world are neither true nor false. So it must be a
false belief about the particular sensation or belief I accept as true, namely, that that
sensation or belief is true. This higher-order belief is false if my sensation or belief is
either false or (as on Bett’s reading of the claim in section [D]) neither true nor false.
Bett must think that if I trust, and so accept as true, the sensation or belief that p,
then I believe not only that p but also, at least implicitly, that my sensation or belief
p is true and so the kind of thing that is either true or false. These implicit beliefs
are the content of the misconception to which, according to Bett, I am subject if I
trust my sensations and beliefs. It seems to me just possible that this line of thought
underwrites the inference from section [D] to section [E]. But, if it does, it makes
that inference much less straightforward than it is on a reading of the claim in
section [D] as (2) rather than (2*).
PYRRHO AND TIMON 31
In section [E] we are told that, in addition to not having any beliefs, we should be
“without inclinations and without wavering” (καὶ ἀκλινεῖς καὶ ἀκραδάντους). It’s
plausible to think that we are being told here that the doxastic state or condition we
are told we should be in—not having any beliefs—should itself be stable in certain
ways. We should be “without inclinations” in the sense that we should lack not only
beliefs but also any inclination to believe anything or to accept any sensation as true
or to reject it as false. If this is so, then there is a sense in which we will be, as we
are told we should, “without wavering.” In the absence of any inclination to believe,
we won’t vacillate between doxastic options by first believing one thing and then
rejecting it in favor of another. Moreover, we won’t vacillate between having and
not having beliefs. The state or condition in which we do not have any beliefs will
be (to borrow terminology from Aristotle) firm and unchanging.12
7 MAKING UTTERANCES WITHOUT

MAKING ASSERTIONS
In section [E] we are also told that we should say “about each thing that it no more
is than is not or both is and is not or neither is nor is not.” Either Pyrrho or Timon
or both recommend that we take a certain disposition or stance toward things.
This disposition consists not only in lacking beliefs but also, perhaps surprisingly,
in making utterances. And, crucially, these utterances contain the locution ‘no
more’. Here it is important to distinguish two questions we might ask about these
utterances. The first question is syntactic: it concerns the scope of the locution ‘no
more’. The second question is semantic: it concerns the meaning of that locution
and so of the utterances in which it occurs.
These utterances will seem slightly less baffling if we assume, as commentators
frequently do, that some predicate is understood to follow the occurrences of ‘is’
and ‘is not’ in them (e.g., Bett 2000: 30). So to say that something no more is than
is not is to say that it no more is than is not, for example, yellow or round or sweet.
Now there are two ways to construe the syntax of the utterance we are told to
make about each thing (Stopper 1983: 272–274; Bett 2000: 30). That utterance has
three disjuncts. One possibility is that the locution ‘no more’ is embedded in and so
governs only one of these disjuncts. On this reading (call it R1), we are being told
to say about each thing that it
A. no more is than is not; or
B. both is and is not; or
C. neither is nor is not.
This complex disjunctive utterance has a ‘no more’ statement (A) as one and but
only one of its disjuncts. Moreover, we might understand disjuncts (B) and (C) as
alternatives to disjunct (A). We would then be saying about each thing that one of
three things is true of it: either it no more is than is not F (for some predicate ‘F’),
or it both is and is not F, or it neither is nor is not F. But we might understand the
disjuncts (B) and (C) not as alternatives to (A) but as equivalent formulations of it.
32 SKEPTICISM
On this version of R1, to say about each thing that it no more is than is not F is just
to say that it both is and is not F or that it neither is nor is not F.
The other syntactic possibility is that the locution ‘no more’ is not embedded in
any of the disjuncts but governs all three. On this reading (call it R2), we are being
told to utter about each thing one complex ‘no more’ statement. That is, we are to
say about each thing that it is no more than
A. it is not, or
B. it both is and is not, or
C. it neither is nor is not.
On R2 we would be saying about each thing that it is F no more than it is not F or
it both is and is not F or it neither is nor is not F.
As far as I can see, no independent consideration (i.e., no consideration independent
of the interpretative commitments one has already made in reading the Aristocles
passage) gives us good reason to favor R1 over R2, or vice versa. In this context, it
is worth nothing that in section [F] Aristocles reports Timon as saying that anyone
who adopts the disposition toward things described in section [E] will experience
ἀφασία. Since this disposition consists in part in making meaningful utterances of
a certain type, ἀφασία can’t be speechlessness or quietude. What might it be then?
One plausible suggestion is that ἀφασία is a condition in which one refrains from
making not utterances but assertions or at least sincere assertions that express beliefs.
We are told in section [E] that we should be disposed toward things in such a way
that we have no beliefs. And if we are so disposed, then we have no beliefs we can
express in our assertions. Hence, we are in a state in which we refrain from making
sincere assertions because we are incapable of making them. However, on either
reading R1 or R2, the disjunctive utterances we are told we should make about each
thing seem to constitute assertions. We are asserting that one of three disjuncts is true
of a thing, or that something of which the three disjuncts are equivalent formulations
is true of a thing (versions of reading R1), or that a thing has a certain property or
feature no more than it lacks it or both has it and lacks it or neither has it nor lacks
it (reading R2). So the prohibition on assertions does not give us reason to favor one
syntactical reading of the utterances over the other.13
But how can we respect this prohibition (and so be in a state of ἀφασία) while
also making utterances that, regardless of the way we construe their syntax, seem to
constitute assertions? The answer we give to this question will depend on the answer
we give to the question about the meaning of the locution ‘no more’. And here, again,
there are two options that correspond to the epistemic and metaphysical readings
of Pyrrho’s claim in section [C]. Partisans of the former will take their cue from the
way Sextus Empiricus (PH 1.192) uses the locution ‘no more’ and understand it as
expressing suspension of judgment. To say something of the form ‘x no more is than is
not F’ is not to make any assertion about x, but rather to express one’s suspension of
judgment about whether x is F. Utterances of this sort do not violate the prohibition
on assertions. Now suppose we understand the meaning of ‘no more’ in this way.
Suppose, further, that we accept reading R2 of the syntax of the utterances that contain
this locution. We will then think that in section [E] we are being told that one part
PYRRHO AND TIMON 33
of the disposition we should take toward things consists in making utterances that
do nothing more than express the attitude or state of mind (having no beliefs, but
suspending judgment, about things) that constitutes the other part of that disposition.14
But suppose, instead, that we accept the metaphysical reading of Pyrrho’s claim
and interpret it as the claim that things have no determinate properties. We will
think that, unlike Sextus Empiricus, either Pyrrho or Timon or both uses ‘no more’
to assert about some particular thing x that it no more has than lacks a particular
property F-ness (see Bett 2000: 32–33). We will also think that, for one reason or
another, the prohibition on assertions does not apply to either assertions of this kind
or Pyrrho’s more general assertion that things have no determinate properties.
Finally, in section [F], Aristocles records Timon’s (if not also Pyrrho’s) answer to the
third of the three questions relevant to happiness. Timon claims that those who adopt
the correct disposition toward things—where that is a matter of living without beliefs
and making utterances of a certain kind—will experience tranquility (ἀταραξία). And
this tranquility just is, or at least is a principal component of, happiness.
8 PYRRHO THE SKEPTIC

There is, then, little that is clear or uncontroversial about the passage from Aristocles
that constitutes our best evidence for Pyrrho’s philosophical views. I have tried to
sketch the range of difficult interpretative questions that passage raises. The answers
we give to these questions (if we are confident enough to answer them at all) will
determine what sort of skeptic, if any, we take Pyrrho and his apparently faithful
student Timon to be. On the epistemic reading of his central claim in section [C]
of the passage, Pyrrho is a skeptic in the following sense. He, unlike the skeptic
described by Sextus Empiricus, has the view that nothing can be known. Moreover,
he, or at least Timon, has the view that since this is so, a person ought not to have
any beliefs. Hence, Pyrrho, or at least Timon, is also a skeptic in the sense that he
adopts the attitude of suspension of judgment toward things. On the metaphysical
reading, in contrast, there is little sense in which Pyrrho is a skeptic. He holds the
general metaphysical view that things have no determinate properties. This view, in
turn, commits him to the view that there are an indefinitely large number of truths
we can know. These are truths of the form ‘x is no more F than not F’—for example,
“the lemon is no more yellow than not” or “the honey is no more sweet than not.”
Yet his general metaphysical view also commits him to the view that we know little of
what we ordinarily take ourselves to know, for example, that the lemon is yellow or
that the honey is sweet. And this is because, on that view, lemons are not yellow and
honey is not sweet. So if, on the metaphysical reading, Pyrrho is less a skeptic than
a dogmatic metaphysician, his metaphysical view still has skeptical implications.15
NOTES
1. For the former view, see Brunschwig (1994: 193–195), for the latter, Brennan (1998:
431–432).
2. For an extended defense of this assumption, see Bett (1994).
34 SKEPTICISM
3. The Greek text is most easily found in Long and Sedley (1987: II 5, text 1F).
4. For versions of the epistemic reading of Pyrrho’s claim, see especially Stopper
(1983: 274 and 292 n.50), Brennan (1998), and Castagnoli (2002: 447). If we give
“indeterminable” an epistemic reading, then we will feel pressure to read the other
two adjectives in Pyrrho’s claim epistemically as well. Hence, we’ll take ἀδιάφορα to
mean “indistinguishable” by us and ἀστάθμητα to mean “unmeasurable” by us. One
possibility is that, on either the epistemic or the metaphysical of the adjectives, both
occurrences of καί in Pyrrho’s claim are epexegetical. If this is so, then ἀστάθμητα
gives the meaning of ἀδιάφορα, and ἀνεπίκριτα gives the meaning of ἀστάθμητα
and so of ἀδιάφορα.
5. See also Long and Sedley (1987: I 16–17). Bett’s book offers the most extensive
presentation and defense of the metaphysical reading of Pyrrho’s claim. It is,
by some wide margin, the most important book on Pyrrho published in several
decades. Some scholars have objected that the Greek adjective ἀνεπίκριτα can’t
bear a metaphysical reading and can only mean, as on the epistemic reading,
“indeterminable” by us. I assume here that a metaphysical reading of each of the
trio of adjectives in Pyrrho’s claim is at least linguistically possible.
6. Even if Pyrrho’s claim is one about what things are like, on the epistemic reading it is
a claim about what things are like in relation to us and, more precisely, our cognitive
abilities. Moreover, it is reasonable to think that Aristocles at least understands
Pyrrho’s claim as equivalent to, or at least as clearly implying, the claim he attributes
to Pyrrho at the outset of the passage, namely, that our nature is such that we can’t
know anything.
7. But see Svavarsson (2004) for what I take to be a version of the epistemic reading
that does not rely on emending the text in this (or any other) way.
8. First, the phrase Zeller emends (διὰ τοῦτο) occurs again in section [E] of Aristocles’s
text. That fact might explain the corruption from διὰ τὸ to διὰ τοῦτο. (Though, as
Bett (2000: 25) notes, there is no trace of this corruption in the manuscript tradition.
All manuscripts have διὰ τοῦτο here.) Second, the phrase Zeller emends appears to
be asyndotic, that is, it lacks a particle where it is thought it should have one. Third,
the forms of the negatives employed in section [D] are thought to be incorrect
given the grammatical construction in which they occur. Bett (2000: 25–26) claims
that these linguistic considerations do not constitute reasons to emend Aristocles’s
text. I’m inclined to agree with Brennan (1998: 432–433) that Bett shows only that
the text as we have it is linguistically possible, but not that it is correct or that the
linguistic considerations in question aren’t good reasons to emend.
9. See especially the useful remarks in Castagnoli (2002: 447) and Brennan
(1998: 421–22).
10. Thanks to Baron Reed for making this point clear to me.
11. Thanks, again, to Baron Reed for suggesting this conception of a misconception
to me.
12. For this way of reading the injunction, see Bett (2000: 30). Brunschwig (1994:
209), following Ausland (1989), reads the injunction more broadly to cover
practical as well as doxastic inclinations. He thinks we are being told to have no
inclination to choose or pursue anything as choiceworthy or to reject or avoid it as
not choiceworthy.
PYRRHO AND TIMON 35
13. See Stopper (1983: 274) for the claim that ἀφασία is non-assertion. (In making
this claim he appeals to the gloss Sextus Empiricus gives the term at PH 1.192.)
14. That is, for any x and any predicate F, we suspend judgment about whether x is F
or not F or both F and not F or neither F nor not F.
15. Many thanks to Diego Machuca and Baron Reed for their comments on the initial
version of this chapter.
REFERENCES
Ausland, Hayden. 1989. “On the Moral Origin of the Pyrrhonian Philosophy,” Elenchos
10: 359–434.
Bett, Richard. 1994. “Aristocles on Timon on Pyrrho: The Text, Its Logic and Its
Credibility,” Oxford Studies in Ancient Philosophy 12: 137–181.
Press.
Brennan, Tad. 1998. “Pyrrho on the Criterion,” Ancient Philosophy 18: 417–434.
Brunschwig, Jacques. 1994. “Once Again on Eusebius on Aristocles on Timon on Pyrrho.”
In his Papers in Hellenistic Philosophy, translated by J. Lloyd, 190–211. Cambridge:
Castagnoli, Luca. 2002. “Review of Pyrrho, His Antecedents, and His Legacy.” Ancient
Philosophy 22: 443–457.
Long, Anthony and David Sedley. 1987. The Hellenistic Philosophers. 2 vols. Cambridge:
Stopper, M. R. 1983. “Schizzi Pirroniani,” Phronesis 28: 265–297.
Svavarsson, Svavar Hrafn. 2004. “Pyrrho’s Undecidable Nature,” Oxford Studies in Ancient
CHAPTER THREE
Arcesilaus
ANNA MARIA IOPPOLO
1 INTRODUCTION
Arcesilaus of Pitane (316/5–241/0) became the head of Plato’s Academy in 268
or 266 BCE and oriented the school toward skepticism. Since he did not write
any philosophical work, we have to rely on secondary sources to reconstruct his
philosophy. The ancient sources are not unanimous on the nature and origin of
Arcesilaus’s skepticism, but offer conflicting testimonies. Arcesilaus himself claimed
Socrates, Plato, and certain Presocratics as his predecessors. Some sources attribute
his skepticism to the influence of Pyrrho of Elis, while others maintain that he
only posed as a skeptic in order to conceal the true doctrine of Plato that he was
secretly teaching. The main extensive sources, Cicero and Sextus Empiricus, are
not unbiased and innocent. Both have a philosophical perspective they wish to
defend, and both inform us about Arcesilaus’s philosophy by closely connecting it
to his epistemological dispute with the Stoic, Zeno. Since the debate between the
Academic skeptics and the Stoics did not end with Arcesilaus and Zeno in the third
century BCE but raged through the following centuries down to Cicero’s day, the
reconstruction of Arcesilaus’s philosophy is rather controversial, for the positions of
individual representatives of the two schools would inevitably have changed in the
course of this epistemological debate. The risk, then, is that Arcesilaus’s philosophy
may have been distorted by the way that his successors elaborated on it. In order
to produce a coherent account of Arcesilaus’s doctrine, we need to examine the
testimonies, not in isolation, but by setting them within the philosophical and
argumentative context of the sources recording them. We must take account of their
date, philosophical preferences, and—last, but not least—of their polemical aims.
2 THE DIALECTICAL METHOD OF

“ARGUING THE OPPOSITE” AT THE CORE
OF ARCESILAUS’S PHILOSOPHY
The most ancient testimonies trace Arcesilaus’s philosophy back to two opposing
traditions, which in a way imply one another: one, originating with Arcesilaus
himself, claims a strong link with his forerunners, Socrates and Plato, as well as with
ARCESILAUS 37
other illustrious philosophers of the past; the other, which is favored by Arcesilaus’s
contemporaries, accuses him of plagiarizing other philosophers’ doctrines. In his Life
of Arcesilaus, Diogenes Laertius states: “It seems that Arcesilaus admired Plato, whose
books he possessed. Some say that he emulated (ἐζηλώκει) Pyrrho as well. He was
very keen on dialectic and made use of the arguments of the Eretrians” (DL IV 32–
33, trans. Long and Sedley 1987 modified).1 The philosophical continuity between
Arcesilaus and Plato was questioned by the philosophers of his day, and in particular
by Aristo of Chios2 and Timon of Phlius, who both accused him of being a plagiarist.
This charge was based on Arcesilaus’s own admission that he was “not saying anything
new,” as he simply wished to refer to his predecessors’ doctrines (Plut. Adv. Col. 1121
F). Arcesilaus thus became a target for Aristo’s and Timon’s witty remarks: while
exploiting Arcesilaus’s own claim not to be an original philosopher, they did not
see in his teaching a conformity to Plato’s doctrines but instead attributed Pyrrho’s
views to him. Timon implied that Arcesilaus lacked any originality as a thinker,
since he made use of Menedemus’s eristic, and sought refuge under Diodorus’s
dialectic and Pyrrho’s philosophy (“all flesh”). In a parody of a famous Homeric
verse—“Plato in front, Pyrrho behind, Diodorus in the middle”—Aristo described
Arcesilaus’s philosophy as a chimera: an eclectic monstrosity that seeks to reconcile
what is irreconcilable, to wit Plato and Pyrrho via Diodorus.3 Arcesilaus’s detractors
also included the Epicurean Colotes, who accused him “of rubbing off his doctrines
about suspension of judgment and non-cognition (ἀκαταληψία) on Socrates, Plato,
Parmenides, and Heraclitus, who did not need them” (Plut. Adv. Col. 26, 1121 F;
trans. Long and Sedley 1987).4 It is clear, then, that the charge of unoriginality, if not
of downright plagiarism, leveled against Arcesilaus by his contemporaries stemmed
in particular from his use of philosophers of the past: Arcesilaus sought to “rub off”
his own doctrines on others with the aim of strengthening them. By proclaiming
the identity between his own philosophy and that of his predecessors, Arcesilaus
sought to lend credit to his skepticism, which in the eyes of his contemporaries
represented a betrayal of his teachers, Socrates and Plato. Timon, Aristo, Colotes,
and other unnamed detractors agreed in challenging Arcesilaus’s attempt to attribute
his own “doctrines” to his forerunners5; in doing so, they also disputed his claim to
be carrying on the philosophical tradition of his school. Aristo’s verse sums up his
contemporaries’ refusal to grant Arcesilaus’s philosophy the legitimation it aspired to,
on the grounds that through dialectical arguments Arcesilaus used Plato as a screen
to conceal his Pyrrhonism.6
Still, given Colotes’s silence with regard to Pyrrho, we may rule out any connection
between Arcesilaus and this philosopher; otherwise Colotes would no doubt have
included Pyrrho among the authorities invoked by Arcesilaus. The lack of any
philosophical connection between Arcesilaus and Pyrrho is further confirmed by the
fact that Cicero knew nothing about Pyrrho’s role as a skeptic and presented him
simply as a moralist.7 More generally, the whole Academic tradition, both dogmatic
and skeptical, ignored Pyrrho completely. In order to understand the reason why
Aristo attributed Arcesilaus’s skeptical turn to his plagiarism of Pyrrho’s philosophy,
we must examine how Arcesilaus justified his connection with his predecessors in the
school, and especially Socrates, whom both Zeno8 and Aristo9 claimed to be following.
38 SKEPTICISM
What is most revealing in this respect is the fact that, according to the tradition
of the skeptical Academy (Acad. I 44–46), an antecedent for Arcesilaus’s skepticism
is to be found in Socrates’s confession of ignorance, based on the idea that all things
are hidden in darkness—an approach also common to the Presocratic philosophers
Heraclitus, Democritus, Anaxagoras, and Empedocles.10 By drawing upon the
aporetic Socrates of Plato’s dialogues, Arcesilaus saw Socrates’s practice of ἔλεγχος
(refutation) as the chief instrument for purifying the mind of false beliefs. The
central role he attributed to Socrates’s dialectical method enabled him to justify his
own particular interpretation of Platonic philosophy, as well as to claim a degree of
continuity between his own teaching and that of Plato (Acad. I 46; cf. De or. III 67).
The ἔλεγχος was able to bring people who wrongly imagined that they possessed
knowledge to realize their own ignorance. Unlike Socrates, however, Arcesilaus
even denied that he could claim to know that he did not know: he therefore turned
the aporetic stance against itself, yet without discarding it:
[Arcesilaus] used to act consistently with this philosophy, and by arguing against
everyone’s opinions he led most of them to his own, so that when reasons of
equal weight were found on opposite sides on the same subject, the easier course
was to withhold assent from either side. (Acad. I 45, trans. Long and Sedley
1987 modified)11
Arcesilaus would allow his interlocutor to speak first and only afterwards put
forward the argument opposite to the one already advanced. The result was that he
would speak in such a way as to present a counterargument, but without appearing
to be involved in any assertion or denial (see Fin. II 2; De or. III 67; ND I 11). For
Arcesilaus, arguing on the opposite side—that is, by opposing the thesis put forward
by one’s interlocutor—represented the very essence of philosophizing, which
fulfills itself and comes to a conclusion through this process, since the ἰσοσθένεια
(equipollence) of contrary arguments can never be overcome: the suspension of
judgment simply emerges from their opposition. From Diogenes Laertius we learn
that “Arcesilaus was the first to suspend his assertions owing to the contrarieties
of arguments, he was also the first to argue pro and contra and the first to change
the traditional Platonic discourse, and by question and answer, to make it more
of a debate contest” (DL IV 28, trans. Long and Sedley 1987). One may question
whether the method of arguing pro and contra was first used by Arcesilaus, given
that this method may be traced back to Protagoras—as Diogenes Laertius himself
records (DL IX 51)—and it constituted an integral part of the dialectical training
offered in the Peripatos.12 Arcesilaus was certainly the first to practice it in a different
way from that common in the philosophical schools of his day, such as the Peripatos
and the early Academy. In a sense—especially at a superficial glance—Arcesilaus’s
dialectical method might have seemed little different from that of arguing pro and
contra, given that the interlocutor would always be the first to state his view, with
Arcesilaus then offering a counterview. A more in-depth philosophical analysis,
however, reveals that Arcesilaus’s method of argument sharply differs from that of
arguing for and against a given thesis, as witnessed by Chrysippus, who had carefully
studied Arcesilaus’s arguments in an attempt to refute them.13
ARCESILAUS 39
Plutarch informs us that, according to Chrysippus, “arguing the opposite” (πρὸς

τὰ ἐναντία διαλέγεσθαι) is only useful for those who suspend all judgment, with
the aim of turning people away from cognition (κατάληψις).14 Chrysippus had thus
developed a strategy to neutralize the effectiveness of the method of arguing the
opposite. First of all, he warned his disciples not to let themselves be diverted by
the persuasiveness of the arguments of those who suspend judgment; secondly, he
encouraged his disciples to overturn the dialectical method of those suspending
all judgment by presenting a counterargument without defending it, but only
weakening it as a lawyer in a courtroom might do (see Ioppolo 1986: 109–120).
Chrysippus’s words clearly confirm the idea that Arcesilaus’s dialectical method
consisted in arguing against the thesis presented by one’s interlocutor, rather than
arguing pro and contra a given thesis. Plutarch further informs us that Chrysippus
feared this method because it has the power of turning away from cognition even
the kind of people who base their knowledge on common sense.15 The aspect of
Arcesilaus’s dialectical method that Chrysippus feared the most was not its capacity
to make one argument prevail over another, but the fact that, by presenting two
opposite arguments on a given thesis that have the same weight, so that neither can
prevail, it prevents listeners from understanding the solutions.16
Chrysippus’s testimony thus agrees with Cicero’s view of Arcesilaus’s dialectical
procedure, as presented in Academica I 45: faced with the equipollence of opposite
theses on the same argument, the interlocutors will fail to find any solution and
fall into aporia. Hence, the suspension of judgment does not dialectically depend
upon the Stoic premises but rather invariably follows the aporia as a conclusion,
due to the mutual contrariety of the arguments.17 In other words, ἐποχή is not the
goal toward which Arcesilaus strives—for this would make it the preset aim of his
dialectical method—but constitutes the unavoidable manifestation of aporia at an
epistemological and linguistic level.18 In his Academica, Cicero records that the τέλος
for Arcesilaus is not ἐποχή, but verum invenire velle (Acad. II 76). It is in the light
of the search after truth that Cicero interprets the dispute with Zeno: “Arcesilaus
did not fight with Zeno for the sake of criticizing him but from a wish to discover
the truth” (Acad. II 76, trans. Long and Sedley 1987). With these words Cicero
clears Arcesilaus of the charge of wishing to argue with Zeno simply for the sake
of competitiveness or a taste for victory (see Acad. I 44, II 16). Cicero also tells us:
“None of Zeno’s predecessors (nemo umquam superiorum) had clearly expounded,
or even stated, that a human being can refrain from opining and that a wise man
not only can but must do so. To Arcesilaus this idea seemed true and honorable and
worthy of a wise man” (Acad. II 77, trans. Long and Sedley 1987). Cicero’s emphasis
in presenting the agreement between the two philosophers on the thesis that “the
wise man must absolutely never formulate opinions” suggests that this thesis played
a pivotal role in the debate and at the same time that Arcesilaus’s agreement was
sincere.19 Arcesilaus used this, however, as the premise for an argument from which
he then drew a completely opposite conclusion. Cicero recounts the debate between
Zeno and Arcesilaus in dialogue form. Arcesilaus asks Zeno what would happen
if the wise man could cognize nothing at all and were absolutely required not to
formulate any opinions. Zeno replies that he will not opine because there exists
40 SKEPTICISM
something cognitive, namely the presentation (φαντασία).20 He defines this as that

which is “stamped and reproduced from something which is, exactly as it is” (Acad.
II 77, trans. Long and Sedley 1987). But Arcesilaus gets Zeno to admit that if a true
presentation does not possess that mark whereby it can be told from a false one,
then, since it is indistinguishable therefrom, it cannot be grasped. Arcesilaus then
focuses his criticism on the third clause (“such as could not arise from what is not”),
which raises the problem of establishing how it is possible to distinguish a cognitive
presentation from a noncognitive one, given that, in terms of cognizance, a person
can have access only to his own presentation.21 Once Zeno has been forced to accept
the third clause, Arcesilaus establishes the nonexistence of the Stoic criterion. This
is not enough to conclude, however, that Arcesilaus upheld the opposite thesis
to Zeno’s, that “nothing can be known”—that is, that he was a proponent of
ἀκαταληψία. In this regard, it will be useful to compare Cicero’s account of the
debate between Zeno and Arcesilaus to Sextus’s version.
3 ἈΚΑΤΑΛΗΨΊΑ AS A DIALECTICAL THESIS

AGAINST THE STOICS
Sextus’s testimony on Arcesilaus is rather problematic. Sextus speaks of Arcesilaus
at AM VII 150–158 and PH I 232–234, but at first sight the two accounts appear
to paint two different pictures of Arcesilaus. In PH I Sextus represents Arcesilaus
as a philosopher close to Pyrrhonism: he does not assign him to any of the three
currents into which he has divided philosophy (see PH I 3–4), nor in the course
of his exposition does he ever refer to him as an “Academic”; rather, he includes
Arcesilaus’s philosophy in his discussion of so-called neighboring philosophies.
Moreover, although Sextus’s professed aim is to bring out the differences between
Pyrrhonian skepticism and its neighboring philosophies, he personally undertakes to
associate Arcesilaus with the Pyrrhonists (PH I 232). In AM VII, by contrast, when
distinguishing skeptics and dogmatists in relation to the criterion of truth, Sextus
includes not just Carneades but also Arcesilaus among those dogmatists who posit
the criterion (AM VII 1, 46).
Sextus introduces Arcesilaus’s criticism of the criterion by stressing that he did not
determine any criterion, and does so through the use of a rather loaded and typically
Pyrrhonian verb, ὁρίζειν (AM VII 150).22 However, he then reduces Arcesilaus’s
perspective to his dialectical counterarguments, without explaining why Arcesilaus
did not “determine” any criterion. Sextus maintains that the main difference between
Academic and Pyrrhonian skepticism is to be found in the use of language: for the
Academics it expresses a commitment to the truth of their assertions, whereas for
the skeptics it simply means speaking without holding opinions, ἀπαγγέλλειν.23
Consequently, the terminology used is particularly significant. The procedure Sextus
adopts in his account is also peculiar: before illustrating Arcesilaus’s criticism of the
criterion, he presents the Stoic doctrine; this is followed by Arcesilaus’s arguments,
which are clearly distinguished from the former in such a way as to bring out
their dialectical and ad hominem character (AM VII 151–155).24 This procedure is
ARCESILAUS 41
perfectly in line with the dialectical method of arguing the opposite, which the most
ancient and reliable sources attribute to Arcesilaus. The arguments that Arcesilaus
raises against the Stoic epistemological doctrine are of Platonic inspiration in form
as much as substance; hence, they are quite independent of the Stoic premises.25 The
argument that κατάληψις is not a μεταξύ, since it is a mere name without any real
determination, is not the dialectical reversal of a Stoic thesis, but a counterargument;
and the same is true of the objection that κατάληψις does not exist if, as the Stoics
argue, it means “assent to a cognitive presentation”: for the assent is given to the
proposition, whereas presentation, according to Zeno’s definition, is a τύπωσις,
which is to say simply an affection of the soul. After attacking the role of assent
as a constitutive part of the definition of κατάληψις by questioning its existence,
Arcesilaus questions the existence of the cognitive presentation that is claimed to
be its object (AM VII 154). Once the Stoic has been forced to accept the thesis that
alongside any purported cognitive presentation another false one may be found
that is indistinguishable from it, Arcesilaus establishes the nonexistence of the Stoic
criterion. Note that Arcesilaus infers ἀκαταληψία from his demonstration of the
nonexistence of the Stoic criterion, not of a criterion in general. So the thesis that
all things are non-cognitive does not concern him, but only the Stoics, according to
whom κατάληψις exists. After having demonstrated the nonexistence of κατάληψις,
Arcesilaus concludes: “If cognition does not exist, everything will be noncognitive.
And if everything is noncognitive, it will follow that according to the Stoics too
(καί), the wise man suspends judgment” (AM VII 155, trans. Long and Sedley 1987).
Up until this point, Arcesilaus has only been disputing against, opposing Stoic
arguments by counterarguments, according to the dialectical method of arguing the
opposite. Arcesilaus’s argument is structured in such a way as to lead the Stoic sage
to accept the thesis of ἀκαταληψία, which is precisely the opposite thesis to the Stoic
one; yet he does not positively establish its truth. So, if ἀκαταληψία is a dialectical
thesis, what still remains to be determined is whether ἐποχή is a conclusion to
which Arcesilaus is committed, or whether he is resorting to the demonstration of
ἀκαταληψία merely as a reductio ad absurdum of the stance of the Stoic wise man,
in such a way as to lead him to universal suspension of judgment. In the latter part
of his refutation (AM VII 156–157), starting from a reversal of the Stoic theses,
Arcesilaus then turns them into their contradictory opposites. It is in this latter
section that he demonstrates that, if the Stoic sage does not wish to opine, he must
withhold assent, that is, that he is forced to ἀσυγκαταθετεῖν. If Arcesilaus merely
wished to provide a reductio ad absurdum of the Stoic position, it would be enough
for him to demonstrate that, since everything is non-cognitive, the only option the
Stoic wise man has is to opine. Arcesilaus does not ignore this argument: in fact, he
adopts it as an intermediate step before leading the Stoic sage to ἐποχή. Arcesilaus
forces the sage to acknowledge that “assent to the non-cognitive is opinion,”
drawing this definition from a reductio ad absurdum of Zenonian cognition. Hence,
the above definition of opinion holds only against the Stoics. “But the wise man is
certainly not one of those who opine”: this is the premise that Arcesilaus shares with
Zeno and that they had agreed upon before engaging in the debate, as Cicero makes
clear in his account (Acad. II 76–77). Sextus does not mention the two philosophers’
42 SKEPTICISM
agreement with regard to this premise, as he is not interested in presenting

Arcesilaus’s point of view. If “assent to the non-cognitive is opinion” and “the wise
is certainly not one of those who opine,” then the Stoic sage has no other option but
to withhold assent. Yet, “to withhold assent (ἀσυγκαταθετεῖν) is no different from
suspending judgment (ἐπέχειν)” (AM VII 157). From the shared premise that “the
wise is certainly not one of those who opine,” two consequences follow: the Stoic
sage is forced to withhold assent because κατάληψις does not exist, whereas in the
face of the equipollence of two opposite theses—the Stoic thesis of the existence
of cognition and the demonstration of its nonexistence—the Academic is forced to
suspend judgment, since he finds himself in aporia, that is, incapable of asserting
or denying that there is knowledge. The ἐποχὴ περὶ πάντων (universal suspension
of judgment), that is, the attitude of the Academic wise man, is thus what the Stoic
sage is left with, too—where “too” stresses the fact that it is the Stoic sage who
must conform to the behavior of the Academic sage, if he wishes not to formulate
any opinions. If “to withhold assent” (ἀσυγκαταθετεῖν) had the same meaning as
“to suspend judgment” (ἐπέχειν), there would be no point in Arcesilaus forcing the
Stoics to accept the equivalence of the two terms, since the Stoics themselves would
be treating them as synonyms. Moreover the verb ἀσυγκαταθετεῖν is composed
of α-privativ + the verb συγκαταθετεῖν, clearly referring to the Stoic concept of
assent, to which Arcesilaus is not committed, as has been shown before, while the
verb ἐπέχειν does not need necessarily assent as its object.26 The introduction of the
two verbs makes sense only if we admit that they describe two attitudes that differ
fundamentally, not just linguistically. The use of a different term to denote the
attitude of the Stoic sage compared to that of the Academic sage reveals that ἐποχή
is Arcesilaus’s own point of view, and cannot be reduced to a dialectical thesis.
If then, as it seems, Arcesilaus does not infer ἐποχή from the nonexistence of the
Stoic criterion, but upholds it as his own view, this must be on the basis of other
considerations that are not mentioned here by Sextus.
4 UNIVERSAL SUSPENSION OF JUDGMENT AS THE

“NATURAL” AND UNAVOIDABLE OUTCOME OF
THE EQUAL STRENGTH OF OPPOSITE ARGUMENTS
It remains to be established whether the previous considerations are compatible with
skepticism. Sextus himself in PH I provides the considerations that he omits in AM
VII: “[Arcesilaus] seems to me to have very much in common with the Pyrrhonian
discourses, so that his school is practically the same as ours” (PH I 232, trans. Long
and Sedley 1987).27 Sextus justifies his assertion in the following terms: Arcesilaus
“is not found making assertions about the existence or nonexistence of anything,
nor does he prefer one thing to another by way of credibility or incredibility, but
suspends judgment about everything. He also says that suspension of judgment
is the end” (PH I 232, trans. Long and Sedley 1987). Sextus provides two more
possible interpretations of Arcesilaus’s suspension of judgment. According to one
interpretation, attributed to an anonymous “someone” (τις), Arcesilaus, unlike
ARCESILAUS 43
the Pyrrhonists, speaks of suspension of judgment with regard to the real nature
of things. But while Arcesilaus does not speak in favor of either the existence or
nonexistence of things, there is no evidence for the idea that the expression “in
reference to nature” (ὡς πρὸς τὴν φύσιν) may refer to a cognitive-ontological
pronouncement on the nature of the external world. The third interpretation, again
drawn from an anonymous source, presents Arcesilaus as being covertly Platonic.
Yet Aristo’s parody of the Homeric verse, which is quoted in support of the esoteric
interpretation, suggests the very opposite of what the charge of esotericism directed
against Arcesilaus implies: for πρόσθε Πλάτων indicates that Arcesilaus was openly
Platonic, not Pyrrhonian.28 While Sextus’s discussion of Arcesilaus’s philosophy is
certainly based on διαφωνία (disagreement)—Arcesilaus is presented as a genuine
skeptic first, then as a negative dogmatist, and finally as a covert Platonist—ultimately
the three accounts are not set on the same level: not just because the account that
assimilates Arcesilaus within Pyrrhonism is explicitly endorsed by Sextus himself,29
but also because the other two are only weakly justified. It is noteworthy, moreover,
that Sextus does not ground Arcesilaus’s skepticism on his polemic against the
Stoics, but rather presents ἐποχή as Arcesilaus’s point of view, in agreement with the
most ancient sources, predating Sextus himself—from Chrysippus30 to Cicero. But
if ἐποχή emerges from the opposition between two contradictory theses, neither
of which prevails over the other (Acad. I 45; DL IV 28), then it simply describes
the state of mind in which the Academic finds himself; as such, it cannot be the
τέλος of Arcesilaus’s philosophy, as Sextus would have it. Besides, in his account
on the criterion of conduct at AM VII 158, Sextus himself argues that according
to Arcesilaus the goal of life is happiness, thus leaving open the possibility that the
identification of ἐποχή as the τέλος may simply be a polemical expedient.
In order to grasp the meaning of Arcesilaus’s ἐποχή, it is worth dwelling on
the words with which Cicero ends his account of the debate between Zeno and
Arcesilaus. He states that the thesis that “the wise man will assent to nothing” was
no longer taken into consideration in the dispute, since the wise man “might ‘grasp
nothing and yet opine’—a thesis Carneades is said to have accepted” (Acad. II 78).
Indeed, the thesis of the opinionless31 sage may only be upheld provided one practices
universal ἐποχή, as Arcesilaus did. However, this thesis has serious implications
with regard to the issue of conduct, so much so that later Academics abandoned
it.32 In Clitomachus’s day, Academics were still debating how to practice universal
suspension of judgment. Clitomachus distinguishes between two ways of conceiving
suspension of judgment: “one, when it means that he assents to nothing at all; the
other, when he checks himself from responding in such a way as to accept or reject
something with the result that he neither denies nor asserts something” (Acad. II
104; trans. Long and Sedley 1987).33 The latter way of conceiving suspension agrees
with the meaning of ἐποχή that other sources attribute to Arcesilaus, as ἐπέχειν τὰς
ἀποφάσεις.
On Clitomachus’s account the action does not require assent: the skeptic may
react positively or negatively provided he does not assent (dum sine adsensu).34 This
means that Clitomachus, by distinguishing two modes of suspension, is offering
not so much a theory of qualified assent as a theory of “qualified non-assent” (see
44 SKEPTICISM
Burnyeat 1997). Clitomachus rejects the notion of suspension as checking oneself

from responding “Yes” or “No” to questions, which was implied by Arcesilaus’s
ἐποχή περὶ πάντων, since it does not allow one to provide a rational justification
for one’s actions, and hence to answer the charge of ἀπραξία (inaction) leveled
against skeptics by dogmatists, as is illustrated by Arcesilaus’s difficulty in defining
a criterion of conduct.
5 THE CRITERION OF CONDUCT AND THE

LANGUAGE OF “NON-ASSERTION”
Sextus focuses on the criterion of conduct in the last section of his account on
Arcesilaus at AM VII 158. The first thing one notices here is the sudden change
of tone compared to the previous section. Sextus directly presents Arcesilaus’s
position, without first introducing the Stoic view, in open contrast to the dialectical
procedure he has attributed to Arcesilaus so far. Furthermore, the criterion of
conduct is presented as something that the person suspending his judgment must
“investigate” after the criterion of truth has been abolished. Most importantly, the
language used by Arcesilaus undergoes a shift of tone, as attested by the fact that all
the finite verbs are expressed in the future tense to indicate that the philosopher is
delivering an exhortation rather than arguing dialectically:
Since after this it was necessary (ἔδει) to investigate (ζητεῖν) the conduct of life
too, which is not of a nature (πέφυκεν) to be explained without a criterion,
on which happiness too, i.e., the end of life, has its trust dependent, Arcesilaus
says that one who suspends judgment about everything will regulate (κανονιεῖ)
choice and avoidance and actions in general by the reasonable (εὔλογον); and by
proceeding in accordance with this criterion, he will have success (κατορθώσει).
(AM VII 158, trans. Long and Sedley 1987 modified.)
It is significant that the expression οὐδὲν ὁρίζειν that Sextus had used to introduce
the criterion of truth is matched here by the verb ζητεῖν, which introduces the
criterion of conduct. Sextus’s stress on language, as a way of distinguishing
Pyrrhonian skepticism from Academic skepticism, makes it unlikely that the choice
of the two verbs, both belonging to the jargon of Pyrrhonian skepticism (see PH I
1–3, 7), is merely a casual one. The verb “to investigate” emphasizes that Arcesilaus
is suggesting the adoption of a criterion of conduct, the εὔλογον, as an open,
temporary solution, compatible with the universal suspension of judgment.35 The
argument by which Arcesilaus recommends the εὔλογον is based on three premises:
(1) happiness is attained through prudence; (2) prudence resides in successes (κατο-
ρθώματα); and (3) a success is whatever, once it has been done, has a reasonable
defense. Arcesilaus concludes: “Therefore one who attends to the reasonable will
have success and be happy.” This argument draws upon some concepts that are
apparently of Stoic origin, such as εὔλογον, καθῆκον, and κατόρθωμα: what
is especially crucial is Arcesilaus’s definition of κατόρθωμα, which is identical
to Zeno’s definition of καθῆκον.36 And since the term καθῆκον is used in Stoic
ARCESILAUS 45
doctrine to refer to the action of the ordinary man, as opposed to κατόρθωμα,

which describes the wise man’s action, the supporters of the dialectical interpretation
have concluded that Arcesilaus was being ironical: that he sought to prove that once
deprived of his knowledge, the Stoic sage is forced to act as an ordinary man, which
is to say as a fool. We would be compelled to accept this interpretation if Arcesilaus
had borrowed his notion of εὔλογον from the Stoa. But apart from the fact that the
concept of εὔλογον was far from new in Greek thought, having previously been
widely employed by Aristotle, in the debate between the Stoa and the skeptical
Academy, it is not always easy to ascertain which school borrowed certain technical
terms and concepts from the other (Glucker 1978: 33 with n.78; Görler 1994:
817). The use of a common technical terminology, therefore, is not enough in itself
to prove Arcesilaus’s dialectical reading, unless evidence is found for the Stoic use
of the same notions. Moreover, if Arcesilaus had wished to suggest the idea of an
ironic exchange between καθῆκον and κατόρθωμα, he ought to have made at least
some allusion in his arguments to the difference between the two kinds of actions
according to the Stoics and to have recalled Zeno’s position—as he had indeed done
in order to refute the criterion of truth. But unlike in the first section, where the noun
“Stoics” occurs twice between 155 and 157, highlighting the ad hominem nature of
the reply, in the section on the criterion of conduct that noun does not appear at all.
Nor does Arcesilaus mention it in the three arguments by which he establishes the
εὔλογον as the criterion to be adopted by the person who suspends his judgment in
order to regulate (κανονιεῖ) his actions. The choice of the verb κανονίζειν likewise
suggests a regulatory norm, a κάνων, with an epistemological and justificatory
function, rather than a prescriptive one. The definition of κατόρθωμα—identical
to Zeno’s definition of καθῆκον—is a formal one, since it stresses the a posteriori
justification of action, yet without enunciating the principles on which this action
is based. The past participle πραχθέν is used to emphasize that “rational defense”
comes after the action has been performed, not before.37 Hence, the εὔλογον cannot
be identified with the judgment “It is reasonable that p,” if this is expressed before
performing the action, since the εὔλογον is not taken in this sense in the definition
of κατόρθωμα. Either Arcesilaus had expressed himself unclearly, or Sextus has
omitted some details, to shorten his account. Favoring the latter hypothesis is Sextus’s
complete silence with regard to Arcesilaus’s answer to the charge of ἀπραξία. It
is rather strange that Sextus ignores it, not just because the εὔλογον acquires a
far clearer meaning when viewed in its light, but especially because this accusation
represents a problematic issue that we know for sure Arcesilaus had to address, since
it concerns the very credibility of skepticism, as witnessed by the accusation raised
by his contemporary Colotes (see Adv. Col. 1121F–1124B). Faced with the charge
leveled by Stoics and Epicureans that by removing assent the Academic can neither
act nor avoid committing absurd or random actions, Arcesilaus replies that neither
the Academic’s faculty of presentation nor his impulsive faculty are impaired. His
impulse will thus naturally (φυσικῶς) lead him toward what is appropriate (οἰκεῖον),
without assent ever coming into play. For error derives from opinion and hence
from assent. Following nature means being guided by impulse without assenting
to the presentations by which one is involuntarily affected. It is nature that ensures
46 SKEPTICISM
both the possibility of action and the possibility of not committing any errors. The
importance of nature in providing guidelines for one’s conduct links the Academic’s
description of the mechanism of action to the rational justification suggested by the
εὔλογον, since the conduct of life “‘is not of a nature’ (οὐ πέφυκεν) to be accounted
for without a criterion on which happiness too, i.e., the end of life, has its trust
dependent.”
Plutarch’s account, which explains how Arcesilaus describes the mechanism of
action, not only answers the charge of total inaction,38 but also gives meaning to the
εὔλογον, which otherwise could not be accounted for—be it as a dialectical reversal
of a Stoic view or as a criterion of conduct. Only by integrating Plutarch’s testimony
with Sextus’s is it possible to answer both objections contained in the charge of
ἀπραξία, and—most importantly—to grasp the connection drawn by Arcesilaus
between the criterion of conduct and nature, as asserted in both accounts. Sextus
himself, in PH I as well as in AM XI, provides an explanation for the Pyrrhonist’s
action without assent, which stresses the crucial role of nature and is very similar
to Arcesilaus’s explanation for action without assent.39 What is also suspicious is
Sextus’s silence with regard to Arcesilaus’s criterion of conduct in PH I, since he
discusses and refutes the criterion of conduct of Carneades and Clitomachus at
PH I 231, distinguishing it from the Pyrrhonists’ criterion of conduct, yet makes
no reference to Arcesilaus’s formulation of a criterion. It cannot be ruled out that
Sextus sought to conceal the role played by nature in Arcesilaus’s philosophy, as
Pyrrhonian skepticism was greatly indebted to it.
A comparative analysis of the sources shows that Arcesilaus’s skepticism does
not involve any commitment to the doctrine of ἀκαταληψία but is compatible with
universal ἐποχή. The latter is not the final goal for Arcesilaus, and hence does not
represent a substantive belief. Rather, it is the practical expression of the theoretical
difficulty that the Academic experiences in the face of aporia, which invariably emerges
from the equipollence of opposite theses. The problem that Arcesilaus must address
is one common to all forms of skepticism, namely: whether it is legitimate for the
skeptic to communicate his own non-position to others by drawing upon concepts
such as reason, opinion, and truth. The only way of communicating without making
assertions that will concern oneself is to develop a language that is distinct from that
of dogmatists, but that at the same time does not reject the use of their concepts.
Traces of this reflection on language are to be found in various accounts of Arcesilaus’s
philosophy, ranging from Cicero (Acad. I 45) to Diogenes Laertius (DL IV 36)40 and
Sextus Empiricus himself. The quest for a language of “non-assertion” indicates that
Arcesilaus foreshadowed Pyrrhonian skepticism in his awareness that every assertion,
by implying the presence of a subject, can ultimately be turned against itself. Sextus
himself, who chiefly perceives the difference between Pyrrhonian skepticism and
Academic dogmatism as a matter of language, goes so far as to admit that Arcesilaus
does not assert anything διαβεβαιωτικῶς,41 while at AM VII 150 he claims that
Arcesilaus and his followers οὐδὲν ὥρισαν κριτήριον. More generally, both accounts,
in PH I and AM VII, suggest that Arcesilaus employed not just a dialectical terminology
drawn from a refutation of Stoic theses—which, as such, cannot be attributed to
him—but also a language of his own to convey his skeptical attitude.
ARCESILAUS 47
NOTES
1. Cf. Antigonus of Carystus apud Philodemus, Academicorum Historia col. 19,
11–16 (154 Dor.= fr. 20A). As Long (2006: 96–114) has persuasively suggested,
Diogenes’s Life of Arcesilaus almost certainly derives from Antigonus of Carystus—a
reliable source, since he was a contemporary of Arcesilaus and “was still being read
as late as the second century AD” (Long 2006: 101).
2. On the epistemological dispute between Aristo of Chios and Arcesilaus, see DL
VII 162–163, as well as Ioppolo (1980: 26–33; 2013: 188–190) and Long
(2006: 106–108).
3. See DL IV 33; PH I 234; Numenius apud Eusebius, Praep. evang. XIV 6, 4–5 = fr.
25 Des Places.
4. As Colotes’s work—Plutarch’s source—was written around 260 BCE, its doctrinal
and polemical context can only coincide with the height of Arcesilaus’s career.
On the reliability of Adversus Colotem as a source for Arcesilaus’s philosophy,
see Ioppolo (1986: 121–156; 2000: 333–360; 2009: 123 n. 98); also Warren
(2002: 333–360).
5. The same charge is leveled against Arcesilaus at Acad. II 15.
6. PH I 234 quotes the verse by intentionally interpreting it as meaning the very
opposite (see below).
7. Cf. Acad. II 130; Fin. II 35, III 31, 43, IV 40, etc.
8. See DL VII 2; 31; Cicero, De or. III 61. On the Stoics’ claim to be following Socrates,
see Philodemus, De Stoicis XIII 3.
9. See Ioppolo (1980: 78–90).
10. Some scholars interpret Acad. I 44–45 as evidence that ἀκαταληψία (non-
cognition) was endorsed by Arcesilaus. See Maconi (1988: 246–247). As Brittain
and Palmer (2001: 44) note, however, “It is not clear . . . how Arcesilaus’s appeal to
the Presocratics supports the akatalepsia premise because his appeal is susceptible to
both a ‘dialectical’ and a ‘non-dialectical’ reading.”
11. huic rationi quod erat consentaneum faciebat, ut contra omnium sententias
dicens, in eam plerosque deduceret ut cum in eadem re paria contrariis in partibus
momenta rationum invenirentur, facilius ab utraque parte adsensio sustineretur.
I follow Madvig’s reading dicens in eam, as this explains how Arcesilaus would
lead his interlocutors to share his point of view, that is, to suspend all judgment.
This controversial passage is crucially important for any attempt to understand
Arcesilaus’s philosophy.
12. Cicero first attributes the method of in utramque partem disserere to Aristotle at
Tusc. II 9. Cf. Fin. V 10–11 and De or. III 80. Arcesilaus had been a pupil of
Theophrastus (DL IV 29).
13. See Plutarch, St. Rep. 1037A, ironically noting with regard to Chrysippus: “in
the ambition to outdo Arcesilaus” (ὑπερβαλέσθαι φιλοτιμούμενος τὸν Ἀρκεσί-
λαον). This confirms the hypothesis formulated by Long (2006: 97): “Traces of
Arcesilaus’s actual words, we may presume, were transmitted in writing through the
Academy’s Stoic opponents, especially Chrysippus (DL 7.183), who was his student
for a time.”
48 SKEPTICISM
14. St. Rep. 1036A: τοῖς μὲν γὰρ ἐποχὴν ἄγουσιν περὶ πάντων ἐπιβάλλει, φησί, τοῦτο
ποιεῖν. Regardless of which of Chrysippus’s works the quotation may be taken
from, it confirms that the πρὸς τὰ ἐναντία διαλέγεσθαι method was at the core of
Arcesilaus’s philosophy.
15. St. Rep. 10, 1035F–1037C, faithfully records Chrysippus’s words, as well as the
titles of his works.
16. St. Rep. 1036 D–Ε: οὐχ ὡς ἔτυχεν δ’οὐδὲ τοὺς ἐναντίους ὑποδεικτέον
λόγους . . . ἀλλ’εὐλαβουμένους μὴ καὶ περισπασθέντες ὑπ’αὐτῶν τὰς
καταλήψεις ἀφῶσιν, οὔτε τῶν λύσεων ἱκανῶς ἂν ἀκοῦσαι δυνάμενοι
καταλαμβάνοντες τ’εὐαποσείστως.
17. DL IV 28: ἐπισχὼν τὰς ἀποφάσεις διὰ τῆς ἐναντιότητας τῶν λόγων (suspending
his assertions owing to the contrarieties of arguments).
18. The dialectical procedure cannot be cut off from its aim, which according to
Platonists consists in the “search after truth”: see Plat. Apol. 21 B–23 B, 38 A; Gorg.
505 E.
19. The words vera . . . honesta . . . digna sapienti reveal that Arcesilaus shared this
premise. It is worth recalling that the precondition for Socratic ἔλεγχος is to be
found in the sincerity with which the two interlocutors express their views: see Plat.
Gorg. 457D–458B; Prot. 331C–D; Rep. 346A; Theaet. 167D–168C.
20. In Ioppolo (1989: 232–233), I explain why I prefer to translate φαντασία as
“presentation” instead of “impression.”
21. This is the problem that Carneades was to formulate more clearly by drawing a
distinction between the two σχέσεις of presentation: see AM VII 168–169.
22. προηγουμένως μὲν οὐδὲν ὥρισαν κριτήριον. On the various possible ways of
translating this passage, see Ioppolo (2009: 81–82).
23. See PH I 15; also PH I 197, 203, and Spinelli (2005: 114–126).
24. See AM VII 159–166, on the different dialectical procedures that Sextus adopts in
his account on Carneades.
25. For a detailed analysis of Arcesilaus’s arguments, partly drawn from Plato’s
Theaetetus, see Ioppolo (2009: 82–93).
26. See Arcesilaus’s explanation of how action is possible in the absence of assent in
Plut. Adv. Col. 1121F-1124B.
27. The importance of this κοινωνία with Arcesilaus, acknowledged by Sextus himself,
is further witnessed by the limits he sets to the κοινωνία with Protagoras at PH
I 217.
28. Note that in the transition from his enunciation of ἐποχή as a good to that of its
interpretation from a dogmatic perspective—ὡς πρὸς τὴν φύσιν—Sextus switches
from the first person to the third person of his anonymous source: πλὴν εἰ μὲν λέγοι
τις (PH I 233). Sextus mentions the charge of esotericism in even more hypothetical
terms: εἰ δὲ καὶ τοῖς περὶ αὐτοῦ λεγομένοις πιστεύειν, φασίν (PH I 234). See also
Ioppolo (2009: 42–52).
29. Note that Sextus uses the first person very rarely and usually only when his own
view is not subject to διαφωνία: see PH I 237, 239; II 9, 10, 22, 98, 204, 212.
30. See Plutarch, St. Rep. 1036A. See also Colotes in Plut. Adv. Col. 1120C: τοὺς περὶ
Ἀρκεσίλαον Ἀκαδημαικούς. Οὗτοι γὰρ ἦσαν οἱ περὶ πάντων ἐπέχοντες.
ARCESILAUS 49
31. See Acad. II 67 on the divergence between Arcesilaus and Carneades with regard to
the thesis nihil opinari sapientem. See also Ioppolo (2007: 238–241).
32. Numenius apud. Eusebius, Praep. Evang. XIV 7, 15 (fr. 26 Des Places) distinguishes
Arcesilaus’s position from that of Carneades precisely with reference to the
ἐποχὴ περὶ πάντων, which Carneades does not maintain, since it abolishes the
distinction between “uncertain” (ἄδηλον) and “noncognitive” (ἀκατάληπτον).
Regarding ἄδηλον as a technical term of Arcesilaus’s philosophy, see Ioppolo
(2009: 193–208).
33. Note that the verb ἐπέχειν does not always require συγκατάθεσις as its object: see
Cicero, Ad Att. XIII 21, 3. Cicero discusses the matter in relation to the dispute
between Chrysippus and Carneades on the sorites paradox at Acad. II 93 (see also
Ioppolo 2007: 227–267).
34. I have discussed this topic extensively; see Ioppolo 2007.
35. The practice of ζητεῖν is a distinguishing feature of Arcesilaus’s philosophy, which
was modeled after that of Socrates: see Plat. Apol. 21B–23B, 38A; DL II 22. Cf.
Acad. I 44–46, II 76–77 on verum invenire velle.
36. See DL VII 107–108. On the peculiar meaning of Arcesilaus’s κατόρθωμα, see
Ioppolo (2009: 118–124).
37. With regard to the word κατόρθωμα, Tsekourakis (1974: 121) observes: “Terms
in -μα . . . seem to have presented the act as a whole, as finished and with its results
present. . . . We often find a perfect tense in the definition of the concepts expressed
by these terms.”
38. See Plut. Adv. Col. 1122B: ὥσπερ Γοργόνα τὴν ἀπραξίαν ἐπάγοντες.
39. See Adv. Col. 26 1122B–1122 F and PH I 23–24. See also AM XI 162–168 and
Spinelli (1995: 325–335). Cf. AM VII 158 with AM VII 30.
40. Φυσικῶς δέ πως ἐν τῷ διαλέγεσθαι ἐχρῆτο τῷ Φημ᾽ἐγώ, καί, οὐ συγκαταθήσεται
τούτοις ὁ δεῖνα, εἰπὼν τοὔνομα. Φημ᾽ἐγώ. “I say” translates “Yes” in the context
of an ordinary exchange during a discussion and may be explained as a natural
impulse that would drive him to speak, even though this did not imply any claim to
truthfulness with regard to the content of his words, thereby freeing the expression
from any assertive meaning.
41. From the very opening of PH I 4 Sextus uses two specific verbs in order to
distinguish dogmatic διαβεβαιοῦσθαι from the ἀπαγγέλλειν. Note that Sextus does
not name Arcesilaus among the Academics, implicitly including him among the
anonymous skeptics.
REFERENCES
Brittain, Charles and John Palmer. 2001. “The New Academy’s Appeal to the Presocratics,”
Phronesis 46: 38–72.
Burnyeat, Myles. 1997. “Antipater and Self-Refutation: Elusive Argument in Cicero’s
Academica.” In B. Inwood and J. Mansfeld (eds.), Assent and Argument: Studies in
Cicero’s Academic Books, 277–310. Leiden.
Ruprecht.
50 SKEPTICISM
Görler, Woldemar. 1994. “Älterer Pyrrhonismus. Jüngere Akademie. Antiochos aus

Askalon.” In H. Flashar (ed.), Die Philosophie der Antike 4: Die hellenistische Philosophie,
717–989. Basel: Schwabe.
Ioppolo, Anna Maria. 1980. Aristone di Chio e lo Stoicismo antico. Napoli: Bibliopolis.
Ioppolo, Anna Maria. 1986. Opinione e scienza. Il dibattito tra Stoici e Accademici nel terzo
e secondo secolo a.C. Napoli: Bibliopolis.
Ioppolo, Anna Maria. 1989. “Stoici e Accademici sul ruolo dell’assenso,” Elenchos 10:
231–246.
Ioppolo, Anna Maria. 2000. “A proposito di alcune recenti interpretazioni dello scetticismo
dell’Accademia,” Elenchos 21: 333–360.
Ioppolo, Anna Maria. 2007. L’assenso in Clitomaco:un problema di linguaggio? In A. M.
Ioppolo and D. Sedley (eds.), Pyrrhonists, Patricians, Platonizers. Hellenistic Philosophy
in the Period 155–86 B.C., 225–267. Napoli: Bibliopolis.
Ioppolo, Anna Maria. 2009. La testimonianza di Sesto Empirico sull’Accademia scettica.
Napoli: Bibliopolis.
Long, Anthony. 2006. “Arcesilaus in His Time and Place.” In his From Epicurus to Epictetus:
Studies in Hellenistic and Roman Philosophy, 96–114. Oxford: Oxford University Press.
Long, Antony, and David Sedley. 1987. The Hellenistic Philosophers, 2 vols. Cambridge:
Maconi, Henry. 1988. “Nova Non Philosophandi Philosophia,” Oxford Studies in Ancient
Spinelli, Emidio. 1995. Sesto Empirico: Contro gli Etici. Napoli: Bibliopolis.
Spinelli, Emidio. 2005. Questioni scettiche: Letture introduttive al Pirronismo antico.
Roma: Lithos.
Tsekourakis, Damianos. 1974. Studies in the Terminology of the Early Stoa. Wiesbaden:
Franz Steiner Verlag.
Warren, James. 2002. “Socratic Scepticism in Plutarch’s ‘Adversus Colotem,’” Elenchos
23: 333–356.
CHAPTER FOUR
Carneades
HARALD THORSRUD
1 INTRODUCTION
Carneades of Cyrene (214–129/8 BCE) became head of the Academy sometime
before 155 BCE, a little more than one hundred years after Arcesilaus initiated
a skeptical movement in the school. Inspired by his reading of Plato’s dialogues,
Carneades adopted a version of Socrates’s dialectical project—recognizing his own
ignorance and the supreme importance of knowledge, he set about testing, and
ultimately refuting, others.1 Carneades followed in Arcesilaus’s footsteps, continuing
and strengthening this method.2 Like his predecessors, he wrote nothing,3 and was
formidable if not unconquerable in debate.4 And as head of the Academy, he too
must have acknowledged some connection with Socrates. So even though he focused
on the refutation of the Stoics (see DL IV 62, St. Rep. 1035F), we should not lose
sight of the Platonic or Socratic context for his argumentative practice.
According to one version of the history of the Academy, Socrates initiated a
dialectical method aimed at (i) relieving others of falsely believing they know what
they don’t, and (ii) revealing the most probable solutions to the problems discussed,
while (iii) concealing his own opinion so as not to interfere with the rational
autonomy of the participants (Tusc. V 11).5 There is no doubt that Carneades
devoted himself to the first of these three aims. He produced powerful objections
to the Stoic view of divination,6 to both Stoic and Epicurean views of freedom and
determinism,7 and he argued forcefully against Stoic theology.8 He also expanded
the range of targets in ethics and epistemology, arguing not only against all the
actual views held on the ethical τέλος and on the criterion of truth, but against all
possible views, given certain restrictions on the candidates (Fin. V 16, AM VII 159).9
Whether, or in what sense, Carneades embraced (ii) and (iii) is controversial.10
He is credited with developing an account of living in accordance with the probable
or persuasive impression11 to explain how, while suspending judgment, one might
remain active and even attain wisdom and happiness. But it is not clear whether he
himself relied on such impressions in day-to-day life and in the pursuit of truth, or
whether this account was merely part of a dialectical strategy aimed exclusively at
refuting the Stoics. In either case, the proper understanding of Carneades’s view of
probability must still be determined.
52 SKEPTICISM
While the question about the content of his view should be kept distinct from
the question about his attitude toward it, they are related in at least one important
way. If Carneades’s theory successfully shows how one may live well in the absence
of certainty, then it shows this for everyone, not just the Stoics; and in that case he
too would have a powerful motive to accept his own account and live accordingly
(Obdrzalek 2006: 253).
2 THE DIALECTICAL CONTEXT

Like Arcesilaus before him, Carneades focused a great deal of effort on refuting the
Stoic view of wisdom and its epistemic foundation in apprehension (κατάληψις). If
the Stoic sage is to be infallible, there must be something worthy of his assent. Zeno
defined such an apprehensible, or kataleptic, impression (καταληπτικὴ φαντασία)
as (i) one that arises from what is, (ii) is stamped, impressed, and molded just as it
is, and (iii) is of such a kind as could not arise from what is not (Acad. I 42, II 77;
AM VII 252).
One of the arguments Carneades deployed against the third, modal condition is
an epistemological sorites. We begin with a false impression that is persuasive, for
example, that some counterfeit is genuine. Suppose this initial impression is relatively
easy to correct by careful examination. We may next imagine an impression of a more
skillful counterfeit whose deceit is more difficult to discern. And we may increase
this difficulty until we arrive at a false impression that is, for us, indistinguishable
from an impression of the genuine article. For any level of skillful detection we care
to name, we can always imagine an even more skillful counterfeit. In the end, two
equally persuasive impressions, one false and one true, differ in nothing but number
(Acad. II 47–49; AM VII 415–422).
The consequences of Carneades’s arguments, if successful, would be devastating
for the Stoics. For if there is no apprehension, the sage will either have to withhold
assent or assent to what is not apprehended. Neither is acceptable: the former would
render the sage inactive since the Stoics claimed assent is necessary for intentional
action, and the latter would render the sage fallible (Acad. II 67). Without the
possibility of apprehension, it seems the Stoics must choose between an infallible
but inactive sage and an active but fallible one.
In an attempt to remove the source of this dilemma, the Stoics provided some
reductiones of the Academic denial of κατάληψις (i.e., ἀκαταληψία). These
are referred to as apraxia arguments since they are supposed to show that if the
Academic attack succeeds, certain sorts of action would be impossible.12 If there
might be no discernible differences between true and false impressions, we can
have no confidence in the truth of any of our impressions, and the worldviews we
construct might always be systematically deceptive. We would be right in that case
to withhold judgment, but that would paralyze us. Since it is absurd to claim that it
is right to render ourselves utterly inactive, we must reject ἀκαταληψία. Similarly,
if we suppose there is no apprehension, the sage can have no confidence in his
moral judgments. This is especially problematic when morality demands we make
CARNEADES 53
a great sacrifice (Acad. II 23). So if there is no apprehension, there is no wisdom.

But there is (or at least may be) wisdom according to the Stoics, so again we must
reject ἀκαταληψία.
Why would Carneades bother to respond to these apraxia objections? From a
dialectical standpoint, ἀκαταληψία is not his view, and thus he should feel no need
to defend it. But if the Stoics also maintained that no account other than their own
could adequately explain wisdom, then Carneades may have developed a view in
response to that challenge (Allen 1994: 87). In fact, Carneades is supposed to have
offered two distinct alternatives. First, the sage might follow a persuasive impression
without assent and the resulting opinion, and thus without committing himself to
the truth, or even probable truth, of the impression (radical skepticism). And second,
the sage might assent to the truth of persuasive impressions with the appropriate
modesty or caution due to recognizing the possibility of error (mitigated skepticism).
While it might be the case that neither of these is Carneades’s view, I will argue
that the evidence strongly favors mitigated skepticism. First, note that there are two
major problems with attributing radical skepticism to him. In order for the radical
skeptic to be free from opinion, he takes no stand on the truth of what strikes him
as probable (or better: plausible). On this view, Carneades’s radical skeptic appears
no different from Sextus’s Pyrrhonist, who follows appearances without any
commitment as to whether things really are as they appear. By suspending judgment
the Pyrrhonist expects that the tranquility he desires will follow fortuitously. But
since there is no reason to think Carneades’s radical skeptic has a similar expectation,
or even prizes tranquility in the way the Pyrrhonist does, it is hard to see why such
a life would be thought worth living. As a Socratic, he might cherish his intellectual
integrity, but that alone seems an unacceptably thin account of wisdom and happiness.
Without the payoff of tranquility, radical skepticism does not provide Carneades
with an effective response to the apraxia objection in its entirety. At a minimum, his
Stoic interlocutors would reject the claim that one may live a flourishing life in the
absence not only of knowledge, but even reasoned opinion.13
The second problem arises from the incompatibility of radical and mitigated
skepticism: one maintains that the sage is infallible (by holding no opinions), and the
other that the sage is fallible. And yet Cicero appears to endorse both of these as the
proper interpretation of Carneades in the Academica, one of our most important and
detailed sources. Since Cicero explicitly intends to explain his Academic allegiance
in this work (ND I 11), and since he consistently endorses and practices a mitigated
skepticism in every other philosophical dialogue, we should seek an interpretation
of the Academica that is consistent with such a mitigated skepticism (Thorsrud 2002,
2012). To the extent that we succeed, much of the support for Carneades’s radically
skeptical proposal will vanish, and we will be left with the mitigated variety, even if
the possibility remains that Carneades endorses nothing.14
In the remainder of this chapter, I will pursue this approach by discussing
Carneades’s distinction between kinds of impression, the kinds of assent given to
each, the varieties of cognitive state, or opinion, that arises from such assents, and
the compatibility of wisdom with opinion.
54 SKEPTICISM
3 VARIETIES OF IMPRESSION
Carneades argued that no impression could serve as the criterion of truth.
Generalizing on the Academic arguments against Stoic epistemology, he claimed that
for any true impression, an indistinguishable false one could be found. And he argued
that impressions cannot reveal both themselves and their objects or states of affairs in
such a way that we can grasp the correspondence between them with certainty. So all
such criteria will “consist in an appearance [impression] that is common to both true
and false. But the appearance that is common to these two is not apprehensive, and
not being apprehensive, neither will it be a criterion” (AM VII 164).15
So it is surprising to find that, after refuting his predecessors, Carneades
introduces the probable impression as yet another account of the criterion of truth,
especially since he admits that what is probably, or persuasively, true is common to
both what is in fact true and false. It appears that his earlier argument preemptively
refutes this new proposal.16 But this will only be the case if he agreed with his Stoic
opponents that a criterion of truth must infallibly indicate truth.
Carneades’s innovation is to reject that view. “We should not,” he says, “because
of the rare occurrence of this [an apparently and vividly true impression being false]
distrust the one that for the most part tells the truth. For both our judgments and
our actions are, as a matter of fact, regulated by what applies for the most part” (AM
VII 175). Obviously there is no sophisticated or mathematical appeal to frequency
in the notion of what happens for the most part. But it is just as clear that it is an
appeal to an ordinary understanding of frequency—we can comfortably rely on
such persuasive impressions because they have a good track record, they typically
get things right.
Carneades is proposing, apparently for the first time, the notion that probability
is the very guide to life.17 In the absence of certainty, we are not left with the
desperate and impractical view that everything is equally uncertain (Acad. II 99,
103), as if there were no basis whatsoever to prefer one impression to another.18
Such a view would undermine life in the way the Stoics objected. But this is not
Carneades’s position. He holds that some impressions are persuasive because they
appear to be true (AM VII 169), and that among these some are more persuasive
and vivid than others, even though the most vivid may still turn out to be false (AM
VII 173, PH I 227).
Furthermore, he seems to have more than a purely dialectical interest in providing
this account. Sextus reports that “since he [sc. Carneades] too requires some criterion
for the conduct of life and for the achievement of happiness, he is in effect compelled
for his own part to take a stand on this” (AM VII 166). Sextus does not tell us
why he was so compelled. But insofar as there was broad agreement that happiness
(εὐδαιμονία) is the proper specification of our τέλος, we may plausibly suppose that
all Hellenistic philosophers would have been expected to offer guidance in attaining
it (see Hadot 1995; Cooper 2013).
There are additional distinctions to be made among persuasive impressions. In
trivial matters, or when we simply don’t have the time or opportunity for further
scrutiny, we rely on the merely persuasive impression without further ado. But
CARNEADES 55
in more important matters we examine the neighboring impressions, looking

for confirmation or disconfirmation. For the extent to which an impression is
convincing is, in part, a function of the coherence and mutual support provided by
related impressions. So, for example, a good physician will not diagnose a patient
on the basis of one symptom, but will rather look for corroborating evidence. And
in matters of the greatest importance, that is, those pertaining to our happiness, we
subject each of the associated impressions to additional scrutiny, rather than taking
them at face value (AM VII 176–189; cf. PH I 227–229). But what might be the
content of a persuasive impression pertaining to happiness?
The examples typically involve some proposed course of action: for example,
whether to avoid the object on the floor because it appears to be a snake, whether to
set out on a sea voyage, sow crops, get married, etc. (Acad. II 109, AM VII 176–189,
PH I 227). Successfully avoiding coiled snakes or safely crossing the sea is certainly
important, but in what sense, if any, do these actions pertain to my happiness?
It might turn out, unlikely though it seems, that I will be happier as a result of
snakebite, shipwreck, failed crops, or a ruined marriage. If Carneades intended to
limit persuasive impressions to those revealing expected outcomes, it would provide
only the barest minimum of guidance with respect to attaining happiness. For living
in accordance with such impressions will only make it more likely that we satisfy our
desires, and not that we are more likely to be happy by doing so.
Fortunately, there is good evidence that persuasive impressions extend well
beyond the merely perceptual and predictive.19 According to Sextus, members of
Carneades’s Academy affirm that what persuasively appears to be good is really
good (PH I 226). Whether or not this claim is entirely accurate, it does suggest the
existence of persuasive, evaluative impressions—for example, that it would be good
to do something and not merely that the proposed action will succeed. Furthermore,
Cicero tells us that when the sage is “asked about what it is appropriate to do
[de officio]. . . . He won’t say he doesn’t know, as when asked about whether
the number of the stars is even or odd” (Acad. II 110).20 This allows for deontic
impressions, which we should expect given that “the wise person will use whatever
strikes him as persuasive . . . and the whole structure of his life will be governed in this
way (omnis ratio vitae gubernabitur)” (Acad. II 99; see also Acad. II 104, PH I 231).21
But since it would be frivolous to act on the basis of whatever happens to strike
one as plausibly worth doing or as what one is plausibly required to do, the sage’s
moral judgment must be informed by rational principles (Acad. II 35–36). Now if,
as the Stoics think, the sage can have no principles (decreta = δόγματα) that he fails
to securely grasp as true, Carneades would have to admit that his Academic sage
securely grasps this much: that nothing is apprehensible. Cicero responds, on behalf
of Carneades: “As if the wise person has no other principles and could live his life
without principles! But just as he holds those as persuasive rather than apprehended
principles, so with this one, that nothing is apprehensible” (Acad. II 109–110; see
also Acad. II 29, PH I 226). Accordingly, Carneades seems to allow for philosophical
impressions as principles governing the life of the sage.22
The only plausible explanation of how Carneades’s sage could arrive at such
principles is philosophical investigation: the sage “will deliberate about what to
56 SKEPTICISM
do or not to do on the basis of [persuasive] impressions” (Acad. II 100). Indeed,

Carneades and his followers are said to have been especially inspired by their
claim that persuasive or truth-like impressions provide them with a rule for both
conducting their lives and for philosophical investigation and disputation (regula et
in agenda vita et in quaerendo ac disserendo) (Acad. II 32).23 Cicero tells us that the
whole point of Carneades’s refutative practice was to discover the truth, or what
is most like the truth, the most probable solution (ND I 11–12, Tusc. V 10–11; see
also Acad. II 7–9, ND I 4, Div. II 150).24
If Carneades develops his account of persuasive impressions to show not only
how we may navigate through the world but also how we may achieve wisdom,
there will have to be a wide variety of persuasive impressions for the sage to rely
on. And his dialectical practice of arguing for and against must, as Cicero repeatedly
insists, be aimed not only at refutation but also at revealing which position is most
truth-like, or rationally defensible; otherwise Carneades’s wisdom will merely
enable the sage to navigate through the world and satisfy his desires.
But how then can the method of subjecting sense-impressions to increasing
degrees of scrutiny and the method of arguing pro and con both rely on the same
criterion? The ordinary sense of frequency invoked in the former seems entirely
out of place in the latter—our philosophical judgments aren’t regulated by what
happens for the most part. While none of our sources address this problem, there is
a ready solution. In both cases the reigning assumption is that truth is what survives
refutation.25 When we subject an impression to closer scrutiny by first checking for
overall coherence with its associated impressions and then by examining each of
those impressions independently, we are seeking to falsify the target impression.
If we fail to do so, the target impression becomes even more persuasive—it has
survived refutation.
The Academics do not deny that some impressions are in fact true, but only that
we can be certain of which are true (Acad. II 73, 111). And the possibility that what
is true can, at least in some cases, constrain our judgments is implicit in a remark
that Carneades used to make:
the sons of the wealthy and sons of kings do learn to ride on horseback, but they
learn nothing else well and properly; for in their studies their teacher flatters
them with praise, and their opponent in wrestling does the same by submitting to
be thrown, whereas the horse, having no knowledge or concern even as to who is
private citizen or ruler, or rich or poor, throws headlong those who cannot ride
him. (Plutarch, Adulator 58f5, tr. Babbit 1927)
Similarly, only life itself may correct those philosophical views that are left untouched
by overly polite or incompetent interlocutors. In the dispute over Stoic epistemology,
where we have competent opponents, the extent to which the claim that nothing is
apprehensible survives refutation is the extent to which we will, and should, find it
persuasive. The dialectical testing of philosophical impressions is thus analogous to
the scrutiny of perceptual impressions. And while the notion of dialectical survival
is admittedly vague, it does show how one might reasonably think that τὸ πιθανόν
provides a rule for both the conduct of life and philosophical investigation.
CARNEADES 57
4 VARIETIES OF ASSENT
Just as Carneades offers the persuasive impression as an alternative to the kataleptic
impression as the sage’s guide to life, he also provides an alternative to the Stoic
view of assent. According to the Stoics, strong assent is given only to kataleptic
impressions (or more precisely, only to the true propositions that articulate the
content of kataleptic impressions) and in such a way as to be irreversible. To be
capable of assenting strongly requires the virtue of dialectic, which “conveys a
method that guards us from giving assent to any falsehood or ever being deceived
by specious probability, and enables us to retain and defend the truths that we have
learned about good and evil” (Fin. III 72, tr. Rackham 1914). This dialectical virtue is
also described as freedom from precipitant or rash judgment: “knowledge [of] when
to give or withhold the mind’s assent to impressions,” or “the disposition not to
assent in advance of apprehension” (DL VII 46–48; see also Plutarch, St. rep. 1056F,
1057B–C). Having this virtue, the Stoic sage never believes a falsehood, and never
assents to anything that is not apprehended (Stobaeus, Anthologium II 7, 11m).
By contrast, a weak assent is one that would not have been given had the
circumstances been even slightly different. The foolish are too easily swayed by
factors that are irrelevant to the truth or falsity of the proposition in question; when
they get things right, it is to some extent just a matter of good fortune. The foolish
lack the systematic knowledge that secures each individual piece of knowledge, so
even when they assent to a true proposition, the assent is weak because changeable.
These distinctions rest squarely on the possibility of apprehension. If there are
no impressions worthy of strong assent, then all assent will be weak and rash on the
Stoic view, and we will be back once again at the conclusion that we must withhold
all assent. To avoid this conclusion, Carneades offers a very different distinction:
The wise person is said to suspend assent in two senses: in one sense, when this
means he won’t assent to anything entirely; in another, when it means he will restrain
himself even from giving responses showing that he approves or disapproves of
something, so that he won’t say “yes” or “no” to anything. Given this distinction,
the wise person accepts the suspension of assent in the first sense, with the result
that by following what is persuasive wherever that is present or deficient, he is
able to reply “yes” or “no.” Since the person who keeps himself from assenting
to anything nevertheless wants to move and act, Clitomachus maintained there
are still impressions of the kind that excite us to action; and likewise, there are
still responses we can use when questioned on either side, by just following our
impressions on the matter, provided we do so without assent (Acad. II 104).26
Despite the lack of any explicit reference to the Stoics in this passage, it is reasonable
to expect that, having eliminated kataleptic impressions, Carneades would finish the
job by forbidding the sort of assent that is only properly given to such impressions—
this is nicely captured by the expression “assenting entirely.” If so, we may take
Carneades’s first kind of assent to correspond to the Stoic’s strong assent. Only now
the point is that the sage refrains from such assent since there is nothing worthy of
that absolute degree of confidence.
58 SKEPTICISM
The description of the second type of assent is brief and obscure—replying “yes”
to what is persuasive and “no” to what is not. But if the distinction is offered by
way of contrast, it will be a matter of not assenting entirely, but rather to a lesser
degree, yet still with enough confidence to motivate action and provide guidance.
The notion that such assent comes in degrees of confidence correlates perfectly
with the degrees of persuasiveness among impressions. So we may say that, since
no impression is apprehensible, the sage does not assent entirely to anything, but
that he does assent to the extent that the persuasive impression has been thoroughly
examined and is not contradicted by associated impressions.
This amounts to a consciously fallible judgment of truth.27 Unlike the Pyrrhonists,
Carneades holds that some impressions are more convincing than others (PH I 227).
Since what we are convinced of, typically, is that something is the case, this amounts
to saying that some impressions appear true to a greater degree than others. So the
reason Carneades prefers persuasive impressions that have been subjected to the
highest degree of scrutiny is that they are more likely to get things right and provide
good guidance (PH I 229). It is precisely on this point that Sextus differentiates
Carneades’s New Academy from Pyrrhonism.28 Academics assent in the way a
dissolute man agrees with someone who urges extravagant living. What Carneades
actually desires is not extravagance, of course, but truth.29 So the charge is that
Carneades knows better than to assent to the apparent truth of merely persuasive
impressions, but cannot resist doing so. The Pyrrhonists, by contrast, yields to some
impressions, not because they are apparently true, but simply because they move
him. What Sextus probably finds objectionable about the former sort of assent is
that, unlike the Pyrrhonist’s yielding, it leaves one vulnerable to rational challenges
and the disturbance that follows when those challenges are not met.
To reply on behalf of Carneades, we may say that it would be truly blameworthy
only if one were to assent entirely. This provides a good way to understand his
almost Herculean labor of driving “assent—i.e., opinion and rashness—from our
minds, as one would drive out a wild and savage monster” (Acad. II 108). Since it
would be flagrantly inconsistent to insist that one should never assent to anything
whatsoever and that one should assent in a more qualified manner to persuasive
impressions, we may conclude that what Carneades drives out is assenting entirely.
And we should provide this qualification as needed in Cicero’s discussion of the
rashness of assent or the view that the sage withholds assent. For example: “Even if
anything is apprehensible, the very habit of assent [i.e., assenting entirely] is slippery
and dangerous. Hence, since it’s agreed that it is extremely vicious to assent [entirely]
to anything false or unknown, it’s better to restrain all assent [i.e., assenting entirely]
so the wise person doesn’t go awry by advancing rashly” (Acad. II 68). Similarly,
what is most shameful is to approve of something false in place of the truth, or as
if it were true [pro veris probare false] (Acad. II 66). By supposing that what we
are assenting to is true, rather than probably true or persuasive, we give our assent
entirely. Likewise, it is rash to assent to the unknown in place of, or as if it were,
known (ne incognita pro cognitis habeamus iisque temere assentiamur) (Off. I 18).30
Once again, by supposing that what we are assenting to is securely known, we give
our assent entirely—and this is rash.
CARNEADES 59
Consequently, the error Carneades seeks to eliminate is not simply assenting

to what is false, but rather being credulous or gullible, especially with respect to
the rational support of our own convictions. He would in that case be seeking,
in good Socratic fashion, to relieve us of the rashness of thinking that we know
what we do not and of obstinately defending the views we have grown attached
to (Acad. II 9).
5 VARIETIES OF OPINION AND

THE FALLIBLE SAGE
There are, however, a number of passages in which both Carneades and Cicero
seem to reject the view that it is ever wise to hold mere opinions. For example,
Cicero openly approves of Zeno’s view that the sage will hold no opinions (Acad.
II 113). On the basis of what I have argued so far, we may take him to mean that
the sage never has the blameworthy mental state that arises from assenting rashly,
that is, entirely, in the absence of apprehension. The fact that he chooses to call this
“opinion” is potentially misleading since it may be confused with the cognitive state
that guides the life of the Academic sage. Both sorts of opinion involve assenting to
what is, or may be, false. The crucial difference is that from the Stoic perspective
any such assent is necessarily a moral and epistemic failing, while from Carneades’s
Academic perspective only assenting entirely is blameworthy.31
This explains why Carneades would not vehemently combat the Peripatetics
when they affirm the sage will hold opinions. For the Peripatetics similarly think
that fulfilling one’s epistemic obligation does not make one infallible, so assenting
to what turns out to be false is not necessarily a transgression (Acad. II 112). This
also illuminates Carneades’s dialectical strategy in combating the Stoics. When
Carneades proposes, for the sake of argument, that the sage will sometimes have
opinions, he is relying on the Stoic view of opinion. This is why the Stoic spokesman
in the Academica describes his strategy this way: “Carneades was occasionally liable
to sink so low as to say that the wise person would have opinions, i.e. that he would
err” (Acad. II 59, see also II 67). In order to drive the Stoics to the unacceptable
conclusion that the (Stoic) sage would transgress, Carneades had to rely on their
own views of assent and opinion.32
Finally, we also find that the Academics identify rashness as either assenting to
what is false or to what has not been sufficiently examined or comprehended.33
The disjunction is puzzling. Is it blameworthy to assent to what is false because it
is shameful to have false opinions? Or is it blameworthy, as I have been arguing, to
assent to what has not been adequately examined because this is contrary to what
wisdom demands? These two options are potentially at odds. Suppose one assents
to an adequately examined persuasive impression that turns out false. It would
then be both rash to assent to it (because false) and not rash (because sufficiently
examined). On the other hand, suppose one assents to a true impression that has
not been sufficiently examined. Now it will be both rash to assent to it (because not
sufficiently examined) and not rash (because true).34
60 SKEPTICISM
We may avoid this problem by taking the second disjunct as an emphatic

alternative. In that case, the claim would be that it is rash to assent to what is false,
or at least (since we are often not in a position to know whether the proposition in
question is false) to what has not been sufficiently examined.35
In any case, we must maintain the fundamental point that, since Carneades’s sage
assents to persuasive impressions and since such impressions may always turn out
to be false, it necessarily follows that the sage will sometimes assent to what is false.
So the Academic sage’s wisdom must be constituted by the distinctive manner in
which he gives his assent. In particular, he will proportion the strength of his assent
to the persuasiveness of the impression, whether through systematic scrutiny of his
perceptual impressions or through argument pro and con in the case of philosophical
propositions, and he will remain perpetually open to being corrected by subsequent
experience and additional argument. On those rare occasions in which a thoroughly
examined persuasive impression turns out to be false, or dialectically defeated, the
sage’s opinion that it was probably true is not blameworthy.36
NOTES
1. See Cicero, De Or. III 67, Cooper (2004: 92–95), Ioppolo (1986), and Thorsrud
(2010: 58–62).
2. See De Or. III 68, ND I 11, Acad. II 16. See also Acad. II 46, and Eusebius, Praep.
Evang. 14.7.15.
3. See DL I 16, IV 65; Plutarch, De Alex. fort. 328A–B; Gellius, NA 17.15.1–3.
Carneades’s student Clitomachus wrote over 400 books documenting the arguments
of his teacher (DL IV 67; see Acad. II 16, Or. 51). None of these survive except in
the form of quotations and paraphrases, which are collected by Mette (1985). Cicero
and Sextus are the most important sources; however, Cicero is more sympathetic and
closer in time, and over half of the lines of text collected by Mette are his or derive
from his writing. Many of the central texts are also collected and translated in Long
and Sedley (1987: 438-460).
4. See De Or. I 45, III 68; Fin. III 41; DL I 16, IV 65; De Alex. fort. 328A–B; Gellius NA
17.15.1–3; Praep. Evang. 14.8. Carneades, along with two other philosophers acting
as Athenian emissaries in 155, convinced the Romans to reduce a fine levied against
Athens for having sacked the city of Oropus. Powell (2013) has convincingly shown
that the famous arguments for and against justice that Carneades is reported to have
delivered on this occasion are likely to be a fabrication of Cicero’s. Had he really
sought to show the Romans that they did not know what justice is, he would likely
have betrayed his diplomatic charge.
5. Since it was common practice among Academic philosophers to locate the origin
of their views in Plato and Socrates, there were a number of competing histories of
the Academy—see Brittain (2001: 169–254). I am not presupposing that this is the
correct account. I mention it here because it neatly contains the major features that
one might find exemplified by Carneades, as indeed Cicero did.
6. See Cicero, Div.; also Hankinson (1988).
7. See Cicero, Fat.; also Bobzien (2002) and O’Keefe (2005: 138–149, 153–162).
CARNEADES 61
8. See Cicero, ND III; Sextus, AM IX. See also Burnyeat (1982a) and Long (1990).
9. On Carneades’s epistemological arguments, see Frede (1987b), Schofield (1999:
338–344), and Perin (2006). On his ethical arguments, see Long (1967), Algra
(1997), and Annas (2007).
10. The current consensus is that he did not—see n. 32 below.
11. The proper translation of the property expressed by these impressions, τὸ πιθανόν,
is also controversial. Cicero renders it probabile and veri simile (Acad. II 7–8, 99,
107; Tusc. I 17, II 5). See Glucker (1995).
12. Vogt (2010) illuminates the variety of apraxia arguments.
13. Obdrzalek (2006: 257) convincingly argues that in order for Carneades’s response
to the Stoic apraxia charge to succeed, it must show how one can lead a coherent
and happy life while refraining from dogmatic assent.
14. It is clear that there was some dispute among Carneades’s students regarding
his position (Acad. II 78). I maintain that whatever the nature of that dispute,
Cicero’s testimony does not support a radically skeptical Carneadean alternative.
Görler (1994: 816–819) also rejects the view that Carneades was a radical skeptic
insofar as he shares some basic convictions with the Stoics—for example, that
sense-impressions exist and that some are true. For more on the extraordinarily
complicated history of the post-Carneadean Academy, see Tarrant (1985), Glucker
(1978), Brittain (2001), and Lévy (2010).
15. All translations of AM VII are from Bett (2005).
16. Ioppolo (2009) argues that Sextus relies on inconsistent accounts of Carneades’s
criterion: on the one hand, exclusively practical, having no connection to truth, and
on the other, epistemic.
17. This is hardly a novel proposal. Indeed, Cicero seems to have understood Carneades
this way, as does Reid (1885) in his notes to the text of the Academica. More
recently, see Stough (1969: 50–64) and Grundmann (2003). Jeffrey (1984: 73–75)
and Niiniluoto (1984: 594) find in Carneades the historical precedent for their
favored views of probability. In opposition, Burnyeat (1982b) argues, with reference
to Hacking (1984), that Carneades does not, and in fact could not, share or even
anticipate our contemporary view of probability (that X is good evidence for Y in
proportion to the frequency with which X and Y go together). But even without any
mathematical or statistical interpretation of frequency, one may still think that we
are habituated by the regularity of experience to anticipate a certain effect from a
certain cause. Furthermore, the fact that the Academic sage is guided by probable
impressions indicates that it plays a normative role, showing not merely how people
live, but also how they should live.
18. The one key issue on which Carneades is said to have departed from Arcesilaus
is in claiming that it is impossible to suspend judgment on everything, but that
fortunately while all things are inapprehensible, not all things are uncertain (Praep.
Evang. 14.7.15; cf. Acad. II 59).
19. Görler (1994: 874–879) argues to the contrary that Carneades’s probable
impressions are exclusively sensory and that the approval of philosophical views as
probable is Cicero’s distinctively Roman innovation.
20. All translations of the Academica are from Brittain (2006).
62 SKEPTICISM
21. Compare ND I 12: persuasive impressions direct or regulate the life of the sage
(sapientis vita regeretur).
22. Cicero reports that the Stoics and Peripatetics used to choose Carneades to
adjudicate their disagreement over the proper understanding of the τέλος (Tusc.
V 120), presumably because of his great oratorical skill. If so, they would have
wanted him to judge which side was more persuasively argued, which suggests a
straightforwardly philosophical application of Carneades’s skeptical criterion.
23. Brittain (2001: 95–96, 103–108) accounts for such passages as an intrusion
of mitigated skepticism into Carneades’s real position of universal ἐποχή
and an exclusively dialectical engagement with dogmatists. But see Glucker
(1994: 147–152).
24. If Carneades really did argue pro and con for the sake of drawing out the most
probable philosophical views, we may reasonably ask why there is no record
of what he judged to be the dialectical winners. Cicero provides the answer:
Academics conceal their views so that students may be guided by their own reason
rather than by authority (Acad. II 60, Tusc. V 11, de Or. I 84). This would explain
why Clitomachus used to say that he had never been able to understand what
Carneades did accept (quid probaretur, Acad. I 139); it would make no sense for
him to try to figure out what Carneades accepted if he had been confident that he
accepted nothing.
25. Waterfield (1989) makes a persuasive case for attributing to the Socrates of Plato’s
Socratic dialogues the view that truth is what survives refutation.
26. I have made a slight but crucially important modification to Brittain’s (2006)
translation. For omnino eum rei nulli adsentiri he provides, “[The sage] won’t
assent to anything at all,” whereas I follow Reid (1885: 300) in taking omnino as an
adverb: “[The sage] won’t assent to anything entirely.”
27. For a very different interpretation see Bett (1990), and similarly: Frede (1987c),
Striker (1996; 2010: 200), Schofield (1999: 334–337), and Ioppolo (2007). Bett
argues that Carneades’s response to the apraxia objection involves distinguishing
between an impression being convincing from our taking it to be true. On this view,
Carneades proposes we may act in accordance with former while withholding assent
to the latter. I believe, however, that finding something convincing is inextricably
connected with taking it to be true, in some sense.
28. It is interesting to note that this is virtually the same argument that Sextus makes
against those who would count Plato as a skeptic (PH I 222–223).
29. Carneades argued against Stoic theology “in such a manner as to arouse in persons
of active mind a keen desire to discover the truth” (ND I 4; see also Acad. II 7,
66, 76).
30. For similar uses of pro, see Acad. II 53, Part. 44, ad. Brut. 1.5.2.
31. Similarly, when Cicero admits that he cannot resist assenting to some persuasive
impressions and thereby holding opinions (Acad. II 66, 113), we should not suppose
that he is sincerely convicting himself of foolishness—see Görler (1997). For if he
sincerely believed that wisdom is incompatible with opinion, he could not endorse his
Academic method, the whole point of which is to draw out the most probable views,
that is, the most rationally defensible opinions. Elsewhere, he is quite happy to insist
on the compatibility of wisdom and mere opinion (e.g., Mur. 63).
CARNEADES 63
32. However, we might also seek to explain Carneades’s rejection of the compatibility
of wisdom with opinion by claiming, as some have, that the mitigated skeptic is
merely a dialectical invention, which was later attributed to Carneades as his own
positive view. With the inspiration of Couissin (1929), this has been developed and
employed in a variety of ways by Burnyeat (1982b), Sedley (1983), Frede (1987c),
Bett (1989; 1990), Allen (1994; 1997), Striker (1996; 2010), Schofield (1999), and
Brittain (2001). Perhaps the most important piece of evidence in support of this
view of Carneades is a reported dispute between his students: “[That the sage] could
fail to apprehend anything and yet still have opinions … is said to have been the
position approved by Carneades—although, since I [sc. Cicero] trust Clitomachus
rather than Philo or Metrodorus, I consider it a position he argued for rather than
approved” (Acad. II 78; see Brittain 2001: 76–82). For an interpretation of this
passage that is consistent with the position I take here see Thorsrud (2012).
33. There are no texts that attribute this sort of disjunction explicitly to Carneades. The
following three, however, could plausibly be linked to him. At Acad. II 68, Cicero,
on behalf of the New Academy is describing our view. And Div. I 7 and ND I 1 are
general descriptions and justifications of Cicero’s Academic methodology, which he
does identify explicitly as Carneades’s elsewhere (Tusc. V 11).
34. Augustine explores some similar possibilities in C. Acad. III xv 34.
35. A prime example of this is Acad. II 7: the sole purpose of the Academics’ arguing on
both sides is to draw out that which is either true or at least the closest approximation
to the truth (aliquid quod aut verum sit aut ad id quam proxime accedat). See also
Off. III 54, Tusc. V 51.
36. The inspiration for this view may have come from the Stoics themselves. In order
to maintain the sage’s infallibility, some Stoics argued that, when the sage assents to
some proposition p that turns out false, we should suppose that what he assents to is
the proposition that it is reasonable that p. Regardless of whether p is true or false, it
will still be the case that it was reasonable that p, and the sage will not have made a
mistake (see DL VII 177, Brennan 1996). Similarly, when Carneades’s sage assents to
the proposition that it is probable that p, he is probably right in doing so, and again
even if p turns out to be false, he will not have made a mistake.
REFERENCES
Algra, Keimpe. 1997. “Chrysippus, Carneades, Cicero: The Ethical Divisions in Cicero’s
Lucullus.” In B. Inwood and J. Mansfeld (eds.), Assent and Argument: Studies in Cicero’s
Academic Books, 107–139. Leiden: Brill.
Allen, James. 1994. “Academic Probabilism and Stoic Epistemology,” Classical Quarterly
44: 85–113.
Allen, James. 1997. “Carneadean Argument in Cicero’s Academic Books.” In B. Inwood
and J. Mansfeld (eds.), Assent and Argument: Studies in Cicero’s Academic Books, 217–
256. Leiden: Brill.
Annas, Julia. 2007. “Carneades’ Classification of Ethical Theories.” In A. M. Ioppolo and
D. Sedley (eds.), Pyrrhonists, Patricians, Platonizers. Hellenistic Philosophy in the Period
155–86 BC, 187–223. Naples: Bibliopolis.
Babbit, Frank. 1927. Plutarch: Moralia, vol. 1. Cambridge, MA: Harvard University Press.
64 SKEPTICISM
Bett, Richard. 1989. “Carneades’ Pithanon: A Reappraisal of Its Role and Status,” Oxford
Bett, Richard. 1990. “Carneades’ Distinction between Assent and Approval,” Monist 73:
3–20.
Bett, Richard. 2005. Sextus Empiricus: Against the Logicians. Cambridge: Cambridge
University Press.
Bobzien, Susanne. 2002. Determinism and Freedom in Stoic Philosophy. Oxford: Oxford
University Press.
Brennan, Tad. 1996. “Reasonable Impressions in Stoicism,” Phronesis 41: 318–334.
University Press.
Brittain, Charles. 2006. Cicero: On Academic Scepticism. Indianapolis: Hackett.
Burnyeat, Myles. 1982a. “Gods and Heaps.” In M. Schofield and M. Nussbaum (eds.),
Language and Logos: Studies in Ancient Greek Philosophy, 315–338. Cambridge:
Burnyeat, Myles. 1982b. “Carneades Was No Probabilist,” unpublished manuscript.
Cooper, John. 2004. “Arcesilaus: Socratic and Sceptic.” In his Knowledge, Nature, and the
Good, Essays on Ancient Philosophy, 81–103. Princeton: Princeton University Press.
Cooper, John. 2013. Pursuits of Wisdom: Six Ways of Life in Ancient Philosophy from
Socrates to Plotinus. Princeton: Princeton University Press.
Couissin, Pierre. 1929. “Le stoicisme de la nouvelle Academie,” Revue d’historie de
la philosophie 3: 241–276. Translated and reprinted as “The Stoicism of the New
Academy.” In M. Burnyeat (ed.), The Skeptical Tradition, 31–63. Berkeley: University
of California Press, 1983.
Frede, Michael. 1987a. Essays in Ancient Philosophy. Minneapolis: University of Minnesota
Press.
Frede, Michael. 1987b. “Stoics and Sceptics on Clear and Distinct Impressions.” In his
Essays in Ancient Philosophy, 151–178, Minneapolis: University of Minnesota Press.
Frede, Michael. 1987c. “The Sceptic’s Two Kinds of Assent and the Question of the
Possibility of Knowledge.” In his Essays in Ancient Philosophy, 201–224, Minneapolis:
University of Minnesota Press.
Glucker, John. 1978. Antiochus and the Late Academy. Göttingen: Vandenhoeck und
Ruprecht.
Glucker, John. 1995. “Probabile, Verisimile, and Related Terms.” In J. Powell (ed.), Cicero
the Philosopher: Twelve Papers, 115–143. Oxford: Clarendon Press.
Görler, Woldemar. 1994. “Ältere Pyrrhonismus, Jüngere Akademie, Antiochos aus
Askalon.” In F. Ueberweg (ed.), Grundriss der Geschichte der Philosophie, Die Philosophie
der Antike, Bd. 4. Die hellenistische Philosophie, 732–989. Basel: Schwabe.
Görler, Woldemar. 1997. “Cicero’s Philosophical Stance in the Lucullus,” B. Inwood and
J. Mansfeld (eds.), Assent and Argument: Studies in Cicero’s Academic Books, 36–57.
Leiden: Brill.
CARNEADES 65
Grundmann, Thomas. 2003. Der Wahrheit auf der Spur. Eine Verteidigung des
erkenntistheoretischen Externalismus. Paderborn: Mentis.
Hacking, Ian. 1984. The Emergence of Probability. Cambridge: Cambridge University
Press.
Hadot, Pierre. 1995. Philosophy as a Way of Life, translated by M. Chase. Oxford:
Blackwell.
Hankinson, Robert James. 1988. “Stoicism, Science and Divination,” Apeiron 21: 123–160.
Inwood, Brad and Jaap Mansfeld (eds), 1997. Assent and Argument: Studies in Cicero’s
Academic Books. Leiden: Brill.
Ioppolo, Anna Maria. 1986. Opinione e Scienza. Naples: Bibliopolis.
Ioppolo, Anna Maria. 2007. “L’assenso nella filosfia di Clitomaco; un problema di linguaggio?”
In A. M. Ioppolo and D. Sedley (eds.), Pyrrhonists, Patricians, Platonizers. Hellenistic
Philosophy in the period 155–86 BC, 225–267. Naples: Bibliopolis.
Ioppolo, Anna Maria. 2009. La testimonianza di Sesto Empirico sull’accademia scettica.
Naples: Bibliopolis.
Ioppolo, Anna Maria and David Sedley (eds.). 2007. Pyrrhonists, Patricians, Platonizers.
Hellenistic Philosophy in the period 155–86 BC. Naples: Bibliopolis.
Jeffrey, Richard. 1984. “De Finetti’s Probabilism,” Synthese 60: 73–90.
Lévy, Carlos. 2010. “The Sceptical Academy: Decline and Afterlife.” In R. Bett
The Cambridge Companion to Ancient Scepticism, 81–104. Cambridge: Cambridge
University Press.
Long, Arthur Anthony. 1967. “Carneades and the Stoic telos,” Phronesis 12: 59–90.
Long, Arthur Anthony. 1990. “Scepticism about the Gods in Hellenistic Philosophy.” In
M. Griffith and D. J. Mastronarde (eds.), Cabinet of the Muses, 279–91. Atlanta, GA:
Scholars Press.
Long, Arthur Anthony and David Sedley. 1987. The Hellenistic Philosophers, vol. 1.
Mette, Hans Joachim. 1985. “Weitere Akademiker heute Von Lakydes bis zu Kleitomachos,”
Lustrum 27: 39–148.
Niiniluoto, Ilkka. 1984. “The Significance of Verisimilitude,” PSA: Proceedings of the
Biennial Meeting of the Philosophy of Science Association 2: 591–613.
Obdrzalek, Suzanne. 2006. “Living in Doubt: Carneades’ Pithanon Reconsidered,” Oxford
O’Keefe, Tim. 2005. Epicurus on Freedom. Cambridge: Cambridge University Press.
Perin, Casey. 2006. “Academic Arguments for the Indiscernibility Thesis,” Pacific
Philosophical Quarterly 86: 493–517.
Powell, Jonathan 2013. “The Embassy of the Three Philosophers to Rome in 155BC.” In
C. Kremmydas and K. Tempest (eds.), Hellenistic Oratory: Continuity and Change,
219–248. Oxford: Oxford University Press.
Rackham, Harris. 1914. Cicero. De finibus bonorum et malorum. Cambridge, MA: Harvard
University Press.
Reid, James. 1885. M. Tulli Ciceronis Academica. London: Macmillan.
Schofield, Malcolm. 1999. “Academic Epistemology.” In K. Algra, J. Barnes, J. Mansfeld,
and M. Schofield (eds.), The Cambridge History of Hellenistic Philosophy, 323–354.
66 SKEPTICISM
Sedley, David. 1983 “The Motivation of Greek Skepticism.” In M. Burnyeat (ed.), The
Skeptical Tradition, 9–29. Berkeley: University of California Press.
Stough, Charlotte. 1969. Greek Skepticism. Berkeley: University of California Press.
Striker, Gisela. 1996. “Sceptical Strategies.” In her Essays in Hellenistic Epistemology and
Ethics, 92–115. Cambridge: Cambridge University Press.
Striker, Gisela. 2010. “Academics versus Pyrrhonists, Reconsidered.” In R. Bett (ed.),
University Press.
Thorsrud, Harald. 2002. “Cicero on His Academic Predecessors: The Fallibilism of
Arcesilaus and Carneades,” Journal of the History of Philosophy 40: 1–18.
Thorsrud, Harald. 2010. “Arcesilaus and Carneades.” In R. Bett (ed.), The Cambridge
Companion to Ancient Scepticism, 58–80. Cambridge: Cambridge University Press.
Thorsrud, Harald. 2012. “Radical and Mitigated Skepticism in Cicero’s Academica.” In
W. Nicgorski (ed.), Cicero’s Practical Philosophy, 168–192. Notre Dame: Notre Dame
University Press.
Vogt, Katja. 2010. “Scepticism and Action.” In R. Bett (ed.), The Cambridge Companion to
Ancient Scepticism, 165–180. Cambridge: Cambridge University Press.
Waterfield, Robin. 1989. “Truth and the Elenchus in Plato.” In P. Huby and G. Neal (eds.),
The Criterion of Truth, 39–56. Liverpool: Liverpool University Press.
CHAPTER FIVE
Aenesidemus
LUCA CASTAGNOLI
1 INTRODUCTION
Unlike many other philosophers who flourished at the dawn of the Hellenistic age,
Pyrrho did not found a school. Probably as a result of this and of his choice not
to write, his philosophy, and the fascination that his personality exercised during
his life, failed to outlive his pupil Timon and to produce any lasting legacy beyond
the third century BCE.1 In line with their standard practice, some post-Hellenistic
doxographers reconstructed an uninterrupted “succession” of philosophers
seamlessly linking Pyrrho and Timon with the later Pyrrhonian tradition (see, e.g.,
DL IX 116). But on the basis of other sources (e.g., Aristocles apud Eusebius Praep.
Evang. 14.18.29) scholars tend to agree that this is a largely fictional construction,
and that the history of Pyrrhonism actually lapsed after Timon and resumed only in
the first century BCE. That is the time when Aenesidemus appropriated, revived, and
(in some measure to be determined) reinterpreted and developed Pyrrho’s ideas and
attitudes, thus initiating a new brand of skepticism that he chose to baptize after him.
The terms σκέψις (“skepticism”) and σκεπτικοί (“skeptics”) were not yet in
currency at Aenesidemus’s time, nor do they seem to have been introduced in his
works either. Later on, however, others, both within and outside the Pyrrhonian
school, adopted these terms to refer to the whole tradition from Pyrrho onwards and
(but only outside the Pyrrhonian school) to the skeptical phase of Plato’s Academy
between the third and first centuries BCE, that is, from Arcesilaus to Philo of Larissa.2
As we will see, Aenesidemus’s Pyrrhonian revival set itself in overt opposition to
what he took to be contemporary forms of positive (e.g., Stoic) and negative (e.g.,
Academic) “dogmatism.” According to the scholarly vulgata, Aenesidemus was
himself a member of the Academy before defecting (but see Decleva Caizzi 1992
and Polito 2004), and perhaps founded his own school precisely in reaction to
the unacceptable slackening of skepticism under the two leading first-century BCE
Academics, Philo and Antiochus.
Although we can establish with a certain degree of confidence that he lived in
the first century BCE, virtually nothing else is certain about Aenesidemus’s dates
and life, except for the fact that at some point he worked in Alexandria. According
68 SKEPTICISM
to Diogenes Laertius (DL IX 115), he was originally from Cnossos in Crete, but,
according to Photius (Bibl. 212, 170a41), he was from Aigai in Asia Minor.
2 PYRRHO’S DISPUTED LEGACY

Because of Aenesidemus’s appropriation of Pyrrho, who was by his time a rather
obscure figure, as the “noble father” of his newly founded school, any reconstruction
of his brand of skepticism should not dispense with the question of how exactly it
was related to Pyrrho’s original outlook. But since antiquity Pyrrho has been an
exegetical puzzle, and especially within the last three decades, since fully informed
research on Pyrrho has become possible thanks to the publication of a complete
collection of testimonies (Decleva Caizzi 1981), the number of rival reconstructions
has proliferated. Of these, the two most plausible and influential, and also most
directly relevant to our study of Aenesidemus, are (1) the “epistemological” and (2)
the “metaphysical” interpretations:
(1) According to the former interpretation, which used to be the standard
one, Pyrrho was essentially a skeptic in the same mold as Sextus Empiricus
(second century CE),3 our key source for later Pyrrhonism. Having witnessed
the weakness and unreliability of our senses and reason, which had led
others, before him, to despair of the possibility of human knowledge, Pyrrho
concluded that the correct attitude toward reality was withholding assent
from any belief and claim, whether positive or negative. While registering
how things appear to us, he suspended his judgment on how things really
are (for defenses of this traditional interpretation, see, e.g., Stopper 1983;
Thorsrud 2009; and Castagnoli 2013).
(2) According to the latter interpretation, Pyrrho was impressed by the
metaphysical insight of some earlier thinkers, like the Eleatics and Plato, who
had made stability and absoluteness the mark of reality and truth, relegating
change and relativity to the realm of appearance and mere opinion. He
thus concluded that things cannot really be in any of the ever-changing
ways in which they manifest themselves to our senses and minds. Because
of the pervasiveness of the phenomenon of variability, that conclusion
amounted to a nihilist metaphysics according to which reality itself, beneath
the appearances, is completely indeterminate, and we should thus deny
that things really are how they appear to be (see, e.g., Long-Sedley 1987;
Hankinson 1995; and Bett 2000).4
The deep sensitivity to the pervasiveness of conflict and disagreement was Pyrrho’s
starting point on both interpretations, but his diagnoses of the phenomenon would
have been different, and indeed incompatible, on either interpretation. Finally, the
two interpretations agree in attributing to Pyrrho the view that a correct understanding
of reality, or of our prospects of knowing it, will transform our lives for the better:
both (1) suspension of judgment and (2) recognition of the intrinsic indeterminacy of
reality were supposed to induce in us a desirable state of psychological “tranquility”
AENESIDEMUS 69
or “peace of mind” (ἀταραξία).5 This tranquility was presumably the result of

relinquishing any positive commitment to the good and the truth, thus avoiding the
anxiety, frustration, and fear related to the pursuit and preservation of them.
3 AENESIDEMUS: APPROACHES AND SOURCES

While the number and divergence of the interpretations of Aenesidemus is not as
wide as for Pyrrho, attempts to reconstruct his philosophy face analogous and related
methodological issues. Unlike Pyrrho, Aenesidemus wrote several works, but none
of them is preserved, and our later sources on Pyrrhonism, whether Pyrrhonian
or hostile, tend either to be silent on Aenesidemus’s own distinctive position or
to conflate it with Pyrrhonian skepticism as represented by later exponents like
Sextus. The identification of original Aenesidemean fragments within our second-
hand testimonia on his philosophy, or on Pyrrhonian philosophy as a whole, is often
impossible or deeply conjectural. In light of the nature of the scant information
available, it is difficult to establish the exact nature of the connection with Pyrrho’s
earlier thought and to disentangle Aenesidemus’s position from the later Pyrrhonian
(or “Neo-Pyrrhonian”) tradition which he initiated.
Most interpreters have focused on the task of identifying and explaining those
contributions that appear to be more distinctively Aenesidemean, especially on the
basis of important evidence provided by Photius’s late (ninth century CE) summary of
Aenesidemus’s Arguments of the Pyrrhonists (in eight books), but also by examining
several passages scattered in book 9 of Diogenes Laertius’s Lives of Philosophers
(see Barnes 1992); in excerpts of Aristocles of Messene’s On Philosophy, preserved
by the father of the church Eusebius of Caesarea (see Chiesara 2001); in Sextus
Empiricus’s extant corpus; and in Tertullian’s On the Soul.6
4 AENESIDEMUS AGAINST THE ACADEMICS

From our sources a decidedly polemical focus against the Academics emerges clearly
as an Aenesidemean trait: their position is singled out from the beginning of the
first book of the Arguments of the Pyrrhonists as the polemical foil against which
it is possible to define the correct Pyrrhonian attitude. The question of how the
Academics and the Pyrrhonists differ would become a topos in the first and second
centuries CE (see, e.g., Gellius, NA 9.5.6), but it was clearly put on the map already
by Aenesidemus. At the end of the first book of his Outlines of Pyrrhonism, Sextus
will also include various Academics in his long list of “neighboring” philosophies
not to be conflated with Pyrrhonian skepticism (PH I 220–235). While Sextus’s
discussion is extensive, its position does not however show the same level of urgency
originally displayed by Aenesidemus. Even the choice of expounding his philosophy
in writing poses Aenesidemus in contrast with some of the leading Hellenistic
Academics, such as Arcesilaus and Carneades, who had adopted the Socratic stance
of not writing (Polito 2014). The point should not be overstated, however, since
that choice also marks an interesting discontinuity with Pyrrho himself.
70 SKEPTICISM
According to Aenesidemus, the Academics are barely disguised dogmatists:

far from being aporetic and suspending judgment about everything, they have all
kinds of opinions. They confidently affirm certain things—for example, when they
admit that some views or arguments are more persuasive than others (clearly a
reference to the Carneadean doctrine of the πιθανόν)7—and unambiguously deny
others—for example, when they claim, against the Stoics, that “nothing can be
apprehended,” that is, grasped with infallible certainty. Contemporary Academics,
like Philo and Antiochus, sometimes are even in agreement with Stoic doctrines,
and are essentially “Stoics fighting against Stoics.” Moreover, they are guilty
of contradicting themselves: they give their assent to a number of things, while
denying the possibility of knowledge as firm apprehension (κατάληψις) of reality
(Bibl. 212.169b36–170a17). Aenesidemus’s polemics against the Academy might
have had not only a clear philosophical rationale, but also “political” roots: as we
have seen, it is possible that Aenesidemus was himself an Academic renegade, and
certainly he placed his school as a form of genuine anti-dogmatism in competition
with it (see Decleva Caizzi 1992; Mansfeld 1995). Aenesidemus’s anti-Academic
spin might also be reflected in his predilection for the Socratic/Platonic language of
aporia to describe the Pyrrhonian stance, and in the quasi-Socratic flavor (cf. Plato,
Apol. 21b4–5) of his claim that “he who philosophizes after Pyrrho is happy not
only in general but also, and especially, in the wisdom of knowing that he has no
firm cognition of anything,” as reported by Photius (Bibl. 212.169b28–29).8 We are
also informed that Aenesidemus took a position on whether Plato was a skeptic,
something which had become a focus of controversy within the late Hellenistic
Academy. Unfortunately, textual corruption in our source (PH I 222) makes that
position impossible to determine with full certainty, but it seems more likely that
Aenesidemus, like Sextus after him, rejected the skeptical credentials that some
Academics had attributed to the founder of their school (Spinelli 2000).
5 AENESIDEMUS’S “HERACLITEANISM”
Aenesidemus’s so-called Heracliteanism is another distinctive, and more puzzling,
aspect of his thought, and as such has attracted ample scholarly attention (in
recent years, Polito 2004; Pérez-Jean 2005; and Schofield 2007). Sextus reports,
reproachfully, that “Aenesidemus and his followers said that the skeptical way is a
path which leads to Heraclitean philosophy, since the idea that contraries appear
to hold of the same thing leads to the idea that they actually hold of it” (PH I 210).
In other words, the Pyrrhonian epistemological awareness of the pervasiveness of
conflicting appearances would lead, somehow, to the Heraclitean metaphysical
tenet of the unity of opposites. On a few occasions Sextus and other sources also
mysteriously attribute dogmatic views—for example, about the soul or time—to
“Aenesidemus κατά Heraclitus” (Aenesidemus “citing Heraclitus”? Aenesidemus “in
accordance with Heraclitus,” that is, following him? Aenesidemus “in his Heraclitean
mode”?). Although these testimonies have been taken by some scholars to attest to
the existence of an early or late Heraclitean phase in Aenesidemus’s thought, this
reading has been convincingly shown by others to be implausible. Aenesidemus’s
AENESIDEMUS 71
interest in Heraclitus is more likely to have been instrumental in his dialectic with
contemporary philosophical rivals: in particular, we know that the Stoics, who
are with the Academics one of Aenesidemus’s primary targets, had appropriated
Heraclitus as a precursor, and Aenesidemus could have engaged in Heraclitean
exegesis to show that the Stoic interpretation of Heraclitus was tendentious, and
that un-Stoic consequences actually followed from his doctrines. The exegesis might
have gone as far as to claim that, if anything, Heracliteanism had more affinities with
Pyrrhonism than with Stoicism—for example, in the measure in which the Pyrrhonian
emphasis on the all-pervasive existence of conflicting appearances (and possibly on
the undecidability of conflict: see below) is logically prior to, and explanatory of, the
Heraclitean postulation of the unity of opposites (see Polito 2004).
6 AENESIDEMUS’S “MODES” OF
SUSPENSION OF JUDGMENT
The “invention” or compilation of the ten “modes” (τρόποι) of suspension of
judgment (ἐποχή),9 which form the backbone of the Pyrrhonian argumentative
strategies along with the “five modes” later introduced by Agrippa, is Aenesidemus’s
most celebrated contribution to the history of Western skepticism (see, e.g.,
Striker 1983; Annas and Barnes 1985). A large body of conflicting appearances,
observations, beliefs, and arguments was collected by Aenesidemus and arranged
into ten broad groups. The material often derived from the previous philosophical
tradition, but Aenesidemus systematized its collection and put it to a specific use
within his skeptical project (for the scholarly controversy surrounding this use, see
below). Following the order in which the modes appear in Sextus’s report (PH I
35–163), things appear in different ways relative to
(1) different animals: for example, sea water is unpleasant and poisonous for
human beings, but not for sea fish;
(2) different human beings: for example, most human beings die if they drink
hemlock, but some do not;
(3) different sense organs: for example, perfume is pleasant to smell, unpleasant
to taste;
(4) different circumstances: for example, honey tastes sweet to the healthy,
bitter to the sick;
(5) different positions, distances, and places: for example, the same tower will
appear round from far off, square from nearby;
(6) different mixtures: for example, smells are pungent in the warm and humid
air of a bath-house, faint in the cold air outside;
(7) different quantities and constitutions: for example, a single grain of sand
feels rough to the touch, but as part of a heap grains of sand feel smooth;
(8) relativity: the mode, not exemplified, can be seen as a “super-mode”
encompassing all the others;
72 SKEPTICISM
(9) different frequencies of experience: for example, the sun is not striking
because we see it daily, while a comet appears impressive because it is
observed only very rarely;
(10) different ways of life, customs, laws, mythical beliefs, and philosophical
views: for example, some people believe that the soul is immortal, some
believe that it is mortal.
The modes are presented in a slightly different way, number, and order in other
reports by Philo of Alexandria (On Drunkenness 169–202), Aristocles of Messene
(Praep. Evang. 14.18.11), and Diogenes Laertius (DL IX 78–88).
7 AENESIDEMUS’S MODES AGAINST

CAUSAL EXPLANATIONS
Aenesidemus argued extensively against dogmatic theories of causation and
introduced eight “causal modes” specifically tailored to expose the weakness and
inconclusiveness of any possible causal explanation (αἰτιολογία) devised by the
dogmatists (Barnes 1983; Hankinson 1999). According to Sextus’s presentation of
these modes (PH I 180–185), a causal explanation can be rejected on the grounds
that
(1) it is not itself evident and is not corroborated by agreed-upon evidence;
(2) it is arbitrarily selected from a number of possible explanations available;
(3) it is too simple to account for the order and complexity of the explanandum;
(4) it extends without warranty to the non-evident what has been observed as
evident;
(5) it is based on “hypotheses,” that is principles or assumptions, which are not
universally agreed upon;
(6) it fails to take into due account opposite hypotheses;
(7) it is incompatible with the hypotheses of its own proponents;
(8) it is no less puzzling than the explanandum.
8 AENESIDEMUS AS A SEXTAN-STYLE SKEPTIC

Apart from the distinctive views and contributions which can be attributed directly
to Aenesidemus in the four areas mentioned above, traditionally interpreters have
tended to presuppose that we should credit Aenesidemus with essentially the same
form of skepticism as that presented in detail in Sextus Empiricus’s extant corpus.
From the limited and second-hand evidence available in our sources, we know that
Aenesidemus put forward a battery of arguments against all the main theories of
dogmatic schools, systematically, along the lines of the standard Hellenistic division
of philosophical discourse into physics, logic, and ethics, just as Sextus would later do
(although in a different order). According to Photius (212.170b3–36), for example,
in books 2 to 8 of his Arguments of the Pyrrhonists, Aenesidemus attacked dogmatic
AENESIDEMUS 73
theories concerning first principles, passions, change, generation and corruption,

thought, sensation and their properties (i.e., presumably, truth and falsehood),
signs, nature, the world, gods, causal explanations (see above), what is good and
bad, what is choiceworthy and to be avoided, virtue, happiness, and the goal of
life. The assimilation of Aenesidemus’s skepticism with Sextus’s would also seem
to be justified by those testimonies according to which Aenesidemus “determined
absolutely nothing, not even this very claim, that nothing is determined” (Bibl.
212.170a11–12). Moreover, it seems that Aenesidemus was aware of the problem of
self-refutation which the Pyrrhonian pronouncements risk incurring, a problem later
addressed extensively by Sextus, when he explained that the Pyrrhonists utter the
apparently assertoric formula “Nothing is determined” simply “because [they] have
no other means of blurting out what [they] have in mind” (Bibl. 212.170a12–14).
On this standard interpretation, Aenesidemus suspended judgment about
everything. Suspension of judgment resulted from the realization that “there is no
firm basis of cognition,” presumably as the result of Aenesidemus’s “collection of
appearances or thoughts, of any kind,” which are “brought into confrontation with
each other and, when compared, are found to present much disparity (ἀνωμαλίαν)
and confusion (ταραχήν)” (DL IX 78), clearly a reference to the ten modes. From
ἐποχή a happy state of tranquility will follow “like a shadow.” According to
Aristocles’s report at Praep. Evang. 14.18.4, Aenesidemus somehow identified this
state as “pleasure” (ἡδονή), but this need not be taken as a sign that he espoused a
form of philosophical hedonism over and above the pursuit of ἀταραξία itself.
That various forms of skepticism and other similarly unorthodox philosophical
stances should lead to ἀπραξία (“inaction”), in the form of either vegetable-style
paralysis or random, crazy, or self-destructive behavior, was a standard criticism in
antiquity (see Vogt 2010). The targets of such criticism devised a variety of strategies to
defend the “livability” of their position. Aenesidemus made “appearances” (φαινόμενα)
the Pyrrhonists’ “criterion of action”: while suspending judgment on whether things are
as they appear to be, they noncommittally “went along with” appearances in their day-
by-day life, thus living an outwardly ordinary existence despite their alternative frame
of mind (PH I 21–24). It is probable that for Aenesidemus the scope of appearances
was broad and included cultural conventions and social norms (DL IX 107–108),
just as for Sextus. Addressing criticism leveled at his teacher, Timon had claimed that
Pyrrho “did not depart from normal practice” (DL IX 105), in contrast with several
ancient anecdotes which depict Pyrrho’s behavior as wildly eccentric, unattractively
non-humane, and dangerously erratic. Aenesidemus too offered a “normalizing”
picture of Pyrrho, and maintained that “although he practiced philosophy on the
principles of suspension of judgment, [Pyrrho] did not act carelessly in the details of
daily life” (DL IX 62), because he “determined nothing dogmatically on account of
opposite arguments, but followed appearances instead” (DL IX 106). A typical trait
of late Hellenistic and late-antique philosophy is that formulating and clarifying one’s
philosophical position was part and parcel of offering the correct interpretation and
explication of the thought of the founders of the tradition or school to which one
declared one’s allegiance. Aenesidemus was no exception, although in his case the
interpretation could not be, of course, a matter of textual exegesis.
74 SKEPTICISM
No source reports Aenesidemus’s reasons for maintaining that the conflict of

appearances, views, and reasons should induce us to apply the skeptical formula “no
more” (οὐ μᾶλλον) to everything, and produce a mental state of ἐποχή.10 The idea
that the conflict is not decidable (or, at least, has not been decided) because of the
“equipollence” (ἰσοσθένεια), that is, the equal persuasive force, of the opposites is
pivotal in Sextus’s account (PH I 8–10) and has traditionally been imported into the
equation for Aenesidemus, too:
(1) x appears F relative to S. [Material provided by the ten modes.]
(2) x appears F* relative to R. [Material provided by the ten modes.]
(3) We cannot prefer S and F to R and F*, or vice versa, because of equipollence.
(4) We can neither affirm nor deny, neither believe nor disbelieve, that x is F, or
that x is F* (i.e., we suspend judgment): x is no more F/F* than it is not.11
9 AENESIDEMUS AS A NEGATIVE
OR “APORETIC” SKEPTIC
The Sextan-style interpretation of Aenesidemus sketched above is by no means the
only one on the market. It would be impossible to present here all of the available
interpretations. I will focus on an alternative reconstruction of Aenesidemus’s outlook
that has recently gained currency. It rejects the assumption that what we know from
Sextus can be safely used to fill the gaps in our picture of Aenesidemus’s original
Pyrrhonism. Woodruff (1988) and Bett (2000: 189–222), the two most notable
supporters of this approach, have maintained that unprejudiced reassessment of our
evidence will reveal that Aenesidemus’s brand of Pyrrhonism differed substantially
from Sextus’s. Aenesidemus’s arguments did not aim to induce ἐποχή on the basis of
the undecidable equipollence of conflicting appearances and arguments; they were
meant to establish, for any object x and predicate F, the negative conclusion that x
is not really or “by nature” F, on the basis of an “invariability condition” ultimately
equivalent to the one endorsed by Pyrrho, according to some supporters of the
metaphysical interpretation (see Section 2 above), and having the same Platonic
pedigree:
For an object to be a certain definite way “by nature,” it should be and appear
that way invariably or without qualifications.12
Hence, something that manifests itself as F only to some people (but not to others),
or only in some circumstances (but not in others), or only in certain respects (but not
in others), cannot be F “by nature.” In virtue of this principle, and of the widespread
conflict and variability catalogued in the ten modes, Aenesidemus concluded that
not only should we refrain from making positive dogmatic claims determining how
things are, but we should deny that anything really is any particular way:
(1) x appears F relative to S. [Material provided by the ten modes]
(2) x appears F* relative to R. [Material provided by the ten modes]
AENESIDEMUS 75
(3) Unless x appears invariably F (F*), x cannot be F (F*) “by nature.”

[Invariability condition]
(4) x is “by nature” neither F nor F*.
On this reading, Aenesidemus’s negative or “aporetic” conclusions about the nature
of things had as their counterpart a form of phenomenal relativism: he allowed
relativized claims such as “x is F at time t,” “x is F for y,” or “x is F in circumstances
C,” which, exactly because relative (or “ambiguous”), do not purport to determine
the intrinsic, essential nature of things (Bibl. 212.169b38–170a11). Such claims
would have qualified in Aenesidemus’s own terms as assertions about appearances
and would have been allowed as objects of non-dogmatic belief and, perhaps, even
(a mundane form) of knowledge, and not only as a practical criterion.13
The revisionary interpretation promoted by Woodruff and Bett was prompted by
the observation that, in several reports of Pyrrhonian views and arguments which
either are directly attributed to Aenesidemus or could have an Aenesideman pedigree
(e.g., Sextus’s Against the Ethicists: see Bett 1997), we often find unqualified negative
conclusions (e.g., signs do not exist at all, the various goals of life celebrated by
dogmatic philosophers do not exist [Bibl. 212.170b12–14], nothing is good or
bad by nature [AM XI 71]) and apparent formulations and uses of the invariability
condition.14 These occurrences are insufficient, however, to establish conclusively
that there must have been a form of Aenesidemean “aporetic” Pyrrhonism
essentially distinct from Sextan “ephectic” (i.e., suspensive) Pyrrhonism. In Sextus’s
works there are plenty of cases in which, although the context unequivocally shows
that inducing ἐποχή on the basis of the equipollence is the goal, Sextus limits
himself to providing only evidence and arguments for the negative conclusion that
something is not the case or does not exist. This is typically explained by the fact
that the equipollent positive evidence and arguments require no rehearsing, since
they represent the default commonsense or philosophical position. That sources
like Photius could have misinterpreted and misrepresented these typical Pyrrhonian
strategies is not implausible; after all, Photius himself also attributes to Aenesidemus
views to the effect that truths, causes, affections, motion, generation, destruction,
nature, the world, gods, good, and bad are beyond our grasp or impossible to know.
If taken as the ultimate conclusions of Aenesidemus’s arguments, these views would
clearly be different from, and possibly inconsistent with, the negative conclusions of
“aporetic” Pyrrhonism (although still unacceptable from the point of view of Sextan
Pyrrhonism) and would be surprisingly at odds with Aenesidemus’s criticism of the
Academics’ inapprehensibility conclusion.15 As for the invariability condition, we can
conjecture that Aenesidemus used it only dialectically, to establish a negative anti-
dogmatic conclusion starting from a dogmatic principle to which the Pyrrhonists
themselves had no commitment, following, again, an argumentative trope pervasive
in later Pyrrhonism (Thorsrud 2009: 102–122; Hankinson 2010).16
If not imposed by our evidence, is the revisionary reading of Aenesidemus at
least consistent with it? Something especially problematic to account for on this
reading is Aenesidemus’s vocal attitude toward contemporary Academics: how
could he criticize Plato’s heirs for making negative and positive claims, and then
76 SKEPTICISM
adopt a position which, on the basis of positive adherence to a dogmatic principle

with a clear Platonic pedigree, committed him to countless negative conclusions?
One could also object that stressing Aenesidemus’s relativistic tendencies is difficult
to reconcile with the therapeutic vocation of Pyrrhonism as a route to ἀταραξία,
explicitly attested for Pyrrho and Sextus. If Pyrrhonian arguments conclude that
nothing is good or bad in its own nature, while admitting that things can be correctly
judged to be good or bad in relation to certain individuals, at certain times, and
in certain circumstances, then the disturbing psychological effects of the dogmatic
belief that certain things are good or bad “by nature” will be replaced by the no less
disturbing effects of the Pyrrhonian belief that certain things are good or bad for me,
at this time, and in my present circumstances. At the same time, however, a different
objection can be leveled against Sextus’s own adoption of suspension of judgment
as a route to ἀταραξία: the awareness of the possibility that there are truths about
what is good and bad, which can in principle be known but which we have failed to
grasp so far, could be thought to engender a psychological state of constant anxiety
and fear rather than ἀταραξία (see Striker 1990).
10 CONCLUSION
Any satisfactory interpretation of Aenesidemus must take into account the individual
and cumulative force of arguments focusing on the terminology and logic of specific
passages, on the nature and reliability of the relevant sources, on issues of intrinsic
philosophical consistency, and on how to account for the dialectical relationship
between Aenesidemus and his philosophical sources and targets. Hopefully, the
recent publication of the first commented edition of Aenesidemus’s testimonia (Polito
2014) will give fresh impetus to the re-examination of such arguments and their
relative weight. However, no satisfactory reconstruction of Aenesidemus’s thought
can sidestep the additional question whether and how it can make sense of its pivotal
position within the Pyrrhonian tradition as a whole. After all, Aenesidemus adopted
Pyrrho as the noble father of his philosophy and was seen by later Pyrrhonists like
Sextus as a bona fide representative of the “Pyrrhonian way.”
As I have explained, the question is complicated, however, by the fact that
Pyrrho’s outlook is itself subject to deep scholarly controversy, and thus our
equation contains two variables. While Pyrrho is never attacked as a skeptic in
pre-Aenesidemean sources, this argument from silence appears less urgent than the
objection that, on the metaphysical interpretation, Pyrrho would seem to have little
in common with what is supposed to be his tradition as represented to us by Sextus.
For those who assume that Aenesidemus was himself a Sextan-style skeptic, the
challenge would then be to explain why he appropriated the obscure metaphysician
Pyrrho in the first place.
Coordinating the metaphysical Pyrrho with the “aporetic” Aenesidemus would
appear to offer an easy way out of the impasse: their philosophical core is essentially
identical, despite the fact that the consequence of the “invariability condition” is
cashed out slightly differently, in the former as a claim of indeterminacy of reality,
in the latter as a series of definite denials concerning the nature of things. But while
AENESIDEMUS 77
this exegetical move would make Aenesidemus’s adoption of Pyrrho as a figurehead

easy to explain, the transition from Aenesidemus to Sextus would turn out to look
problematic. According to Bett (2000), it originated from Sextus’s decision to drop
Aenesidemus’s invariability condition. It is clear why a Sextan-style Pyrrhonist
should have rejected as intolerably dogmatic that condition and its consequences: it
is harder to imagine how he could still have thought to belong to the same school
as someone who had made that condition the foundation of his philosophy. One
would expect that the “aporetic” Aenesidemus should have deserved a place in the
list of philosophers whom Sextus discards as bona fide skeptics (this consideration
applies, of course, to the metaphysical Pyrrho as well). Was Sextus hypocritical,
saving Aenesidemus (and Pyrrho) from well-deserved bashing? The aporetic reading
also requires us to postulate that Aenesidemus understood and used most of the key
Pyrrhonian concepts and terminology differently from Sextus. Έποχή cannot, on
that reading, be a “standstill of the intellect, owing to which we neither deny nor
affirm anything” (PH I 10, emphasis added). Consequently, all the skeptical slogans
must carry different meanings than in Sextus (Bett 2000: 198–199). As we have
seen, the concept of equipollence would have been extraneous to Aenesidemus’s
strategy (just as to Pyrrho’s on the metaphysical reading), but the suggestion that
this key concept might have been such a late addition is perplexing and possibly
contradicted by the Anonymous commentary on Plato’s Theaetetus (61.26–30),
whose dating is uncertain, but perhaps not much later than Aenesidemus. Moreover,
as Polito (2014) suggests, the reference to the “equal persuasiveness” of the opposed
positions at DL IX 79 might be part of the report of Aenesidemus’s account of the
Pyrrhonian logos which begins at DL IX 78.
I have also explained how the ten modes would themselves have had different
functions and conclusions in Aenesidemus and Sextus according to the revisionary
interpretation of the former. It has been suggested that the “aporetic” reading of
Aenesidemus dovetails neatly with the moderate or “urbane” reading (see Burnyeat
and Frede 1997) of the scope of Sextus’s skepticism (Hankinson 2010: 110), but
this too is only partially correct. For the two agree on the target of the skeptical
arguments, that is, dogmatic declarations about the non-evident nature of things,
but not on the attitude adopted toward them: the “aporetic” Aenesidemus resolutely
“refutes” the claim that x is F by nature, while the “urbane” Sextus suspends judgment
on the very same claim. From this point of view, the label σκεπτικός (literally,
“inquirer”), and Sextus’s emphasis on the fact that Pyrrhonism is an open-ended
inquiry after the truth, which has not been shown to be inexistent or ungraspable by
Pyrrhonian argument (see, e.g., PH I 1–4), would have also been difficult to retroject
onto Aenesidemus.
That such radical changes may have occurred at some point after Aenesidemus,
in the later Pyrrhonian tradition preceding Sextus, despite our lack of any ancient
record of this doctrinal earthquake, is not impossible but appears at first sight
implausible. Someone might object that this impression is the result of our weighing
plausibility anachronistically, on the basis of our own standards and expectations.
The ancients could be much more flexible than we tend to be in their way of
appropriating and interpreting philosophical antecedents, emphasizing continuity,
78 SKEPTICISM
and passing over differences within those traditions they constructed, whether for
themselves or others. The shared core insight that happiness, as tranquility, will
come from abandoning any positive belief (and not—as many other philosophers
had suggested—from knowledge of reality) might be just enough to explain why
Aenesidemus looked back at Pyrrho, and Sextus was comfortable in calling himself
a “Pyrrhonist,” even if we decided that Sextus’s route to the abandonment of belief
was very different from that traveled by Pyrrho, Aenesidemus, or both. But before
we decide to stretch ourselves in such a way, and to reach for the very boundaries of
plausible exegesis, we will need extremely compelling reasons to accept that this is
the best way of making sense of the extant evidence on Aenesidemus (and Pyrrho) in
its own right. It is questionable whether revisionary interpretations have succeeded
in providing such reasons.
NOTES
1. On Pyrrho and Timon, see Casey Perin’s chapter in this book.
2. On the skeptical phase of the Academy, see the chapters in this book on Arcesilaus,
Carneades, Philo, and Cicero.
3. On Sextus Empiricus, see chapter 10 of this book.
4. For discussion of these two interpretations, see chapter 2 of this book.
5. For the influence of the Democritean tradition on Pyrrho’s adoption of ἀταραξία
as a desirable goal, see Warren (2002).
6. For very useful discussion of the nature and reliability of all these sources, see
Polito (2014).
7. See Harald Thorsrud’s discussion of this topic in chapter 4.
8. On the question whether this and other pronouncements by Aenesidemus
show awareness of the dangers of self-refutation, see below and Castagnoli
(2010: 329–352).
9. The term ἐποχή, which probably had never been used by Pyrrho, was central in
Academic skepticism.
10. For the ancient philosophical history of the formula οὐ μᾶλλον, see De
Lacy (1958).
11. For a different, unorthodox interpretation of the logic of Sextus’s use of the
Aenesidemean tropes, see Morison (2011).
12. This principle could also be connected to Aenesidemus’s “Heraclitean” definition of
the true as that which appears to all in common at AM VIII 8.
13. Some narrower kind of “phenomenalism” restricted to perception has been associated
by some (e.g., Polito 2007) with a possible influence of medical empiricism upon the
development of Aenesidemus’s skepticism: the ancient reconstruction of the whole
succession of Pyrrhonists conjoining Timon to Sextus’s pupil, Saturninus, however
fanciful, is interesting in that it includes several Empiricist doctors and makes the
Empiricist Heraclides of Tarentum Aenesidemus’s teacher. On medical Empiricism,
see Allen (1993) and chapter 8 in this book.
AENESIDEMUS 79
14. E.g. AM VIII 8: “Aenesidemus says that . . . the things that appear in common to
everyone are true, while the ones not like this are false”; XI 69–71: “If there is
anything good by nature, this is good in relation to all men, and if there is anything
evil by nature, that is evil in relation to all.”
15. Some interpreters (e.g., Stough 1969; Hankinson 1995: 120–128; and Polito
2004) have suggested that from phenomenal variability Aenesidemus did draw the
conclusion that the underlying nature of things is beyond our grasp.
16. For a different interpretation, which attributes to Aenesidemus some qualified
commitment to a version of the condition, but relativized to ordinary experience,
see Polito (2004: 40–101).
REFERENCES
Allen, James. 1993. “Pyrrhonism and Medical Empiricism.” In W. Haase (ed.), Aufstieg und
Niedergang der Römischen Welt II 37.1, 646–690. Berlin and New York: de Gruyter.
Annas, Julia and Jonathan Barnes. 1985. The Modes of Scepticism: Ancient Texts and
Modern Interpretations. Cambridge: Cambridge University Press.
Barnes, Jonathan. 1983. “Ancient Skepticism and Causation.” In M. Burnyeat (ed.), The
Barnes, Jonathan. 1992. “Diogenes Laertius IX 61–116: The Philosophy of Pyrrhonism.”
In W. Haase (ed.), Aufstieg und Niedergang der Römischen Welt II, 36.6, 4241–4301.
Berlin and New York: de Gruyter.
Bett, Richard. 1997. Sextus Empiricus: Against the Ethicists. Oxford: Oxford University
Press.
Press.
Burnyeat, Myles and Michael Frede (eds.) 1997. The Original Sceptics. Indianapolis:
Hackett.
Castagnoli, Luca. 2010. Ancient Self-Refutation: The Logic and History of the Self-Refutation
Argument from Democritus to Augustine. Cambridge: Cambridge University Press.
Castagnoli, Luca. 2013. “Early Pyrrhonism: Pyrrho to Aenesidemus.” In F. Sheffield and J.
Warren (eds.), The Routledge Companion to Ancient Philosophy, 496–510. London and
New York: Routledge.
Chiesara, Maria Lorenza. 2001. Aristocles of Messene. Testimonia and Fragments. Oxford:
Decleva Caizzi, Fernanda. 1981. Pirrone: Testimonianze. Napoli: Bibliopolis.
Decleva Caizzi, Fernanda. 1992. “Aenesidemus and the Academy,” Classical Quarterly 42:
176–189.
De Lacy, Phillip. 1958. “οὐ μᾶλλον and the Antecedents of Ancient Scepticism,” Phronesis
3: 59–71.
Hankinson, Robert. 1995. The Sceptics. London and New York: Routledge.
Hankinson, Robert. 1999. Cause and Explanation in Ancient Greek Thought. Oxford:
80 SKEPTICISM
Hankinson, Robert. 2010. “Aenesidemus and the Rebirth of Pyrrhonism.” In R. Bett (ed.),
University Press.
Long, Anthony and David Sedley. 1987. The Hellenistic Philosophers. 2 vols. Cambridge:
Mansfeld, Jaap. 1995. “Aenesidemus and the Academics.” In L. Ayres (ed.), The Passionate
Intellect: Essays on the Transformation of Classical Tradition, 235–248. New Brunswick,
NJ and London: Transaction Publishers.
Morison, Benjamin. 2011. “The Logical Structure of the Sceptic’s Opposition,” Oxford
Pérez-Jean, Brigitte. 2005. Dogmatisme et scepticisme: L’héraclitisme d’Énésidème. Lille:
Septentrion.
Polito, Roberto. 2004. The Sceptical Road: Aenesidemus’ Appropriation of Heraclitus.
Polito, Roberto. 2007. “Was Skepticism a Philosophy? Reception, Self-Definition, Internal
Conflicts,” Classical Philology 102: 333–362.
Polito, Roberto. 2014. Aenesidemus of Cnossus: Testimonia. Cambridge: Cambridge
University Press.
Schofield, Malcolm. 2007. “Aenesidemus: Pyrrhonist and Heraclitean.” In A. M. Ioppolo
and D. Sedley (eds.), Pyrrhonists, Patricians and Platonizers: Hellenistic Philosophy in
the Period 155–86 BC, 271–338. Naples: Bibliopolis.
Spinelli, Emidio. 2000. “Sextus Empiricus, the Neighbouring Philosophies and the Sceptical
Tradition.” In J. Sihvola (eds.), Ancient Scepticism and the Sceptical Tradition (Acta
Philosophica Fennica 66), 35–61. Helsinki: The Philosophical Society of Finland.
Stopper, Malachi Rudolph. 1983. “Schizzi Pirroniani,” Phronesis 28: 265–297.
Striker, Gisela. 1983. “The Ten Tropes of Aenesidemus.” In M. Burnyeat (ed.), The
Skeptical Tradition, 95–115. Berkeley and Los Angeles: University of California Press.
Striker, Gisela. 1990. “Ataraxia: Happiness as Tranquillity,” The Monist 73: 97–110.
Thorsrud, Harald. 2009. Ancient Scepticism. Durham: Acumen.
Vogt, Katja Maria. 2010. “Scepticism and Action.” In R. Bett (ed.), The Cambridge
Warren, James. 2002. Epicurus and Democritean Ethics: An Archaeology of Ataraxia.
Woodruff, Paul. 1988. “Aporetic Pyrrhonism,” Oxford Studies in Ancient Philosophy 6:
139–168.
CHAPTER SIX
Philo of Larissa
HAROLD TARRANT
1 LIFE AND EVIDENCE

Philo of Larissa was an adherent of the so-called New Academy, that is, one of those
who saw themselves as philosophizing in the tradition of Arcesilaus as well as of
Plato. Born between 159 and 153 BCE,1 he had apparently developed an interest in
current Academic philosophy while still in his native town and set off to study at the
Academy in Athens at the age of about twenty-four. His mentor for fourteen years
was Clitomachus, apparently the strictest of the followers of Carneades, whose
arguments against the criterion of knowledge were the most comprehensive and
wide-ranging. While there is no evidence that the terms “skeptic” and “dogmatic,”
so familiar from Sextus Empiricus, were yet in regular philosophical use, it is clearly
appropriate for us to regard the Academy under which Philo had studied, and
his teacher Clitomachus, as “skeptical” in important ways—as deserving of that
description as any school up until that time.
Philo became the scholarch (Head) himself on the death of Clitomachus
in 110/109 BCE, and the evidence (Numenius fr. 29 = T. XXIII) suggests that
there was at first little change in the official activities or stance of the school. It is
therefore generally agreed among his interpreters that in the first instance Philo
continued to philosophize in a manner compatible with his Carneadean heritage as
normally interpreted. Ultimately, however, after the dispersal of the school in 88
BCE following the second Mithridatic War, Philo left for Rome where he already
had admirers, and may have lived there for four to five years before his death. At any
rate, soon after his arrival there, he wrote his two so-called Roman Books, which
rapidly provoked considerable controversy among others who had studied with
his school.2 That controversy is an important part of the background of Cicero’s
Academica and, through Cicero, of Augustine’s work Against the Academics.
It is important to realize that virtually the whole of the evidence for Philo is
colored by this controversy, and that there are almost no neutral reports. Cicero
himself was writing a dialogue, and in extant parts of it he may represent himself
as speaking either for Philo or for the traditional Academic position attributed to
Carneades; the anti-skeptical position is defended by his character Lucullus in the
first version of the work (Academica Priora or Lucullus) and by Varro in a revised
82 SKEPTICISM
version (Academica Posteriora).3 The controversy was led by Antiochus of Ascalon,

a former pupil of Philo who had begun to attract pupils of his own while the
Academy still functioned at Athens, and who then claimed to teach the doctrines of
the Old Academy, and particularly of Xenocrates and Polemo, the third and fourth
scholarchs, respectively.4
2 ACADEMIC FREEDOM
The Old Academy had been quite open about offering positive teaching and
embracing doctrine, though given the variety of doctrine that one might claim to
trace in Plato, it is not surprising that the early Academy seems to have been very
flexible in its beliefs. Indeed, Cicero (Acad. II 7–9), in offering his own account of
continuing Academic ideology, does not give a picture of a system of tenets, but of
the acknowledgment that much is unclear and that human beings are consequently
fallible, leading to the habit of giving every point of view a hearing and arguing
both for and against positions. Philosophy’s ultimate goal will still be the truth or
the closest possible approximation to it (Acad. II 7), but one that gives them the
freedom to judge what is “probable” without fear that their judgments will be seen
as involving disloyalty to their teacher and their school (Acad. II 8). While this
position is not openly attributed to Philo, Cicero had been strongly influenced by
his studies with Philo after he had come to Rome (Brutus 306 = T. II), and it would
be surprising if the picture had differed greatly from that which Philo had painted
when in Rome.5
It is vital to note two things in this account. First, the Academic is given the
freedom to make judgments that appear to conflict with his loyalties to his school
or to his personal teacher. Much is made by Philo’s enemies of his change(s) of
stance, and disloyalty to what they saw as the Academic tradition (Acad. II 11–12 =
T. XXIX, Acad. II 18 = XXVII). It is here clear that his own picture of the Academic
tradition imposes no obligation of philosophical loyalty, disloyalty involving
anything that approached a school doctrine. No doubt it would not excuse him
if he were guilty of historical disloyalty involving the willful misrepresentation of
earlier Academic positions, but it did excuse deviations from those positions. The
Academy’s identity was determined by its culture of philosophic investigation, not
by anything to be believed or provisionally accepted.
Secondly, the Academic is expected to search without obstinacy for what can be
maintained with most consistency (constantissime, Acad. II 9). It looks almost as
if there is a kind of criterion for tentative assent being proposed here, whether it
concerns the agreement of the claims with the facts, agreement among the claims
themselves, or the wider agreement of trustworthy human beings. We know that the
term “probable” here refers to the Greek πιθανόν that Carneades had employed as
his basis for choice of action, that is, rather to what is capable of being believed or
approved rather than to anything whose probability could be calculated objectively.
A link between the Carneadean criterion and consistency may be intended, but if
so the focus has shifted from trying to find single “presentations” (φαντασίαι) that
have such startling clarity that they must be totally reliable, and toward having
PHILO OF LARISSA 83
presentations that cohere with all other pertinent presentations that may occur
either to the individual or indeed to any human being who examines the relevant
considerations objectively. Has the single cognitive act lost its overriding importance
in epistemology at this time? Can systems of knowledge be built on aggregations of
cognitive experiences that are legitimately convincing when taken together, but not
wholly convincing when taken singly?
3 PHILO’S FINAL EPISTEMOLOGY

At this point, we should consider the position that is attributed to Philo regarding
the Stoic criterion of truth, the “cognitive presentation” (καταληπτικὴ φαντασία).
Sextus Empiricus (PH I 235 = T XXVIII) tells us that Philo’s adherents “say that
things are non-graspable as regards the Stoic criterion, but graspable as regards the
nature of things themselves.” It seems that the established New Academic theme
that things were non-graspable (ἀκαταληψία) was being blamed not upon nature
itself but upon a Stoic requirement that was so rigorous that no cognitive act could
conform with it. Human beings simply do not have cognitive presentations such
that there could not be another that was identical in all respects except insofar as it
did not correctly present what had given rise to it. Since Sextus does not mention
human beings here, it seems that Philo’s intention was to blame the lack of such
presentations on the Stoics, who devised such a requirement, rather than on serious
cognitive incapacity on the part of human beings in general.6 As if to confirm this,
Sextus goes on to accuse Antiochus of introducing Stoic doctrine into the Academy,
and Cicero presents the whole epistemological debate between Philo and Antiochus
as being centered on the need for a criterion of truth such as the Stoics had conceived
it, tracing accordingly the New Academic endeavor to Arcesilaus’s genuine problems
with Zeno’s concept of a cognitive presentation (Acad. I 44).
Sextus does not seek to limit his report to Philo’s final years in Rome, and taken
together with what he had just said about the multiplicity of Academic positions,
with Philo allegedly introducing a Fourth Academy and Antiochus a Fifth (PH
I 220 = T V), it seems clear that he thought this position on the Stoic criterion had
been shared by Philo’s Academic colleague Charmadas, who died three years before
Philo came to Rome. Brittain agrees that this is what Sextus thought but claims that
“he must have been mistaken” (2001: 54), since it does not agree with Brittain’s
view of three clearly delineated phases of Philo’s thought. Yet Charmadas’s name
here needs some explanation, for he is being treated as if he were virtually a co-
scholarch; perhaps it indicates that Philo had himself controversially made appeal
to Charmadas, who as a former pupil of Carneades might have commanded greater
authority regarding Academic traditions. In that case, it seems strange that Cicero
did not give any hint that Charmadas had anticipated Philo’s position, except in one
rather important respect.
There have been various explanations of how the Roman Books had been
doctrinally innovative. All of them face the difficulties that neither version of
Cicero’s Academica has survived complete, the position of Philo is not explained
in the extant parts, and Cicero himself seems to have found it unconvincing.
84 SKEPTICISM
Tarrant (1981 and 1985) was perhaps over-reliant on Numenius’s evidence (fr. 28
= T XXIII), where it was claimed that Philo was swayed “by the agreement and
perspicuity of our experiences,” suspecting that Philo found little reason to doubt
the evidence of the senses (and ultimately of empirically derived concepts), even if
attempts to prove what is perspicuous were doomed to fail.7
Striker (1997: 274) appears uncertain that Philo ever specified how cognition
might be possible without the existence of a cognitive presentation. She considers
the possibility that Philo had claimed Carneades’s maximally probable presentations,
which had been thoroughly tested, were enough to justify the inference to a
proposition’s truth; but this is rejected in favor of the view that Philo distinguished,
in his own mind at least, between (i) the belief that one has a reliable sensory
indication that p (which he could allow) and (ii) the inference from that presentation
that p is true (which he would have resisted). This would have given him “what
could be called a criterion of truth,” but not an infallible one (1997: 274–275).
Agreement between presentations would surely have been pertinent to the rejected
possibility, while perspicuity would have been pertinent to the one chosen.
Brittain (2001: 18–19) thinks that Philo simply dropped the Stoic requirement
that a presentation should be such that it had to be true of its object in the relevant
way. It seems to me that this option makes disappointingly little attempt to make
Philo sound convincing. It seeks to understand Philo very much in terms of the
traditional debate between Academics and Stoics, and seems not to consider
something fundamental: that the traditional debate, which privileged the discussion
of Stoic concepts in particular, had worked itself out (cf. Augustine, C. Acad. III
41 = T XXXI), and stood like an obstacle in the way of those who thought that
philosophy had a salutary role to play in the shaping of human life—too important
to be neglected because epistemological issues remained unsolved.
In a briefer discussion, albeit informed by decades of working on these texts,
Lévy (2010: 88) affirms that Philo’s originality was not in affirming an alternative
to the Stoic criterion, but “in affirming that epochê, in a restricted version, could be
put in the perspective of a knowledge rooted in the very nature of things.” While
the claim seems to embody contradictions (affirmation and ἐποχή, ἐποχή and firmly
grounded knowledge), it has the merit of conveying a little of the complexity of
Philo’s project. Titles embody something of the different conceptions of what Philo’s
contribution was like. In Lévy’s the key word seems to be “decline,” Brittain’s still
calls Philo an “Academic skeptic,” Striker’s a “dogmatist fighting dogmatists,” while
Tarrant (1985) preferred to ask whether he was a “skeptic” or a “Platonist.” If I
had answered this directly I would perhaps have said “both, in some sense,” but
“neither” might have been equally appropriate.
4. THE ONE-ACADEMY THESIS

Stating an epistemological position was almost certainly not the principal purpose
of the Roman Books. It was in fact not normal for Academic scholarchs to record
their epistemological views in writing.8 Cicero announces what seems to have
been the main function of these two books at the commencement of his Academica
PHILO OF LARISSA 85
(I 13 = T XXX), where he says that they maintained the same thesis as Cicero
had heard from him orally, that there was only one Academy, attacking those
who believed in two separate schools, an Old and a New Academy. This may
reflect the actual title of the books, given that a similar title was also employed
by Plutarch,9 who, of the various Platonists of the Roman Imperial period, was
certainly one of the most accommodating toward Arcesilaus and the so-called
New Academic heritage. The books must have been aimed at Antiochus, who had
already apparently acquired followers of his own (Acad. II 69–70),10 and whose
raison d’être appears to have been the promotion of the “Old Academy” as it
had functioned until the death of Polemo. Any scholarch of the school that Plato
founded must have been offended at the suggestion that its allegiance to Plato had
ceased with Arcesilaus, but Antiochus seems to have assumed, like Gerson (2013:
13–14), that anti-skepticism was a defining characteristic of Platonism proper.
Antiochus’s offense at the content of the Roman Books when he read them in
Alexandria in 87/6 BCE (Acad. II 11–12 = T XXIX) all but guarantees that they
were written in direct opposition to him.
In the course of arguing against the idea of a gulf between Old and New
Academies, Philo made claims that were unfamiliar to all Academics present to the
extent that they might not have believed these to be Philo’s books if it could not be
confirmed by the presence of some who had heard him in Rome. The passage accuses
Philo of telling “lies” (II 12: mentitur), a word usually used of historical rather than
philosophical claims, and the topic obviously required an account of why Plato and
the Academy had once seemed so much more willing to embrace their preferred
theses than Arcesilaus and his successors. Somehow the gulf between Plato and the
recent Academy had to be bridged, and there needed to be a single approach to
epistemology that could plausibly be credited to both. In this way Philo was forced
to reveal his current position on the criterion of truth, a position that he had not
revealed before, even though he might long have been leaning in this direction. It
is here that I must resist Brittain’s distinction between Philo’s Metrodoran phase
and his Roman Book position; I think it more plausible that Philo moved toward
his final position over a period of many years, which could better explain why
Antiochus’s “Old Academic” teaching had initially been tolerated.
It is quite possible that Philo’s One-Academy thesis was not a simple construction,
especially given that it was expounded in a two-book work. It seems likely that the
Academics had long had some kind of story to tell about Socrates and Plato being
authorities for skepticism, and chapter 10 of the anonymous Prolegomena to Plato’s
Philosophy lists five arguments designed to show that Plato was a “suspensionist”
(ἐφεκτικός, i.e., a champion of suspension of judgment), which appealed to the
Theaetetus and several other dialogues. Some such arguments were discussed earlier,
in Cicero and others.11 In Cicero (Acad. I 44–45, cf. II 74) Socrates is supposed to
have been led by the obscurity of certain matters to his profession of ignorance
(at least in the course of any given investigation),12 in which only the knowledge
of his own ignorance was made an exception. What Cicero’s character Varro says
about Socrates at Acad. I 15–16 shows that even a spokesman for Antiochus’s
side would have no great quarrel with this picture of Socrates, and that it could
86 SKEPTICISM
never be described as a “lie.” Correspondingly, Plato’s dialogues are supposed to

have affirmed nothing, argued for both sides of many questions, inquired on all
subjects, and resisted certainty (Acad. I 46). Such claims might easily have been
used to support a more orthodox Academic skepticism that privileged argument
on both sides of the question and the suspension of judgment, but it cannot have
been new since it has Plato himself go beyond the requirements of the “Fourth
Academy” position as stated by Sextus. This position accepts that things are not
naturally shrouded in obscurity after all (as Socrates had allegedly thought), and
that they would in fact be graspable as long as one does not make cognition depend
on the Stoic cognitive presentation.13 In order to align Plato (and preferably Plato’s
Socrates, too) with Academic thought and practice as Philo was now depicting it,
there were really only two requirements:
1. Neither Plato nor Socrates could be depicted as accepting the Stoic cognitive
presentation, though they did not need to be disowning every explanation of
a cognitive grasp (κατάληψις);
2. Both should be seen as resorting on occasions to such Academic practices as
suspension of judgment and argument on both sides of the question.
Both of these requirements are easily met, given the variety of Platonic dialogues
and the variety of the representation of Socrates therein. No great effort had to be
made to present either Plato or Socrates as a genuine skeptic, for genuine skepticism
was now no longer being attributed to the Academy (cf. Striker 1997: 276). If Philo
was lying, it was not about the school’s founder but about his successors. Here two
possibilities present themselves: either he gave a false account of the Old Academy
after Plato, presumably including Xenocrates and Polemo whom Antiochus had
appeared to favor, or he misrepresented the New Academy from Arcesilaus. The
former alternative would have the advantage of tackling Antiochus’s new authorities
for the true Academic discipline. Cicero’s Varro, speaking for Antiochus at Acad. I 17,
claims that not only Aristotle but also Xenocrates had complete doctrinal systems,
following Plato in the essentials, and abandoned the Socratic style of inquiry with its
widespread doubts and resistance to final conclusions; later he claims that Polemo,
Crates, and Crantor had defended diligently the tradition that they inherited
(Acad. I 34). The overall picture is one of increasing certainty, to which Zeno would
apply the final touches with his theory of a cognitive presentation, laying the ground
for certain knowledge (Acad. I 41–42). Yet there is no suggestion in what survives
that Philo had tried to argue over any of these Old Academics, and “Cicero” at I 46
simply judges the proper “Old Academic” position from what Plato seemed to have
believed.
But how could Philo have “lied” about the New Academy? There was nothing
novel in the view that Arcesilaus devised his armory for an ad hominem attack
against the Stoics, as that had been Metrodorus’s view; likewise Metrodoran was the
view that the sage might on occasion “opine” owing to the lack of a secure cognitive
act such as the Stoics demanded (Acad. II 78 = T XXV, cf. II 59, 148). Furthermore,
Brittain (2001: 104–108) would assign the acceptance of “perspicuous” presentations
PHILO OF LARISSA 87
at Acad. II 34 to the Philonian/Metrodoran position, that is, the one preceding the
Roman Book position, so that the suggestion that Academics accepted “perspicuity”
could not have been the Roman Book lie. Could he have attributed to the New
Academy the acceptance of presentations that are true and “caused by the object in
the relevant way,” dismissing only the clause that requires them to be true in such a
way that a similar false one could not exist? This is the Roman position that Brittain
would attribute to Philo (2001: 18–19, 129–168), but, though Philo himself may
well have been thinking along such lines, it is most unlikely that he would have
sought to attribute it to Academics from Arcesilaus to Clitomachus as anything like
their official view (i) because they did not espouse official views and (ii) because
Cicero explicitly regards this position as Peripatetic (Acad. II 112),14 suggesting
perhaps that the Old Academy too might not have resisted it (Acad. II 113), but
giving not the slightest hint that it had been attributed to any New Academic—or
even openly adopted as his own by Philo.
Philo’s “lies” resided neither in the attribution of New Academic doctrine to
Plato and the Old Academy, nor in the attribution of Platonic, Old Academic, or
even Peripatetic doctrine to the New. Rather, it lay in a theory of two strands always
(or almost always) coexisting. That does not require a theory of esoteric dogmatism,
as usually believed.15 The evidence sometimes understood as pointing to claims of
esoteric dogmatism is as follows:
1. Augustine (C. Acad. III 43 = Cic. Acad. fr. 21) reports that according to
Cicero “the Academics had been in the habit of concealing their view, and of
only revealing it to those who had lived with them until old age”;
2. Sextus (PH I 224) mentions that some thought Arcesilaus, when he used to
test aporetically whether his colleagues were well suited to the reception of
Platonic doctrine, appeared to remain aporetic himself, but was still putting
the teaching of Plato into the hands of the gifted among his colleagues;
3. Cicero’s Lucullus (Acad. II 60), presumably responding to some claim
already made before the beginning of book II concerning the Academics’
practice of concealing the truth that they discover, asks “What are
those mysteries?”
One looks in vain for any word for teaching, or for anything indicating doctrine
in texts 1 and 3. Text 1 seems to refer to no more than one’s preferred option on
a given issue, but the long qualifying period agrees well with the esoteric doctrine
of [Plato] Epistle VII 342c5–d2 (cf. II 314a5–7);16 text 2 suggests no commitment
on Arcesilaus’s part to Platonic doctrines or writings, but rather an openness to
communicating them so that suitable persons were entrusted with Plato’s written
views17; text 3 perhaps seems sarcastic, and uses “mysteries” in part because of the
claim that any fruits of Academic searches were not to be revealed, but it remains
possible that Philo had somewhere used this term, whose application to Platonic
studies is attested in early imperial Platonism (Theon of Smyrna, Expositio rerum
mathematicarum ad legendum Platonem utilium XIV 18). However, mysteries
typically involve not only a revelation for the initiate, but also some kind of ritual
88 SKEPTICISM
in the course of which that private revelation takes place. One was guided through
the mysteries, not taught them. For the Platonist the model would be the revelation
that is presented by Socrates at Theaetetus 155d–157c, where Socrates remains
uncommitted and regards himself as a partner in Theaetetus’s search for a hidden
truth—for ideas that are presented as “mysteries” such that they must be kept from
the uninitiated and uncultured materialists. In both text 2 and text 3, the Academic
teacher is being assimilated to “Socrates” in the Theaetetus.18
These texts suggest the existence of a view of the Academy that allowed
individuals to hold more beliefs and to discover more truths than could be made
public, but they still did not prescribe what beliefs might be held, provisionally
or otherwise. Text 2 in particular suggests that Plato continued to be taught, not
as scripture or dogma, but in the sense that he is taught today. Indeed, we find
Charmadas teaching the Gorgias in the Academy close to the time when Philo took
over as scholarch.19 We do not find Philo himself teaching Platonic writings, and
those who had most to do with Plato’s writings under the Old Academy included
Philip of Opus and Crantor; this last a figure of comparable stature with Polemo
in the Old Academy just as Charmadas had comparable stature with Philo in the
New. Plato was kept alive, but there may have been a separation of roles, so that
the scholarch would not be constrained by the need to defend Plato.20 At any rate,
the high regard in which Plato was held at that time cannot easily be disputed. It is
entirely possible that Philo, in attempting to close the gap between Plato and the
Academy’s recent past, had exaggerated the ongoing commitment of the school to
its founder and the significance of the kind of Platonic “teaching” that was now
being undertaken.
5 ETHICS
There is only one substantial testimony about moral philosophy, which concerns
the division of the types of discourse that it requires. The fragment is preserved in
Stobaeus (Eclogues II 7 = T XXXII), via some intermediary21 who clearly admired
Philo greatly and considered this contribution valuable. It pursues an analogy between
philosophy and medicine. Like the doctor, philosophy must persuade the student to
take the cure and argue against contrary advice, and this comes under the heading
“protreptic.” At stage two the therapy will be introduced, involving the removal of
false opinions, which have a deleterious effect on our faculties of judgment, and the
imparting of healthy opinions. This requires discourse on the topic of things good
and bad. A further stage will require discourse on the end or goal, that is, happiness,
analogous to health in the case of medicine. This discourse involves discussion of
types of life and many moral and political questions that will assist the preservation
of one’s psychic health. Since the goal is essentially the privilege of the wise, Philo
also demands ethical writings of a more summary kind for those of more modest
ability. The obvious presence of a presumption that philosophical progress may be
made and the talk of imparting opinions make it difficult to regard this division as
the work of anybody who should be called “skeptic.” Rather, it bears out the view
that Philo was anxious for philosophy to move on.
PHILO OF LARISSA 89
6 RHETORIC
The ethical testimony just discussed, insofar as it discusses types of discourse, has
a link with rhetorical theory. It is clear from Cicero (Tusc. II 9, 26, = T XXXV–
XXXVI) that Philo taught rhetoric as a separate discipline from philosophy and also
had special skills in the recitation of poetry. None of this has much bearing on his
relation to skepticism.
7 INFLUENCE
In Tarrant (1985) my primary concern had been to further weaken the evidence
that Antiochus was the force behind the Platonic revival, which followed under
the early Roman Empire, by showing that his opponent Philo was producing an
epistemological position that granted one considerable latitude and was closer to
being something that could be used by subsequent generations of Platonists. The
Stoic theory of knowledge and the Stoic cognitive presentation were not something
that were taken up and used by those known as Middle Platonists, whereas Philo’s
stance was not inimical to constructive philosophical inquiry in the Plato tradition.
Dealing with those I thought to have been influenced by Philo, particularly the
anonymous Theaetetus-commentator and Plutarch,22 both of whom seem to have
held some kind of One-Academy thesis, I perhaps blurred the lines that separated
them from Philo, thus giving somewhat too “Platonist” a picture of Fourth Academic
activity—especially given uncertainty over when various Platonic themes were
reintroduced as Platonist doctrine. Opsomer (1998) has clarified many of the issues
involved. Antiochus is still looked upon by some as a key figure in the transmission
of Platonic doctrine (see Bonazzi 2012), though there is an increasing feeling that
no single re-inventor of Platonism should be sought. The version of Philo described
by Brittain (2001) would certainly not have had much influence, either on Platonism
or on skepticism, though however we describe Cicero it is clear that Philo, if not
his Roman Books, did influence him greatly, and as a result Augustine and the
western tradition have become more familiar with the issues surrounding Academic
skepticism than they otherwise would have been.23
NOTES
1. According to T. I, i.e., testimony I in Brittain (2001), which is now more convenient
to use for the testimony than Mette (1986), Philo was born in the archon-year
of Aristaechmus, and accordingly Brittain (2001: 41) and Lévy (2010: 85) offer
ca. 159/8 BCE. Dorandi (1999: 34) had given 154/3 BCE. Discrepancies arise
from many uncertainties regarding the reconstruction of numbers of years in col.
XXXIII of our papyrus source (Philodemus’s Academic Index from his Syntaxis
Philosophorum).
2. Glucker (1978) was a seminal work in modern scholarship on this controversy.
3. It is important to note that, since an earlier part of the second version and a later part
of the original version are extant, Acad. I conventionally refers to the later version
and Acad. II to the earlier version.
90 SKEPTICISM
4. Sedley (2012) is an important and wide-ranging collection of articles on Antiochus,

but the contributors are by no means agreed on several controversial issues. On
Antiochus’s life, see particularly Hatzimichali (2012).
5. We are told by Cicero (ND I 59 = T XXII) that Cotta was advised by Philo to hear
Zeno the Epicurean, a notable instance of the Academic policy of giving everybody
a hearing.
6. To this extent I may be in disagreement with Sedley (1981: 72), who speaks of
“human cognitive incapacity,” though he too sees that “incapacity” in terms of the
lack of a Stoic criterion.
7. See Tarrant (1985: 52–53). Note particularly anonymous in Theaetetum, LXX,
a work best consulted in Bastianini and Sedley (1995), on Academic attacks on
growth, allegedly designed to demonstrate the inability to improve on perspicuity
by rational proofs that growth happens.
8. Rather, versions of the lectures of Arcesilaus and Carneades were recorded by
others (Philodemus Acad. XX 3–4, XXII 37–XXIII 6); the accusation that Philo’s
position in the Roman Books was novel does not seem to have been supported by
reference to any earlier epistemological text that he had written, so that I do not
find it “probable that Philo wrote at least one work on the Philonian/Metrodoran
position” (Brittain 2001: 71). Stobaeus Ecl. 2.7 (= T XXXI) gives his division of
ethical discourse, and the verb πεπραγματεύεται suggests that this was laid out
schematically, and hence probably in writing, though what we are told does not
necessarily reflect its title. Nothing gives the impression that writing came naturally
to Philo.
9. Lamprias Catalogue no. 63: On There Being One Academy since the Time of Plato.
Nos. 64 and 134 seem also to testify to his Academic interests, and the extant
Against Colotes defends Arcesilaus.
10. There is some discussion about when this happened, and how formal Antiochus’s
group of followers was; I follow Hatzimichali (2012: 12–16) rather than Polito
(2012: 32–34), and feel that Antiochus may well have taken over the role that
Charmadas had been playing in 91 BCE.
11. One argument appeals to the repeated use of words like “perhaps” in the dialogues,
and is found also in the anonymous Theaetetus-commentator (col. LIV).
12. This important qualification is proposed by Burnyeat (1997: 299).
13. Augustine (C. Acad. III 41 = T XXXI) adds that Metrodorus had previously denied
that the Academics had adopted as their own position the theme that knowledge was
impossible, but that it had emerged rather from their dispute with the Stoa (cf. Acad.
I 44). On Metrodorus, see Lévy (2005).
14. “If I were dealing with a Peripatetic, who stated that this thing can be perceived that
has the impression of something true, but doesn’t [append]* that important extra
‘in such a manner that it could not come from something false.’” The verb marked
* cannot reliably be ascertained.
15. In Tarrant (1981: 70), I failed to recognize the limited nature of such reports,
influenced perhaps by the One-Academy thesis of anon. in Tht. LIV–LV. Brittain
(2001: 34) sees no merit in the theories of esoteric Platonism but wrongly attributes
what I had said to Tarrant (1985).
PHILO OF LARISSA 91
16. It would be perverse to tackle the vexed question of the authorship of this passage
here; Cicero seems to be extending this idea of “living with” philosophers to
living with their heritage at Acad. II 74: “I seem to have lived with them [Socrates
and Plato].”
17. The passage needs careful translation. The aporetic testing referred to, in which
suitable persons are given a taste of the doctrines of Plato, is reminiscent of
Platonic midwifery in the Theaetetus, particularly 150b–151b for testing, the
suspension of judgment by the tester, and the rejection of unsuitable candidates,
and 157c–d for testing, continued suspension of judgment, and the offer of a taste
of the thoughts of the wise. For a new study of Arcesilaus’s use of midwifery, see
Snyder (2014).
18. On the early importance of this dialogue, see Tarrant (2013).
19. Evidence in Cicero, de Oratore I 47 where Crassus says: “Then at Athens I read
the Gorgias rather carefully with Charmadas”; the influence of Plato’s Gorgias and
Phaedrus is evident in Charmadas’s arguments against non-philosophical rhetors
at De Or. I 84–94, though Charmadas did not of course openly reveal his own
views in the course of these arguments (De Or. I 84). For a view of Charmadas as a
conservative Carneadean, see Lévy (2010: 84–85).
20. I suspect that Arcesilaus became an exception, and that he had taken over his close
friend Crantor’s role, “acquiring Plato’s books” before unexpectedly becoming
scholarch, too.
21. Stylistically, the passage that we have resembles other material that follows, at least
some of which derives from Eudorus of Alexandria.
22. Plutarch refers to Philo in his Lucullus (42 = T XIV), where he praises Cicero’s
Lucullus highly, and his Cicero (3, 4 = T XX, XV), in which the praise of Philo’s
character contrasts with the suggestions that Antiochus was swayed rather by
ambition than by any genuine perspicuity that he discovered through the senses.
23. For some account of the “afterlife” of the Academy, see Lévy (2010: 89–102).
REFERENCES
Bastianini, Guido and David Sedley (eds). 1995. Edition of anonymous In Theaetetum in
Corpus dei papiri filosofici greci e latini, iii: Commentari, 227–562. Firenze: Olschki.
Bonazzi, Mauro. 2012. “Antiochus and Platonism.” In D. Sedley (ed.), The Philosophy of
Antiochus, 307–333. Cambridge: Cambridge University Press.
University Press.
Burnyeat, Myles. 1997. “Antipater and Self-refutation: Elusive Arguments in Cicero’s
Academica.” In B. Inwood and J. Mansfeld (eds.), Assent and Argument: Studies in
Cicero’s Academic Books, 277–310. Leiden: Brill.
Dorandi, Tiziano. 1999. “Chronology: The Academy.” In K. Algra, J. Barnes, J. Mansfeld,
and M. Schofield (eds.), The Cambridge History of Hellenistic Philosophy, 31–35.
Gerson, Lloyd. 2013. From Plato to Platonism. Ithaca: Cornell University Press.
92 SKEPTICISM
Glucker, John. 1978. Antiochus and the Late Academy. Göttingen: Vandenhoeck and
Ruprecht.
Hatzimichali, Myrto. 2012. “Antiochus’ Biography.” In D. Sedley (ed.), The Philosophy of
Lévy, Carlos. 2005. “Les petits Académiciens: Lacyde, Charmadas, Métrodore de
Stratonice.” In M. Bonazzi and V. Celluprica (eds.), L’Eredità Platonica: Studi sul
Platonismo da Arcesilao a Proclo, 51–77. Naples: Bibliopolis.
Lévy, Carlos. 2010. “The Sceptical Academy: Decline and Afterlife.” In R. Bett (ed.),
University Press.
Mette, Hans-Joachim. 1986. “Philon von Larisa und Antiochos von Askalon,” Lustrum
28: 9–63.
Opsomer, Jan. 1998. In Search of the Truth: Academic Tendencies in Middle Platonism.
Brussels: Koninklijke Akademie voor Wetenschappen, Letteren en Schone Kunsten.
Polito, Roberto. 2012. “Antiochus and the Academy.” In D. Sedley (ed.), The Philosophy of
Sedley, David. 1981. “The end of the Academy,” Phronesis 26: 67–75.
Sedley, David (ed.). 2012. The Philosophy of Antiochus. Cambridge: Cambridge University
Press.
Snyder, Charles E. 2014. “The Socratic Benevolence of Arcesilaus’ Dialectic,” Ancient
Philosopy 34: 341–361.
Striker, Gisela. 1997. “Academics Fighting Academics.” In B. Inwood and J. Mansfeld
(eds.), Assent and Argument: Studies in Cicero’s Academic Books, 257–276. Leiden:
Brill.
Tarrant, Harold. 1981. “Agreement and the Self-Evident in Philo of Larissa,” Dionysius
5: 66–97.
Tarrant, Harold. 1985. Scepticism or Platonism: A Study of the Fourth Academy. Cambridge:
Tarrant, Harold. 2013. “L’importance du Théétète avant Thrasylle.” In D. el Murr (ed.),
La Mesure du savoir: Études sur le Théétète, 243–266. Paris: J. Vrin.
CHAPTER SEVEN
Cicero
J. P. F. WYNNE
1 CICERO’S SKEPTICISM
The Roman orator and politician, Marcus Tullius Cicero (106–43 BC), earned
his place in this book with a sequence of skeptical dialogues he wrote late in life
(Hortensius, Academica, On ends, Tusculan disputations, On the nature of the gods,
On divination, On fate). These dialogues pioneered attractive writing about the
whole of philosophy in Latin, and they did so from a skeptical perspective. For
Cicero, who had met, and often studied with, most of the leading philosophers of
his day, had been won over to the skeptical “new” Academy at about the age of
twenty by the visit of Philo of Larissa to Rome, and he was to retain his Academic
affiliation throughout his life.1 But he was not, I shall argue, a follower of Philo’s
brand of skepticism.
When we study Cicero’s skepticism in these writings, we face a complex task
of interpretation. For from his dialogue dedicated to Academic philosophy (the
Academica), we learn that the recent history of the skeptical Academy had been
tangled.2 Usually, then, to characterize Cicero’s own skepticism, we must first offer
an interpretation of the historical background he describes, and then try to place
him in it.3 But in this book the first stage would duplicate the preceding chapters.4 So
I shall simply give my answer to the controversial question of where Cicero suggests
that his affiliation lay.5
In the Academica, it is plain that Cicero follows Clitomachus’s interpretation of
Carneades, the dominant figure in the later skeptical Academy. Clitomachus was
by birth a Carthaginian who moved to Athens to pursue a philosophical career.
He became Carneades’s student and successor as the head of the Academy, and
was succeeded by Cicero’s teacher, Philo, in turn.6 Like the Academics in general,
Clitomachus argued against the Stoic “cognitive” or “grasping impression,” the
sort of impression the Stoics claimed could not be false. He argued that for any
impression at all, even one which seems to offer a cognitive “grasp” on its content
because it seems that it cannot be false, there could be an identical, but false,
impression (Acad. II 40, 83).
A further Stoic premise was that a wise person would never opine, that is,
would never assent to impressions which could be false (Acad. II 66–67). Thus,
94 SKEPTICISM
if the Academics were right and there were no such impressions, a Stoic would
have to conclude that the wise person would never assent—and that neither should
anyone else. Now sometimes Carneades seemed to depart from the Stoic premise,
and said that the wise person might assent, but sometimes he seemed to endorse it
and conclude that the wise person would never assent. His own thoughts on the
matter were therefore open to interpretation. Clitomachus’s interpretation was to
take Carneades’s deployment of the Stoic premise to heart, and to conclude that
indeed, Stoic or not, one should never assent (Acad. II 59, 67, 78). By contrast,
other, “mitigated” skeptics, including Cicero’s teacher Philo, thought Carneades
was rightly not in earnest about the sage’s universal suspension of assent. They
decided that we may assent to an impression—take it to be true—even though it
might be false (Acad. II 78, 112, 148).
Clitomachus faced the “inactivity” objection to his skepticism: how could he
act if he never took anything to be true? Clitomachus replied that commitment to
the truth of any impression does not seem necessary to act. For impressions differ
in how plausible or persuasive they seem or, as Cicero often puts it in Latin, how
“like the truth.”7 Clitomachus says that a wise Academic will live by “following” or
“accepting” the impressions that seem plausible, rather than by “assenting” to them,
that is, rather than rationally committing himself to them by taking them to be true
(Acad. II 103–105).
In the Academica, Cicero follows Clitomachus on both points: he interprets the
proper recommendation of the skeptical Academy he follows to be the avoidance
of mere opinion, and aims to follow what is plausible instead.8 He says, “I trust
Clitomachus when he writes that Carneades endured a Herculean labor because
he dragged assent, which is to say opinion and rashness, from our souls as though
it were a wild and savage monster. . . . [W]hat keeps somebody from acting, who
follows things plausible, so long as nothing impedes him?”9
Cicero’s and Clitomachus’s skepticism, by contrast with the mitigated kind, is
sometimes called “radical.”10 But it might seem not so radical. For, you might say,
if someone does philosophy in order to encounter and delight in what seems like
the truth, is she so radical a skeptic? But notice that, epistemically speaking, what
seems like the truth is a world away from the truth. For example, what seems like
the truth can be false, as what is true cannot. It seems like the truth that the Earth
does not move—at least, it seems so when I look out my window—but it moves. To
have found only what seems like the truth, then, is to have made no progress toward
finding the truth, so far as you can tell. Again, you might say, a radical skeptic would
not consider the goal of philosophy to be the truth, or what seems like the truth, or
discovery, at all. She would take as her goal (you might say) suspension of judgment
and emancipation from the tyranny of the truth. Yet notice how little Cicero’s pursuit
of the truth commits him to. He takes to be true no theory of truth, nor any scheme of
logic. He has no idea what, in truth, discovery would be like. He shares the dogmatist’s
conception of the goal of the philosophical enterprise only in the vaguest and most
abstract way. Nevertheless, he shares that conception of the goal: discovery.
On this interpretation, Cicero has no way to tell when, if ever, he has encountered
the truth. You might ask, then how can he tell which impressions are like the truth,
CICERO 95
when he has no truth with which to compare them? The answer must be that the
phrase “like the truth” is potentially misleading. When Cicero accepts that something
seems like the truth, he neither accepts that it approximates the truth nor that it is likely
to be true, for he has no way to assess such claims. Rather, some impressions seem to
him such that he is inclined to accept them, in ways that are presumably familiar: they
seem perceptually clear and distinct, for example, or their content seems to follow
from a seemingly sound argument. It seems to Cicero that this is what the truth would
be like, if he found it. But he has no way to tell if he is right about that.
It is my view that Cicero presents himself in all his skeptical dialogues, and
not just in the Academica, as an occupant of Clitomachus’s position.11 (Where he
stood earlier in life is harder to say.)12 I shall say more to support this thesis in the
next section.
2 CICERO AND THE ETHICS OF BELIEF

In Section 1, I sided Cicero with Clitomachus in the internal debates of the skeptical
Academy. I shall now try to bring Cicero’s skepticism to life: why did Cicero choose
to be a skeptic of this kind, when his charismatic teacher, Philo, was not? In answer
to this question, I shall suggest that Cicero reports a strong intuition about what we
call today the ethics of belief debate, an intuition for which he found a theoretical
model in some Stoic psychology.
A focus of the debate over the ethics of belief is this question: ought we, or
may we, ever form a belief in the absence of sufficient evidence for its truth? For
example, suppose that there is insufficient evidence for the truth of the proposition
that God exists. But suppose also that those who believe in Him act more justly and
have a unique source of comfort. In such circumstances, would we be permitted, or
obliged, to form the belief that God exists? The contemporary version of this debate
was begun by W. K. Clifford, who held firmly the “evidentialist” line that we ought
never to form a belief in the absence of sufficient evidence (Clifford 1877). I wish to
suggest that Cicero describes a similar evidentialist intuition. To see this, let us look
first at a passage where he tells his friends about his personal attitude to skepticism.
Cicero begins his defense of skepticism in the second book of the Academica
as follows:
So I shall get on with it, if I may first say a few things about my reputation, so to
speak. For if I was led to apply myself to this philosophy in particular by some
sort of need to show off, or by eagerness to pick a fight, I think that not only
my stupidity should be condemned, but also my character and nature. For if we
reprove stubbornness in small matters and punish misrepresentation, would I
wish, when the condition and design of my whole life is at stake, to contend with
others just for the sake of the fight, and to cheat them, and myself too? Thus, if
I didn’t think it unfitting to do in this speech what it is sometimes customary to
do in political debate, I would swear by Jupiter and by the gods of the city, both
that I am aflame with zeal to discover the truth, and that I state my own views.
For how could I not desire to find the truth, when I rejoice if I find something like
96 SKEPTICISM
the truth? But just as I judge that this is most beautiful—to see the truth—just so
it is most vicious to accept falsehoods in place of truths.
Nevertheless, I am not the sort of person who never approves anything false,
who never assents, who opines not at all. But we are inquiring about the wise
person. I am indeed a great opiner, for I am not wise. I steer my thoughts not by
the little Cynosure [the constellation of the Little Bear] . . . but by Helice [the
Great Bear], that is, by reasons that are bigger in the mind’s eye, not filed down
to a fine point.13 Thus it comes about that I err and wander more widely. But as I
said, our inquiry is about the wise man. For I accept those impressions that strike
the mind or the sense sharply, and sometimes I even assent to them—but I do not
grasp them, for I judge that nothing can be grasped. I am not wise, thus I yield to
impressions, nor can I resist. (Acad. II 65–66)
Cicero’s latter point is that he does not conduct his mental life according to the
skeptical standard he accepts, in that he sometimes assents. But he says that a wise
person would never assent. To be wise is to have succeeded at philosophy, to be
the ideal for one who pursues wisdom. Thus I think it is clear that Cicero regrets
his frequent lapses into assent, even while he is proud that what skeptical discipline
he manages to maintain affords him some measure of guidance.14 Why does Cicero
wish to be the sort of wise man who does not assent?
The key to this question is in the sentence I have underlined in Acad. II 66
above. Cicero desires to find the truth. But, crucially, his desire is discriminating.
The difference is like that between somebody who is looking for love and thus is
ready to go on a date with anybody promising, and somebody who is already in
love and thus wishes to date his beloved and to avoid dating anybody else. Cicero’s
burning desire for the truth is of such a sort that he judges it “most vicious to accept
falsehoods in place of truths.” To “accept in place of truths” is to take to be true, to
assent. “Most vicious” translates the superlative of Cicero’s term of art turpe, which
in everyday use means “foul.” Cicero uses turpe for what is ethically repulsive, the
opposite of what is virtuous.15 Thus Cicero says here that it is ethically vicious to
assent to what is false.
Now, this particular ethical judgment of Cicero’s is not Clifford’s evidentialist
judgment: Cicero says that it is vicious to assent to what is false, Clifford to assent
where there is insufficient evidence. Clifford’s judgment is more stringent, because it
rules out assent even to what is true when there is not enough evidence. But in fact
Cicero is prone to deliver Clifford’s judgment, too. For example, in On Divination
I 4, he says, “rashness and error in assent are vicious in all matters.”16 So, does
Cicero’s judgment that assent to what is false is vicious motivate his further view
that rash assent, assent to what might be false, is vicious? I think so, as follows.
The truth, if it were available, would be available to Cicero through his impressions.
To find the truth would involve the recognition of the truth of a true impression,
that is, assent to the impression. But the Academics have made it seem to Cicero that
he cannot tell which, if any, of his impressions are true. Now one reaction—even
for somebody repulsed at the thought of assent to what is false—would be to assent
to everything, or, more realistically, to any impression that seemed like the truth.
CICERO 97
For in so assenting one might expect to have assented to some truths, albeit perhaps
at the repulsive cost of assent also to some falsehoods. Cicero rejects this wager.
Thus he must see some asymmetry between the gain of assent to the truth and the
loss of assent to the false, such that the risk of the latter trumps the chance of the
former. I suggest that, for Cicero, to assent to a truth which one cannot tell is true
is not to find the truth, but rather to fool oneself that an impression is true which,
coincidentally, happens to be true. Thus there is nothing virtuous in having a true
impression, or in taking it to be true.17 For either thing can happen to any fool. By
contrast, assenting to what is false in place of what is true is most vicious, because
that is precisely the wage of foolishness: taking to be true what might, so far as she
could tell, be false, the fool has taken to be true what was false. Thus we have found
Cicero’s asymmetry. Where there are no cognitive impressions, to assent to what is
true is not ethically virtuous, but to assent to what is false is ethically vicious. Thus
the risk of assent to what is false trumps the chance of assent to the truth. In this
way, Cicero’s judgment that it is vicious to assent to what is false, combined with his
skepticism, leads to his judgment that rash assent, assent to what might be false, is
itself vicious, the equivalent of Clifford’s contention.
But why does Cicero think that rash assent is an ethical failing? Why is it not
merely a stumbling block in the pursuit of knowledge or wisdom? It is easy to see why
a mistake in the pursuit of truth is ethically vicious for thinkers like the Stoics, who
hold that virtue is identical with knowledge. But why so for Cicero, who was not a
Stoic? We can answer this, and make better sense of many of Cicero’s pronouncements
quoted in this chapter, if Cicero cherished an aspect of human psychology described
theoretically in a text on Stoic ethics he wrote for his son, On Duties. Cicero there
presents four basic drives in the human soul, each of which is the basis for one of four
cardinal virtues. One of these is the natural drive for truth or knowledge, which is the
basis for the virtue of wisdom (see also Off. I 15–16; cf. Fin. III 17–18):
The search for and inquiry into the truth, above all, are proper to humanity. Thus
when we are free of pressing business and care, we yearn to see something, to
hear, to learn, and we suppose that the grasp of matters hidden or wonderful is
necessary for the happy life. (Off. I 13)
Of these four topics, it is the one that has to do with the grasp of the truth that
most of all gets at human nature. For we are all dragged about and led on to a
desire for cognition and knowledge, in which we think it beautiful to excel, but
we say that to slip, to err, to be ignorant, to be deceived is both bad and vicious.
While this kind of drive is both natural and virtuous, we must avoid two flaws
in it. One is that we might hold what we have not grasped in place of what
is grasped, and assent to these impressions. Anyone who wishes to escape this
flaw—and everyone should wish to—must use both time and care in pondering
matters. The second flaw is that some people bestow too much enthusiasm and
effort on subjects difficult, obscure and indeed not necessary. (Fin. 1.18–19)
It is the former of these two flaws that Cicero worries about in his skepticism:
the desire to grasp can itself lead one to assent to what one does not grasp. The
point, I imagine, is that faced with a plausible but possibly false impression, we
98 SKEPTICISM
wish to decide that it cannot be false, so that we may assent to it. If we yield to
this temptation, we placate our desire for a grasp of the truth, but only by gulling
ourselves with a counterfeit.
Cicero reports similar views of his own in the Academica. First, on the point that
we have a natural urge for the truth, here Cicero concludes his arguments against
dogmatism in physics with the remark that physical inquiry should continue:
Investigation into these most important but most hidden matters holds a delight
of its own. If something turns up which seems like the truth, the soul is filled
with the most human of pleasures. Thus both your [dogmatic] sage and our own
[skeptical sage] inquire into these things, but yours with the result that he assents,
he trusts, he makes sure, ours with the result that he fears that he might rashly
opine, and so that he thinks he has done famously, if he has found something like
the truth in matters of this kind. (Acad. II 127–128)
The pleasure that a skeptic feels on finding something even like the truth is the
“most human of pleasures.” Both dogmatist and skeptic, it appears, are responding
to the sort of drive we saw Cicero describe in On duties.
Second, on the point that dogmatists yield to a sort of temptation to assent, here
is the conclusion to Cicero’s arguments against dogmatism in ethics:
So do you gentlemen think that, when I hear these things and countless others
like them, I am entirely unmoved? I am moved just as you are, Lucullus, nor
should you think me less human than you. The difference is only that when you
are moved, you give in, you assent, you agree, you wish it to be true and sure
and grasped and perceived and reasoned and firm and fixed, and you cannot be
moved nor pushed out of it by any argument. But I judge that nothing is of such
a sort that, should I assent to it, I would not often assent to something false, since
truths are distinguished from falsehoods by no mark, especially since there are no
criteria of logic. (Acad. II 141)
Cicero’s diagnosis of dogmatism here is that dogmatists wish to grasp that their
views are true, and give in to this temptation, while the skeptic holds firm against it.
Thus in the Academica Cicero shares views with the sorts of Stoics he was to
describe in On duties, that humans have a natural and laudable drive for the truth
but that this very drive tempts people into rash assent. I suggest this explains why
Cicero considers rash assent to be an ethical failure. To allow the most distinctive
aspiration of our rational and valuable nature to trip itself up is to betray that nature.
In this light, we can understand the whole sequence of thought in Cicero’s remarks
on his attraction to skepticism (Acad. II 65–66, quoted above). Cicero’s love for
the truth leads him to despise assent as ethically vicious, just as the same love leads
people at large to despise mendacity as ethically vicious.18
I shall now return to the question of Cicero’s affiliations. For you might see
in Cicero’s repeated declaration of his devotion to the truth a sign that he is a
mitigated skeptic, aligned with Philo, after all.19 At times, you might point out, he
even says that in skeptical discourse, “[his] goal is to speak both for and against all
philosophers, for the purpose of discovering the truth” (N.D. I 11, emphasis added).20
CICERO 99
But I have shown that, on the contrary, it is exactly Cicero’s discriminating love of
the truth that leads him to disagree with the mitigated skeptics, because it leads him
to think that assent to what might be false is always vicious. While discovery of truth
is indeed his fond hope in skeptical inquiry, a more realistic statement of his goal is
to find, “either what is true or what approaches as closely as possible to the truth”
(Acad. II 7).21 So far he has found at best the latter, to which he must not assent
because it might not be true.
We might see tragedy in Cicero’s situation. He is the man who loved truth so
faithfully that he lost it altogether. But Cicero does not seem downcast. On the
contrary, he presents himself as confident and content in his skepticism. He is free to
navigate away from danger, while the dogmatist clings to whatever rock life’s storms
drove her onto (Acad. II 8). As Cicero remarks when challenged in the Tusculans
that he has contradicted one of his own arguments in On Ends, “We live for the
day. We say whatever strikes our minds as plausible. Thus only we are free” (V 33).
NOTES
1. For Cicero’s philosophical studies see, e.g., Cicero’s Brut. 306309, Fam. XIII 1.2,
N.D. I 6–7, and Plutarch, Life of Cicero IV 24. For his attraction to Philo, see Brut.
306, De Inv. II 10 (cf. De Or. I 5), N.D. I 11–12 and 17. See also n. 11.
2. For an accessible translation, with help on the work’s confusing history, see
Brittain (2006).
3. I substantially follow the interpretation of these debates in Brittain (2001; 2006:
xxii–xiii). For some recent alternatives, see Glucker (2004) (a critical review of
Brittain), Thorsrud (2009: chs. 4–5; 2010), and Lévy (2010).
4. See the chapters on Arcesilaus, Carneades, and Philo of Larissa in this book.
5. For different conclusions about Cicero’s skepticism from much of the same
material as I examine in this chapter, see Thorsrud (2009: ch. 5; 2010: 76–78)
and Görler (1997).
6. Diogenes Laërtius, Lives of the Philosophers IV 67; Cicero, De Or. I 45. There is
also information about Clitomachus in a papyrus discovered at Herculaneum, the
Index Academicorum. A synthesis of the evidence on Clitomachus’s life, in German,
is Görler (1994). See also the chapters on Carneades and Philo in this book.
7. “Like the truth” translates veri simile and “plausible” probabile, both of which
reflect Carneades’s term. For the data on what these various terms might mean, see
Glucker (1995), who draws different conclusions than I do.
8. Here I alert readers to a difficulty thrown up by the form in which Cicero writes,
narrated conversations in which “he” plays a part as a character. It is possible that
what “Cicero” in the drama says is not meant to express the views of Cicero the
author. But for simplicity’s sake I shall refer to both Cicero the author and “Cicero”
the character as Cicero in this chapter. Readers who rightly wish to be wary on this
point should consult the context of each quotation.
9. Acad. II 108. For Cicero’s other Clitomachean pronouncements in the Academica,
see II 65–66, 78, 127–128, 141.
10. See Brittain (2006: xxvii).
100 SKEPTICISM
11. Some passages in other dialogues which support my view are: Fin. II 43; Tusc. I 8,
IV 47, V 32–33; N.D. I 12, III 95; Div. I 4, II 150. Brittain (2016) is an example
of how one might interpret a dialogue other than the Academica as radically
skeptical. Woolf (2015) gives an extended and accessible picture of a more radically
skeptical Cicero.
12. Cicero seems to have been a skeptic at about the age of twenty, after his encounter
with Philo (see On Invention II 10; cf. De Or. I 5). But at Leg. I 39, written about
a decade before the skeptical dialogues, Cicero’s character invites the skeptical
Academy to step aside from the discussion, which could ground an argument
that Cicero was then in a dogmatic phase (see Glucker 1988). But the latter
passage does not require that interpretation, and arguably does not support it
(see Görler 1995).
13. In Cicero’s day, there was no pole star, because the north celestial pole lay in a dark
part of the sky between the “backs” of the Great and Little Bears. The Little Bear
was thus a useful guide to true north and more precisely so than the Great Bear,
because the Little Bear swept a circle of smaller radius around the pole. Navigation
by the Little Bear was associated with the seafaring Phoenician peoples, of whom
Clitomachus, as a Carthaginian, was one (ND II 105–106, translating Aratus,
Phaenomena 39).
14. Thorsrud (2010: 77) and Görler (1997: 38, 56) think Cicero is proud that he has
mere opinions. Thorsrud interprets Cicero as a mitigated skeptic and thus supposes
that Cicero means that his opinions are the result of responsible inquiry.
15. See, e.g., Fin. III 38, Tusc. V 27, Off. I 10.
16. See also: N.D. I 1, Acad. I 45, Off. II 8. Cicero acknowledges the Stoics as allies in
this: Fin. III 72.
17. You might object: Cicero says it “is most beautiful to see the truth,” so must he
not think it is virtuous to see the truth? But I do not think Cicero means that
it is virtuous to see the truth. He uses the adjective pulcher, “beautiful,” which
he habitually keeps distinct from honestus, “ethically attractive” or “noble” (for
pulcher see, e.g., Tusc. IV 31, for honestus, e.g., On Ends III 27–28). Certainly,
the beauty of “seeing the truth” cannot be something distinctive of the experience
of having or assenting to true impressions, because then true impressions would
offer to us a mark of their own truth, the beauty of having or assenting to
them, and thus true impressions would be cognitive, such that they could not
be false. The beauty of somebody “seeing the truth,” whether this means having
or assenting to a true impression, must either be a beauty that is also to be had
from false impressions, or, if distinctive of an encounter with the truth, it must
be something Cicero imagines from the outside: it is fitting that a rational mind
should encounter, perhaps even take to be true, the truth that it desires, even
though it cannot tell that it has done so. But if so, this is a beauty that both wise
men and fools could exemplify.
18. In On Ends, Cicero makes this sort of parallel explicit in a related argument against
Epicurean ethics: “Since [n]ature has implanted in humanity a desire to see the
truth, as is easily apparent when, once freed from care, we wish to know what
happens even in heaven, we are drawn on by these first stirrings to love everything
true, i.e. what is trustworthy, simple and consistent, while we hate what is empty,
false and deceptive, like fraud, perjury, malice and injustice” (Fin. II 46).
CICERO 101
19. See Thorsrud (2010: 76–78).

20. For the discovery of truth as Cicero’s goal in skeptical inquiry, see also Fin. I 13.
21. See also: Acad. I 45; Off. II 8. Cicero acknowledges the Stoics as allies: Fin. III 72.
REFERENCES
Bret, Richard. 2010. The Cambridge Companion to Scepticism. Cambridge: Cambridge
University Press.
Brittain, Charles. 2001. Philo of Larissa. Oxford: Oxford University Press.
Brittain, Charles (trans.). 2006. Cicero: On Academic Scepticism. Indianapolis, IN: Hackett.
Brittain, Charles. 2016. “Cicero’s Sceptical Methods: The Example of De finibus.” In J.
Annas and G. Betegh (eds.), Cicero’s De Finibus: Philosophical Approaches, Cambridge:
Cambridge University Press, 12–40.
Clifford, W. K. 1877. “The Ethics of Belief,” Contemporary Review 29: 289–309.
Glucker, John. 1988. “Cicero’s Philosophical Affiliations.” In J. Dillon and A. A. Long (eds.),
The Question of “Eclecticism”. Berkeley, CA: University of California Press, 34–69.
Glucker, John. 1995. “Probabile, Veri Simile, and Related Terms.” In J. G. F. Powell (ed.),
Cicero the Philosopher: Twelve Papers. Oxford: Oxford University Press, 115–143.
Görler, Woldemar. 1994. “Die Akademie zwischen Karneades und Philon.” In H. Flashar
(ed.), Grundriss der Geschichte der Philosophie v. 4.2, 898–914. Basel: Schwabe.
Görler, Woldemar. 1995. “Silencing the Troublemaker: De Legibus 1.39 and the Continuity
of Cicero’s Scepticism.” In J. G. F. Powell (ed.), Cicero the Philosopher: Twelve Papers,
Görler, Woldemar. 1997. “Cicero’s Philosophical Stance in the Lucullus.” In Brad Inwood
and Jaap Mansfeld (eds.), Assent and Argument, 36–57. Leiden: Brill.
Inwood, Brad and Jaap Mansfeld (eds.) 1997. Assent and Argument. Leiden: Brill.
Lévy, Carlos. 2010. “The Sceptical Academy: Decline and Afterlife.” In R. Bett (ed.),
The Cambridge Companion to Scepticism, 81–104. Cambridge: Cambridge University
Press.
Powell, J. G. F. (ed.). 1995. Cicero the Philosopher: Twelve Papers. Oxford: Oxford
University Press.
Schofield, Malcolm. 2002. “Academic Therapy: Philo of Larissa and Cicero’s Project in the
Tusculans.” In G. Clark and T. Rajak (eds.), Philosophy and Power in the Graeco-Roman
World: Essays in Honour of Miriam Griffin, 91–109. Oxford: Oxford University Press.
Thorsrud, Harald. 2009. Ancient Scepticism. Durham: Acumen.
Thorsrud, Harald. 2010. “Arcesilaus and Carneades.” In R. Bett (ed.), The Cambridge
Companion to Scepticism, 58–80. Cambridge: Cambridge University Press.
Woolf, Raphael. 2015. Cicero: The Philosophy of a Roman Sceptic. London: Routledge.
CHAPTER EIGHT
Menodotus and
Medical Empiricism
JAMES ALLEN
1 INTRODUCTION
Menodotus of Nicomedia was a physician and philosopher, who probably
flourished in the first half of the second century CE (cf. Capelle 1931: 901–902).
The Empirical school of medicine to which he belonged arose in the third century
BCE, defining itself in opposition to the rationalist tendency then dominant in the
profession. Though our sources often speak of a rationalist “school” or “sect” (a
logikē hairesis), rationalism was not a proper school in the way Empiricism was,
that is, a collective with common doctrines and/or projects and methods, leaders or
recognized authorities, and a felt group identity (cf. von Staden 1982). The term was
applied polemically by the Empiricists to members of different schools—properly
so-called—who were often sharply at variance with each other.
The contrast between reason and experience on which the opposition is based
is an old one. In Plato’s Gorgias, citing medicine as an example, Socrates maintains
that the practitioner of a genuine art (technē) selects the measures he will employ
on the basis of an understanding of the nature of the object upon which he acts, and
he insists that a true artist must be able to give an account (a logos) of his actions
in the light of this understanding by specifying the causes because of which that
object is as it is and those because of which the measures he applies will have the
desired effects (501a; cf. 465a, Phaedrus 270b, Laws IV 720b ff.; cf. 857c). In the
terminology favored by later rationalist physicians, who accepted Plato’s view, the
pathological condition indicates its own treatment. From the perspective of Socrates
and those sharing his outlook, mere experience can by contrast never constitute an
art because it is confined to recording conjunctions and sequences among observed
events without grasping the nature of the beings that act or are acted on in them,
and therefore without being able to get at the cause or explanation, which must
refer to the underlying nature of what is observed.
According to Galen’s testimony, in the view of the rationalists, reason was above
all a faculty that enables us to grasp natural or rational relations of consequence
(akolouthia) and exclusion (machē) (Subfig. emp. 87, 7; 89, 12, 15 Deichgräber).
MENODOTUS AND MEDICAL EMPIRICISM 103
The grasp of indication is one application of this faculty, which also explains one of
the distinctive marks of rationalist medicine: its reliance on (alleged) truths about
so-called non-evident matters, which, because they are forever beyond the reach of
observation, can become objects of knowledge only with the aid of such a power (cf.
Galen, De sect. ingred. SM III 11, 16–19; Subfig. emp. 62, 24–31 Deichgräber; On
Medical Experience 100, 102, 103, 104–105, 107, 111, 132–133, 135, 136, 137,
139, 141, 153 Walzer).
In effect, medical Empiricists came to the defense of the position attacked in
the Gorgias. Though they developed their position with an eye on medicine, they
held that it applied to other arts as well (Celsus, Proem. 31–32; Galen, MM X 126,
15–127, 1 K; Subfig. emp. 86, 23–87, 4 Deichgräber). They strove first to show that
experience, which had its origin in observation and was confined to objects accessible
to observation, was equal to the task of constituting an art without the aid of reason
and the unobservable entities allegedly discovered by it. But the Empiricists also
took the battle to their opponents by trying to show that the difficulties in which
the rationalists and their theories were entangled undermined the claims they made
for reason.
2 SKEPTICISM AND EMPIRICISM

Menodotus owes his place in the history of skepticism, and his presence in this book,
to the intriguing fact that, like several other Empiricists, he was also a Pyrrhonian
skeptic.1 The list of Pyrrhonists in Diogenes Laertius includes at least three more
Empiricists apart from Menodotus: his fellow student, Theodas, Sextus Empiricus,
who is identified by his surname as an Empiricist, and Sextus’s student Saturninus
(IX 115 = fr. 9 Deichgräber). Others on the list and Pyrrhonists not mentioned in
it may also have been Empiricists. Sextus and Menodotus are also included in the
list of Empiricists in the pseudo-Galenic medical work, Introductio seu Medicus
(9, 23–24 Petit = XIV 683, 11 Kühn = fr. 6 Deichgräber).
Like the Academics, the Pyrrhonists were dogged from the outset by the charge
that their philosophy, which was expressed in skeptical slogans like “Everything
is inapprehensible” and “Suspension of judgment about all things,” made life
impossible. Sextus stoutly rejects this charge, maintaining that skeptics live and
act under the guidance of a criterion, namely what appears (to phainomenon),
which amounts to living in accordance with everyday observance, one of whose
four elements is the teaching of the arts (PH I 21–23). Like the members of other
medical schools, Empirical physicians diagnose, make prognoses, and prescribe
and administer treatments and regimens—all of which would seem to require
a high level of discrimination and judgment. Yet they, or those of them who
were also Pyrrhonists, plainly thought they could discharge their responsibilities
as doctors without compromising skeptical suspension of judgment. An
understanding of the art practiced by so many Pyrrhonists may give us some
insight into the skeptical form of life they claimed to lead or at the very least, by
affording us a fuller picture of such a life, give us a better understanding of the
explanandum.
104 SKEPTICISM
Sextus wrote a work on Empirical medicine, which has not survived (AM I
61).2 The extant philosophical works contain expressions of sympathy for
the Empiricists and their ideas (PH II 246, 254; AM V 104, VIII 288, 291; cf.
Spinelli 2010). At one point, Sextus even treats Empiricists and Pyrrhonists as
allies who join forces to argue that non-evident matters are not apprehended (AM
VIII 191), and at another, without mentioning the Empiricists, he presents the
skeptics as disagreeing not only with dogmatic philosophers, but also with the
Empiricists’ principal opponents, rationalist physicians (AM VIII 157).3 When he
tackles the question whether as “some people say” medical Empiricism is the same
as skepticism, however, Sextus takes Empiricism severely to task for dogmatic
tendencies unacceptable to true skeptics, and he suggests that Pyrrhonists
interested in medicine might be better advised to practice as members of the so-
called Methodist school of medicine (PH I 236–241). Although it can be argued
that Sextus did not intend to condemn Empiricism, but to criticize tendencies in
it (cf. Frede and Walzer 1985: xxv–xxvi; Allen 2010), we must look elsewhere in
order to learn how someone could be both a Pyrrhonian skeptic and an Empirical
physician. Attention to Menodotus is likely to prove especially rewarding. Owing
to Galen’s interest, we know a good deal about his views on the issues that divided
the Empiricists from their rivals in medicine as well as the stands he took in
controversies internal to Empiricism.
3 ART AND EXPERIENCE

The form the Empiricists’ positive account of their art took was shaped in part by
a challenge put to them by their rationalist opponents: experience, based as it is on
observation, is available to laypeople as well as to experts; therefore, it cannot be
the art that sets the latter apart from the former. The Empiricists acknowledge that
the capacity for observation that is the basis of experience and a certain amount
of experience itself are common to everyone, but insist that there is such a thing
as expert experience (cf. Galen, On Medical Experience 98–99 Walzer; MM 126,
10 ff. K = fr. 45 Deichgräber). And it is experience conceived in this way, they
maintain, that constitutes an art worthy of the name. Galen devotes much attention
to the Empiricists’ ideas about this issue, especially in the Subfiguratio empirica.4
Two terms, peira and empeiria, which are translated as “experience,” both derived
from the verbal root, peirān, meaning “to put to the test.” They are not always
kept apart, and the latter especially has several related senses that must be clearly
distinguished. Peira is typically used to mean the knowledge one gains from a single
episode of observation (Subfig. emp. 44, 6 ff. Deichgräber). Such episodes can be
by chance; result from an intervention by an observer, who tries something to see
what will happen; and they can be the result of efforts to reproduce a previous
observation. Individual episodes like this or the pieces of knowledge one has as a
result of each taken singly do not qualify as experience strictly so-called. Experience
(empeiria) in this privileged sense is knowledge or memory of what has been
observed to happen many times in the same way, which can now be formulated
as a theorem, which specifies, whether, for example, bleeding is followed by the
cessation of fever always, for the most part, roughly half the time, or rarely (Subfig.
emp. 45, 24–46, 14; 50, 23 Deichgräber).
In another sense, “experience” means the art composed of such theorems as a
whole (47, 26; 54, 10–13; cf. De sect. ingred. SM III 3). Experience has its origin
in observation, but as the Empiricists acknowledge, we can profit not only from
our own observations, but also from those of others. Hence they distinguish two
sources of empirical knowledge, autopsia, one’s own observation, and historia, the
experience of others as preserved in books or oral teaching—one practitioner’s
autopsy becomes history for another (De sect. ingred. SM III 3, 17–20).
So much was common ground, but two related issues especially were subject to
controversy in the school and figured in important developments in the school’s
position. One concerned the status of the so-called transition to the similar, the
other, how the Empiricists understood the kind of reason that they reject. Over
the course of their school’s history, Empiricists adopted different stances toward
reason. Some, especially early on, took a very hard line, while other, mostly later,
Empiricists admitted that there was a place in the art for something they were willing
to call reason, but they insisted that it did not qualify as reason as the rationalists
understood it, and to mark the difference they employed two terms, epilogismos for
the form of reason they found acceptable and analogismos for the kind peculiar to
rationalism (Subfig. emp. 62, 22–63, 1 Deichgräber). Theodas maintained that the
method through similarity was a form of rational experience (Subfig. Emp. 50, 2–4).
The transition to the similar is the Empiricists’ response to an obvious challenge,
namely, what will the Empirical physician do when faced with a case of a new type?
The answer is that he will use the transition to the similar by applying a therapy
that has already been effective when applied to one part of the body to another
part similar to it, or by applying the same therapy to diseases that are similar to one
against which it has proven effective (and if the disease is familiar from experience
but the known remedy is unavailable, he will use a similar remedy). But what part
can be played in an empirical art by the transition to the similar, whose point of
departure is, to be sure, something that has been observed and remembered by the
practitioner, but that leads him beyond his experience?
A difficult passage in the Subfiguratio empirica reviews some of the opinions held
by Empiricists (49, 23 ff.; cf. Perilli 2004: 154–157). It begins with a scholarly question
about what an early Empiricist, Serapion, may have thought and becomes—though
it is unclear precisely where—a discussion of the merits of different views about the
issue itself. “Did Serapion judge that transition to the similar is a third constitutive
part of medicine [together with autopsy and history]?” Menodotus is cited for his
view: “No, he only uses it.” Probably still following Menodotus, Galen says: “There
is a difference between using something and using it as a part.” Menodotus, then,
appears to have held that Serapion used transition to the similar without using it as
a constitutive part of medicine, that is, as a reliable source of medical knowledge.
Others, Galen notes, held that it was better viewed as an instrument (organon). It
seems that an Empiricist cannot put the results of the transition on a level with those
of autopsy and history without compromising his empiricism. When Galen returns
to the subject, he again appeals to the authority of Menodotus, who held that the
106 SKEPTICISM
transition to the similar is not a true criterion, but rather a criterion of the possible
(69, 33 ff., 70, 7) and “a road to experience” (70, 6).
Experience remains the ultimate authority. Nonetheless, the distinction between
two moments that is implicit in this account of the nature and value of the transition
to the similar is noteworthy. On the one hand, there is the task of testing or confirming
the potential theorems suggested by the transition; on the other, that of inventing
or “thinking up” what is submitted to the test of experience. An empirical account
of the latter is not impossible, and there are traces of one in Galen’s testimony. In
past experience similarities in certain respects, for example, in the taste and smell
of remedies, have proved more reliable predictors of therapeutic efficacy than in
others, for example, color. But it is also easy to see how the choice of similarities to
which to attend could provide scope for reasoning.
4 THE ROLE OF REASON IN THE DEBATE

Everything depends on what counted as reason in the debate. There are three
relevant ways of understanding reason, which can be roughly ordered in terms of
logical strength, so that it is fair to speak of three degrees of rationalist commitment
to which correspond three levels of antirationalism. What is entailed by the first
and weakest degree of commitment is revealed in an argument brought against
the Empiricists by Asclepiades of Bithynia, a prominent physician and medical
theorist active in Rome in the first century BCE. He espoused a paradigmatically
rationalist theory, according to which health is ultimately the flow at the proper rate
of microscopic bodies through passages in the body and diseases are due to their
excessive or deficient flow.
According to Asclepiades, the finite number of types under which infinitely
various experiences can be classified are not given in experience, but discovered by
reason. He gives as an example the twenty-four letters of the Greek alphabet, with
the aid of which infinitely various verbal sounds can be apprehended (On Medical
Experience 88 Walzer). And he argued that the observation of what happens many
times in the same way cannot begin without the prior contribution of reason—the
reason exercised in concept-formation (Galen Subfig. emp. 88, 25 Deichgräber; De
sect. ingred. SM III 9, 9–13; On Medical Experience 85 ff. Walzer).
The argument is ingenious and touches on important issues. The problem is that
this is not what the Empiricists, or most of their opponents, meant by reason. And
it begs the question to assume that it is impossible to possess the capacities that
Asclepiades counts as rational for the purpose of his argument without also having
the rational faculties that the Empiricists claimed to do without.
The second degree of rational commitment covers reasoning as inference, the
drawing of a conclusion on the basis of grounds or evidence, arguably a practice
in which, in some form or to some degree, almost all human beings engage. The
third and last degree of commitment is reason as characterized in the account of
the debate between rationalism and Empiricism, namely, as a power to apprehend
relations of consequence and exclusion, which is the peculiar property, at least in
its refined and developed form, of highly trained experts and whose most striking
and distinctive manifestation is the discovery of non-evident entities and processes

beyond the reach of observation and unknown to laypeople. Whether and how
this power is related to reasoning corresponding to the second degree of rational
commitment could be a matter of controversy.
No Empiricist embraced reasoning of the last kind. None rejected reason of the
first degree, though they may have questioned how useful it is to speak of “reason”
in connection with it. It was the second degree of rational commitment about
which there was controversy and their attitude to which seems to have undergone
development over time.
How far some Empiricists were willing to go in their rejection of reason is plain
from Galen’s summary of the account some of them gave of the discovery of the
“cephalicum” drug (MM X 163, 14 ff. Kühn = fr. 105 Deichgräber; cf. Comp. med.
gen. XIII 366, 2 ff. Kühn = fr. 106 Deichgräber; cf. Frede 1990: 233). It is a mixture
of ingredients that have been effective treating the same ailment when applied singly
to different patients. Some Empiricists, Galen maintains, say that discoveries of this
sort could be suggested by a dream, or that, after the ingredients had been mixed
by chance, someone might have ventured to try the mixture. This, says Galen, is
so much nonsense, and he contrasts this account with the epilogistical method
employed by other Empiricists. In his words, a physician employing this method,
having experience of the curative effect of each of the simple uncompounded
ingredients, may discover that not every one of them suits every patient’s nature. He
will then use epilogismos to conclude that he should mix together as many simple
remedies for the affliction as he can in order that each nature will be supplied with
the remedy that suits it.
It is not easy to see how a human being can lead a life, let alone practice an art,
without doing anything of the kind that we or the Empiricists who countenance
epilogismos would describe as inferring, drawing a conclusion, and the like. Yet
it seems that radically “anti-rational” Empiricists refused to agree to the second
degree of rational commitment and tried to explain as an exercise of the faculty
of memory what others would call reasoning. According to this way of thinking,
repeated observation of occurrences that coincide, precede, or follow one another,
supplemented by history based on the experience of other people, forms dispositions
in us to be reminded of one thing by another. When one sees smoke, on this account,
one does not infer that there must be fire by appreciating the grounds that favor
this conclusion, let alone by apprehending relations of consequence and exclusion
imposed by the nature of fire. Rather, one is simply put in mind of fire. It would
not be wrong to borrow talk of custom from Hume and like-minded philosophers
and regard the Empiricists’ view as a variety of associationism. A contrast like this
one is the basis for the distinction between indicative and commemorative signs, to
which Sextus Empiricus devotes two lengthy discussions (PH II 97–133; AM VIII
141–299). The former signify that of which they are signs by their own nature and
constitution; the latter by reminding us or recalling to mind matters with which
they have been co-observed in past experience, as, for example, smoke signifies fire.
Though the matter remains controversial, it has been argued that the conception of
commemorative signification is due to the Empiricists.5
108 SKEPTICISM
We can trace the progress in the Empiricists’ attitudes toward reason through
the way in which they handled certain oppositions. As part of his discussion of the
transition to the similar in the Subfiguratio empirica, Galen compares two ways of
understanding the term “transition” (metabasis). It looks as if metabasis is being
presented as a genus whose two main species are inference or rational inference, on
the one hand, and the Empiricists’ alternative, on the other:
Rational transition derives knowledge indicatively from the nature of the matter;
Empirical transition follows matters that are known from experience, not because
it is plausible that like things are productive of like effects, or that they stand in
need of like measures or are similarly affected—it is not for this reason that [the
Empiricists] insist on using the transition, but because we have experienced like
things behaving in this way. (Subfig. emp. 70, 10 ff. Deichgräber)
Note that “rational,” and by implication “reason,” belong on the rationalist side
of the opposition between two varieties of transition. “Rational” also attaches to
the “consequence” (assecutio, akolouthia) that figures in Galen’s characterization of
indication (Subfig. emp. 44, 9; 64, 7 Deichgräber), and as we have seen consequence
together with conflict are the objects grasped by the special faculty of reason,
commitment to which was in Galen’s view the defining characteristic of medical
rationalism (Subfig. emp. 87, 4–7).
5 EMPIRICAL REASONING: EPILOGISMOS

A different picture emerges from the work that has the most to say about epilogismos,
On Empirical Medicine. Though we are fortunate to have it, apart from two short
fragments, it survives only in an Arabic translation of a Syriac translation of the
Greek, edited and translated into English in 1944 by Richard Walzer (reprinted in
Frede and Walzer 1985). My interpretation is based on Walzer’s English translation,
and the reader should be aware of the layers standing between the lost original and
the present author.
Galen first presents the anti-Empirical polemic due to Asclepiades mentioned
above and then an extended response from the Empiricists. In the part relevant to
our present purpose, the spokesman for Empiricism takes the opportunity afforded
by the rationalists’ charge that “[the Empiricists] have no method of drawing a
conclusion, with the help of which inferences are drawn from the visible to the
invisible” to expound “[his] opinion [that] there are two kinds of logoi: the one
called epilogismos, the other called analogismos” (On Empirical Medicine 131–
132). The extent to which the Empiricist is willing to take over the language of his
rationalist opponents is remarkable. Thus we find: “The conclusion from visible
to visible that is known as epilogismos,” which is contrasted with “the conclusion
from the visible to the invisible” (141); “the method of reasoning from visible to
visible” (149); “Epilogismos is an inference from common and universally used by
all mankind” (133); “the conclusion and logos known as epilogismos” (135, 139);
and “in affirming what is necessitated by something that is said or done, you will find
two kinds of affirmation, one after the manner of the logos known as epilogismos,
the other after the manner of the logos known as analogismos” (135–136). And there
is no missing the fact that the battle lines in the case made by Galen’s Empiricist
have been redrawn. The dispute is no longer simply between the partisans of reason
and those of memory and observation, but between advocates of two kinds or two
understandings of rational inference.
According to the Empiricists, epilogismos is confined to evident matters; it does
not, as analogismos purports to do, extend the reach of human knowledge beyond
the realm of observable items. Unlike analogismos, it is not the possession of a few
specialists but common to all humankind. And in sharp contrast to analogismos, it
yields results on which all can agree instead of giving rise to endless and apparently
irresolvable controversy (De sect. ingred. SM III 11, 8; Subfig. emp. 62, 24–31
Deichgräber; On Medical Experience 133–135, 140 Walzer). It covers what we call
inference based on empirical generalizations (cf. De sect. ingred. SM III 10, 23–24;
11, 9–10), but it also seems to cover the invention or thinking-up of new ideas that
I mentioned earlier in connection with the transition to the similar.
It is perfectly clear that a difference exists between each logos and conclusion
used by the dogmatists [i.e., rationalists] and the logos and conclusion known
as epilogismos, which is universally used by everybody, namely that epilogismos
seeks the guidance of visible things—and it is from this that it seeks confirmation
of its truth and rightness. (On medical experience 139, cf. 133–134 Walzer)
The specimen of epilogistical reasoning about the cephalicum discussed above may
furnish an example (MM X 163, 14 ff. Kühn = fr. 105 Deichgräber; cf. Comp.
Med. gen. XIII 366, 2 ff. Kühn = fr. 106 Deichgräber). Although it appeals to the
“natures” of the patients’ bodies, it does so in a completely general way: in effect
“the nature, whatever it is, that responds to an ingredient in the drug,” and the
upshot must of course be submitted to the test of experience.
Most controversial is the use of reason by the Empiricists in anti-rationalist
polemic. They argued plausibly that the conflicting results reached by the rationalists,
using their version of the rational method, undermined the claims they—the
rationalists—made for it. This and the attempts by some Empiricists to show where
the rationalists’ anti-Empirical arguments go wrong may invite the charge that the
Empiricists were more invested in reason than they allowed or than can be explained
as simple epilogismos.
Menodotus stands accused of yet another offense, however. According to Galen,
he contradicts himself:
Menodotus often introduces something third, in addition to memory and
perception; he calls this third thing “epilogism”; often though, he does not posit
anything in addition to memory except perception. (Subfig. emp. 87, 25–88, 1)
According to one attractive suggestion, in taking different and incompatible
positions at different times, Menodotus was acting as a skeptic (cf. Frede 1990: 248–
249). Academics like Carneades were known to argue for incompatible positions
about, for example, the question whether the wise person will have opinions, at
most one of which can have been their real view (cf. Cicero, Acad. II 67, 78).6
110 SKEPTICISM
By arguing in this way, they could show that their dogmatic opponents were not
entitled to certain conclusions even granted certain concessions. Following this
strategy, Menodotus could, for example, have argued that an art of medicine was
possible on the basis of experience even if, contrary to fact, they foreswore the use
of reason, including epilogismos.
A close look at the polemical context of this charge, leveled in the twelfth and
last chapter of the Subfiguratio, suggests another possibility, however. The favorable
attitude toward Empiricism on display in most of the Subfiguratio is conditional:
provided they stick to what—in Galen’s view—they did best. In chapter XI, to
which we shall return because it faults Menodotus for failing to live up to the
ideal of Pyrrho, Galen says that the account so far covers Empirical medicine that
is sufficient for the good treatment of patients. He turns now to the criticism of
superfluous additions made by leaders of the Empirical school (Subfig. emp. 80,
16–81, 20 Deichgräber). In Galen’s view, the ideal Empiricist is a man of deeds and
very few words, who makes his mark by curing patients, not by engaging in disputes.
Chapter XII, which belongs to the same polemic, rejects the view, apparently
held by some Empiricists, that medical experience includes parts that resolve the
difficulties raised against Empiricism by its opponents and furnish arguments
against the opponents’ own positions (Subfig. emp. 86, 15–23 Deichgräber). Galen
maintains that people who claim to base an art solely on evident perception and
memory have no business attempting to show what is wrong with the rationalists’
case against Empiricism, let alone arguing against their theories. To do this, one
would have to accept that human beings have a faculty capable of evaluating and
judging relations of consequence and exclusion, that is, the rational faculty as Galen
and the rationalists understood it (Subfig. emp. 87, 4–12; cf. 89, 5–23, 44, 9–10,
63, 23–26 Deichgräber). To accept this, however, places one under an obligation
to develop the power by the study of logic according to Galen. Only then will one
have earned the right to participate in the debates from which he thinks Empiricists
should otherwise remain aloof.
This is the occasion for Galen to mention the Empiricists who found a place for
reason. “[Heraclides] knows that there is [such a faculty] and is seen to use it on many
occasions” (Subfig. emp. 87, 19–21). The problem is that he never did what he needed
to do to develop it. For this reason, “he is as much worse a doctor than Hippocrates
as he is a better one than Menodotus” (Subfig. emp. 87, 22–25), who is now accused
of contradicting himself. Galen, then, is looking for the thin end of the wedge,
evidence that some of the Empiricists were committed to the existence of a faculty of
reason, which he can then use to fault them for failing to develop this faculty as they
would have to have done to make their arguments against the rationalists worthy of
serious consideration. His evidence is Heraclides’s self-conscious use of reason and
Menodotus’s acceptance of a form of reasoning that he styled “epilogismos.”
The argument would surely have seemed tendentious to the Empiricists,
however. They were at pains to distinguish the form of reasoning they accepted
from the rationalists’ alternative. There is no reason to think that they would have
agreed that the use of epilogismos committed them to accepting the existence of
a special rational faculty answering to the description supplied by Galen or to
acknowledging the attendant obligation to study logic. It may be doubted, then,

whether they viewed the adoption of epilogismos as a sharp break with the past.
The interpretation of Serapion’s views about the transition to the similar shows
that, in the manner of other long-established schools, many of them wanted to
see themselves as in fundamental agreement with their distinguished predecessors.
Despite his wholehearted endorsement of epilogismos, the Empirical spokesman
in On Medical Experience describes Empiricists as proponents of memory and
observation and recognizes a sense of “reason” in which it is the exclusive possession
of the rationalists (138, 154 Walzer). If this proposal is on the right lines, Menodotus
will have regarded the acceptance of epilogismos as compatible with a commitment
to memory and observation as the sole ultimate bases of knowledge. Epilogismos,
as he conceived of it, is reasoning based on and not departing from what we know
through these powers. The innovation, if it amounted to that, will simply have been
to describe better what the Empiricists had been doing all along.
6 WAS MENODOTUS AN
INCONSISTENT PYRRHONIST?
In chapter XI of the Subfiguratio, Galen levels another charge of inconsistency. He
tells us that as the Pyrrhonist is in the whole of life so the Empiricist is in the medical
sphere; he informs us that Menodotus had held up Pyrrho as a model for his fellow
Empiricists to follow; and he faults him for failing to live up to the ideal that Pyrrho
represented. This is powerful evidence that Menodotus, together with at least some
other Empiricists, was a Pyrrhonist and regarded himself as one. Galen’s criticism of
Menodotus contains two charges that are not clearly distinguished. The first, that
unlike Pyrrho, Menodotus is a man of many words (multiloquus) and that, far from
avoiding disputation, he is constantly engaged in it, is easily answered. The evidence
furnished by Sextus Empiricus about the Pyrrhonian tradition to which both he and
Menodotus belonged shows that disputation was the life blood of Pyrrhonism and
that the Pyrrhonists were aware that they likely differed from Pyrrho in important
ways while they continued to find aspects of his life exemplary (see PH I 7).
According to Galen, Menodotus’s too many words were also often vituperative
and buffoonish, and worst of all—this is the second main charge—he was guilty of
making rash and dogmatic assertions. Thus according to Galen, in his writings against
Asclepiades of Bithynia, Menodotus declares that he knows certainly that all of his
doctrines are false—despite the fact that “in many of his writings, he has demanded
that one should approach everything non-evident as if perhaps it is true and perhaps
it is not” (Subfig. emp. 84, 18–31)—the true Pyrrhonian attitude. If Menodotus
did really claim to know for certain that Asclepiades’s theories were false, he was
guilty of the second inconsistency of which Galen accuses him. Galen’s charge bears
some resemblance to Sextus’s complaints about the unskeptical firmness with which
Empiricists speak about non-evident matters, with the difference that Sextus thinks
Empiricists are too ready to pronounce non-evident matters to be inapprehensible
while, if Galen is right, Menodotus claimed to know that one theory was false.
112 SKEPTICISM
Nonetheless, this shows not that Menodotus was not a Pyrrhonist, but at worst
that he was an inconsistent one. We do not have the evidence on which Galen
based his charge, but in view of his penchant for uncharitable interpretations of his
opponents, it is at least possible that it might support a different and lesser charge:
in his impatience with Asclepiades, Menodotus so far forgot himself as to say, or
to say things that imply, that he knew Asclepiades was all wrong. Though Galen
may have been too hard on Menodotus, that there could be a tension between
Empiricism and Pyrrhonism is undeniable. The Empirical position, stated simply
and without qualifications, is dogmatic. The denial that knowledge of non-evident
matters is possible, of which Sextus complains, can be combined with the equally
unskeptical conviction that knowledge of evident matters is. According to Galen,
Empiricists hold that the fact that “each thing has appeared as it is confirms the
correctness of those who are right and refutes those who are not [about evident
matters by contrast with non-evident ones]” (De sect. ingred. SM III 1120–1112,
4; cf. On Medical Experience 103, 133–135 Walzer). Pyrrhonists acknowledge that
things appear a certain way, to be sure, but of course “investigate whether the real
object is such as it appears” and cannot “be sure regarding the phenomena whether
they are as they appear” (PH I 19; AM VIII 142, cf. 357).
If some Pyrrhonists embraced Empiricism, they will not have endorsed the
conviction about the phenomena or appearances that Galen ascribes to the
Empiricists. This will not have prevented them from following the same appearances
as the Empiricists. They will have done so with the proviso that things may not be as
they appear, however. Such a person would be a skeptical Empiricist, differing from
his dogmatic colleagues not necessarily by following appearances with a different
content, but in the character of his attitude to them. Empiricism will most likely
have recommended itself to a Pyrrhonist of this kind because of the way in which
it purported to be nothing more than a refinement of everyday observance and
so a form of artistic teaching that had a place in the Pyrrhonists’ criterion. Even
if Menodotus is guilty of the lapses of which Galen accuses him, he likely was, or
strove to be, a Pyrrhonian Empiricist of this kind.7
NOTES
1. If a corrupt passage in Sextus (PH I 222) contains a reference to Menodotus, he may
have taken an interest in the much debated question whether Plato was a skeptic. See
Spinelli (2000), who favors this view, and Perilli (2005), who rejects it.
2. Thought by scholars to be the same as the “Medical Commentaries” cited at AM VII
202.
3. Consider, however, AM VIII 327–328, where the Empiricists are pitted against
dogmatic philosophers and rationalist physicians as parties to a dispute—about the
existence of demonstration—regarding which skeptics suspend judgment.
4. Latin text in Deichgräber (1965); English translation in Frede and Walzer (1985).
5. See Allen (2001: 106–139), with references to the literature on page 107 n. 19. Ebert
(1987) argues for an entirely different origin of the theory of signs discussed in Sextus.
6. On Carneades, see Harald Thorsrud’s chapter in this book.

7. Medical Empiricism is discussed in standard histories of skepticism: Brochard (1923:
book IV, ch. 1), Robin (1944: part IV, ch. 1), and Hankinson (1995: ch. XIII).
Menodotus is the subject of specialized monographs: Favier (1906) and Perilli (2004).
REFERENCES
Allen, James. 2001. Inference from Signs: Ancient Debates about the Nature of Evidence.
Allen, James. 2010. “Pyrrhonism and Medicine.” In R. Bett (ed.), The Cambridge
Brochard, Pierre. 1923. Les sceptiques grecs, 2nd edn. Paris: Vrin.
Capelle, Wilhelm. 1931. s.v. “Menodotos (2).” RE XXIX, cols. 901–916.
Deichgräber, Karl. 1965. Die griechische Empirikerschule: Sammlung der Fragmente und
Darstellung der Lehre. 2nd edn. Berlin and Zürich: Weidmann. (First published in
1930).
Ebert, Theodor. 1987. “The Origin of the Stoic Theory of Signs in Sextus Empiricus.”
Oxford Studies in Ancient Philosophy 5: 83–126.
Favier, Albert. 1906. Un médecin Grec du II siècle ap. J.-C. précurseur de la méthode
expérimentale moderne: Ménodote de Nicomédie. Paris: Jules Rousset.
Frede, Michael and Richard Walzer. 1985. Galen: Three Treatises on the Nature of Science.
Indianapolis: Hackett.
Frede, Michael. 1990. “An Empiricist View of Knowledge: Memorism.” In S. Everson
(ed.), Epistemology, 225–250. Cambridge: Cambridge University Press.
Hankinson, Robert James. 1995. The Sceptics. London and New York: Routledge.
Perilli, Lorenzo. 2004. Menodoto di Nicomedia: Contributo a una storia galeniana della
medicina empirica. Leipzig: Saur.
Perilli, Lorenzo. 2005. “Quantum coniectare (non) licet: Menodotus between Sextus
Empiricus (P 1.222) and Diogenes Laertius (9.116),” Mnemosyne 58: 286–293.
Robin, Léon. 1944. Pyrrhon et le scepticisme grec. Paris: Presses Universitaires de France.
Spinelli, Emidio. 2000. “Sextus Empiricus, the Neighbouring Philosophies and the
Skeptical Tradition (again on Pyr. 220–225).” In J. Sihvola (ed.), Ancient Scepticism
and the Sceptical tradition. Acta Philosophica Fennica 66: 36–61.
Spinelli, Emidio. 2010. “Pyrrhonism and the Specialized Sciences.” In R. Bett (ed.), The
Cambridge Companion to Ancient Scepticism, 249–264. Cambridge: Cambridge
University Press.
von Staden, Heinrich. 1982. “Hairesis and Heresy: The Case of haireseis iatrikai.” In
B. F. Meyer and E. P. Sanders (eds.) Jewish and Christian Self-Definition, vol. III: Self-
definition in the Graeco-Roman World. 76–100, 199–206. London: Fortress Press.
CHAPTER NINE
Middle Platonism and

Skepticism: Plutarch
and Favorinus
CARLOS LÉVY
1 INTRODUCTION
The concept of skepticism has at least two different meanings. The first is general
(M1), the second is more restricted (M2). A tendency to doubt the things commonly
considered as clear was present in many ancient philosophical doctrines (M1). That
was the reason why the neo-Academics were so eager to show that their absolute
ἐποχή, far from being an invention of their own, was the result of a long tradition
that included thinkers as different as Democritus, Xenophanes, and Socrates (Cicero,
Acad. II 72–74). But generally the term “skepticism” is used in studies on ancient
philosophy to designate only two schools (M2), the New Academy, inaugurated by
Arcesilaus, and neo-Pyrrhonism, which was the creation of Aenesidemus in the first
century BCE.1 One of the reasons for this duality is that skepticism as a technical
concept did not appear before Aenesidemus. For example, Arcesilaus, scholarch
of the Academy in the third century BCE, never used this concept. He defined his
own philosophical identity as that of an Academic practicing a radical suspension of
judgment (ἐποχὴ περὶ πάντων), not as that of a skeptic belonging to the Academy.
Inside these two schools, the problem of continuity versus rupture was a permanent
one, with a complex net of strategies to differentiate themselves from each other, as
we will see in the case of Plutarch. Academics seem to have ignored contemptuously
Aenesidemus, and two centuries later, Sextus Empiricus, defining himself as a skeptic
and Pyrrhonist, worked hard in his Outlines of Pyrrhonism to demonstrate that the
neo-Academics were not authentic skeptics (see Bonazzi 2011).
The problem I will tackle in this chapter is how to understand the presence
of skeptical terms and themes in Middle Platonism, a rich and confused period
between the Hellenistic systems and Neo-Platonism, focusing my analysis on
Plutarch and Favorinus.2
MIDDLE PLATONISM AND SKEPTICISM: PLUTARCH AND FAVORINUS 115
2 THE MIDDLE PLATONIC BACKGROUND

2.1 Problems of Method
There are two possible interpretations of the presence of skeptical terms and
themes in Middle Platonism. First, one may consider skepticism and dogmatic
Middle Platonism as radically different. On this interpretation, the skeptical themes
are only traces, without any essential meaning, of the previous period. Second,
Platonism may be deemed a paradoxical product of the evolution of skepticism, as
the response to some of its aporias; and so there would be no deep rupture between
the two periods.
When examining the history of the New Academy, one realizes that skepticism
seems to have become more and more moderate from Arcesilaus to Philo of Larissa,
who, while continuing to fight Stoic dogmatism, asserted that things were knowable
by nature, but without saying what the criterion of such a knowledge could do (see
Brittain 2001: 138–145). Philo opened the door to an interpretation of Platonic
thought that would not be entirely determined by the traditional Academic hostility
toward Stoicism. The best example of this continuity can be found in the Commentary
on the Theaetetus, the date of which is still at the center of many discussions.3 The
author criticizes the Stoic idea that justice has a natural foundation. In order to
demonstrate the falsity of this doctrine, he uses an example found in the famous
antilogy of the third book of Cicero’s De re publica. In his dialectic refutation of
the Stoic ethical immanentism, Carneades said that if a ship was wrecked and two
shipwrecked sailors had to compete for holding on to a board, the struggle for life
would be stronger than any ethical concern (Cicero, De rep. 3. 26). The solution to
this Stoic aporia was, in the opinion of the author of the Commentary, the return
to a dogmatic version of Platonism. It was based on the famous so-called digression
of the Theaetetus 176B, namely, the view that justice could not be founded on the
idea that the world is the city of gods and mankind, but in the φυγή, the flight away
from the world of sensation. The difference between the position of Carneades
and that of author of the Commentary is clear. For the former, at least from what
we can deduce from the complex Ciceronian metaphrasis, the essential thing was
to demonstrate that the premises of Stoic ethics lead to the exact opposite of what
they pretended to prove, but he did not offer any alternative solution of his own.
The Theaetetus, and more precisely the concept of ὁμοίωσις κατὰ τὸ δύνατον,
indicates the way to avoid the difficulties inherent in naturalistic ethics. It is likely
that the Theaetetus, a dialogue certainly frequently read in the New Academy that
was aporetic and contained a digression with an assertive tone, was the best means
to try a kind of conciliation between Carneadean skepticism and a much more
dogmatic form of Platonism.
2.2 A Pioneer: Transcendence and Skepticism in Philo of Alexandria

Plutarch was not the first philosopher of the Middle Platonic period to have
interpreted skepticism as both an access road to transcendence and a consequence
of it. In the first decades of the first century CE, Philo of Alexandria, who certainly
116 SKEPTICISM
was not a philosopher in the technical sense of the word but had an excellent
knowledge of the philosophical milieu and texts, opened this way. He was the
first, at least in our testimonies, to have expounded the tropes of Aenesidemus,
not in order to affirm his own skeptical affiliation, but as a path to his rather
fideistic theology.
In Philo’s treatises, there are some quite interesting allusions to Academics and
skeptics of his time, but their interpretation is complex, both because the allusion
that seems to have been the most explicit one has reached us in an Armenian
translation and because Philo has a tendency to transform the skeptics in an
allegory of the resurgence of Sophistic (Lévy 2008: 105–110).4 But we find in De
fuga et inventione (136), in the exegesis of the sacrifice of Isaac, an articulation
between the concept of ἐποχή and the almightiness of God, which is not only
interesting from the perspective of Biblical studies, but as the sign of a change
in the intellectual climate, making it possible to create a strong link between
doubt and transcendence. Above all, Philo’s De ebrietate has been considered by
prominent scholars like von Arnim and Bréhier as being probably the closest to
the classification elaborated by Aenesidemus (von Arnim 1888: 79; Bréhier 1907:
210). More recent scholars expressed a different opinion about Philo’s testimony,
but without convincing evidence.5 In any case, it is important to notice that in
about 30 CE, the approximate date of the redaction of most Philo’s treatises, a
quite learned man with a real philosophic formation was able to bring together
Academic concepts (ἐποχή, πιθανόν, εὔλογον), neo-Pyrrhonian modes, and, last
but not least, the motto of Middle Platonic thought, the Platonic recommendation
to become “like God so far as it is possible” (Theaet. 176b). The conditions under
which this mixture took place are far from clear, but it is important that some
decades before Plutarch the structures of the intellectual environment in which he
was going to write his own philosophical treatises were already present.
3 PLUTARCH
3a General Remarks
John Glucker (1978: 257–279) demonstrated, rather convincingly in my opinion,
that the claim that Plutarch of Chaeronea (ca. 45–120 CE) studied in an organized
Academy lacks evidence. The passages in which he mentions the Academy do not allow
us to say firmly that, after having been Ammonius’s pupil, he went to the Platonic
school, where he would have received all his knowledge about Academic skepticism.
One of the most interesting elements about Plutarch’s relation to skepticism is that
he had a precise knowledge not only of the tradition represented by Arcesilaus,
Carneades, and Clitomachus, but also of the neo-Pyrrhonian tradition. He wrote
books in the anti-dogmatic, and especially anti-Stoic tradition of the New Academy,
such as On the Contradictions of the Stoics and Against the Stoics on Common
Notions, but he was also the author of On the difference between the Pyrrhonists
and the Academics and of On the Pyrrhonian Ten Tropes. Plutarch’s literary and
philosophical production is enormous, and any attempt to give a complete account
of it in a few pages is impossible. However, it seems that Plutarch’s relation to

skepticism has different aspects that it is necessary to present separately.
3b The Critique of Stoicism

The first aspect is his anti-Stoicism, or more exactly his hostility toward the Hellenistic
dogmatisms, since he did not spare Epicureanism, against which he wrote many
treatises (see Hershbell 1992). However, since the refutation of Stoicism was much
more fundamental in the history of the New Academy than that of Epicureanism,
I will focus on this aspect of Plutarch’s thought. In fact, his relation to Stoicism
is quite complex, as Daniel Babut (1967) showed, but that does not prevent him
from carrying on and extending the long tradition of criticism of the Stoic doctrine
elaborated by Arcesilaus and his successors in the New Academy.
In the Catalogue of Lamprias, an exhaustive list of the titles of Plutarch’s books,
eight explicitly anti-Stoic treatises are mentioned, three of which have reached us,
the third one being a short and polemic treatise in which Stoics are accused of saying
things more absurd than those said by the poets. However, many other treatises are
attacks against particular aspects of the Stoic doctrine, even if neither the name of
this school nor any specific Stoic concept are present in the title. As an example, the
treatise “Is rhetoric a virtue?” was certainly anti-Stoic, since these philosophers were
those who had defined true rhetoric as a component of wisdom.
The two main treatises, On the Contradictions of the Stoics and Against the
Stoics on Common Notions, though being rich in references and often explicit
about sources, are full of mysteries. It is probable that these anti-Stoic treatises were
written more or less at the same time as the anti-Epicurean ones, when Plutarch
was still influenced by his Platonic formation under the direction of Ammonius.
Babut’s argument that these treatises are full of too much philosophical information
to have been written by a young scholar is not entirely convincing, given that
Plutarch could have used material carefully elaborated in the New Academy.6 Von
Arnim remarked that no Stoic philosopher later than Antipater, who was more or
less contemporary with Carneades, is quoted in the On the Contradictions of the
Stoics. He deduced from this terminus ante quem that Plutarch used a book of
the prolific Clitomachus, the successor of Carneades at the head of the Academy,
as a source for his own treatise (von Arnim 1903: XI–XIV). In my opinion, it is
quite probable that this neo-Academic material was the basis of the refutation by
Plutarch of his main adversary, Chrysippus. However, the idea of a single source
must be rejected, since his deep knowledge of the Stoa allowed him to use his
own material, at the risk of creating a great deal of disorder in the composition
of these treatises.
Plutarch remains in the dialectical tradition of Carneades, about whom the
testimonies are fewer and often more difficult to understand.7 Indeed, in many
places Carneades says that he does not want to bring a dogmatic response to the
philosophical problems tackled by the Stoics, just to highlight the contradictions
inherent in their argumentation. Actually, his purpose was somewhat more perverse
than he declares, since Stoicism argued that it was impossible to remove one element
118 SKEPTICISM
(Cato says one letter) without destroying the whole system (Cicero, Fin. 3.74).
What Jacques Brunschwig (1995) called “the conjunctive model” was both the
structure and the essence of the doctrine. To highlight contradictions was not only
a dialectical ploy inside the Stoic system, it aimed to demonstrate that Stoicism had
no legitimacy to give satisfactory responses to any philosophical problem.
In the Contradictions, Plato is never presented as the holder of the truth. At
the same time, Plutarch wants to show how unfair was Chrysippus’s criticism of
many aspects of Plato’s dialogues. This βοήθεια, this effort to protect Plato against
inequitable attacks was already present in the ancient Academy, though there were
many differences between Speusippus or Xenocrates and the founder of their
school. There is no conclusive part in the Contradictions. The last paragraphs are a
discussion of Chrysippus’s dogma that it is impossible to act and to follow impulse
without assenting to it, a problem involving concepts like liberty, God, and the wise,
but without any attempt to go outside of the anti-Stoic debate.
The same polemical context characterizes the Common Notions. Though Stoic
texts can seem nowadays quite difficult to understand, even for a learned reader,
Zeno or Chrysippus did not want to write abstruse treatises. On the contrary, they
intended to build their system on common notions, formed from representations of
reality that are the same for everyone. For example, Plutarch says that it is contrary to
common sense to have two τέλη, two supreme goals in life (Common Notions 1070F).
That would be true if the Stoics had really asserted the existence of this duality. In
fact, by saying that the τέλος was to be found in doing one’s best for reaching the
goods that conform to nature, they tried to avoid the intrusion of Fortune in the field
of ethics (Soreth 1968). Plutarch resumes here in a rather confused way Carneades’s
polemics against Diogenes of Babylon’s and Antipater’s teleological theories, which
tried to render precise Zeno’s definition of the τέλος, namely, to live in harmony.
Indeed, if we only knew the On the Contradictions of the Stoics and the Against
the Stoics on Common Notions, we would consider Plutarch as a philosopher in the
tradition of the New Academy, differing from Carneades by his ability to openly
defend Plato, without however giving a dogmatic interpretation of his thought. This
could lead us to identify him as a thinker in the tradition of Philo of Larissa’s Roman
Books. But things are more complex. Plutarch is not only the defender of Academic
skepticism, someone who highlights the contradictions of the Stoics in their attacks
against Plato, but also someone with positive Platonic beliefs.
3c Plutarch and Plato

What Jan Opsomer (1998: 127) called the “interplay between Academic and
Platonic themes in Plutarch” is a very complex problem. Plutarch wrote a treatise
On the Unity of the Academy from Plato Onwards, a theme developed by Philo of
Larissa at the end of his life (Boys-Stones 1997: 55 n. 2), but it did not reach us. If
we knew with some precision the chronology of the Moralia, it would perhaps be
easier to interpret this articulation as the result of an intellectual evolution.
Plutarch tackled the controversial question of the origin of Academic skepticism
in the Adversus Colotem, his reply to a highly polemical work written by a disciple
of Epicurus against both Socrates and Arcesilaus. One of the main elements of
Plutarch’s argumentation is that, far from being a foolish invention of Arcesilaus,
Academic skepticism was continuous with the philosophy of doubt practiced by
prestigious thinkers such as Socrates, Plato, Parmenides, and Heraclitus. Actually,
Plutarch says to be grateful to Colotes and to everyone showing, though in a
hostile way, that the Academic outlook came to Arcesilaus as an ancient tradition.
Plutarch’s position is not fundamentally different from the one we find in Cicero,
when he evokes at length the quite honorable genealogy of Academic skepticism.8
Colotes charged Socrates with being a Sophist who criticized the senses while acting
in contradiction with this criticism, especially when he said that he was seeking to
discover what human nature is (Plutarch, Adv. Col. 1118C). Plutarch’s reply to the
Epicurean attacks against Socrates and Plato does not refute the interpretation of
them as skeptics. One of the main arguments developed at length in the treatise and
attributed to Arcesilaus is that which aims to demonstrate that assent is not necessary
for action. The interpretation of this argument is too controversial to be discussed
here. Was it defended by Arcesilaus in propria persona or only dialectically? Be that
as it may, Plutarch seems to consider it a fully convincing argument.
The real problem is to determine how he, the author of a treatise aiming to
demonstrate the unity of the Academy throughout its complex history, could be both
a Platonist and a skeptic. Plutarch, who actually defined himself as an Academic,
never as a Platonist, knew about the many differences between philosophers who all
referred to Plato as the inspirer of their thought. What could be for him the logic
of this unity? If one accepts the concept of “metaphysical skepticism,” coined by
Pierluigi Donini, it is possible to say that Plutarch’s position presents three main
aspects (see Donini 1986 and Bonazzi 2016). The first is the refutation of all the
doctrines proclaiming the truth of all sensations (Epicureanism) or of most of them
(Stoicism). The second is that truth must be searched for in the principles identified
with God, a frequent feature of Middle Platonism. The third attributes the truth only
to God, so that men are reduced to looking for it carefully, without ever yielding to
the temptation of asserting, as the Stoics did, that a human sophia was exactly the
same as that of God, and even greater than the divine perfection, since conquered by
free will. Dualism and the perfection of God are for Plutarch principles about which
doubt is not allowed. That is one of the reasons why Babut suggests that his thought
can be defined as a kind of fideism, speaking of “a conscious tension between the
need of rationality and the divine transcendence whose manifestations go beyond
the capacity to understand of our reason” (Babut 2007: 93). This is very close to
the conclusion of the On the E at Delphi, where Ammonius shows that the eternity
of God and human weakness, despite seeming to be contradictory, are in reality
complementary (Plutarch, De E 393BC).
An alternative to Babut’s fideist interpretation is Boys-Stones’s, for whom
“Plutarch felt himself able to draw positive conclusions from his skepticism, and
even his most Academic works—as notably the de Stoic. rep.—should still be read
as vehicles for an expression of his Platonism” (Boys-Stones 1997: 55). In any case,
scholarship now seems to be unanimous in having a unitarian interpretation of
Plutarch’s thought, even if the formulations of this unity are different.
120 SKEPTICISM
3.4 Plutarch and Pyrrhonism

The third aspect of Plutarch’s relation to skepticism is his attitude toward Pyrrhonism,
a subject about which the loss of his treatise on the differences between Pyrrhonists
and Academics leads us to a quasi-silence. It can seem quite strange that he could
have written his anti-Stoic and anti-Epicurean treatises without any reference to
neo-Pyrrhonism, though this doctrine was much closer to him from a chronological
point of view. This only means that the strength of a philosophical affiliation was
so compelling that discussions that were almost four centuries old could seem to
Plutarch of a greater topicality than a skeptical doctrine that appeared less than two
centuries earlier. He did not ignore Pyrrhonism, but he was not interested in it; he
seems to want to use it as an instrument of his own reflection. In his extant writings
at least, there is no allusion to the modes elaborated by Aenesidemus, though they
were certainly extensively diffused at the beginning of the second century CE.
The only reference to a specifically major Pyrrhonian concept—since ἐποχή was
common to the Academics and the Pyrrhonists—namely, that of ἀφασία, is in the
Adversus Colotem, without any precise reference to Pyrrhonism (Plutarch, Adv. Col.
1123C; see Ioppolo 1994, 2004; Bonazzi 2016). Plutarch says that the Epicurean
claim that all sensations are true leads to the claim that sensation has the same
degree of reliability as that ascribed to dreams and illusions. Here he establishes an
implicit link between Epicureanism and Pyrrhonism, considering the later as a sort
of consequence of the former. It seems that for an Academic philosopher, as he
perceived himself, this intra-Hellenistic link between two forms of philosophy that
rejected all forms of transcendence was of no interest.
4 FAVORINUS AND THE ACADEMY

4a A Strange Personality
Favorinus (ca. 80–160 CE) was not unknown to Plutarch, who dedicated the
treatise on the principle of cold to him (Plutarch, De Prim. Frig. 945F–946A). He
wrote a book on the Academic method, the title of which was Plutarch. He seems
to have been an interesting, contradictory, and somewhat mysterious person. He
said himself that there were three contradictions in his life (Philostratus, Vit. Soph.,
I, 8=test. 6 Barigazzi): (i) despite having been born in Arles, in the South of Gaul,
he lived in a Greek way; (ii) despite being an eunuch, actually an hermaphrodite, he
was accused of adultery; and (iii) despite having been hostile toward the emperor
Hadrian, he was still alive.
The testimonies about his philosophical identity are of two kinds. First, the Suda
says that he was a very learned man (πολυμαθής), but more interested in rhetoric
than in philosophy (Suda s.v. Favorinus = Barigazzi 1). It must be added that
Eusebius evokes a polemic between him and Polemo the rhetor (Chronic. Can. 166
Schöne = Barigazzi 2). Second, Philostratus presents him as a philosopher who was
considered a sophist because he was very eloquent (Vit. Soph. 1.8.1 = Barigazzi 2).
He never says that Favorinus was a Pyrrhonist but affirms that his best writings were
his Pyrrhonian speeches, in which, despite saying that Pyrrhonists were ephectics, he
allowed them to assume the functions of judges (Vit. Soph. 1.8.6). For Lucian, too,
he was a philosopher, if Favorinus is the Celtic eunuch belonging to the Academic
tradition who is mentioned in the Eunuch (Eun. 7 = Barigazzi 4). Gellius, who was
his disciple, mentions him many times and describes him as a philosopher, without
giving any precision on his philosophical affiliation. However, in a passage of the
Noctes Atticae (20, 1, 10), Favorinus presents himself as belonging to a sect that
preferred research over decision (inquirere potius quam decernere), a motto quite
close to the definition of the Academic attitude given by Cicero.
4b Favorinus as Platonist
Favorinus was quite interested in Plato. He wrote Memorabilia, a book on the history
of philosophy in which evidently Plato had a great role, a treatise on Plato (Περὶ
Πλάτωνος) and another one on Forms (Περὶ ἰδεῶν), which some scholars think
was a writing about rhetoric, but which could as well have been a philosophical
writing. If that was the case, Barigazzi’s (1966: 171) thesis that he had criticized
idealism from a skeptical point of view is merely speculative. There is no evidence
that, despite being a friend of Plutarch, he thought that it was impossible to find any
kind of conciliation between Platonism and skepticism.
The most extensive account of his philosophical position is given by Galen, who
says that like other “recent Academics” he was sometimes so fiercely ephectic that he
said that even a representation (φαντασία) of the sun could not be said to be certain,
and sometimes so dogmatic that he transmitted doctrines to his disciples without
having a critical approach to the criteria of knowledge (Opt. Doct. 40 = Barigazzi
28–31). Galen also says that he changed his mind between his writings, since he
said in his Plutarch that something could be known with certainty, while in the
Alcibiades he asserted the contrary. But if we read carefully what Galen wrote, we
realize that, in fact, Favorinus always spoke in a prudent, that is, probabilistic way.
He never went beyond what seemed probable to him, practicing a kind of disputatio
in utramque partem (antilogy) between the possibility and the impossibility of
knowledge.9
Despite being a probabilist, Favorinus had a special interest in Pyrrhonism, since
he wrote a treatise in ten books about the Pyrrhonian tropes, devoting one book
to each of them.10 In front of so many philosophical attitudes, Eugenio Amato
(2005) has defined Favorinus as a brilliant eclectic. I am not sure that was the
case. Actually, Favorinus is in my opinion an extremist of what could be named
“Platonocentrism,” that is to say, the Academics’ tendency to think that their
school was not a school among others, but the source of philosophy as a whole.
In the famous excursus of book 3 of the De oratore, Cicero affirms that Socrates
and Plato were at the origin of most philosophical schools: “Many people virtually
took their origin from Socrates, since different followers seized upon different
aspects of his discussions, which had been varied and diverse and had branched
out in all directions” (De or. 3.61, transl. May and Wisse). It is likely that for
Favorinus to be close to Pyrrhonism was not to adopt a new doctrine but to display
122 SKEPTICISM
an interest in what he thought to be a variant of Academic skepticism, itself a

variant of Platonism. This genetic conception of philosophy is perhaps the key to
what otherwise could appear as a patchwork.
5 CONCLUSION
Is the problem of the relation between skepticism and Platonism in the Middle
Platonist period artificial? What seems strange to us was considered normal by many
prominent philosophers of that time. From a methodological point of view, we are
not bound to study ancient authors with concepts of their time. But at the same
time, we cannot ignore that some of the concepts we use, and especially that of
Middle Platonism, did not exist at their time. To analyze concepts is of course an
essential task of historians of philosophy, but one should realize that these concepts
were elements of dynamic processes of identification, implying quite a number of
factors that are not all explicitly conceptualized. In ancient philosophy, as Foucault
and Hadot argued, there was no clear separation between thought and life. What
the subject of this chapter has shown us is that the life of traditions, often so
disconcerting, is a topic at least as fascinating as the static analysis of the elements
that compose them.
NOTES
1. On the outlooks of Arcesilaus and Aenesidemus, see Ioppolo’s and Castagnoli’s
chapters in this book. On the controversial question of Aenesidemus’s dating, see
Decleva Caizzi (1992).
2. Dillon (1977) chronologically defines this period as the one between 80 BCE and
220 CE. Other delimitations are possible, since such a dating includes Antiochus of
Ascalon, which is controversial. On this, see Sedley (2012).
3. On this point, see Tarrant (1985) and Bonazzi (2003: 79–212). It seems difficult to
adopt for the Commentary a date later than the first century BCE.
4. The text that has reached us in an Armenian version is Quaestiones in
Genesim 3.33.
5. See Striker (1983: 97), who considers it likely that Philo was a contemporary of
Aenesidemus, a quite improbable hypothesis in my opinion. The absence of “the
skeptical framework” in Philo’s exposition of the modes is not a convincing
argument, since we have no proof that Aenesidemus’s framework of the tropes was
the same as the one found in Sextus.
6. See Babut (2004: 18), for whom the most probable date of composition is ca.
80 CE.
7. On Carneades, see Thorsrud’s chapter in this book.
8. The main difference between Cicero and Plutarch is that the former does not
mention Heraclitus.
9. On Favorinus’s use of antilogy, see Amato and Julien (2005: 180–181).
10. On Favorinus and the Pyrrhonists, see Ioppolo (1993).
REFERENCES
Amato, Eugenio. 2005. “L’éclectique.” In E. Amato and Y. Julien (eds.), Favorinos d’Arles:
Œuvres, Tome 1. Introduction générale. Témoignages. Discours aux Corinthiens sur la
Fortune, 167–176. Paris: Les Belles Lettres.
Amato, Eugenio and Yvette Julien. 2005. Favorinos d’Arles: Œuvres, Tome 1. Introduction
générale. Témoignages. Discours aux Corinthiens sur la Fortune. Paris: Les Belles Lettres.
Babut, Daniel. 1967. Plutarque et le stoïcisme. Paris: PUF.
Babut, Daniel. 2002. Plutarque: Œuvres morales. Traité 72: Sur les notions communes
contre les Stoïciens. Paris: Les Belles Lettres.
Babut, Daniel. 2004. Plutarque: Œuvres morales. Traité 70: Sur les contradictions des
Stoïciens. Paris: Les Belles Lettres.
Babut, Daniel. 2007. “L’unité de l’Académie selon Plutarque. Notes en marge d’un débat ancien
et toujours actuel.” In M. Bonazzi, C. Lévy, and C. Steel (eds.), A Platonic Pythagoras:
Platonism and Pythagoreanism in the Imperial Age, 63–98. Turnhout: Brepols.
Barigazzi, Adelmo. 1966. Favorino d’Arelate. Opere. Introduzione, testo critico e commento.
Florence: Le Monnier.
Bonazzi, Mauro. 2003. Academici e Platonici. Il debattito antico sullo scetticismo di
Platone. Milano: LED.
Bonazzi, Mauro. 2008. “L’offerta di Plutarco. Teologia e filosofia nel De E apud Delphos
(capitoli 1–2),” Philologus 152: 205–211.
Bonazzi, Mauro. 2011. “Again on Sextus on Aenesidemus and Plato.” In D. Machuca (ed.),
New Essays on Ancient Pyrrhonism, 11–20. Leiden: Brill.
Bonazzi, Mauro. 2012. “Plutarch on the Difference between the Pyrrhonists and the
Academics,” Oxford Studies in Ancient Philosophy 43: 271–298.
Bonazzi, Mauro. 2016. “Le platonisme de Plutarque de Cheronée entre scepticisme,
théologie et métaphysique.” In A.-I. Bouton-Touboulic and C. Lévy (eds.). Scepticisme
et religion: Constantes et évolutions, de la philosophie hellénistique à la philosophie
médiévale, 75–88. Turnhout: Brepols.
Boys-Stones, George. 1997. “Thyrsus-Bearer of the Academy or Enthusiast for Plato?
Plutarch’s De Stoicorum repugnantiis.” In J. Mossman (ed.), Plutarch and His Intellectual
World, 41–58. London: Duckworth.
Bréhier, Émile. 1907. Les idées philosophiques et religieuses de Philon d’Alexandrie. Paris:
Picard.
University Press.
Brunschwig, Jacques. 1995. “Le modèle conjonctif.” In his Études sur les philosophies
hellénistiques, 161–187. Paris: PUF. (First published in 1978.)
Decleva Caizzi, Fernanda. 1992. “Aenesidemus and the Academy,” Classical Quarterly 42:
176–189.
Dillon, John. 1996. The Middle Platonists: 80 B.C. to A.D. 220. Ithaca: Cornell University
Press. (First published in 1977.)
Donini, Pierluigi. 1986. “Lo scetticismo accademico, Aristotele e l’unità della tradizione
platonica secondo Plutarco.” In G. Cambiano (ed.), Storiografia e dossografia nella
filosofia antica, 203–214. Torino: Tirrenia stampatori.
124 SKEPTICISM
Frazier, Françoise. 2008. “Philosophie et religion dans la pensée de Plutarque. Quelques

réflexions autour des emplois du mot πίστις,” Études platoniciennes 5: 41–62.
Ruprecht.
Hershbell, Jackson P. 1992. “Plutarch and Epicureanism.” In W. Haase (ed.), Aufstieg und
Niedergang der römischen Welt 36 (5), 3353–3383. Berlin: de Gruyter.
Ioppolo, Anna Maria. 1993. “The Academic Position of Favorinus of Arelate,” Phronesis
38: 183–213.
Ioppolo, Anna Maria. 1994. “Accademici e Pirroniani nel II secolo d.C.” In A. Alberti (ed.),
Realtà e ragione. Studi di filosofica antica, 85–103. Florence: L.S. Olschki.
Isnardi Parente, Margherita. 1988. “Plutarco contro Colote.” In I. Gallo (ed.), Aspetti dello
stoicismo e dell’epicureismo in Plutarco, 65–88. Ferrara.
Lévy, Carlos. 2005. “Deux problèmes doxographiques chez Philon d’Alexandrie:
Posidonius et Énésidème.” In A. Brancacci (ed.), Philosophy and Doxography in the
Imperial Age, 103–120. Florence: Leo S. Olschi.
Lévy, Carlos. 2008. “La conversion du scepticisme chez Philon d’Alexandrie.” In F. Alesse
(ed.), Philo of Alexandria and Post-Aristotelian Philosophy, 103–120. Leiden: Brill.
Lévy, Carlos. 2009. “Favorinus et les Academica,” Revue des Études Anciennes 111: 45–54.
Lévy, Carlos. 2013. “Plutarque juge et partie: à propos des débats entre l’Académie,
le Jardin et le Portique,” Aitia 3.
May, M. James and Jakob Wisse. 2001. Cicero on the Ideal Orator. New York and Oxford:
Opsomer, Jan. 1998. In Search of the Truth. Academic Tendencies in Middle Platonism.
Bruxelles: Koninklijke Academie voor Wetenschappen.
Sedley, David. 2012. Antiochus of Ascalon. Cambridge: Cambridge University Press.
Soreth, Marion. 1968. “Der zweite Telosformel des Antipater vor Tarsos,” Archiv für
Geschichte der Philosophie 50: 48–72.
Striker, Gisela. 1983. “The Ten Tropes of Aenesidemus.” In M. Burnyeat (ed.), The
Skeptical Tradition, 95–116. Berkeley and Los Angeles: University of California Press.
von Arnim, Hans. 1888. Quellenstudien zu Philon. Berlin: Weidmann.
von Arnim, Hans. 1901. Stoicorum Veterum Fragmenta. Leipzig: Teubner.
CHAPTER TEN
Sextus Empiricus
TAD BRENNAN AND CLIFF ROBERTS
1 INTRODUCTION
There is no evidence that, in his own day, Sextus Empiricus was a figure of any
great prominence. If we look to sources outside his writings, we find only a few
contemporaries who mention him, in lists of teachers and students, and they credit
him with no philosophical innovations (DL IX 116; Ps-Galen 14.683). If we look
inside his writings, we find a clever and competent compiler of other people’s
arguments, who claims no originality for either the structure or contents of his
works. For all that we know, it may be that in his own day Sextus was a very
minor figure.
But accidents of transmission have led to his becoming a very major figure indeed
for the modern era. In the history of skepticism in the West, Sextus is central and
indispensable: he is the conduit through which the most copious streams of Ancient
Greek skepticism flooded into Renaissance Europe, with effects both cataclysmic
and fertilizing.1 In St. Paul’s Cathedral in London, a modest plaque marks the
tomb of Christopher Wren, its architect. It reads “Lector, si monumentum requiris,
circumspice.” So too, the reader looking for a fitting monument to the influence of
Sextus Empiricus may look around this book as a whole, and see something of his
influence in nearly every chapter.
In the present chapter, we devote Section 2 to an overview of the life and works
of Sextus Empiricus. We then introduce the problem of how to characterize the
philosophical project of Pyrrhonism, and in particular its claim to offer practical
guidance in living (Section 3). This problem has been discussed extensively
in the literature, and we focus on two especially influential proponents of
opposing viewpoints, Michael Frede and Myles Burnyeat (Section 4). In Section
5, we consider controversies about how to understand Sextus’s use of Modes of
Skeptical argumentation.
126 SKEPTICISM
2 OVERVIEW OF THE LIFE AND WORKS

OF SEXTUS EMPIRICUS
The evidence for Sextus’s life is “frightfully meager,” as Burnyeat (1984: 242) put
it—estimates of his dates range anywhere within a century on either side of 200 CE.
His location can be specified with comparable imprecision—he probably lived in a
major city somewhere in the Mediterranean basin. Scholarly arguments for locating
him in Rome or Alexandria turn out to be tissue-thin, as House (1980) has shown.
He is clearly a follower of the Neo-Pyrrhonian School of Aenesidemus. He is very
well informed about the main philosophical schools of the period from 300 BCE
to 100 BCE, but he seems curiously unaware of philosophical developments in the
centuries closer to him. We must imagine that he or his sources had access to a good
library, given the many texts that he quotes; but nothing about his works suggest
that he is in daily contact with a community of philosophers, whether of his own
school or opponents.
The reasons to think he was a medical doctor begin with the name “Empiricus,”
attached to him in his own manuscripts and in the two contemporary references,
which suggest that he was connected with the Empiricist school of medicine. The
picture is somewhat confused by his stated preference for the Methodical school of
medicine, but not in a way that speaks against his having been a trained physician
(PH I 237–241). He also refers at two points to treatises that he wrote on medicine2;
these and a treatise “On the Soul” have not survived (ΑΜ X 84).
The works that have survived are three: a treatise in three books called Outlines
of Pyrrhonism (cited as PH from the initials of its Latin title); a treatise in six books
called Against the Professors (AM, likewise from its Latin title), and a treatise in five
books of the same name (AM VII–XI), but composed independently of the second,
six-book treatise.
The Outlines begins with an overview of the Skeptical system (PH I), before
turning to the detailed refutations of dogmatic theories of logic and epistemology
(PH II), and physics, theology, and ethics (PH III).
In AM I–VI, Sextus offers arguments against the advocates of the liberal arts, sc.
grammar, rhetoric, geometry, arithmetic, astronomy, and music (in that order). In
AM VII–VIII, he investigates theories of logic and epistemology; in AM IX–X he
investigates physics and theology; and in AM XI he investigates ethics.
As that brief account indicates, Sextus gives us two separate treatments of each
of the main topics in philosophy: logic and epistemology are discussed both in AM
VII–VIII and in PH II; physics in AM VIII–IX and in PH III, and ethics in AM XI and
PH III. Much material is reused, though seldom through verbatim copying from one
treatise to another. And indeed, scholars disagree about which of the two sources was
written earlier. Janáček (1963) argued that the shorter, more summary treatments in
PH were written before the fuller treatments in AM, but Bett (commentary to Sextus
1997) has argued for the opposing view.
Most of the material in these treatises is, by design, quoted from dogmatic
philosophers. Sextus begins his investigation of a given dogmatic tenet by presenting
the view at length before arguing against it. Often his counterarguments, too,
SEXTUS EMPIRICUS 127
are drawn from dogmatic sources. As a result, his works are a treasure-trove of
dogmatic arguments, as well as skeptical arguments. His reporting of earlier schools
is generally fair and accurate, so far as we are in a position to judge the matter, and
accordingly his evidentiary value is very high.
However, precisely because so much of his work consists in quotations, it is
also not directly useful for understanding his own Skeptical outlook. Sextus had no
philosophical dogmata—so he claims—but he did practice a way of life. And he does
offer us a variety of comments and explanations about what he does as a Skeptic.
3 SEXTUS’S SKEPTICAL OUTLOOK: SKEPTICAL

VERSUS DOGMATIC ASSENT
The clearest and most direct statement of his own understanding of the Skeptical
outlook and life is given by Sextus in the first book of PH, in particular, §§1–35. The
outlook there portrayed is a (perhaps turbulent) marriage of two separate aspects.
The first aspect captures our familiar contemporary view of skepticism as
exclusively destructive.3 Sextus defines the Skeptic as someone who possesses
the ability, when faced with any claim, to generate an equally strong case for its
contrary. The Skeptic, and anyone else exposed to the competing arguments and
considerations pro and con, then has no better reason to assent to the given claim
than to its contrary. Since the Skeptic cannot assent to both claims (they are, after
all, contraries), her most natural recourse is the suspension of assent (ἐποχή) with
respect to claims: to assent neither to the claim nor to its contrary. Yet a virtue
can be made of what appears, at first, to be mere rational necessity, for the Skeptic
discovers, upon suspending belief, that she has achieved some measure of peace of
mind, what the Skeptic calls “quietude” or “freedom from disturbance” (ἀταραξία),
which is the final end of the Skeptical philosophy. It is no wonder, then, that Sextus
devotes the lion’s share of the first book of PH to characterizing the elaborate and
various forms of argument—called “Modes” (τρόποι)—by means of which ἐποχή is
to be secured; for without a sufficiently robust armory of such arguments, there is
little hope of securing ἐποχή, and beyond it, the grail of ἀταραξία.4
Neither ἐποχή nor ἀταραξία makes a life; the former, in particular, is a merely
negative stance of withholding assent. Yet the Skeptic, in Sextus’s words, “cannot
help but be active” (PH I 23), and a distinctively human life, one given structure and
content by reasons, seems to demand assent. Without assent a human is reduced
to a merely vegetative life, a life without action. This apparent conflict need not
perturb the Skeptic, for the Skeptic’s commitment to ἐποχή is fully consistent—so
Sextus insists—with the Skeptic’s giving assent, when this assent is limited to the
standard or criterion of “appearances” or “what is apparent.” This is the second
or positive aspect of Sextus’s Skepticism, which is less familiar to contemporary
readers (PH I 19–24, 29–30). It is appearances to which the Skeptic assents in living
her life, and it is appearances that determine a quite ordinary life lived in light of
the natural human capacities for thought and sensation, the constraint of feelings
(e.g., of thirst and hunger), the familiar customs and laws, and the teachings of
the crafts (PH I 23). While it is true that these appearances are objects of Skeptical
128 SKEPTICISM
assent, her assent on these matters is consistent with her advocacy of ἐποχή, since
she withholds assent only to the “things underlying” (ὑποκείμενα) appearances,
not to the appearances themselves. These ὑποκείμενα are precisely the objects of
that form of assent, dogmatic assent, which the Skeptic avoids at all costs. Thus the
Skeptic aspires to navigate the twin perils of inaction and dogmatic assent.
Whether this aspiration can be realized has been the subject of much scholarly
controversy. A verdict on this issue presupposes answers to specific questions about
the difference between dogmatic and Skeptical assent: is it a difference in the object
of assent or a difference in the kind of assent? Is Skeptical assent to be understood,
like dogmatic assent, as assent to the truth of a claim? Does it constitute what we
would now think of as belief? What are the “apparent” or “evident” matters to
which the Skeptic may assent, and how are they to be distinguished from the topics
about which the Skeptic suspends? Very different answers to these questions have
been given by interpreters, and it is to those answers that we now turn.
4 INTERPRETING SEXTUS ON ASSENT: FREDE

AND BURNYEAT
For the last four decades, philosophical discussions of how the Pyrrhonist lives
have been dominated by the work of Michael Frede and Myles Burnyeat, and the
reactions that their work has provoked. In two papers, Frede (1979, 1984) argued
for a radical revision of the standard understanding of Sextus’s philosophy and
accompanied this rereading of Sextus with a novel account of the development of
ancient skepticism as a whole. In two other papers, Burnyeat (1980, 1984) argued
for a more traditional, Humean understanding of Sextus.
These papers are brilliant examples of how to write the history of philosophy,
engaging with questions of the broadest philosophical scope as well as the most
minute philological detail. They deserve and repay repeated study. That said, they
can be frustrating because of their difficulty and obscurity.5 In the next two sections,
we will discuss the work of Frede and of Burnyeat.
4a Frede’s Sextus Empiricus

In his first paper on skepticism, Frede (1979) drew attention to the fact that Sextus
claims not to eschew all beliefs in general, but only dogmatic beliefs, that is, beliefs
about “non-evident” or “non-apparent” matters, matters of dogmatic investigation.
Frede proposed that the scope of the “evident” is very inclusive; Sextus possesses
an extensive collection of beliefs that are quite ordinary in their character and
subject matter. The Pyrrhonian Skeptic goes through his or her day believing most
of the things that anyone else believes about the ordinary objects and events in their
world. What the Pyrrhonist feels no inclination either to accept or to deny are the
theories of the Dogmatic schools, and their elaborate, non-evident stories about
Democritean atoms, Platonic forms, Aristotelian entelechies, and so on. Believing in
those things is what makes one a Dogmatist. But through his investigations of these
theoretical posits, and through his ability to construct arguments in opposition to
the Democritean, Platonic, and so on arguments for special entities and theories, the
Pyrrhonist has come to feel that the case for and against all of these things is perfectly
equipollent. As a result, he suspends judgment about their existence. But doing so
is fully compatible with believing all of the ordinary things that non-philosophers
believe about the world around them. And since the Skeptic has very nearly the same
number and kind of beliefs that any ordinary person has, the traditional question
of how a Skeptic can live without any beliefs simply never arises. So far as their
ordinary interactions with tables, chairs, food, cliffs, and the other people in their
lives, the Skeptic lives just as anyone else lives.
That, in broad outline, is the reading that Frede advanced. When we move to
the details of Frede’s reading, however, it becomes more complex and obscure. The
complexity is not in itself a defect in Frede’s reading; any adequate interpretation of
a complex philosophical text will have to be a complex interpretation. The obscurity,
however, is a lamentable artifact of Frede’s own failure to make his views clear.
The crux of the problem is his characterization of the difference between the beliefs
that Pyrrhonists may have (about “evident” or “apparent” matters), and those that they
reject as dogmatic (concerning the “non-evident”). At first sight, it may seem that the
difference is one of content—the Pyrrhonist may believe in the evident chairs around
them, but not in the non-evident Form of the Chair. But it is fairly clear that this could
not have been what Sextus meant, and that it is not what Frede thought he meant. We
can see this because Sextus sometimes attacks beliefs as dogmatic even when they have
the same content as beliefs he elsewhere accepts. As Frede says, “it is not the content
of theoretical views . . . that makes them dogmatic views; it is, rather, the attitude of
the dogmatist” (1979: 195). So it is clear that Frede did not think that the difference
between evident and non-evident matters is one of disjoint subject matters.
However, it can be very difficult to tell where precisely Frede does locate the
crucial difference. In parts of his first paper, he suggests that the difference between
the Skeptic’s beliefs and the Dogmatist’s beliefs consists in the difference between
believing that P and believing that it is true that P; or passively believing P and
actively believing P; or implicitly and explicitly believing P; or unconsciously and
consciously believing P. Fine (2000) does a thorough and excellent job of showing
insurmountable problems with all of these glosses. In his second paper, Frede
distinguished between “holding a view” that P and “taking a position” that P. Frede
clearly wanted these phrases to convey some distinction, but on their own they get
us no further toward understanding what he meant.
We are inclined to find the core of Frede’s view in a different set of glosses that
he sometimes uses, according to which the Dogmatic beliefs present themselves as
part of an ultimate scientific theory of the world:
[Dogmatic beliefs claim to contain] deeper insights into the true nature of things.
(Frede 1979: 179)
[I]t is characteristic of the dogmatists that they believe it is possible to go behind
the surface phenomena to the essence of things, to the nature of things, to true
reality. (Frede 1979: 187)
[The Skeptic does not think that his impressions are] such that they will come out
true on the true theory of things. (Frede 1984: 210)
130 SKEPTICISM
[In dogmatic assent, one] commits oneself to a belief about what will come out
as true on the true theory of things, about what would turn out to be true if one
really knew what things are like. (Frede 1984: 215)
Dogmatists think that the philosophical project, the ζήτησις that they began
along with the Pyrrhonists, aims at the ultimate possession of a deep, explanatory,
comprehensive, rationally justified theory of the world; a scientific knowledge of the
nature of things, expressed in the language of science, whose nouns name the real
entities and whose predicates map the real properties of the world. This is the kind
of theory whose possession will purge our lives of anomaly and disturbance. Once
one has it, one will be able to say individual true things, of course, but far more than
that: one will have scientific knowledge of the true things one says and be able to
show, for any truth one utters, exactly how one’s grasp of it counts as knowledge,
how it is justified by its place in the total system.
It is possible that the Pyrrhonists share or shared (in the beginning) some parts of
this characterization of the project, the ultimate aim of their ongoing ζήτησις. But
the Dogmatists think that their school already has attained it, or some large coherent
fragment of it, and that the pronouncements they make as members of that school
are already deliverances of that ultimate theory. They are not merely true; they are a
part of the ultimate true theory, and so inherit from it those features of consistency,
coherence, explanatory adequacy, ideal vocabulary, and so on. This, so far as we can
tell, is the reason that Frede sometimes wrote as though the essence of the Dogmatic
stance was the additional belief that P is true; he was using “true” as a shorthand for
“is a deliverance of the ultimate true, explanatory, rationally justified, etc. theory,”
or as he says, “is true on the true theory.” “For them [sc. Dogmatists], those things
count as truths which on the true account of things would come out as truths”
(1984: 210, 215). In the modest sense of “true” in which every belief that P entails
believing that it is true that P (which we take to be the normal sense of “true,” as
well as the normal account of belief), it is clearly not the case that “taking something
to be true” means “committing oneself to its being explanatory, expressed in the
ideal vocabulary, and so on”—that is, the whole long range of properties that Frede
abbreviates by calling them “true.” Fine (2000) raised a number of decisive points
against the idea that a Skeptic can believe P without believing that P is true, in the
ordinary sense of “believe” and “true.” We think that Frede’s best response would
be simply to concede that she is right, and that he was using “true” and “truth” in a
nonstandard sense. It is a sense with some parallels among ancient Dogmatists—for
example, Parmenides’s “Way of Truth” or the Stoic idea that only the Stoic Sage has
“truth”—but it is certainly not the normal sense of “truth.”
This is how Sextus understands what it is to be a Dogmatist and what it is to
assent to something dogmatically: to assent to something dogmatically is to arrogate
to one’s belief the status of ultimacy, to present it as part of the ultimate theory
of everything.
When one reflects on the writings of actual Dogmatists such as Plato, Aristotle,
or Chrysippus, this characterization of them may seem unfair. There is much that is
tentative and provisional in their work, places where one can see that their thoughts
are still developing and will continue to develop. If Frede’s Sextus characterizes
dogmatic belief in terms of granting something the status of ultimacy, it seems
that Plato, Aristotle, and Chrysippus turn out not have been uniformly dogmatic
after all, and this outcome may look like an objection to Frede’s characterization
of Sextus’s position. But it is not. For the same mismatch between the figure of the
Dogmatist in Sextus and the actual schools that Sextus opposes is clear from the
opening lines of the work. The Dogmatists of PH I 1 are figures whose searching
has terminated, because they claim to have discovered the truth. It is clear to us as
sensitive readers of Plato (e.g.,) that this misrepresents his attitude toward many of
his own doctrines; his own searching was by no means at an end when he wrote
his dialogues. But Sextus clearly characterizes Plato and other Dogmatists as people
who believe that their answers have put an end to searching.
Frede is not misrepresenting Sextus on this point; but is Sextus misrepresenting
his opponents? Is “the Dogmatist” in Sextus merely an idealization or projection
rather than an accurate report of people whom Sextus had met or whose writings
he had read? The answer is presumably mixed, inasmuch as the historical figures
themselves had mixed attitudes toward their own views. Plato, for instance,
seems to have been willing to try a number of competing characterizations of the
“participation-relation” between Forms and particulars; nothing in his writing
suggests that he thought he had worked out that part of his doctrine in a final form.
By contrast, his attitudes toward such statements as “There are gods” and “Virtue is
good” look to be clear cases of Dogmatism, whether by Sextus’s characterization in
the opening trichotomy of PH I 1, or by Frede’s criterion of ultimacy: Plato thought
his search with respect to those questions had ended with the successful discovery
of truths that would form parts of the ultimate true theory of everything. Whatever
else might change as his doctrine evolved, these claims would remain unaltered.
If Plato’s own attitude toward his doctrine was a mixture of the tentative and the
dogmatic, Epicureanism featured an even higher proportion of the dogmatic, and even
less invitation to future revision or innovation. Finally, when we consider the Stoics,
it is worth remembering that, according to their own theory of assent (which formed
the structural basis for the Skeptic doctrine of assent and suspension), one should
assent to an impression only when it is cataleptic, that is, true, guaranteed to be true
by its content, and of such a sort that if a Sage were to assent to it, her assent would
be unshakeable by any rational argumentation. No future findings of philosophy or
science could give the Sage reason to reverse their assent or revise their views on this
topic. Here the Skeptical equation of dogma with ultimacy is entirely justified.
So the Skeptic conception of dogmatic assent is not a gross distortion of the
ambitions of Sextus’s opponents, though it also does not apply to all of their
practices. Many of his opponents put forward many of their views as though they
were the deliverances of the final, ultimate theory, which would be vindicated by
any future perfected version of philosophy. Such views are the Skeptic’s lawful prey.
True, some Dogmatists put forward some of their views as provisional or tentative
hypotheses awaiting development. But it is not unfair for Sextus to attack these
views, either, given that the attack simply culminates in a request that the Dogmatist
not assent to them in a dogmatic way. And if the view in question was advanced by
132 SKEPTICISM
the Dogmatist in a merely provisional or tentative way, then the Dogmatist does not
need to do anything new in order to comply with the Skeptic’s request.
Because the Skeptic rejects only dogmatic pronouncements, that is, assertions
that arrogate to themselves the status of ultimacy, the Skeptic can consistently have
and use a wide range of other beliefs about ordinary life. The non-philosopher
thinks it is day, and thinks that it is true that it is day, but does not have any view
about whether the ultimate theory of the world will make reference to “days” or
eliminate them in favor of some purified vocabulary. The Pyrrhonists do the same,
both about days and about the rest of their ordinary practical life.
This expansion in the range of the Pyrrhonist’s beliefs leads directly to a contraction
in the range of the Pyrrhonist’s ἀταραξία. All interpretations of Sextus must agree
that he promises tranquility with respect to the prohibited dogmatic opinions and
moderated disturbance with respect to what is necessary.6 But this boundary or surface
between the necessitated things and the realm of dogmatic opinion shifts a great deal
between rival interpretations. The classic Humean reading has the surface shrink to a
minuscule bubble inside the Pyrrhonist’s own mind; within that bubble, Skeptics will
necessarily have appearances and suffer the moderate affections of being appeared
to in various ways. But they can attain tranquility outside that bubble by refusing to
believe that there is a war on, not assenting that their family was just captured by the
enemy, suspending judgment about whether they are being trussed up for execution,
and so on. On Frede’s reading, the realm of moderate affections moves outward to
encompass ordinary life, including families and wars, and the realm of tranquility
shrinks to a sphere around the doctrinal posturings of the Dogmatic schools. The
Pyrrhonist has little choice but to suffer the ordinary injuries, both physical and
mental, that ordinary people are prey to; and they are content if they can share
only moderately in the frailties of flesh. But they can achieve one important kind of
tranquility: they can resist adding on to their inevitable injuries the voluntary and
self-inflicted disturbance that comes from fanatical devotion to a speculative creed.
Much less is being promised by the Pyrrhonist on this reading; they can offer little
advantage over ordinary life. But, as they have learned, no other school can credibly
offer any great advantage, either. And the Skeptic can offer a huge advantage over
the rival Dogmatic schools: they can show you how not to worsen your life with
dogma. They offer no miracle cure for ordinary suffering—injuries, worries, grief,
and pain are all necessary concomitants of life. It is the Dogmatists who offer miracle
cures, and the Pyrrhonists can show you that they are all charlatans and quacks,
whose remedies give you a new disease you did not have before.
That, in outline, is the picture that Frede offers. It allows Sextus to give a very
robust answer to the traditional charge that the life of the Skeptic is not livable:
the Skeptic in this view can live a very ordinary life indeed, with the same beliefs,
actions, attitudes, desires, and so on that the ordinary person has.
4b Burnyeat’s Sextus Empiricus

Burnyeat’s first article was written largely without reference to Frede’s work; only the
sixth of its seven sections (“Controversial Interlude”) acknowledges Frede’s reading
of Skepticism, and it does so in order to consider it as a possible objection to the

paper’s main conclusions and then dismiss it. Otherwise, the paper advances a view
of Skepticism that is traditional in its verdicts (inspired by Hume, a long quotation
from whom begins the paper), although brilliantly sophisticated in its methods.
Those main verdicts are two: first, that if it were possible to live the life that
Sextus describes, it would be horrible, repellent, and sub-human: “The life without
belief is not an achievement of the will, but a paralysis of reason by itself” (Burnyeat
1980: 42). Second, Burnyeat argues that the description of the life is incoherent in
any case: it claims to be a life without belief, but the central elements of the Skeptic’s
life—opposing arguments, equipollence, suspension, and tranquility—all require to
be explained in terms of belief. This is the burden of the last section of the paper.
Burnyeat’s article, too, contains some puzzles. To begin with, he claims early
in the paper that the word “true” in Ancient Greek philosophical contexts means
“‘true of a real objective world’ . . . as distinct from mere appearance . . . . Statements
which merely record how things appear . . . are not called true or false” (1980:
25–26). He then notes that belief involves “the accepting of something as true,”
and then concludes that there can be no beliefs about how things appear, since there
is no truth about such matters. In line with this distinction, Burnyeat argues that
assents to matters of real existence constitute beliefs, and are contrasted with assents
to appearances, which are not beliefs and do not have truth-values (1980: 26–27).
It is quite unclear how Burnyeat intended to establish or argue for his claim
that truth in Greek philosophical contexts is restricted to beliefs about “matters of
real existence.” This is not something that one can look up in a dictionary, nor are
there ancient statements to this effect. Barnes (1982: n. 21) and Fine (2000) have
both rejected this line of argument, and in the article Burnyeat does little more than
assert it. It is unsurprising, therefore, that later interpreters such as Fine (2000) and
Perin (2010a,b) have accepted Burnyeat’s claim that Skeptical assent pertains only
to subjective matters of appearance (not objective matters about the world), while
rejecting his further claim that such assent cannot be viewed as belief, that is, as
assent to the truth of the subjective matters.
The claim that “truth” cannot apply to appearances plays a central role in
Burnyeat’s arguments. It leads immediately to the conclusion that the Skeptic cannot
have any beliefs. For Sextus says he assents only to appearances, but (according to
Burnyeat) appearances have no truth-values, and things without truth-values cannot
be beliefs (Burnyeat 1980: 26).
Since Burnyeat takes the Pyrrhonist to have no beliefs whatsoever, he assumes
that Pyrrhonian suspension and tranquility must also occur without beliefs, on pain
of inconsistency. Accordingly, he imagines that the Pyrrhonist will claim, not that
they believe that opposed arguments are equipollent, but that the opposed arguments
merely appear to them to be equipollent, without their believing that they are. (This
sort of appearance claim is called by both Frede and Burnyeat a “non-epistemic”
appearance claim, but we agree with Fine (2000: 83 fn. 11) that it is more useful
and more accurate to call it a “non-doxastic” appearance claim, since the speaker
disavows even the correlative belief, much less the correlative knowledge). And here
Burnyeat protests: “this appearance, so called, being the effect of argument, is only
134 SKEPTICISM
to be made sense of in terms of reason, belief, and truth—the very notions the sceptic
is most anxious to avoid” (1980: 50). Furthermore, he argues that no tranquility can
be produced by suspension unless the Skeptic is “in some sense satisfied . . . that no
answers are forthcoming, that contrary claims are indeed equal . . . How can Sextus
then deny that this is something he believes?” (1980: 52). So Burnyeat’s Skeptic
is committed to beliefs at two stages in the process: “[b]oth the causes (reasoned
argument) . . . and the effects (tranquillity) are such as to justify us in calling it a
state of belief” (1980: 52).
In pressing the argument against the first stage, Burnyeat anticipates that the
Pyrrhonist will claim that opposed arguments merely appear equipollent. But he
denies they can do so coherently: “In the philosophical case, the impression, when
all is said and done, simply is my assent to the conclusion of an argument, assent to
it as true” (Burnyeat 1980: 53). This argument is rather mysterious. Why can it not
appear to someone that two arguments are of equal weight, without her believing
that they really are? And what does it mean to say that “the impression . . . simply
is my assent”? This last phrase, although it makes an attractive slogan, is surely a
category mistake, given the terms of Hellenistic epistemology. An impression is a sort
of state, a static configuration of one’s soul; an assent is a movement of one’s soul.7
No state can be a movement. And in particular, no impression can ever be anything
more than a possible object for some psychic action such as entertaining, assenting,
rejecting, investigating, and so on. Impressions stand to assents as propositions stand
to propositional attitudes, and to say “the impression just is an assent” is no more
sensible than saying “the proposition just is a wondering whether.”
Burnyeat seems to be aware that he has overstated his case, since in the next
sentences he talks more carefully of the impression “presupposing” or “including”
assent, and of its being not “independent of” assent. So presumably what he meant
was not that the impression at issue “simply is” an assent, but rather that it somehow
presupposes one or more assents. Since the impression at issue is an impression
that two arguments are of equal weight, then perhaps this is his thought: if I have
the impression that A and B are both equally probative arguments, then I must
have assented, previous to having this comparative impression, to the two earlier
impressions that A is a probative argument and B is a probative argument.
But that simply raises a second round of questions: why can I not have the
impression that A is a probative argument, without assenting to that impression
(and likewise for B)?8 Here Burnyeat’s response might be: the impression that A is
a probative argument is not (of course) itself an assent, but it depends, in turn, on
earlier assents, as for instance my assenting to the impressions that A is formally
valid, and that the premises of A are cogent and plausible. So my impression of
equipollence “includes” and “presupposes” my assent to impressions of probative
force, and my impressions of probative force “include” and “presuppose” my
assents to impressions of validity and cogency. The impressions at every higher level
presupposes assents to the impressions at lower levels.
But do they? Why cannot the form appear valid and the premises cogent, without
my assenting to those appearances? Why cannot the entire practice of Skeptical
argument build up from the lowest level with nothing but a sequence of impressions,
to which no assents are forthcoming? I have the impression that certain premises
are cogent, and that certain forms are valid; as to whether they are or not, I suspend
judgment. I have the impression that the argument they constitute is probative,
but I do not assent to its being so. I have the impression that a pair of such forms,
filled out with such premises, is roughly balanced in its force; nonetheless no assent
eventuates. Burnyeat does not seem to have a way to rule this out.
Perhaps at this juncture he switches to the second stage of the argument and
appeals to the “effects” of suspension, instead of its “causes”: if I am not “convinced
by the argument[s]” on both side, and then convinced of their equality of weight,
then I will not be “in some sense satisfied . . . that contrary claims are indeed equal,”
and thus no tranquility will ensue (Burnyeat 1980: 53). Nothing short of assent can
alleviate my worries.
Perhaps this was Burnyeat’s thought in this passage. But if so, then he could have
convicted the Skeptic of having beliefs in much easier ways. Nothing especially
philosophical needs to be introduced; we may simply suppose the Skeptic is sitting
comfortably in a chair. The Skeptic will claim to suspend judgment about whether
the chair will continue to hold him up or will disappear and drop him abruptly to the
floor—there are arguments on both sides, after all. But the comfortable Skeptic will
also manifest no concern about the possibility of being dropped abruptly to the floor.
If, according to Burnyeat, nothing short of assent can alleviate the Skeptic’s worries,
then the unworried chair-sitter must be “in some sense satisfied” that the chair will
not disappear. And how can he then deny that this is something he believes?
Our point is that this problem of tranquility is perfectly general, and has no
special connection to philosophical impressions. There is a general question,
applicable to both chairs and equipollence: “how can you be tranquil about danger
D, when you suspend judgment about whether D impends?” If the Skeptic cannot
answer it for chairs, then Burnyeat can make his point without the complexities
of arguments and equipollence and philosophical impressions. If the Skeptic can
answer it, then the answer will serve just as well for equipollence as for chairs. And
put in terms of chairs, the Skeptic answer should be familiar. When asked why
“someone who suspends judgment does not rush away to a mountain instead of to
the bath,” Plutarch answers on behalf of the Academic Skeptics, “because, of course,
it is not the mountain but the bath that appears a bath to him” (1122E, trans. LS).
The Skeptic is careful around cliffs and comfortable in chairs because it is the cliff,
not the chair, that appears dangerous to him. In the same way, the Skeptic can be
comfortable with equipollence because the arguments appear equipollent. Assent is
not a precondition of tranquility.
So the claim that the Skeptic’s suspensive practice commits him to beliefs, either
in its causes or its effects, does not strike us as successful. And the claim that the
life without belief is a “paralysis of reason” is based on the idea that the Skeptic can
have no beliefs of any sort, which in turn is based on an idiosyncratic claim about
Greek theories of “truth.” It also does not take account of Frede’s suggestion that
the Skeptic has a wide variety of beliefs that can be employed in life.
Burnyeat’s second article “The Skeptic in his Place and Time” is presented as a
direct response to Frede’s reading (1984: fn. 13). He argues that Frede is attributing
136 SKEPTICISM
to Sextus a strategy of “insulating” some areas of discourse from other areas of

discourse. If one “insulates” the claims of ordinary life from the claims of philosophy,
then they cannot clash or contradict; we may claim philosophically that there is no
such thing as time, while claiming in ordinary life that there is enough time to catch
the train, without either claim “affecting or being affected by” the other (Burnyeat
1984: 227). No ancient Skeptic, Burnyeat argues, ever practiced insulation. Thus
Frede’s reading of Sextus is anachronistic, attributing to him a view he could not
possibly have had.
That is a rough outline of Burnyeat’s strategy in the article. But in detail it goes
slightly differently. He introduces at least two kinds of “insulating” readings of
skepticism, one derived from Gassendi and Montaigne that he calls “insulation by
subject matter,” or the “Country Gentleman’s Interpretation,” the other derived
from Kant, which he refers to as “insulation by level,” or “transcendental skepticism”
(Burnyeat 1984: 233, 252). On the Country Gentleman’s Interpretation, a Skeptic
scrutinizes such philosophical posits as Platonic Forms and Epicurean atoms but
accepts as evident our ordinary talk of tables and chairs. On the transcendental reading,
there is no distinction of subject matter, but rather “two ways of understanding a
statement like ‘The stove is warm,’ the plain way and the philosophical way, and it
is only the philosophical claim to an absolute knowledge that the sceptic wants to
question” (Burnyeat 1984: 252).
Burnyeat’s comments about “insulation by level” are clearly intended to be a
sympathetic presentation of Frede’s reading, in which Skeptical enquiry is restricted
to “the dogmatist’s belief that he can achieve a further or deeper kind of knowledge
and truth than the plain man requires for the purposes of ordinary life” (Burnyeat
1984: 233).
He spends the bulk of the article (§§III–IX) looking closely at arguments in Sextus
about the concept of “place” (τόπος) in order to show that Sextus did not insulate
by subject matter in that discussion, and (by implication) that he did not insulate by
subject matter in the rest of his work. He then argues (§§X–XI) that no one ever
insulated by level until Kant. He concludes that it is anachronistic to suppose that
Sextus insulated by level.
But evidently there is a problem here: Burnyeat never examines Sextus directly to
see whether he insulated by level (rather than by subject matter). He has no argument
to that conclusion, other than the claim that no one prior to Kant insulated by
level. And this indirect argument from anachronism simply begs the question against
Frede: Kant was not the first person to insulate by level if Sextus did it before him.
This gap in Burnyeat’s argumentative strategy could be closed if it could be shown
that anyone who insulates by level must also insulate by subject matter. In that case,
the demonstration that Sextus does not insulate by subject matter in his discussions
of place would go some way toward showing that he could not be insulating by
level, either. But in fact, insulation by subject matter does not follow from insulation
by level. Quite to the contrary: the hypothesis on which Sextus insulated by level
actually presupposes that Sextus did not insulate by subject matter.
Here is why. The hypothesis on which Sextus insulated by level claims that he
directed his skeptical scrutiny toward statements and assertions only when they
are read or understood in one way (as dogmatic pronouncements), and not when
they are understood in another way (as claims in ordinary life). But it is also part
of this hypothesis that Sextus is confronting Dogmatists who themselves refuse to
insulate by subject matter. These Dogmatists—idealized rather than historical, as
noted above—think that their dogmatic philosophies are so coherent, so complete,
so all-embracing, that every statement that they make is a deliverance of some part
of the correct ultimate science of nature, the final Theory of Everything. And this
applies just as much to their statements about practical affairs and medium-sized
objects as it does to their more abstruse proclamations.
So the Dogmatist does not insulate by subject matter. Accordingly, the Skeptics
who insulate by level must treat any Dogmatic statement, on any subject matter,
as a theorem of the Dogmatist’s favorite candidate for the Theory of Everything.
If uttered by a Dogmatist, the statement—no matter what its subject matter may
be—is inextricably bound up with the Dogmatist’s favored theory. If the Dogmatist
says “my house is some miles west of here,” then they are staking a claim that the
ultimate theory will vindicate the existence of substances like houses, locations like
“here,” the coherence of directions like “west of,” of metrics like “some miles,” and
so on.
The Skeptic who insulates by level is exactly responding to Dogmatists who
refuse to insulate by subject matter: the Skeptic attacks all of the Dogmatist’s claims
as philosophical claims, no matter what their subject matter may be, because the
Dogmatist’s own ultimate theory is meant to apply to practical affairs as well as
theoretical ones.
Accordingly, the Skeptic distinguishes two levels of discourse about, for example,
stoves, one outside and one inside the Dogmatist’s discourse. But for this very
reason, the Skeptic will not want to distinguish two types of subject matter within
the Dogmatist’s discourse. And this is what Burnyeat shows in the central part of
this paper: the Skeptic does not insulate some kinds of Dogmatic pronouncements
about place from other kinds of Dogmatic pronouncements about place. Both kinds
of Dogmatic talk occur at the same level.
But that means that the paper misfires as a refutation of the claim that the Skeptic
practices insulation by level; that hypothesis is simply never tested. And the results of
the test that is conducted (on insulation by subject matter) are perfectly compatible
with the hypothesis that Sextus himself insulated by level.
5 THE MODES AND THEIR INTERPRETATION

As mentioned earlier, the bulk of PH I (§§36–186) is devoted to a survey of
the Modes (τρόποι) or argument forms essential to generating the opposition
among claims that rationally compels ἐποχή. Sextus considers various
collections of Modes: the Ten Modes of Aenesidemus (§§36–163), the Five
Modes of Agrippa (§§164–177), the Two Modes (§§178–179), and the “Causal”
Modes (§§180–186). The first two collections are most important, not only
because they figure prominently in Sextus’s arguments throughout his work
and have exerted a profound influence on the subsequent history of philosophy
138 SKEPTICISM
after Sextus, but also because they seem to capture the essential character of
Skeptical argument.
The Ten Modes depend upon the familiar fact that things ostensibly the same
will appear differently either to different observers or in different circumstances or
when other factors differ. When these appearances are incompatible, the argument
concludes that one must suspend assent with regard to how the object is.
There seems to be an important connection between the Five Modes and the
Ten, since one of the Five Modes—the Mode from “Relativity” (πρός τι)—seems
to capture the basic form of the Ten Modes as a whole and, indeed, Sextus suggests
precisely that (PH I 167).9 A similar point can be made with regard to the Mode from
Disagreement (διαφωνία), which, while perhaps not as closely related to the Ten
Modes, nevertheless seems to share a similar structure. The remaining three Modes,
by contrast, look to be quite different. Each of them—the Modes from Infinite
Regress, Reciprocity, and Hypothesis—involve showing that a line of argument for
a claim is vitiated either by regress, by circularity, or by lack of any support beyond
bare assertion.
It is quite clear that Sextus conceives of the Five Modes as forming a “loose
system”10 (PH I 170–177), and this suggests that both the Five and the Ten Modes
might be used together, though whether this “togetherness” is more loose collection
or genuine system is a matter of dispute. Indeed, in the next section we will
consider a recent attempt to precisely articulate both the structure of the Ten and
the Five Modes and the nature of their connection with a view to deepening our
understanding of both.
5a Morison’s Understanding of the Modes

Recently, Ben Morison (2011; 2014)11 has argued that all of the Skeptic’s arguments
against some dogmatic thesis P must take the form of arguments in favor of ~P.12
In line with this proposal, Morison argues that the Ten Modes are best analyzed
as “devices for generating an equal and opposing argument in response to a
dogmatist’s attempt to show how things are” (Morison 2014: §3.5.1). When the
Dogmatist argues that P, the Skeptic who has mastered the Ten Modes will be able
to respond with an argument that ~P. Morison’s discussion of his proposal, and his
rejection of the earlier understanding of the Ten Modes, is thoroughly convincing.
Morison has also suggested that we cannot understand the Five Modes of Agrippa
as they have been traditionally construed. Traditionally, critics have assumed that
the point of showing that a Dogmatist’s proof of P rests on an infinite regress or a
reciprocity is to persuade the Dogmatist that their argument for P is a “bad argument,”
and so should be abandoned as a reason for believing P.13 Morison thinks that this
cannot be the correct construal of Sextus’s practice, because showing a Dogmatist that
their argument for P is a bad argument does not give them a reason to believe ~P. And
if the Skeptical argument does not give the Dogmatist a reason for believing ~P, then
it is not an opposed argument of the kind that is essential to the Skeptic’s practice.
In conformity with his thesis that the Skeptic responds to Dogmatic arguments
for P only by producing positive arguments for ~P, Morison has suggested that
the Five Modes should also be understood as “devices for generating an equal and
opposing argument.” So, for instance, when the Dogmatist has argued that P, and
the Skeptic shows the Dogmatist that her argument is guilty of an infinite regress,
that diagnosis of regress is not itself the Skeptical response to the argument for P.
Rather, the Skeptic counters the Dogmatist by constructing a symmetrical argument
for ~P, using the same method: infinite regress is to be used constructively to
provide an infinitely regressive argument for ~P in order to counter the infinitely
regressive argument for P that the Skeptic finds in the Dogmatist’s system.14 This
looks like a very perverse and laborious way for the Skeptic to argue, and such a
construal is certainly not suggested by the text.
But more importantly, this understanding of how the Skeptic sets out to produce
oppositions is unjustifiably narrow, as can be seen by looking at what Sextus says
about equipollence, in his gloss on the definition of Skepticism in PH I 10:
By “equipollence” we mean an equality in respect of credibility and incredibility
(ἰσότης κατὰ πίστιν καὶ ἀπιστίαν), so that none of the conflicting arguments will
stand out from the others as being more credible.
Morison’s thesis gives a good account of what it means to say that the conflicting
arguments have “equality in respect of credibility” (ἰσότης κατὰ πίστιν): there is as
much positive reason to believe P as there is positive reason to believe ~P. But what
is ἰσότης κατ’ ἀπιστίαν, that is, equality in respect of incredibility? This equality,
presumably, is the result of producing reasons not to believe P (as well as reasons
not to believe ~P, should they be needed for balance). Now, one sort of reason not
to believe P is any positive argument in favor of ~P. But that case has already been
covered in the ἰσότης κατὰ πίστιν clause. So the ἰσότης κατ’ ἀπιστίαν must be a
distinct kind of reason to lose faith in P, or to find one’s P-credences diminished.
And undermining arguments are the obvious candidates.
Sextus’s idea of “incredibility” (ἀπιστία) is thus a matter of having a proposition’s
credibility reduced or undermined. P is ἄπιστος when it lacks support, or has
its support weakened or undermined, even if this does not involve an improved
position for ~P, or any active increase in the credences for ~P.
So the text which gives Sextus’s official account of how Skeptical opposition
is supposed to work clearly shows that undermining arguments are envisioned
as playing a role in the production of equipollence. But the same result emerges
when we turn to the texts in which Sextus exercises his own antithetical powers.
For instance, we can see the clear use of undermining arguments in his discussion
of demonstrating P by means of P at PH I 61. If P is used to demonstrate P, Sextus
says, then P will be at once πιστός and ἄπιστος: πιστός insofar as it is being
used as a premise in the demonstration, but ἄπιστος insofar as it is being treated
as a conclusion in need of demonstration. In a related sense, Sextus will call a
method or a source of testimony ἄπιστος in order to undermine its credibility.
So for instance, at PH I 98, the philosophers disagree about what nature is.
Who can decide the question? Not the philosophers, since they are the parties
to the disagreement. But not a layman, either, since he is ἄπιστος (according to
the philosophers). To say that the layman is ἄπιστος does not do anything to
140 SKEPTICISM
establish the contradictory of his account of nature. Instead, it takes away from
the credibility of his testimony, it increases its incredibility: it undermines what
he says.
Similarly, at PH I 117, Sextus argues that we cannot sort out the problem of
the criterion without a good account of demonstration, and we cannot settle the
problem of demonstration without an adequate account of the criterion: “And thus
both criterion and demonstration collapse into the reciprocal mode (διάλληλος
τρόπος), in which both of them are found to be incredible (ἄπιστα). For, since
each of them is dependent on the credibility of the other, each of them is just as
incredible [ἄπιστος] as the other.” The Skeptic here counters demonstrations of P
not by arguing for ~P, but by undermining the credibility of the grounds for P, that
is, by undermining the demonstration directly.
These examples are only the first few of many in Sextus, in which ἀπιστία is
invoked in the course of bringing about equipollence. Sextus’s use of undermining
arguments should not surprise us, since his thoughts about equipollence are at least
in part modeled on the dynamics of a weighing-scale in which arguments on each
side are thought of as “weighty” or “light” (ἐμβριθές, κουφότερον, PH III 281).
If the Dogmatists put a kilogram of sand into one pan of the scale, the Skeptic can
bring the pans to a balance by subtracting some of the sand from the Dogmatist’s
pan as well as by adding some sand to the opposite pan.
Thus, in order to bring the Dogmatist to equipollence, Sextus may well use—as
we see he often does use—a multiplicity of arguments, some of them directly arguing
for ~P, and some of them undermining the Dogmatist’s case for P. This means
that the notion of an argument (λόγος) is doing slightly different work in different
passages of Sextus. It can sometimes refer to a particular piece of evidence (e.g., the
evidence of the sun’s motion, against Zenonian denials of motion) or a particular
sequence of premises and a conclusion (PH I 202). Or, it can refer to the portfolio
or suite of arguments on either side, the total cumulative case that the Skeptic makes
in bringing a question to equipollence. In this sense of “argument,” the cumulative
“argument” against motion will contain many particular “arguments.” It is only
in the cumulative sense that Sextus can declare that, “to every argument there is
an equal argument opposed.” This utterance (φωνή) sums up the state of affairs
that emerges after Sextus has presented a multiplicity of arguments against the
Dogmatist’s position: after a series of particular moves which have supported ~P
or undermined P or done both, the net result is an equality in respect of credibility
and incredibility.
For this reason, then, we can be sure that the Skeptic slogan that “to every
argument there is an equal argument opposed” cannot mean that every particular
Skeptical argument against the Dogmatic advocacy of P must take the form of
an argument advocating ~P. Instead, it must refer to an equality between the
accumulations of particular arguments on either side, that make up the aggregate
cases pro and con. Among the particular arguments, there is no reason why each one
must be an argument directly opposed to the Dogmatic argument as ~P to P, and
indeed there is good reason why each one cannot be of equal and opposite weight
to the Dogmatist’s argument.
It may help to distinguish two claims:

(a) Every equipollence is an equipollence between P and ~P.
(b) Every argument used to produce equipollence is an argument in favor of P
or an argument in favor of ~P.
(a) is correct, but (b) does not follow from it. The equipollence between P and ~P
may be brought about both by arguing for ~P and also by confronting Dogmatic
arguments that raise the credibility of P with Skeptical undermining arguments that
lower the credibility of P.
It is also worth noting that several of the cases cited above, in which Sextus
uses the word ἄπιστος or ἀπιστία in the context of undermining, involve tropes
that Agrippa gathered into the Five Modes. At PH I 117, we saw an invocation of
the Mode of Reciprocity (διάλληλος τρόπος) leading to a verdict of ἀπιστία. At
PH I 61, the allegation of ἀπιστία followed on the charge that the Dogmatist was
attempting to ground P in P, which Sextus may have attacked through the Mode
of Hypothesis.15 This evidence that Agrippan modes produce ἀπιστία supports the
picture above: the Five Modes are intended to undermine the Dogmatist’s case for
P, not to provide constructive arguments for ~P.
Moreover, if this picture of the Five Modes is correct, then we can see a further
way in which the Five Modes complement the Ten. Morison’s picture of the Ten
Modes presents them as means to generate arguments for ~P in response to the
Dogmatist’s arguments for P, while we have argued the Five Modes are meant to
undermine the arguments for P.16
6 CONCLUSION ABOUT
VARIOUS INTERPRETATIONS
In our opinion, the view of Frede is still an open interpretive option, although faced
with many difficulties of formulation—its friends find it suggestive and stimulating,
its critics complain that it has escaped definitive refutation only by remaining
amorphous and underspecified. The more traditional view according to which
Sextus assents to nothing more than his own internal appearances—which Burnyeat,
Fine, and Perin all endorse in various forms—has an easier time with some of the
evidence for Sextus’s epistemic stance but offers very little hope of vindicating
Sextus’s claims to be describing a livable life.17 The aspirations of philosophical
interpretation—either to produce a reading that consistently verifies the whole of
Sextus’s system, or a refutation that clearly shows the conceptual fissures between
its parts—remain to be achieved by future critics, unless they are permanently
frustrated by the limitations of Sextus’s own writings.
NOTES
1. See Popkin (1979), Schmitt (1983), Floridi (2002; 2010).
2. “Medical Notes” (AM VII 202) and “Empirical Notes” (AM I 61). These may have
been alternate titles for one treatise.
142 SKEPTICISM
3. For evidence of this aspect, see PH I 7–10, 12, 26–30.

4. Detailed discussion of the Modes is provided in Section 5 below.
5. Barnes (1982), reprinted along with the two by Frede and the two by Burnyeat, in
Burnyeat and Frede (1997), is equally good and far less obscure than the others—its
laudable clarity perhaps leading to less subsequent literature. Its central polemical
contention, that “urbane” skepticism inevitably collapses into “rustic” skepticism
because of the need for a criterion, was challenged by Brennan (1994), and
definitively rebutted by Schwab (2013).
6. The same phrase, τὰ κατηναγκασμένα, that is, “the necessitated or compelled
things,” is used both to mark out the sphere of things to which the Pyrrhonist will
give their assent in PH I 13 and the sphere of things in which the Pyrrhonist must be
content with “moderate disturbance” in place of total tranquility, in PH I 26 and 30.
7. These are standard claims in Stoic theory, which gave the Skeptics the machinery of
impressions and assent.
8. Fine (2000: 84) gives the example of my considering an argument propounded by
a notorious logical trickster: it appears sound to me, but because of my previous
experience with its author, I do not assent to its really being sound.
9. Whether or not Sextus is referring to the Ten Modes at PH I 167 is something of a
vexed question. For discussion, see Barnes (1990). One of us has argued elsewhere
(Brennan and Lee 2014) that the Agrippan “Mode of Relativity” was originally
intended as a prohibition on epistemic self-grounding and had nothing to do with
the kind of relativity that features in the Ten Modes. That paper also argues that
Sextus conflated the two issues through his own failure to understand Agrippa.
Sextus’s reference at PH I 167 was indeed intended to refer back to the Ten Modes,
but Agrippa would not have assimilated his Relativity to the earlier relativity.
The revisionist proposal of Brennan and Lee has not (yet?) become the standard
view, so in this paper we take the more conventional view about Agrippa’s Mode
of Relativity.
10. In the words of Powers (2010: 157).
11. We are grateful to Morison for discussions about these topics.
12. More precisely: in favor of ~P, or of some other proposition opposed to and
incompatible with P. Sextus explicitly notes in PH I 10 that the opposition is not
restricted to the case of contradiction. What is important is simply that both cannot
be true together. In the main text we use “~P” as a sort of abbreviation for “some
proposition incompatible with P.”
13. Morison (2014) quotes Barnes (1990) for the phrase “bad arguments” as a
summation of the verdict that Sextus intends to convey with the Five Modes.
14. Morison (2011: 293–294): “I take it that all the sets of modes [including the
Five Modes just mentioned] can be made to work in the way the Ten Modes do:
that is, all the sets of modes are devices to be deployed by the sceptic in forming
opposing arguments to the arguments they encounter in the course of their
philosophical investigations.”
15. Brennan and Lee (2014) argue that Agrippa himself would have brought this
under the Mode of Relativity, which (we argue) prohibits grounding a proposition
reflexively in itself.
16. For a different view of the relation between the Five and Ten, see Powers (2010).
17. Fine (2000) adduces one philosopher who thinks that internal appearances may
be sufficient for action. Perin argues that internal appearances allow the Skeptic to
rebut the charge of being reduced to total inaction, but he also concludes that they
are insufficient to provide the Skeptic with actions that are fully human in any richer
sense (see Perin 2010: 113).
REFERENCES
Barnes, Jonathan. 1982. “The Beliefs of a Pyrrhonist,” Proceedings of the Cambridge
Philological Society N.S. 28: 1–29.
Barnes, Jonathan. 1990. The Toils of Scepticism. Cambridge: Cambridge University Press.
Brennan, Tad. 2000. “Criterion and Appearance in Sextus Empiricus,” Acta Philosophica
Fennica 64: 63–92.
Brennan, Tad and Jongsuh James Lee. 2014. “A Relative Improvement,” Phronesis 59:
246–271.
Burnyeat, Myles. 1980. “Can the Sceptic Live His Scepticism?” In M. Schofield, M. Burnyeat,
and J. Barnes (eds.), Doubt and Dogmatism: Studies in Hellenistic Epistemology, 20–53.
Oxford: Clarendon Press.
Burnyeat, Myles. 1983. The Skeptical Tradition. Berkeley: University of California Press.
Burnyeat, Myles. 1984. “The Sceptic in His Place and Time.” In R. Rorty, J. Schneewind,
and Q. Skinner (eds.), Philosophy in History, 225–254. Cambridge: Cambridge
University Press.
Burnyeat, Myles and Michael Frede (eds.). 1997. The Original Sceptics: A Controversy.
Indianapolis: Hackett Press.
Fine, Gail. 2000. “Sceptical Dogmata: Outlines of Pyrrhonism I 13,” Méthexis 13: 81–105.
Floridi, Luciano. 2002. Sextus Empiricus: The Transmission and Recovery of Pyrrhonism.
New York: Oxford University Press.
Floridi, Luciano. 2010. “The Rediscovery and Posthumous Influence of Scepticism.” In
R. Bett (ed.), The Cambridge Companion to Ancient Scepticism, 267–287. Cambridge:
Frede, Michael. 1979. “Des Skeptikers Meinungen,” Neue Hefte für Philosophie, Actualität
der Antike, Heft 15/16: 102–129. (First appeared in English as “The Skeptic’s Beliefs.”
In his Essays in Ancient Philosophy, 179–200. Oxford: Oxford University Press.)
Frede, Michael. 1984. “The Sceptic’s Two Kinds of Assent and the Question of the Possibility
of Knowledge.” In R. Rorty, J. Schneewind, and Q. Skinner (eds.), Philosophy in History,
255–278. Cambridge: Cambridge University Press.
Frede, Michael. 1987. Essays in Ancient Philosophy. Oxford: Oxford University Press.
House, D. K. 1980. “The Life of Sextus Empiricus,” Classical Quarterly 30: 227–238.
Janáček, Karl. 1963. “Die Hauptschrift des Sextus Empiricus als Torso erhalten?” Philologus
107: 271–277.
Morison, Benjamin. 2011. “The Logical Structure of the Sceptic’s Opposition,” Oxford
144 SKEPTICISM
Morison, Benjamin. 2014. “Sextus Empiricus,” The Stanford Encyclopedia of Philosophy.

URL: http://plato.stanford.edu/entries/sextus-empiricus/
Perin, Casey. 2010a. The Demands of Reason: An Essay on Pyrrhonian Scepticism. Oxford
and New York: Oxford University Press.
Perin, Casey. 2010b. “Scepticism and Belief.” In R. Bett (ed.), The Cambridge Companion
to Ancient Scepticism, 145–164. Cambridge: Cambridge University Press.
Popkin, Richard. 1979. The History of Scepticism from Erasmus to Spinoza. Berkeley:
University of California Press.
Powers, Nathan. 2010. “The System of the Sceptical Modes in Sextus Empiricus,” Apeiron
43: 157–172.
Rorty, Richard, Jerome Schneewind, and Quentin Skinner (eds.). 1984. Philosophy in
History. Cambridge: Cambridge University Press.
Schmitt, Charles. 1983. “The Rediscovery of Ancient Scepticism in Modern Times.” In M.
Burnyeat (ed.), The Skeptical Tradition, 225–251. Berkeley: University of California
Press.
Schwab, Whitney. 2013. “Skepticism, Belief, and the Criterion of Truth,” Apeiron 46:
327–344.
Sextus, Empiricus. 1997. Against the Ethicists (Adversus Mathematicos XI), translation and
commentary by Richard Bett. Oxford: Clarendon Press.
CHAPTER ELEVEN
Skepticism in Classical
Indian Philosophy1
MATTHEW R. DASTI
1 INTRODUCTION
There are some tantalizing suggestions that Pyrrhonian skepticism has its roots in
ancient India. Of them, the most important is Diogenes Laertius’s report that Pyrrho
accompanied Alexander to India, where he was deeply impressed by the character of
the “naked sophists” he encountered (DL IX 61). Influenced by these gymnosophists,
Pyrrho is said to have adopted the practices of suspending judgment on matters of
belief and cultivating an indifferent composure amid the vicissitudes of ordinary
life. Such conduct, and the attitudes that it embodied, became inspirations to later
skeptical thinkers. It is a fact that practices of the sort attributed to Pyrrho are richly
evident in a number of ancient Indian philosophical and ascetic movements. On this
basis, attempts have been made to determine the identity of these gymnosophists
and, further, to pinpoint the dialectical tropes within Pyrrhonism that may have
their basis in Indian thought. But despite these attempts, and in the absence of
any new discoveries, these suggestions will likely remain just that. The paucity of
the historical data and the problematic nature of the data itself prevent us from
reconstructing a solid bridge between ancient India and Greek skepticism that may
serve as the basis of robust historical theorizing.2
Classical India does, however, lay claim to a sophisticated and diverse culture of
epistemological reflection, which includes a number of innovative skeptical thinkers
deserving study on their own merits. This chapter is meant to provide such a study,
or perhaps more accurately, a prolegomena for such. After some initial ground-
laying, I will discuss three leading Indian skeptics and situate them historically
and conceptually within the network of competing schools that comprise classical
Indian philosophy. They are Nāgārjuna (c. 150 CE), the founder of Madhyamaka
Buddhism, Jayarāśi (c. 800), associated with the Indian materialist tradition, and
Śrīharṣa (c. 1150), a thinker connected to the school of Nondual (Advaita) Vedānta.
146 SKEPTICISM
2 SKEPTICAL ELEMENTS OF THE PROTO-

PHILOSOPHICAL PERIOD
The proto-philosophical period in Indian thought spans roughly from 700 BCE
to100 CE. The most ancient body of literature in India is the Vedas (c. 1500
BCE for their earliest portions). Though they contain occasional moments of
philosophical, and indeed skeptical, reflection,3 the Vedas are centrally concerned
with the performance of sacrifice and the ritual culture that surrounds it. The seeds
of classical Indian skepticism were laid at a time of departure from the Vedas,
during the middle of the first millennium BCE, amid the tumult associated with the
śramaṇa movement. This was a broad social development connected to the rise of
urban centers and characterized by itinerancy, asceticism, and bold philosophical
speculation. It was motivated by a search for the ultimate solutions to life’s problems
and dissatisfaction with the religious orthodoxy embodied in the culture of Vedic
sacrifice. A number of later traditions, including Buddhism, Jainism, and Upaniṣadic
Hinduism,4 developed out of the śramaṇa milieu.
Certain features of this period lent themselves to skeptical attitudes. Many
of its thinkers express acute concern with our basic cognitive abilities and their
limitations, arguing that whatever the ultimate reality is, it transcends our ordinary
conceptual apparatus. This led naturally to deep strains of apophatic theorizing,
often coupled with the view that meditative gnosis is the only access available to such
higher realities. Two traditions we will consider below, Madhyamaka Buddhism
and Nondual Vedānta, incorporate these ideas into a notion of “two truths,” the
conventional and the ultimate. They do, however, unpack this concept in very
different ways. Tied to such concerns with the limitations of our basic cognitive
abilities was a further contention that the objects identified as the real, medium-
sized dry goods of common experience tend to fall apart under scrutiny. For many
Upaniṣadic thinkers who pave the way for Hinduism, this manifests as a skepticism
about the reality of ordinary objects coupled with an affirmation of unqualified being
itself (the locus classicus for this is the sixth book of the Chāndogya Upaniṣad). For
Buddhists, this manifests in a variety of ways, including a reduction of macro-objects
to sensation-instances that are said to be reified and conceptually constructed into
the objects of ordinary experience.
3 CLASSICAL INDIAN EPISTEMOLOGY AND

THE CENTRALITY OF PRAMĀṆAS
The classical period of Indian philosophy begins soon after the start of the Common
Era and lasts until roughly 1800 CE. On one side, the classical period is bracketed by
the proto-philosophical period, and on the other side, by the modern period, which
is centrally preoccupied by the confrontation with the West and by the problems of
modernity more generally. (Even today, learned pandits carry the classical traditions
onward, though it is an increasingly marginalized feature of Indian intellectual
culture.) In the classical period, we find the coalescence of specific schools and
SKEPTICISM IN CLASSICAL INDIAN PHILOSOPHY 147
sub-schools of thought that have their roots in the proto-philosophical period. A

number of Buddhist, Jaina, and Hindu schools of thought may be traced to various
earlier cultural and religious movements, but they become distinctly philosophical in
the classical period as they become explicitly concerned with both system-building
and defending their holdings through dialectical engagement with rivals. Classical
philosophers typically work within a school, developing their own innovations
through commentaries or sub-commentaries on the foundational texts of their
particular tradition.
Another characteristic of the classical period, of great importance to this study,
is the centrality of pramāṇas, “knowledge sources.” Epistemology in classical India
largely consists in the identification and analysis of irreducible sources of knowledge
(e.g., perception, inference, testimony). Competing schools and thinkers articulate
different lists of fundamental pramāṇas as well as analyses of specific pramāṇas, and
extensive debate over knowledge sources is a key feature of Indian philosophy.5
These knowledge sources are not only taken to be the cognitive resources by
which we as individuals guide our lives. They are also the basis of philosophical
argumentation. Illustrating that my belief or the holding of my school is supported
by genuine pramāṇas places it on firm ground. Indeed, the early history of pramāṇa
theory is intertwined with a robust culture of formal public debate,6 and some of
the first systematic Indian treatments of epistemology are integrated with manuals
on debate theory and dialectics. The very survival of a school typically depended
on patronage, and the primary way to win support was public discourse and debate
with rivals. A particular style of debate, the “purely destructive” mode (vitaṇḍā),
was explicitly recognized in works on dialectics and appropriated by philosophical
skeptics. One arguing in this mode disavows any personal thesis to defend and sets
himself only to destroy other views. This method was controversial and an object
of concern for realists, some of whom argued that it involves an inherent self-
contradiction of some kind or other.7 The skeptics we will study were acutely aware
of this charge and defended themselves against it.
In summation, pramāṇa discourse, as well as an allied culture of formal debate,
provide the conceptual landscape for classical Indian epistemological reflection, and
skeptical attacks on the possibility of knowledge presuppose this landscape.
4 NĀGĀRJUNA AND MADHYAMAKA BUDDHISM

The first Indian skeptic we will consider is the founder of Madhyamaka Buddhism,
Nāgārjuna (c. 150 CE). To understand Nāgārjuna’s project, we need to take note
of some features of Buddhist thought traceable to the historical Buddha himself
(Siddhārta Gautama, fifth century BCE). At the time he began his career as a
philosophical and religious teacher, the Buddha was merely one participant in
the śramaṇa movement. After years of self-abnegation and meditative practice, he
became convinced that he found a solution to the fundamental problem of life,
which he expressed within the Four Noble Truths: life is pervaded by suffering,
we suffer because of our desires, we may uproot suffering by uprooting desire, and
148 SKEPTICISM
there is a way of life (the Eightfold Path) that allows us to do so. Expressing and
developing this simple formula, the Buddha became one of the most influential
thinkers in ancient India and, eventually, much of the Asian world.
A few features of the Buddha’s approach to philosophy are relevant to this study.
In short, it tends to be therapeutic, pragmatic, and anti-metaphysical. Philosophy
is not meant to capture the right theory about the world, but rather to remove
problematic misconceptions. The Buddha thus treats human beings’ natural tendency
to schematize the world as an intellectual analogue of desire: an attempt to harness
the world, a clinging of sorts, which perpetuates suffering. We may notice the way in
which these elements of the Buddha’s thought lead him, and later Buddhist skeptics, to
dismiss elaborate philosophical systematizing as a massive unwholesome distraction.
Moreover, the Buddha is sensitive to the way in which our conceptual and linguistic
frameworks dictate our experiences and understandings. This leads to a general anti-
realist sensibility. Finally, he problematizes the commonsense notion of the world as
populated by enduring, self-standing macro-objects of common experience. These
aspects of the Buddha’s teachings collectively set the stage for various tendencies in
mature philosophical Buddhism, including empiricism, reductionism, idealism, and
skepticism. And with the last, we may now consider Nāgārjuna.
Nāgārjuna flourished while the classical period of Indian philosophy was
taking shape and various systems of thought, Buddhist and otherwise, were being
constructed.8 He opposed what is, from the Madhyamaka9 perspective, a common
practice of philosophers: the enshrinement within philosophical argumentation
of our pre-theoretical tendency to project a world of independent, self-standing
objects. His philosophical opponents are those who espoused various sorts of
realism, including Buddhist scholastics and Hindu realists like the Nyāya (“Logic”)
and Vaiśeṣika (“Particularism”) schools. The central theme of Nāgārjuna’s most
important work, The Fundamental Verses of the Middle Way (MMK), is emptiness
(śūnyatā), the lack of intrinsic existence (svabhāva, “self-nature”), which, he
argues, holds for all phenomena. The term svabhāva takes on various related
meanings within Madhyamaka thought. Fundamentally, it refers to something
whose existence is independent of other conditions or factors. 10 Emptiness, the
denial of intrinsic existence, is taken by Nāgārjuna to be equivalent to dependent
co-origination (pratītyasamutpāda), the inexorable interdependence of all
phenomena.11 In short, Nāgārjuna argues that nothing has an ontologically stable
identity independent of other phenomena. This interdependence is not merely
causal, in the sense that we can never separate an entity’s existence from the rich
network of conditions upon which it depends and which it in turn influences.
More radically, it is conceptual: the nature of any particular thing under scrutiny
is determined by the conventions, concerns, linguistic practices, etc. of those doing
the scrutinizing. Any attempt to isolate a certain moment or node within the rich
causal, conceptual, and linguistic networks that constitute life as experienced, while
claiming that this node has a nature existing independently from such networks, is
mere reification and projection.
The notion of emptiness is thus not merely a report on our own cognitive
limitations, that we have no cognitive access to what lies beneath the phenomena-
as-experienced, as Locke speaks of substance in general or as Kant speaks of the

thing in itself. Nor is it merely semantic, in that there are limits to the capacity
of language to capture extra-linguistic reality. It is ultimately ontological. There is
nothing there beneath the appearances. The notion of underlying, independently
existing real entities that undergird the contents of phenomenal awareness is simply
incoherent. We should underscore that this is not merely a kind of idealism, since
there are no enduring, intrinsically existing minds that ground the world either.
Subjects are as empty as the phenomena they experience.12
Nāgārjuna’s primary method to advance this point is argumentum ad absurdum:
positing intrinsically existing entities leads to inescapable contradictions. He
provides a wide array of arguments to the effect that various phenomena
championed by philosophers (e.g., causation, motion, indwelling selves, substances
that bear properties) lack svabhāva.13 One of his major lines of argumentation may
be summarized as follows14:
1. x enters into relations with other nodes within causal, conceptual, and
linguistic networks.
2. x possesses intrinsic existence (svabhāva).
3. If something possesses intrinsic existence, then its fundamental identity is
isolated from the networks of conditions (causal, conceptual, linguistic)
within which it is embedded.
4. If something’s fundamental identity is isolated from the networks of
conditions (causal, conceptual, linguistic) within which it is embedded, then
it does not enter into relations with other nodes of such networks.
5. Therefore, x does not enter into relations with other nodes of causal,
conceptual, and linguistic networks.
The contradiction between premise 1 and the conclusion leads to the rejection
of premise 2. This lack of independent existence further entails that precisely
discriminating some real entity, independent from the conditions within which it is
embedded, is impossible.15 At best, we merely select certain features of the causal/
cognitive nexus as most relevant to our concerns. The fact that Nāgārjuna preserves
premise 1 underscores the fact that, despite the claims of his opponents, he does not
intend his imputation of emptiness to be nihilistic. In fact, he argues, it is the very
opposite: emptiness is a condition for something’s being part of the common world
of causal interaction.16
Nāgārjuna’s arguments in support of emptiness are meant to undermine realists’
attempts to gain purchase on a substantial external world and are thus skeptical
in spirit. But they are not distinctively epistemological in form. In his Dispeller
of Disputes (VV), however, Nāgārjuna takes direct aim at pramāṇa (knowledge-
source) epistemology. The context of his attack is a response to Nyāya realists.
Naiyāyikas (Nyāya philosophers), as far back as the original Nyāya-sūtras (c. 200
CE), have taken the Madhyamaka position to be a kind of skeptical nihilism that
is guilty of the self-referential incoherence mentioned above. They take emptiness
to mean insubstantiality and claim that the Buddhist must either accept that there
150 SKEPTICISM
are non-empty pramāṇas, which support his assertions, or give up the attempt
to prove his point. Nāgārjuna’s response is twofold. First, he denies that he
advances a robust thesis about the nature of a real external world, claiming “I
have no thesis” (VV 29).17 Because of this, he claims to be innocent of the charge
of self-contradiction. Second, and more aggressively, he turns the argument back
on the realist, by arguing that the notion of pramāṇas as foundational sources of
knowledge is unsustainable (31–51):
If you hold that the various objects are established by pramāṇas, then again,
you must explain how it is that the pramāṇas are themselves established (31).18
If the pramāṇas were established by other pramāṇas, then there would be
an infinite regress. Then, neither the beginning, middle, nor end would be
established (32). But if you think that the pramāṇas are established without
need for other pramāṇas, then your position (that claims about an object’s
existence or nature must be supported by appeal to pramāṇas) has been
abandoned. There is inconsistency in this, and you must provide a specific
reason for it. (33)
So far, Nāgārjuna has presented two lemmas: an infinite regress of pramāṇas and
the mere positing of pramāṇas without epistemic grounding, both of which are
undesirable for the reasons he cites. Next, he presents a third lemma, beginning
with the idea that the pramāṇas may be self-established. As opposed to being
ungrounded, this option suggests that pramāṇas have the special power to provide
epistemic support for themselves by dint of their own functioning. From our
perspective, this option seems closest to a classical formulation of foundationalism:
there are cognitive states that are both self-supporting and able to support other
states. Nāgārjuna’s interlocutor suggests that this may be akin to the way that a fire
not only illuminates various objects but also itself.
After finding fault with the details of the fire analogy, Nāgārjuna argues:
If the pramāṇas are self-established, then the objects of knowledge are also
independent (of the pramāṇas). For self-establishment is independent of any
other thing (40). If the pramāṇas are established independently of their objects,
then they would be the pramāṇas of nothing at all. (41)
If the pramāṇas were truly self-established, this would exclude us from accounting
for the veracity of, for example, perception by appeal to the objects it normally
reveals. Track-record arguments, familiar to reliabilist theories, would be off
the table. The very notion of self-establishment would rule out such an appeal.
Putting aside the practical question of how we could then discriminate legitimate
knowledge sources from flawed imposters (perception vs. mere hallucination),
Nāgārjuna highlights the way in which this approach would sever the intimate
tie between knowledge sources and their proper objects. I would suggest that this
argument may be placed within the family of arguments directed against “the
given” and, more precisely, against strong versions of foundationalism. Nāgārjuna’s
contention is that, if foundational cognitive states were truly walled off from the
support of other inputs, this wall would be impassable in both directions. The
very act of insulating these foundational cognitive states makes them impotent as
sources of epistemic support for any particular domain of putative objects or truth
claims. His argument may be understood in two distinct, but complementary ways.
Epistemically, as we have seen, we would be unable to support the veracity of, for
example, perception by appealing to its track record of success. I could no longer
say that perception reliably provides me cognitive access to “this tree right here”
as I point to a tree in my backyard. Conceptually, if self-establishment is taken as
a kind of logical atomism, then the concept of perception could not include its
capacity to provide cognitive access to a distinct domain of physical objects. In his
auto-commentary, Nāgārjuna remarks: “If they are to be pramāṇas of something,
they cannot be independent of their objects.”
Given the way in which this lemma instantiates Nāgārjuna’s general attack on
intrinsic existence, we now see a close connection between his critique of pramāṇa
epistemology and his general philosophical project.
The fourth and final lemma is acceptance of the interdependence of pramāṇas
and their objects:
If, conversely, you accept that “establishment of pramāṇas always occurs in
reference to their proper objects,” there would be the establishment of what is
already established. (42)
If the realist opts for a more holistic approach to the establishment of pramāṇas,
allowing the objects of knowledge to play a role, then, in the course of justifying his
views, appeal to knowledge sources would merely serve the purpose of reiterating
what he already accepts. The existence of knowledge sources is presupposed, as they
are what provide cognitive access to their proper objects. The existence of objects of
knowledge is also presupposed, as they are what we appeal to in order to establish
the existence and nature of knowledge sources.
In succeeding verses, Nāgārjuna argues that there is no way to escape this
circularity in a way that salvages the realists’ project. They would have to hold that
the objects of knowledge are established independently of the pramāṇas to use them
as a non-circular ground for establishing the pramāṇas (43). But if this is the case,
what need is there for appealing to pramāṇas in order to ground knowledge of the
objects (44)? This strategy would, further, reverse the putative dependence relation
between the sources and objects of knowledge (45). Proposing a complete circularity
between them ends up establishing nothing (46). In conclusion, “the pramāṇas
cannot be self-established, nor established by other pramāṇas, nor established by
appeal to the objects of knowledge, nor established without reason” (51). And thus,
they are unsuitable to play the foundational role assigned to them by the system-
building realist philosophers.
It is important to see that Nāgārjuna’s purpose here is not to refute commonplace
epistemic practice, the use of perception, inference, testimony, and the like
in navigating our daily lives (see Siderits 1980: 320). It is rather to refute the
contention that the pramāṇas have the capacity to justify metaphysical realism.19
And this refutation, in consonance with traditional Buddhist attitudes, is meant to
support a life aimed at enlightenment. The affirmation of emptiness, supported by
152 SKEPTICISM
skeptical argumentation, provides an ideative support for meditative praxis, meant

to dislodge the tangled knots of reification and desire, and ultimately terminate in
peace and freedom from suffering.20
5 JAYARĀŚI AND THE MATERIALIST TRADITION

The materialist tradition (Lokāyata, Cārvāka) is a noteworthy exception to the
typically ascetical flavor of classical Indian philosophy.21 Most of the original
works of the materialists are unfortunately lost to us. We tend to find bits and
pieces preserved by their rivals, who often represent them as interlocutors worthy
of contemptuous dismissal. The materialists themselves dismissed the religious
ambitions of other Indian schools (sometimes with quite humorous rhetoric), and
their skepticism stems from their anti-religious and anti-speculative approach to
the good life. There is no immortal soul. We have one life to live, and the best that
life can offer is pleasure; all else is vanity. In a corresponding movement, personal
experience is all that we can take seriously in matters of knowledge, while inference
and testimony—the primary means to bolster claims about transcendent realities—
are never sources of knowledge proper.22 Against inference, materialists note that
it relies on a recognized connection between an inferential sign and that which is
signified (e.g., between smoke and fire). But for this sign to underwrite knowledge
of the signified, it must be known to be invariably and non-accidentally connected
to it. How do we acquire this knowledge? Not by perception, as it operates upon
particulars in the present moment. Nor by inference, as this would trigger an
infinite regress. Nor testimony; since it too relies on a connection between sign and
signified, it would mirror the problems of inference. Furthermore, we can never
be sure that apparently secure inferential relata are not merely accidentally and
“locally” connected. Our ability to know is limited in scope. We lack the resources
to refute the possibility of unknown causal factors that would engender deviation
at some time or place.
Jayarāśi (c. 800 CE) is connected to the materialist tradition, though he seems
willing to develop his own conclusions.23 One sign of his independence is that
he takes the deep skeptical strand within materialism and expresses it across
the board, attempting to refute leading analyses of knowledge and knowledge
sources, including perception. He abandons the somewhat clumsy materialist
formulation that “perception is the only legitimate knowledge source,” fraught with
the potential for easy refutation, and devotes his work to a more sophisticated,
skeptical vindication of ordinary life: “The philosophers’ technical categories
being subverted, ordinary linguistic practices are appropriate, and should not be
subject to critical examination” (my translation of text in Solomon 2010: 228).
In the Tattvopaplavasiṁha (The Lion that Subverts all Philosophical Categories),
his method is straightforward “destructive” argumentation. For any technical view
put forth by a philosopher (typically analyses of the major knowledge sources
or key metaphysical tenets), he tries to illustrate that it is either patently false or
inadmissible given the interlocutors’ own set of truth commitments.
By way of example, below is a summary of one his arguments against the analysis
of testimony given by Nyāya: “testimony is an assertion by an authoritative speaker”
(āpta-upadeśa śabdaḥ; Nyāya-sūtra 1.1.7), where “authoritative” is meant to convey
a speaker’s own veridical cognition of some fact and an honest willingness to convey
it to another.
1. Granting provisionally that there is a property, authoritativeness, possessed
by some persons, it would underwrite the positive epistemic status of
testimonial beliefs either by (a) its mere existence, devoid of any particular
production of testimonial knowledge, or (b) its actual production of
testimonial knowledge.
2. Option (a) cannot be the case, since authoritativeness cannot be a cause of
positive epistemic status if it is causally unproductive.
3. Regarding (b), authoritativeness would have to produce knowledge either (c)
by itself or (d) with the help of auxiliary causal factors.
(3a) Option (c) is not appropriate, as you (Nyāya) do not accept it. And in
any case, it would contradict the experience of causal fluctuations that
are attributed to the influence of auxiliary causal factors.
(3b) If you hold (d), then authoritativeness is neither necessary nor sufficient
for the production of veridical testimonial cognition. Consider the
statement “the boy has nava blankets” (the Sanskrit term nava could
be understood to mean “new” or “nine”). Seeing a boy with a new
blanket, an authoritative person may utter the above statement. Owing
to auxiliary causal factors that beset the listener, however, his statement
produces the false cognition “the boy has nine blankets.” Similarly, a
person intent on deception may say “the boy has nine blankets,” but
owing to auxiliary causal factors, his statement produces for a listener
the true cognition “the boy has a new blanket.”
4. Therefore, the Nyāya analysis “testimony is an assertion by an authoritative
speaker” fails.
This is but one example. But Jayarāśi’s contention is that such destructive argumentation
is possible for any and every technical definition put forth by dogmatic philosophers.
6 ŚRĪHARṢA AND NONDUAL VEDĀNTA

The final skeptic we will consider is Śrīharṣa (c. 1150), a thinker within the school
of Advaita (“Nondual”) Vedānta, living roughly four centuries after Śaṅkarācārya,
the figure who gave definitive form to the school and whose works serve as its
systematic foundation.24 Vedānta is a Hindu school, or rather a large family of sub-
schools, devoted to interpreting the Upaniṣadic scriptures as well as articulating
a coherent, defensible metaphysics that makes sense of the Upaniṣadic view of
the world. The Upaniṣads are heterogeneous and unsystematic in nature, but a
common theme that emerges is that the various objects of ordinary experience are
154 SKEPTICISM
ontologically dependent upon and find their ultimate value as derivative of the
fundamental ground of being, Brahman. Realizing Brahman—in large measure,
realizing our own intimate relationship with it—is considered the key to achieving
supreme felicity and liberation from the cycle of rebirth.
The various sub-schools of Vedānta fall into two major (and hostile) camps,
realism and nondualism, based on their interpretation of the nature of Brahman and
the dependence relation between Brahman and the world. The realists affirm the
existence of a mind-independent world but suggest that, without reflection upon
Upaniṣadic teachings, we remain ignorant of the way in which it is sustained and
pervaded by Brahman, which is understood as a God-like being. Nondual Vedānta
is contrastingly a monist idealism. For the nondualists, Brahman is a homogeneous,
unified, self-aware consciousness. The world of common experience, which
presupposes multiplicity, is ultimately unreal, a persistent illusion, as is the notion of
a creator God. The world is grounded in Brahman in the sense that everything, even
illusion, is grounded in awareness. The difference between the two camps may also
be understood in their interpretation of Brahman’s totality. The Upaniṣads hold that
Brahman is everything. For realists, this means that Brahman undergirds, unifies,
and makes coherent the multiplicity of real objects in the world; for nondualists,
this means that the only real thing is homogeneous conscious awareness; all else is
ultimately illusion.25 While it is outside the scope of this study, we should also note
there are historical and conceptual continuities between Madhyamaka Buddhism
and Nondual Vedānta in spite of the fact that the former is anti-Vedic and the latter
is among the more orthodox of the Vedic Hindu traditions.
This nondualist view is sometimes advanced through sophisticated epistemological
arguments that combine a notion of cognitive sublation (bādha) with the self-
illumination/justification of conscious awareness. A cognitive state is sublated when
it is defeated or undermined by a contrary awareness (e.g., an illusion of a man in the
distance is sublated by expert testimony that the thing in question is a tree stump).
A key nondualist contention is that all contentful awareness is in principle capable
of being sublated, but unqualified awareness simpliciter is not. This is because all
sublation takes place within the domain of awareness. To put it slightly differently,
while any particular cognitive presentation may be doubted, the fact of conscious
awareness cannot. And such unqualified awareness is taken to be identical to Brahman,
the ultimate ground of being.26 This argument provides a helpful way to conceptualize
Śrīharṣa’s project in his dialectical masterpiece, the Khaṇḍanakhaṇḍakhādya (The
Sweets of Refutation): while Descartes’s skepticism is the first half of a two-part
movement that erodes conviction in the external world, pivots on the indubitability of
conscious self-awareness, and rediscovers the world on more secure footing, Śrīharṣa
simply stops at the end of the first movement. He seeks to undermine any positive,
contentful cognition (belief, proposition, etc.), such that the only thing left standing is
unqualified, homogeneous conscious awareness itself (again, equated with Brahman,
the ground of being revealed in the Upaniṣads). As Phillips (1995: 88) puts it, he
would “back into success” by refuting any view that asserts the reality of multiplicity.
In support of this objective, Śrīharṣa argues against a wide variety of philosophical
views, commitments, and theses (including, for example, a scathing attack on
universals). Like Jayarāśi, he often sets the views of a philosopher or school against
themselves, in support of his contention that that systematic realism cannot be
maintained.27 He succinctly expresses this in the claim that “all technical analyses
are untenable (sarvāṇi lakṣanāṇi anupapannāni)” (see Granoff 1978: 23, and Jha
1986: 82). Harkening back to our discussion of the purely destructive mode of
debate (vitaṇḍā), Nicholson (2009: 94) suggests that it “achieves its most developed
expression” in the work of Śrīharṣa, who contends that any understanding of reality,
pre-theoretic or philosophical, which considers distinctness (bheda) real, collapses
under philosophical scrutiny. Śrīharṣa explicitly states that the basic form of his
arguments may be modified in order to attack against positions that he doesn’t
directly address, and thus used as modes, so to speak, in the continued refutation
of realism.28
We my recall that Indian realists contend that a complete denial of knowledge
sources is impossible, as any thesis put forth (including the skeptical hypothesis)
must derive its probative force from somewhere (see note 6). One way to express
this contention is that one cannot engage in philosophical debate without at least
tacitly accepting that there are knowledge sources that determine the legitimacy of
argumentative moves and provide for the rules by which defeat and victory may
be objectively recognized. This would, of course, undermine Śrīharṣa’s program
from the start. Naturally, he tasks himself with deflecting this charge.29 In short, his
response is as follows: it would be misleading for the realist to cast Śrīharṣa’s own
position as starting from a denial of knowledge sources. He is, rather, uncommitted
to the truth of their existence. All he claims to possess, and all that that is needed
to debate their legitimacy, is an idea or notion of knowledge sources, and indeed,
not prior knowledge of their existence. He need no investment in their reality
or in a theory that ties them to an external world of causal conditions. Clearly,
those who attack the legitimacy of knowledge sources (identified by Śrīharṣa as
Mādhyamika Buddhists, Cārvāka materialists, and Advaita Vedāntins) participate in
philosophical discussion and debate. This gives the lie to the notion that accepting
knowledge sources is a prerequisite for dialectical engagement. And further, the
realists themselves engage in vigorous, extensive debate with skeptics! Clearly they
themselves do not in fact accept the rule that debate must require a preliminary
acceptance of knowledge sources.
How then, is the debate between the skeptics and the realists to be prosecuted?
What undergirds what we take to be the standards of success and failure in debate?
Śrīharṣa argues that it is simply a set of agreed-upon rules. This is the only thing
that both sides must accept for debate to occur. And such rules are a matter of
convention and nothing more; certainly they need not be objects of knowledge.
Accepting those debate-governing conventions that have passed the test of time and
seem to function effectively is the baseline requirement for two parties to engage in
debate effectively. Śrīharṣa deftly supports this line of argumentation by reflection
on our general epistemic condition: our action-guiding cognitive resources are as
a rule usually far below the standards of ultima facie justified knowledge. All of
us, realists included, must rely on cognitions that are themselves unverified and
therefore, provisional, on pain of infinite regress. And relying on these, without any
156 SKEPTICISM
sense of deep conviction or pretense to knowledge, is all the skeptic needs to engage
in debate and, indeed, to leverage his attack on knowledge itself.30
NOTES
1. This chapter was generously supported by a grant from Bridgewater State University’s
Center for the Advancement of Research and Scholarship. I would also like to thank
Diego Machuca and Baron Reed for helpful comments and suggestions.
2. See Frenkian (1957) and Flintoff (1980) for arguments that Pyrrho was indeed
influenced by Indian philosophers. McEvilley (1982) identifies a number of
remarkable similarities between dialectical tropes within Pyrrhonism and
Madhyamaka Buddhist thought, though he ultimately remains dubious of Indian
influence. For concise appraisals of the issue see Hankinson (1995: 58–64) and
Bett (2000: 169–178).
3. The famous creation hymn (Nāsadīya sūkta) of the Ṛg Veda (10.129) contains a
delightfully skeptical conclusion regarding the ultimate nature of the cosmos.
4. “Hinduism” is a problematic term for various reasons, including the fact that it
refers to hundreds of philosophical and religious movements under a single heading.
For our purposes, it refers to persons and traditions that accept the authority of the
Vedas—if only nominally—along with allied texts, cultural norms, and practices.
It primarily serves the purpose of distinguishing these groups from those which
explicitly rejected the authority of the Vedic tradition, including Buddhists, Jainas,
and Cārvāka materialists. By “Upaniṣadic Hinduism” I refer to an influential
strain of Hindu thought that takes many of the motivations and concerns of the
śramaṇa period and incorporates them into the Vedic ritual tradition, in the spirit
of reformation and not outright rejection.
5. See Matilal (1986) and Mohanty (2000: 11–40) for introductions to pramāṇa theory.
6. See Solomon (1976: chs. 1–3) for evidence of the primacy of debate within a number
of ancient Indian traditions of thought. Matilal (1998: 32) suggests that debate was
“a preferred form of rationality” in classical India. See Matilal (1986: 80–93; 1998:
32–59) for further discussion.
7. Considerations of space prevent us from extensively considering responses
to skepticism by Indian realists. This short aside will have to suffice: the self-
contradiction alleged may be couched in various ways, including self-referential
incoherence and pragmatic self-contradiction. In simpler versions, dating at
least to the earliest commentary on the Nyāya-sūtra, the realist argues that
since views cannot be advanced without reliance upon knowledge sources, the
skeptic’s position is abandoned in the very act of making his case. Likewise,
one proffering an argument must have some objective or concern that he holds
to be normatively binding. The skeptic must either admit this, betraying his
skepticism, or be dismissed as a purposeless madman (see Matilal 1986: 87 and
Ganeri 2001: 11). In more sophisticated versions, the realist argues that the
very concept of falsehood is parasitical on truth and therefore doubt cannot be
motivated without an acknowledged background of true cognition (see Dasti
2012). Other anti-skeptical arguments are pragmatic in tone: we appeal to
knowledge sources (and engage in epistemological reflection upon them) because

we want to be guided well in support of our goals in life. A priori skepticism
would stultify us from the outset, undermining this very project (see Phillips
2012: 18–19).
8. Phillips (1995: 15) suggests that Nāgārjuna was a catalyst for the “professionalization
within all Indian schools” of this time, as his dialectical attacks provoked them into
“rethinking [their core positions] and tighter argumentation.”
9. The terms mādhyamika and madhyamaka refer to an individual or group devoted
to the “middle way,” a term commonly used by Buddhists to describe the Buddhist
lifestyle. It has come to be specifically identified with the tradition that traces itself
to Nāgārjuna. In his earliest sermons, the Buddha characterizes his path as a middle
way between excessive renunciation and excessive hedonism. Elsewhere, he describes
his attitude toward the reality of the world as a middle way between the complete
affirmation of being and its complete denial. Nagarjuna (MMK 15.7) directly cites
the sermon where this latter notion is put forth, and, as Garfield (1990: 223) notes,
this usage sets the stage for Nāgārjuna’s rejection of both metaphysical realism and
nihilism in favor of the “middle way” of conventionalism. See Huntington (1989),
Siderits (2007: 180–207), and Westerhoff (2009), for philosophically oriented
introductions to the school. See Garfield (1995), Siderits and Katsura (2013), and
Westerhoff (2010) for translation and commentary on the texts that are central to
this study.
10. MMK 15.2 holds that svabhāva is unconstructed (akṛtrimaḥ) and not dependent on
something else (nirapekṣaḥ paratra). See Westerhoff (2009: 19–52) for a study of the
primary Mādhyamika understandings of svabhāva. A helpful characterization that
Westerhoff borrows from Tibetan commentators on Nāgārjuna is that something
possessing svabhāva would have to exist “from its own side” and thus independently
of our conceptual engagement with it. Tillemans (2001: 9) argues that for Nāgārjuna,
the two central aspects of svabhāva, “findability under analysis” (something like
conceptual irreducibility) and causal independence, are mutually entailing.
11. Dependent co-origination was originally put forth by the Buddha as an account of the
factors that collectively perpetuate a personality stream’s continued misconception
and suffering, binding it to continued rebirth in saṁsāra. According to this notion,
there is no original, fundamental cause of our existence, but a self-perpetuating
cycle of mutually dependent causes. We should underscore that while Nāgārjuna’s
usage is tied to the original notion, he clearly develops it according to his own
philosophical understanding.
12. This is the primary concern of Chapter 8 of MMK.
13. Westerhoff (2009) catalogues a number of these arguments.
14. For examples of the reasoning captured in this argument, see MMK 13.4–6; 15.2;
20.17; 21.17; 22.22; 24.38.
15. MMK 18.10 therefore states: “Whatever comes into being dependent on another is
not identical to that thing. Nor is it different from it” (translation by Garfield 1995:
252). This theme is developed in Chapter 10 of MMK.
16. MMK 1.10; 15.8–9; 24.7–20, 36.
158 SKEPTICISM
17. I will cite specific passages of this text by verse number. See Westerhoff (2010:
63–65) for a concise discussion of this enigmatic disavowal.
18. Translations are mine unless otherwise noted. I have used the Sanskrit manuscript
of the Dispeller of Disputes by Yonezawa (2008). Note that derivations of
the verbal roots sidh and sādh are used in this passage, conveying the notion
of accomplishment or establishment. Here they convey a sense of epistemic
establishment or proof.
19. On these grounds, Siderits (1988; 2003: 150–153) argues against using the label
“skeptic” for Nāgārjuna and that he is better understood as an anti-realist devoted to
undermining correspondence theories of truth and foundationalist epistemologies.
Without disagreeing with his cogent analysis of Nāgārjuna’s motivations or anti-
realist sensibility, I would contest that the convention of calling Nāgārjuna a skeptic
(e.g., Matilal 1986 passim) is appropriate, so long as it is understood within the
context of his philosophical project. In defense of this, a few remarks are in order.
I would suggest that the definition Siderits provides of a paradigmatic skeptic is too
narrow. To paraphrase, he defines the skeptic as one who agrees with the realists that
there is a theory that best captures reality, but denies that we are capable of achieving
it. As an anti-realist conventionalist, Nāgārjuna would then reject the very realist
pretentions that give rise to skepticism. But given Siderits’s definition one could
even call into question whether Sextus Empiricus is a skeptic. By modus tollens,
then, the definition must be too narrow. To elaborate: like Nāgārjuna, Sextus seeks
to undermine confidence that an objective theory of reality is even possible as he
repeatedly calls attention to the fact that particular phenomena are experienced in
radically different ways according to the presuppositions and cognitive backgrounds
of the subject. Anyone arguing in favor of her own perspective would do so from
within it and thus still be part of the dispute (PH 1.112). Sextus further deploys
argumentative tropes akin to those of Nāgārjuna (e.g., the two modes; PH 1.78–79)
that are meant to undermine any standard by which the nature of the world may be
definitely adjudicated. We may, therefore, speak of Nāgārjuna as a skeptic as we do
of Sextus, insofar as he argues that pretentions to objective knowledge of a mind-
independent reality are unjustifiable, fruitless, and incoherent. Knowledge of that
sort is not possible. In this way, just as Sextus allows for individuals to adopt conduct-
governing standards while disavowing opinions regarding the true nature of things
(PH 1.21–24), Nāgārjuna would allow for a limited sort of appeal to “knowledge
sources” for the needs of common life, while explicitly denying that he maintains
any thesis about the real nature of the world (VV 29). I would conclude, therefore,
that given the range of meanings associated with “skeptic,” it is not misleading, and
is indeed, appropriate for Nāgārjuna, as long as it is understood in the context of
his broader project. For a more developed comparison between Madhyamaka and
Pyrrhonian skepticism, see Dreyfus and Garfield (2011).
20. Huntington (1989: 78–83) and Westerhoff (2009: 13) underscore the fact
that in Madhyamaka thought theoretical reflection must be conjoined with
contemplative practice.
21. The first chapter of the Sarvadarśanasaṃgraha, a medieval Indian doxography,
is devoted to a reconstruction of the materialist view: see Agrawal (2002). For
translations of Jayarāśi’s central work, see Franco (1994) and Solomon (2010).
For article-length studies of Jayarāśi, see Balcerowicz (2011) and Mills (2015).
22. While this is a common characterization of the materialist view, examining

extant fragments of commentaries on the Cārvākasūtras (Sūtras on Materialism),
Bhattacharya (2011) illustrates that some materialists are willing to allow inference
to stand as a source of knowledge, so long as it is entirely grounded upon and
derivative of perceptual experience.
23. See Balcerowicz (2011: sect. 1.4) for a succinct summary of available evidence
regarding Jayarāśi’s relationship to the materialist school.
24. See Granoff (1978) and Phillips (1995) for influential studies of Śrīharṣa. See Jha
(1986) for a translation of his central work. Ram-Prasad (2002) has a number of
insightful essays on Śrīharṣa. See Potter (1981) for an introduction to early Advaita
Vedānta. See Ganeri (2011: 127 n. 12), however, for a suggestion that the evidence
for Śrīharṣa’s connection to the Nondualist school is weaker than often claimed.
25. The use of “ultimate” is important here. Classical Nondualism tends to speak of
the objects of ordinary experience as conventionally real, but ultimately unreal. As
opposed to a married bachelor, which is unequivocally false, the computer in my
office is conventionally real, even if ultimately illusory. Nondualists therefore tend
to use a third category (other than “real” or “unreal”) for objects that are merely
conventionally or phenomenally real; they are “indeterminate” (anirvacanīya). See
Ram-Prasad (2002) for a study of this notion, which deftly ties its development to
Indian debates over epistemology and metaphysical realism.
26. Deutsch (1969: 15–26) provides a clear explication of this notion. Phillips (1995:
78–81) discusses its development in Śrīharṣa. For translation of the relevant
passages, see Jha (1986: 25–57).
27. Given this project (and with noteworthy affinities to Nāgārjuna), a central target of
Śrīharṣa’s attacks is the Nyāya school. Phillips (1995) illustrates the way in which
Śrīharṣa’s criticisms were a primary catalyst for the rise of “New Nyāya,” which
defends traditional Nyāya positions with increased analytic rigor and refinement.
28. See Jha (1986: 704). For Śrīharṣa’s explicit appeal to the legitimacy of vitaṇḍā, see
Jha (1986: 30–31). Extended discussion of this feature of Śrīharṣa’s argument is not
possible in this paper, but see Granoff (1978: 1–69) for an example and analysis of
his method, applied to the analyses of knowledge provided by his interlocutors.
29. See Jha (1986: 3–24). Also see Granoff (1978: 71–83).
30. Considerations of space preclude discussion of Śrīharṣa’s use of Gettier-style
counterexamples to undermine Nyāya definitions of knowledge. This would,
of course, be a natural point of resonance with contemporary epistemology. For
further discussion, see Matilal (1986: 135–137) and Ganeri (2011: 127–130).
REFERENCES
Agrawal, Madan Mohan (trans.). 2002. The Sarva-Darśana-Saṃgraha of Mādhavācārya.
Vrajajivan Indological Studies, Vol. 7. Delhi: Chaukhamba Sanskrit Pratishtan.
Balcerowicz, Piotr. 2011. “Jayarāśi,” The Stanford Encyclopedia of Philosophy (Spring
2011 edition), Edward N. Zalta (ed.), http://plato.stanford.edu/archives/spr2011/
entries/jayaraasi/.
160 SKEPTICISM
Barnes, Jonathan and Julia Annas (ed.). 2000. Sextus Empiricus: Outlines of Scepticism.
Bett, Richard. 2000. Pyrrho, his Antecedents and his Legacy. Oxford: Oxford University
Press.
Bhattacharya, Ramkrishna. 2010. “Commentators on the Cārvākasūtra: A Critical Survey,”
Journal of Indian Philosophy 38: 419–430.
Dasti, Matthew R. 2012. “Parasitism and Disjunctivism in Nyāya Epistemology,” Philosophy
East and West 62: 1–15.
Deutsch, Eliot. 1969. Advaita Vedānta: A Philosophical Reconstruction. Honolulu:
University of Hawaii Press.
Dreyfus, Gorges and Jay L. Garfield. 2011. “Madhyamaka and Classical Greek Skepticism.”
In The Cowherds (ed.), Moonshadows: Conventional Truth in Buddhist Philosophy,
115–130. New York: Oxford University Press.
Flintoff, Everard. 1980. “Pyrrho and India,” Phronesis 25: 88–108.
Franco, Eli. 1994. Perception, Knowledge and Disbelief: A Study of Jayarasi’s Scepticism.
2nd ed. Delhi: Motilal Banarsidass.
Frenkian, Aram. 1957. “Sextus Empiricus and Indian Logic,” The Philosophical Quarterly
(India) 30: 115–126.
Ganeri, Jonardon. 2001. Philosophy in Classical India: The Proper Work of Reason. London
and New York: Routledge.
Ganeri, Jonardon. 2011. The Lost Age of Reason: Philosophy in Early Modern India 1450–
1700. New York: Oxford University Press.
Garfield, Jay L. 1990. “Epoche and Śūnyatā: Skepticism East and West,” Philosophy East
and West 40: 285–307.
Garfield, Jay L. 1995. The Fundamental Wisdom of the Middle Way: Nāgārjuna’s
Mūlamadhyamakakārikā. New York: Oxford University Press.
Garfield, Jay L. and Graham Priest. 2003. “Nāgārjuna and the Limits of Thought,”
Philosophy East and West 53: 1–21.
Granoff, P. E. 1978. Philosophy and Argument in Late Vedānta: Śrī Harṣa’s
Khaṇḍanakhaṇḍakhādya. Dordrecht: D. Reidel.
Hankinson, Robert J. 1995. The Sceptics. London and New York: Routledge.
Huntington, C. W. 1989. The Emptiness of Emptiness: An Introduction to Early Indian
Mādhyamika. Honolulu: University of Hawaii Press.
Jha, Ganganatha (trans). 1986. The Khaṇḍanakhaṇḍanakhādya of Śrī-harṣa: An English
Translation. 2 Vols. Delhi: Sri Satguru Publications.
Matilal, Bimal Krishna. 1986. Perception: An Essay on Classical Indian Theories of
Knowledge. Oxford: Clarendon Press.
Matilal, Bimal Krishna. 1988. The Character of Logic in India, edited by Jonardon Ganeri
and Heeraman Tiwari. Albany. SUNY Press.
McEvilley, Thomas. 1982. “Pyrrhonism and Madhyamika,” Philosophy East and West 32:
3–35.
Mills, Ethan. 2015. “Jayarāśi’s Delightful Destruction of Epistemology,” Philosophy East
and West 65: 498–541.
Mohanty, J. N. 2000. Classical Indian Philosophy. Janham, MD: Rowman and Littlefield.
Nicholson, Hugh. 2009. “The Shift from Agonistic to Non-agonistic Debate in Early
Nyāya,” Journal of Indian Philosophy 38: 75–95.
Phillips, Stephen H. 1995. Classical Indian Metaphysics: Refutations of Realism and the
Emergence of “New Logic.” Chicago: Open Court.
Phillips, Stephen H. 2012. Classical Indian Epistemology: The Knowledge Sources of the
Nyāya School. New York and London: Routledge.
Potter, Karl H. 1981. Advaita Vedānta up to Śaṁkara and His Pupils. The Encyclopedia of
Indian Philosophies, vol. 3. Delhi: Motilal Banarsidass Publishers.
Ram-Prasad, Chakravarthi. 2002. Advaita Epistemology and Metaphysics: An Outline of
Indian Non-Realism. London: RoutledgeCurzon.
Siderits, Mark. 1988. “Nāgārjuna as Anti-realist,” Journal of Indian Philosophy 16: 311–
325.
Siderits, Mark. 2003. Personal Identity and Buddhist Philosophy. Hampshire: Ashgate
Publishing.
Siderits, Mark. 2007. Buddhism as Philosophy: An Introduction. Indianapolis: Hackett
Publishing.
Siderits, Mark and Shoryu Katsura. 2013. Nāgārjuna’s Middle Way: Mulamadhyamakakārikā.
Somerville, MA: Wisdom Publications.
Solomon, Esther A. 1976. Indian Dialectics. Vol. 1. Ahmedabad: B. J. Institute of Learning
and Research.
Solomon, Esther A. (trans.). 2010. Jayarāśi Bhaṭṭa’s Tattvopaplavasiṃha, edited by Shuchita
Mehta. Delhi: Parimal Publications.
Tillemans, Tom J. F. 2001. “Trying to Be Fair to Mādhyamika Buddhism.” The Numata
Yehan Lecture in Buddhism. Calgary, Canada: University of Calgary.
Westerhoff, Jan. 2009. Nāgārjuna’s Madhyamaka: A Philosophical Introduction. Oxford:
Westerhoff, Jan. 2010. The Dispeller of Disputes: Nāgārjuna’s Vigrahavyāvartanī. Oxford:
Yonezawa, Yoshiyasu. 2008. “Vigrahavyāvartanī: Sanskrit Transliteration and Tibetan
Translation,” Journal of Aritasan Institute of Buddhist Studies 31: 209–333.
PART TWO
MEDIEVAL AND
RENAISSANCE
SKEPTICISM
Introduction
Medieval and Renaissance Skepticism
DIEGO E. MACHUCA
It is at present widely acknowledged among historians of Western philosophy

that the Renaissance was a key period for both the recovery of ancient skeptical
texts and the revitalization of the Western skeptical tradition. There has been
much less recognition that skeptical problems and arguments were discussed fairly
extensively in Western Europe during the Middle Ages. The limited awareness
of the role of skepticism in medieval philosophy partially explains why historical
scholarship on this topic is still scarce compared to that devoted to skepticism in
ancient, Renaissance, or modern philosophy. The situation has started to change,
though, particularly in the last few years.1 And fortunately so, since the widespread
ignorance of the part played by skepticism during the Middle Ages has prevented
most scholars from realizing that medieval discussions of various forms of
epistemological skepticism must be taken into account in order to fully understand
the history of Renaissance and modern skepticism. For instance, new light is shed
on Descartes’s evil genius and deceiving god arguments, and their differences, when
one becomes aware that the question of the possibility of divine deception and the
question of the extent of this deception in comparison to that of the devil were
topics of considerable discussion among medieval theologians and philosophers.2
Given this background, in Descartes’s time the mere reference to deception by an
evil demon or by God vividly evoked in his interlocutors and readers a topic with
a long and rich history of philosophico-theological debates. But leaving aside this
instrumental value, medieval discussions of skepticism—particularly skepticism
about our knowledge of the nature and existence of the external world—may be of
intrinsic philosophical interest to, for example, those present-day epistemologists
concerned with understanding and refuting skepticism.
If we consider the influence of the two main ancient skeptical traditions on
Western medieval thought, it must first be said that the role played by Pyrrhonism
was minor in relation to other forms of skepticism and insignificant in comparison
with the part it has had from the sixteenth century until the present day. Knowledge
of Pyrrhonian skepticism was extremely limited in the Middle Ages because direct
acquaintance with ancient Pyrrhonian texts was rare. Although three manuscripts
166 SKEPTICISM
of an early-fourteenth-century Latin translation of Sextus Empiricus’s Pyrrhonian

Outlines, probably by Niccolò da Reggio (ca. 1280–1350), survive, this translation
seems to have exerted no influence whatsoever.3 One of these manuscripts also
contains a partial Latin translation of Sextus’s Adversus Mathematicos III–V,
probably by the same author.4 Likewise, a manuscript remains that contain a
fifteenth-century Latin translation, by Giovanni Lorenzi, of the first four books of
the Adversus Mathematicos, but it does not appear to have had any circulation.5 It
does seem, however, that in the fourteenth century there was some knowledge of
Sextus, since it has been argued that, in Nicholas of Autrecourt’s treatment of the
infinite regress argument, it is possible to identify certain Pyrrhonian elements (see
Grellard 2007a).
As for other philosophical traditions, it should be noted that Greek/Byzantine and
Arabic scholars continued to read Pyrrhonian texts during Late Antiquity and the
Middle Ages. For instance, in his Myriobiblon (Library), Photius, the ninth-century
Patriarch of Constantinople, offers a detailed and relatively extensive summary of
Aenesidemus’s lost Pyrrhonian Discourses that is our main source for the latter’s
own brand of Pyrrhonism.
By contrast, Academic skepticism was well known in Western Europe during the
Middle Ages especially through Augustine’s Contra academicos, but also through
Cicero’s Academica, De natura deorum, and Tusculanae.6 In general, when medieval
authors engaged with skepticism, they referred to the position of the “Academics.”
For instance, in the thirteenth century, Academic skepticism figures in the works
of Henry of Ghent (ca. 1217–1293) and John Duns Scotus (1265/6–1308). While
Henry mentions the Academics in his analysis of the ancient skeptical challenges to
certain knowledge or cognition, Scotus does so in the attack he mounted against
Henry’s theory of illumination: he argued that Henry’s arguments in favor of divine
illumination led in the end to the position of the Academics. The debate between
them revolved around the question whether human knowledge is possible and, if
so, how. To put it in rather simplistic terms, they both agreed that humans can have
knowledge, but while Henry put forward arguments to the effect that we are unable
to attain certain knowledge without divine illumination, Scotus claimed that the
natural use of our cognitive faculties is sufficient to attain such knowledge.7
Setting aside the transmission and influence of ancient skeptical arguments, it
should be remarked that medieval philosophico-theological discussions gave rise
to original skeptical arguments, the most important of which is that which appeals
to the notion of a God who, being all-powerful, could easily deceive us. In its most
radical form, this argument—versions of which were discussed by such figures as
William of Ockham (ca. 1287–1347), Robert Holkot (ca. 1290–1349), Gregory of
Rimini (ca. 1300–1358), Nicholas of Autrecourt (ca. 1299–1369), John Buridan
(ca. 1300–1361), and Peter of Ailly (1351–1420)—targets all of our knowledge of
the nature and the very existence of the external world, and even logical principles
such as the principle of non-contradiction. Thus, as medieval philosophy scholars
sometimes remark, it is a mistake to think that external world skepticism was
first formulated by Descartes. However, it is also noted that medieval authors did
not in fact draw the radical skeptical implications of their arguments concerning
INTRODUCTION 167
divine omnipotence, the reason being that, unlike Descartes, their project was not
the examination of the reliability of our cognitive capacities in order to determine
whether infallible or certain knowledge is possible. They were interested in
exploring the implications of the theological thesis about God’s absolute power, but
did not doubt that our cognitive capacities are reliable whenever nature follows its
normal course.8
Unlike what happened before and after the Middle Ages, during this period there
were no skeptics, with the probable exception of John of Salisbury (ca. 1120–1180),
who was influenced by Cicero and considered himself an Academic skeptic. Salisbury
sought to adapt the skeptical stance he found in Cicero to Christianity. According
to him, by means of the sole use of reason, certain ancient philosophers came close
to the divine mysteries, but failed to recognize that their rational capacity was a gift
from God and even dared to rival Him. As a result, God punished them by concealing
the truth from them, which accounts for the plurality of the philosophical sects
and their entrenched disputes. Academic skepticism was the humble recognition
that the truth is beyond our reach. Salisbury claimed, however, that even though
only God has certain knowledge of all things, we have a natural desire for truth
that cannot be vain because it is implanted in us by Him: truth exists and we can
gradually approach it, but without ever being able to fully attain it. Salisbury was not
therefore a radical skeptic advocating universal suspension of judgment or denying
the possibility of all knowledge. Rather, embracing a form of probabilism, he took
skepticism to be a stance of prudence and modesty that avoids rash judgments
concerning all the issues that admit of no definite resolution, and that is compatible
with the holding of reasonable and fallible beliefs that are formed through dialogue
and debate and that can serve as both a guide to life and the foundation of natural
science. Moreover, Salisbury took certain propositions to be immune from doubt
and the basis of the critical and open-minded examination of truth: logical and
mathematical propositions, those concerning immediate perceptions, and those
concerning the foundations of faith.9
Even though a number of medieval thinkers adopted what might be regarded as
a skeptical stance on specific issues, when skepticism was the topic of discussion,
in most cases the aim was either to refute the skeptical argument being considered
or, following a reductio ad absurdum strategy, to refute one’s opponent’s position
by showing that it led to skepticism. With regard to the former aim, one could
cite, for example, the various thirteenth-century strategies to refute the skeptical
argument based on the illusion of the senses (see Grellard 2004). As for the latter
aim, one could mention, in addition to Scotus’s attack on Henry’s theory of divine
illumination referred to above, that Nicholas of Autrecourt argued that Bernard of
Arezzo’s view on the intuitive cognition of non-existents had skeptical implications
that were more absurd than those which followed from the Academics’ position,
and that John Buridan, in turn, argued that Nicholas’s foundationalist anti-skeptical
reply actually led to skepticism while his own fallibilist strategy did successfully
answer the skeptical challenge.10
Bernard of Arezzo was a Franciscan whom Nicholas had met when they were both
theology students in Paris and to whom he wrote nine letters, only two of which
168 SKEPTICISM
survive.11 According to these letters, Bernard held that, in those circumstances in

which God intervenes in the normal course of nature, there may be an act of cognition
even if its object does not exist. In Nicholas’s view, this meant that it is not possible
to determine when our awareness of the existence of external objects is true or false,
and hence that we can never be certain of their existence, for it is not possible to
determine when the act of cognition is produced by a supernatural or a natural cause,
that is, by God or by the perceived object itself. Bernard also claimed that we do not
have direct intuitive cognition of our own mental acts. In response, Nicholas argued
that, if one cannot be certain of the existence of those things of which one has a clear
cognition (viz., external objects), neither can one be certain of the existence of those
things of which one has a less clear cognition (viz., one’s own mental acts), and so one
cannot be certain of one’s own existence as an intellect. Although he offered replies
to different skeptical arguments—such as the argument from sensory illusion and the
dream argument—Nicholas’s general reply to skepticism consisted in maintaining
both (i) that there are evident basic beliefs, both perceptual and intellectual, on which
the justification of other beliefs ultimately rest, and (ii) that “evidentness” (evidentia),12
which admits of no degrees, is not a necessary condition for knowledge, since there
is also probable knowledge, that is, knowledge of that which cannot be demonstrated
but can be defended by means of rational arguments—for example, inductive and
causal knowledge is probable, not evident.
Buridan’s own reply to skepticism about the external world consisted in
distinguishing, pace Nicholas, between two degrees of evidentness that can ground
knowledge: evidentness properly so called and natural or relative evidentness. The
former type of evidentness, which commands our assent immediately, is that of the
propositions that are true by virtue of the semantic relation between their terms (e.g.,
“Every horse is an animal”) or that exemplify the principle of non-contradiction.
Natural evidentness is that of the propositions expressing that which appears to be
the case and which cannot appear to be otherwise for any human reason—except
for sophistical ones. This type of evidentness cannot properly be so called because
the intellect may be deceived with regard to those naturally evident propositions
owing to a supernatural cause, that is, when God intervenes in the normal course
of nature, causing the appearance of an object that is not really present. Buridan
therefore conceded the possibility of divine deception with regard to our knowledge
of external objects, but argued that the skeptic placed an unreasonable demand
for absolute evidentness—and hence for absolute certainty—on our knowledge
of the world. For, even though naturally evident knowledge is fallible, it suffices
for the principles and conclusions of natural science whenever nature follows its
normal course.
Just as one should bear in mind that there were (almost) no real skeptics during the
Middle Ages, so too should one bear in mind that medieval refutations of skepticism
were not produced in the context of a serious skeptical crisis, unlike what later
occurred in the Renaissance and modern periods. Rather, at least a considerable
number of medieval authors made a methodological use of skeptical arguments,
which consisted in using such arguments either to undermine a given conception of
knowledge that was then replaced by another one immune from the challenges posed
INTRODUCTION 169
by those arguments, or to distinguish between types of knowledge and to determine

the kind of certainty proper to each one of them.13 Skepticism seems to have been
a logical construction independent for the most part of the skeptical tradition,
which explains why it was understood as a reservoir of arguments to the effect
that knowledge is impossible (Grellard 2004: 113; 2007a: 328). The skeptic was a
fictional character whose radical stance could teach us something about the nature
and limits of human knowledge. The medieval conception of skepticism is thus
analogous to the way in which skepticism is conceived of by most contemporary
analytic epistemologists, who take for granted that there are not, and there cannot
be, any real skeptics, and who therefore regard skeptical arguments as ingenious and
enlightening heuristic devices.
I noted at the outset that many historians of philosophy are still unaware
that discussion of skeptical problems and arguments is well attested in Western
medieval philosophy. Still less known is the fact that skepticism played a central
part in philosophico-theological discussions in classical Islam—a period that
ranges from the ninth to the thirteenth centuries CE—even though in the Islamic
tradition there were no flesh-and-blood skeptics. Skepticism arose in a dialectical
context in which incompatible views were defended concerning such subjects as
the anthropomorphic nature of God, the nature of belief in Islam, and religious
pluralism. Skepticism was conceived of as an attitude of perplexity and doubt in the
face of conflicting and equipollent assertions, and was hence clearly distinguished
from the attitude of disbelief. Even though the writings of Sextus and Cicero were
not available in Arabic, Islamic scholars made use of concepts that remind us of
Greek skeptical notions such as ἀπορία (aporia) and ἰσοσθένεια (equipollence).
Given the significant role played by skepticism during Islam’s classical period, the
present book includes a chapter on this topic that focuses primarily on al-Ghazālī
(1058–1111), a philosopher and theologian of Sunni Islam. al-Ghazālī made use of
skeptical arguments in order to show that philosophy is unable to attain any kind
of true knowledge unless it is combined or supplemented with mystical insight,
for God is the ultimate guarantor of truth. al-Ghazālī is an interesting figure for
historians of modern Western philosophy due to the striking similarities between
the skeptical arguments he formulated and those found in Descartes (e.g., the
argument from sensory illusion or the dream argument) and Hume (the arguments
against causation), as well as between his and Descartes’s autobiographical records
of their skeptical crises, their rejection of traditional beliefs, and their quest for
certain knowledge.14
As for the part that skepticism played in the Renaissance, it should first be
mentioned that in 1562 Henri Estienne published a Latin translation of Sextus’s
Pyrrhonian Outlines, and in 1569 Gentian Hervet published a Latin translation of
Adversus Mathematicos and of Adversus Dogmaticos, which made widely available
what we today know as Sextus’s corpus. It is safe to say that most historians of
ideas accept at present that the Renaissance rediscovery of those Pyrrhonian texts
exerted a strong influence on the European intellectual life in the sixteenth and
seventeenth centuries, playing a key role in shaping philosophical thought in
particular. Richard Popkin was the first to maintain that the revival of Pyrrhonism
170 SKEPTICISM
triggered a “Pyrrhonian crisis.”15 According to Popkin, this crisis first arose in the
religious sphere, and then spread to cover philosophy, the sciences, and the pseudo-
sciences. In his view, the history of Renaissance and modern philosophy should to
a large extent be construed as the history of the various strategies that Renaissance
and modern thinkers devised to deal with that crisis. What is of note about the
rediscovery and dissemination of Sextus’s writings is the fact that his Pyrrhonism
was put into the service of very different, and sometimes incompatible, aims.
Some scholars have argued that Popkin overstated the part played by Pyrrhonism
in shaping Renaissance and modern thought (Ayers 2004; Perler 2004). However that
may be, it seems incontestable that Pyrrhonism was one of the intellectual forces that
molded Renaissance and modern philosophical discussions. This is to be explained, at
least in part, by the fact that those philosophers who took an interest in Sextus found
in his substantive surviving writings several sophisticated batteries of arguments that
call into question the epistemic justification of our beliefs in a more systematic and
compelling way than anything contained in the other available ancient sources.
The Pyrrhonism expounded in the recovered Sextan writings will then become an
integral part of Renaissance discussions of skepticism. Academic skepticism, already
known from the works of Augustine and Cicero, will of course continue to be
taken into consideration in such discussions, but now with the additional invaluable
information on the Pyrrhonian brand of skepticism provided by Sextus’s works. In
this second part of the book, the reader will find three chapters dealing with the
most important figures of Renaissance skepticism, namely, Michel de Montaigne
(1533–1592), Pierre Charron (1541–1603), and Francisco Sanchez (1550/51–1623).
Unlike the medieval authors who engaged with skepticism (again, with the possible
exception of John of Salisbury), Montaigne, Charron, and Sanchez were skeptics,
even if of a mitigated stripe. As in medieval philosophy, skepticism was discussed by
Renaissance skeptics in close connection with religious and theological matters. In
fact, both Montaigne and Charron have been considered as fideists (Popkin 2003),
embracing a type of skepticism according to which reason by itself is useless for the
attainment of truth or the justification of our beliefs, faith being the only means to
the acquisition of knowledge and certainty.
Montaigne, whose Essays heavily rely and expand on Sextus’s works, is the central
figure both in the dissemination of those recently recovered works and in the impact
of Pyrrhonism on Renaissance and modern philosophy, to the extent that he may be
considered the first non-ancient Pyrrhonist. His importance is seen, for example, in
his being the first author to have translated Sextus’s technical skeptical terms into a
modern language. Montaigne’s own form of Pyrrhonism presents some key differences
from Sextus’s. For instance, he interpreted Pyrrhonism as a form of phenomenalism
according to which our knowledge is restricted to appearances (understood narrowly
as perceptual appearances), since objective reality is unknowable. Besides the
fact that Sextus’s notion of appearance included both perceptual and intellectual
appearances, the main problem with Montaigne’s view concerns, from an ancient
Pyrrhonist’s point of view, his claim about the unknowability of reality, which would
gain him the accusation of negative meta-dogmatism that Sextus leveled against neo-
Academics, Cyrenaics, and Empirical doctors. In addition, Montaigne conceived of
INTRODUCTION 171
the skeptic as someone who is in a state of doubt, which is a state of uneasiness or

discomfort, radically different from the state of undisturbedness that the ancient
Pyrrhonist fortuitously attained by suspending judgment. Montaigne also associated
the Pyrrhonian fourfold observance of everyday life with Catholic traditionalism as a
way of avoiding adherence to the Reformation.16
Charron adopted a more positive stance on the epistemic value of reason than
did Sextus and Montaigne. Unlike Sextus, he regarded suspension of judgment as
the voluntary decision of the inquirer and also allowed the skeptic to have weak
opinions, in line with what he took to be Academic skepticism. His own skeptical
stance is closer to this form of skepticism than to Pyrrhonism, even though he did
not in the end distinguish between them. Charron introduced a sharp distinction
between the internal and the external spheres that may sometimes be in strong
tension: the Charronian skeptic behaves in accordance with the social norms and
confines his doubt to the private sphere, where the sole authority is reason, and
preserves his intellectual autonomy from the external requirements by bringing
everything under critical scrutiny.17
As for Sanchez—the author of That Nothing is Known—he seems to have been a
skeptic of a Pyrrhonian stripe, although he has also been characterized as defending
Academic skepticism—this latter understood either as the view that knowledge
is impossible or as the view that less-than-certain knowledge is achievable but
always revisable in the face of new evidence. Sanchez’s Pyrrhonism can be seen, for
example, in his portrayal of his search for truth: though he has not yet found it, he
keeps seeking it, and though he does not claim that the search for truth is doomed
to failure, he cannot assure that it will ever be found. The target of his skeptical
assault was the scholastic Aristotelianism of his day: following an ad hominem
argumentative strategy, he claimed that, if either of the two dominant Aristotelian
metaphysical views were true, then knowledge would be impossible. It should finally
be noted that Sanchez, who was also a physician, seems to have been influenced
by the skeptical approach to the practice of medicine adopted by ancient medical
Empiricism, a school with close ties with Pyrrhonism.18
NOTES
1. See especially Perler (2006), Lagerlund (2010), Grellard (2013), and Denery, Ghosh
and Zeeman (2014).
2. For medieval authors, while the devil’s deception is restricted to knowledge that
originates in perception—i.e., to the realm of external material objects that act upon
our senses—God is also able to act directly upon the intellect by creating a false
proposition and inducing the intellect to assent to it. On this difference, see Gregory
(1992c).
3. See Schmitt (1983: 227) and Porro (1994: 230–235). For a detailed analysis of these
manuscripts, see Cavini (1977: 1–8) and Floridi (2002: 63–69).
4. See Cavini (1977: 4, 8–9) and Floridi (2002: 79–80).
5. See Schmitt (1976), Porro (1994: 236), and Floridi (2002: 80–84).
6. See Schmitt (1983: 227), Porro (1994: 242–251), and Grellard (2004: 114–115).
172 SKEPTICISM
7. On Henry’s position and Scotus’s attack on it, see especially Perler (2006: ch. 1). It
should here be pointed out that a chapter devoted to this topic and to skepticism in
the thirteenth century in general was commissioned, but its author never fulfilled his
commitment and publication of the book could not be further delayed.
8. On the deceiving God argument in the Middle Ages, and some comparisons with
Descartes’s, see, besides Henrik Lagerlund’s chapter in this book, Perler (2010a)
and especially Tullio Gregory’s two seminal studies: Gregory (1992b,c).
9. On the skepticism of John of Salisbury, in addition to David Bloch’s chapter in this
book, see Grellard (2007b, 2013).
10. On Nicholas of Autrecourt’s and John Buridan’s replies to skepticism, besides
Christophe Grellard’s chapter in this book, see Zupko (1993), Thijssen (2000),
Grellard (2007a, 2010), Karger (2010), and Klima (2010).
11. A critical edition and English translation of Nicholas’s extant correspondence is
provided in de Rijk (1994).
12. I render the Latin evidentia as “evidentness” in lieu of “self-evidence” because the
former translation has become a technical term in scholarship on medieval philosophy.
13. See especially Perler (2010b); cf. Grellard (2004: 128–129; 2007a: 342). Of course,
whether an author considered it necessary to answer a given skeptical argument or
merely made a methodological use of is not always easy to establish.
14. On al-Ghazālī, besides Paul Heck’s contribution to this book, see Halevi (2002)
and Kukkonen (2010). On skepticism in classical Islam in general, see Heck (2014).
On the comparison of al-Ghazālī with Descartes and Hume, see Menn (2003) and
Moad (2009).
15. Popkin (1960) is the pioneering work on the impact of ancient skepticism (especially
in its Pyrrhonian variety) on Renaissance and modern philosophy, covering the
period from Erasmus to Descartes. Popkin (1979) extends the analysis to Spinoza,
and Popkin (2003) reaches back to Savonarola and forward to Bayle.
16. On Montaigne’s skepticism, besides Gianni Paganini’s contribution, see e.g. Brahami
(1997), Giocanti (2001), Popkin (2003: ch. 3), and Paganini (2008: ch. 1).
17. On Charron’s skepticism, besides Gianni Paganini’s contribution, see e.g. Popkin
(2003: ch. 3) and Maia Neto (2014: ch. 2).
18. On Sanchez’s skepticism, besides Damian Caluori’s chapter, see Limbrick (1988),
Popkin (2003: 38–43), Caluori (2007) and Paganini (2016).
REFERENCES
Ayers, Michael. 2004. “Popkin’s Revised Scepticism,” British Journal for the History of
Brahami, Frédéric. 1997. Le scepticisme de Montaigne. Paris: Presses Universitaires de
France.
Caluori, Damian. 2007. “The Skepticism of Francisco Sanchez,” Archiv für Geschichte der
Philosophie 89: 30–46.
Cavini, Walter. 1977. “Appunti sulla prima diffusione in Occidente delle opere di Sesto
Empirico,” Medioevo 3: 1–20.
INTRODUCTION 173
Denery, Dallas, Kantik Ghosh, and Nicolette Zeeman (eds.). 2014. Uncertain Knowledge:
Scepticism, Relativism, and Doubt in the Middle Ages. Turnhout: Brepols.
de Rijk, L. M. 1994. Nicholas of Autrecourt: His Correspondence with Master Giles and
Bernard of Arezzo. Leiden: Brill.
Giocanti, Sylvia. 2001. Penser l’irrésolution: Montaigne, Pascal, La Mothe Le Vayer. Trois
itinéraires sceptiques. Paris: Honoré Champion.
Gregory, Tullio. 1992a. Mundana Sapientia: Forme di conoscenza nella cultura medievale.
Roma: Edizioni di Storia e Letteratura.
Gregory, Tullio. 1992b. “La tromperie divine.” In Gregory 1992a, 389–399. (First
published in 1982.)
Gregory, Tullio. 1992c. “Dio ingannatore e genio maligno. Nota in margine alle
Meditationes di Descartes.” In Gregory 1992a, 401–440. (First published in 1974.)
Grellard, Christophe. 2004. “Comment peut-on se fier à l’expérience? Esquisse d’une
typologie des réponses médiévales au problème sceptique,” Quaestio 4: 113–135.
Grellard, Christophe. 2007a. “Scepticism, Demonstration and the Infinite Regress
Argument (Nicholas of Autrecourt and John Buridan),” Vivarium 45: 328–342.
Grellard, Christophe. 2007b. “Jean de Salisbury: un cas médiéval de scepticisme,”
Freiburger Zeitschrift für Philosophie und Theologie 54: 16–40.
Grellard, Christophe. 2010. “Nicholas of Autrecourt’s Skepticism: The Ambivalence of
Medieval Epistemology.” In H. Lagerlund (ed.), Rethinking the History of Skepticism:
The Missing Medieval Background, 119–143. Leiden: Brill.
Grellard, Christophe. 2013. Jean de Salisbury et la renaissance médiévale du scepticisme.
Paris: Les Belles Lettres.
Halevi, Leor. 2002. “The Theologian’s Doubts: Natural Philosophy and the Skeptical
Games of Ghazālī,” Journal of the History of Ideas 63: 19–29.
Heck, Paul. 2014. Skepticism in Classical Islam: Moments of Confusion. New York:
Routledge.
Karger, Elizabeth. 2010. “A Buridanian Response to a Fourtheenth Century Skeptical
Argument and its Rebuttal by a New Argument in the Early Sixteenth Century.” In H.
Lagerlund (ed.), Rethinking the History of Skepticism: The Missing Medieval Background,
215–232. Leiden: Brill.
Klima, Gyula. 2010. “The Anti-skepticism of John Buridan and Thomas Aquinas: Putting
Skepticism in their Place Versus Stopping them in their Tracks.” In H. Lagerlund (ed.),
Rethinking the History of Skepticism: The Missing Medieval Background, 145–170.
Leiden: Brill.
Kukkonen, Taneli. 2010. “Al-Ghazālī’s Skepticism Revisited.” In H. Lagerlund (ed.),
Rethinking the History of Skepticism: The Missing Medieval Background, 29–59. Leiden:
Brill.
Lagerlund, Henrik (ed.). 2010. Rethinking the History of Skepticism: The Missing Medieval
Background. Leiden: Brill.
Limbrick, Elaine. 1988. “Introduction.” In E. Limbrick and D. Thomson (eds.), Francisco
Sanches: That Nothing is Known, 1–88. Cambridge: Cambridge University Press.
174 SKEPTICISM
Maia Neto, José. 2014. Academic Skepticism in Seventeenth-Century French Philosophy:

The Charronian Legacy 1601–1662. Dordrecht: Springer.
Menn, Stephen 2003. “The Discourse on the Method and the Tradition of Intellectual
Autobiography.” In J. Miller and B. Inwood (eds.), Hellenistic and Early Modern
Moad, Omar. 2009. “Comparing Phases of Skepticism in al-Ghazālī and Descartes: Some
First Meditations on Deliverance from Error,” Philosophy East and West 59: 88–101.
Paganini, Gianni. 2008. Skepsis: Le débat des modernes sur le scepticisme. Paris: Vrin.
Paganini, Gianni. 2016. “Sanches et Descartes. Subjectivité et connaissance réflexive au
temps des sceptiques,” Bulletin de la Société internationale des amis de Montaigne 64:
173–185.
Perler, Dominik. 2004. “Was There a ‘Pyrrhonian Crisis’ in Early Modern Philosophy?
A Critical Notice of Richard H. Popkin,” Archiv für Geschichte der Philosophie 86:
209–220.
Perler, Dominik. 2006. Zweifel und Gewissheit: Skeptische Debatten im Mittelalter.
Frankfurt: Klostermann.
Perler, Dominik. 2010a. “Could God Deceive us? Skeptical Hypotheses in Late Medieval
Epistemology.” In H. Lagerlund (ed.), Rethinking the History of Skepticism: The Missing
Medieval Background, 171–192. Leiden: Brill.
Perler, Dominik. 2010b. “Skepticism.” In R. Pasnau (ed.), The Cambridge History of
Medieval Philosophy, 384–396. Cambridge: Cambridge University Press.
Popkin, Richard. 1960. The History of Scepticism from Erasmus to Descartes. Assen: van
Gorcum.
Popkin, Richard. 1979. The History of Scepticism from Erasmus to Spinoza. Berkeley:
Popkin, Richard. 2003. The History of Scepticism from Savonarola to Bayle. New York:
Porro, Pascuale. 1994. “Il Sextus latinus e l’immagine dello scetticismo antico nel
medioevo,” Elenchos 15: 229–253.
Schmitt, Charles. 1976. “An Unstudied Fifteenth-century Latin Translation of Sextus
Empiricus by Giovanni Lorenzi.” In C. H. Clough (ed.), Cultural Aspects of the
Italian Renaissance: Essays in Honor of Paul Oskar Kristeller, 244–261. Manchester:
Manchester University Press.
Schmitt, Charles. 1983. “The Rediscovery of Ancient Skepticism in Modern Times.” In M.
Burnyeat (ed.), The Skeptical Tradition, 226–251. Berkeley, Los Angeles and London:
Thijssen, Johannes. 2000. “The Quest for Certain Knowledge in the Fourteenth Century:
Nicholas of Autrecourt against the Academics.” In J. Sihvola (ed.), Ancient Scepticism
and the Sceptical Tradition (Acta Philosophica Fennica 64), 199–223. Helsinki:
Philosophical Society of Finland.
Zupko, Jack. 1993. “Buridan on Skepticism,” Journal of the History of Philosophy 31:
191–221.
CHAPTER TWELVE
Augustine and Skepticism

STÉPHANE MARCHAND
1 INTRODUCTION
Augustine (354–430 CE) was certainly not a skeptic. He nevertheless played a crucial
role in the history of philosophical skepticism. Three main reasons place Augustine
at a strategic point in relation to skepticism. First, he is the author of Against the
Academics (Contra Academicos), a philosophical dialogue entirely devoted to the
refutation of skepticism, where he discusses the arguments of the New Academy in
order to go beyond skepticism. But, secondly, Augustine’s contribution to skepticism
is not limited to this work, for he never ceased in considering skepticism a problem
of the highest order. His conception of belief in the Christian faith can also be
interpreted as a response to the skeptical contention that one should live without
belief. Finally, through his reflections on the problem of skepticism, Augustine
conceived of a new form of it: skepticism as a thought experiment of the inner
self, which exhibits a metaphysical and epistemological strength that from Descartes
onward has remained vibrant.1
2 THE CONTRA ACADEMICOS

Augustine is a witness of great importance in the history of ancient skepticism
under its Academic guise. First of all, his reading of Cicero played a crucial role
in the education he received.2 Augustine was probably the most important reader
of Cicero’s Academica in antiquity (Schmitt 1972: 29): he took the arguments
put forward by Cicero in that book very seriously, to the point that he confessed
having temporarily endorsed the skepticism of the New Academy in 383 CE, after
giving up Manichaeism and before his conversion to Christianity (see Confessiones
V x 19; cf. De beata vita 4).3 Therefore, for Augustine skepticism was not just
a theoretical hypothesis, but a real (albeit false) philosophical alternative for
those who are disappointed by how difficult it is to find the truth. Since he felt
that skepticism was a genuine problem, he wrote an entire dialogue with the
sole purpose of refuting it, the Contra Academicos, written in 386, just after his
conversion to Christianity.
176 SKEPTICISM
Contra Academicos is presented as a dialogue in the style of Cicero and

is composed of three books. The global aim of the work is to demonstrate, in
countering skepticism, that certain knowledge is possible and that happiness cannot
be attained without believing that it is possible to find the truth (C. Acad. I ii 5).
At the end of the dialogue, Augustine distinguishes between the real intention of
the Academics, who according to him were not true skeptics, and their arguments,
which were genuinely skeptical. Augustine was convinced that the so-called
skeptical Academy represented, in reality, an esoteric form of Platonism that used
skepticism to hide its dogmatism during the Hellenistic period, when materialism
was flourishing (see C. Acad. III xvii 37–xx 43; see also Letter 1).4 Therefore, when
Augustine criticized the New Academy, he was not attacking the Academy as an
institution (indeed, he shared their Platonist conviction that spirit is superior to
matter). Rather, he was trying to inoculate against skepticism anyone who might
be tempted to take it as a serious philosophical option. Because there were people
who believed the Academy’s skeptical arguments to be valid, he considered it not
only necessary to respond to those arguments, but to restore a genuine relationship
to truth.
Augustine’s line of argument in the Contra Academicos is somewhat tortuous.5
However, it is possible to distinguish between two levels of refutation: (i) an
epistemological and (ii) a moral level.6
The epistemological level (i) is a direct response to the arguments presented
in Cicero’s Academica. Augustine purported to prove that it was possible to
offer a rational response to the New Academy. Among the numerous arguments,
Augustine made use of analytic or a priori knowledge: even if one doubts a
proposition, this doubt cannot remove the true knowledge that this proposition
is either true or false (C. Acad. III ix 21). Similarly, every disjunctive proposition
is necessarily true (C. Acad. III ix 23: the world is organized either by a divine
providence or by a body, it either has a beginning or does not, etc.; see also III
xiii 29). These arguments function as counterexamples, showing that it is easy
to have access to certainty and, therefore, that there is no reason to despair of
knowing the truth: Augustine’s contention against skepticism is that some things
can be known.7
We can focus on a set of arguments that are similar to the dream argument in
Descartes’s First Meditation:
Therefore, I call the whole that contains and sustains us, whatever it is, the
“world”—the whole, I say, that appears before my eyes, which I perceive to
include the heavens and the earth (or the quasi-heavens and quasi-earth).
If you say nothing seems to be so to me, I’ll never be in error. It is the man
who recklessly approves what seems so to him who is in error. You do say that a
falsehood can seem to be so to sentient beings. You don’t say that nothing seems
to be so. Every ground for disputation, where you Academicians enjoy being the
master, is completely taken away if it is true not only that we knew nothing, but
also that nothing seems to be so to us. However, if you deny that what seems
AUGUSTINE AND SKEPTICISM 177
so to me is the world, then you’re making a fuss about a name, since I said I call
this “world.”
You’ll ask me: “Is what you see the world even if you’re asleep?” It has already
been said that I call “world” whatever seems to me to be such. (C. Acad. III
xi 24–25)8
Thus, whatever we might think about the existence of an external world (the
knowledge of which is mediated by our senses and can actually be false), it
remains indubitable that there is something in one’s mind (even if I am dreaming
or suffering from a hallucination) that I can call “world” and about which I can
form true knowledge. Admittedly, Cicero’s Academica (II 47–48) already planted
the seeds of this argument. But Cicero glanced over that hypothesis without any
clear formulation of a world composed only of inner states. Augustine’s own
reflection on the Academic arguments produced a clearer formulation of the
hypothesis of external world skepticism which will be a fundamental feature of
modern skepticism.9
As for the moral level of refutation (ii) as Heil (1972: 109) has claimed, in Contra
Academicos the core of the refutation is not epistemological. Augustine was aware
that the truth and certainty he was looking for was not of the kind we find in
mathematics or analytic knowledge (see C. Acad. II iii 9), but concerned instead
knowledge of the soul and of God, the sole objects of worth for Augustine (see
Soliloques I 7). Indeed, this skepticism is dangerous because it can discourage anyone
from being on the path of the truth, i.e., on the path of God. Hence, the real danger
of skepticism is moral, and it becomes essential to demonstrate that skepticism as a
philosophy is essentially defective.
This demonstration consists in showing that skepticism is not what it seems to be
for Cicero, namely, an open-minded stance that earnestly seeks the truth (see, e.g.,
Acad. II iii 7), but rather a renunciation of truth and philosophy:
“Therefore,” I continued, “don’t you know that up to now there is nothing
I perceive to be certain? I’m prevented from searching for it by the arguments and
debates of the Academicians. They somehow persuaded me of the plausibility—so
as not to give up their word just yet—that man cannot find the truth. Accordingly,
I had become lazy and utterly inactive, not daring to search for what the most
ingenuous and learned men weren’t permitted to find. Unless, therefore, I first
become as convinced that the truth can be found as the Academicians are
convinced that it cannot, I shall not dare to search for it. I don’t have anything to
defend.” (C. Acad. II ix 23)
Here is the danger: by their arguments, the Academics deviate from the natural
path that leads from the desire for truth to knowledge and then to God. And since
the knowledge of God is the only means of achieving our natural desire to be
happy, the moral refutation of skepticism consists of demonstrating that it cannot
be considered a serious philosophical option to achieve happiness. Rather, for
Augustine, the skeptics’ excessive prudence hides existential despair (the despair of
truth, the desperatio veri: see Letter 1, 3; C. Acad. II i 1; Confessiones VI i 1). In fact,
178 SKEPTICISM
the skeptic is frightened by the possibility of error. By privileging the criterion of

plausibility (as King translates the Latin probabile, which is itself a translation of
the greek word πιθανόν), he is trying to avoid risk. But, as Augustine shows, this
position cannot provide any quietude and does not prevent anyone from error:
I seemed to see an entrance through which error would rush in upon those who
felt safe. I think that a man is in error not only when he follows the false path,
but also when he’s not following the true one. (C. Acad. III xv 34)
Therefore, a person can avoid giving his assent, follow the criterion of what is
probable—in sum, a person can act exactly as Carneades recommended—and
still be wrong: being guided by the criterion of what is probable does not prevent
anyone from sinning, and therefore can lead them to err from a moral point of view
(C. Acad. III xvi 35–36). Rashness can be useful and necessary in order to allow an
individual to believe something doubtful that will nevertheless lead to truth. Even
if to doubt can prevent many of our errors, one cannot find the true path without
having a belief.
3 HERMENEUTICS AGAINST SKEPTICISM

If something skeptical remained in Augustine’s position, it should be sought in
connection with his thoughts on the weakness of pure reason.10 Indeed, Augustine
was an acute critic of the arrogance of philosophers, who claimed that it was possible
to know everything through the use of reason alone. Against this claim Augustine
insisted, in a skeptical manner, on the difficulty of finding any indubitable truth by
itself through reason, at least on any theological or metaphysical level.
Thus, Augustine tried to find an escape path between two excesses: an excessive
rationalism that claims to know God by reason (as the Manicheans claimed: see De
utilitate credendi I 2) and an excessive skepticism that will lead to the despair of truth.
Augustine frequently agreed that it is reason—or intellect—whose function it is to
understand the truth; but the problem is knowing whether such an understanding
can begin with reason alone or must instead begin with belief and, more precisely,
with faith.
To introduce the necessity of faith (fides) as a condition for understanding,
Augustine made frequent use of Isaiah 7.9: “Unless you believe, you will not
understand.” Hence, Augustine’s position did not entail fideism in the sense
that faith should replace reason, but rather stressed the necessity of believing in
order to understand and know theological truths, according to the principle crede
ut intelligas: you must believe to understand (see e.g., De libero arbitrio I 4). As
Augustine said, “Faith prepares the ground for understanding” (De ut. cred. XVII
35)11; in the end, it is reason that understands, but reason needs to begin with some
form of belief (Rist 2001: 27).
Hence, to attain knowledge of the truths that really matter to Augustine,
knowledge of God and of our soul, it is impossible to rely on pure reason; unlike
formal knowledge, this kind of knowledge concerns objects that are hidden and
requires one to begin with belief. Since there is no direct access to this knowledge,
it is necessary to believe in some authority in order to be on the path of truth. This

contention does not entail, as fideism does, the confusion between knowledge and
faith, but rather the distinction between two epistemological processes: “We must
hold what we understand as coming from reason, what we believe as coming from
authority” (De ut. cred. XI 25).12
Actually, it seems that, in Augustine’s view, belief is necessary not only in
theological or metaphysical matters, for almost all aspects of the human condition
are determined by our confidence in propositions that we cannot verify through
scrutiny; propositions about our relation to the world and to other men, even to
our own history, rely on belief. Belief is a necessary feature of the human condition:
without believing witnesses, we could know nothing of history or about the thoughts
and feelings of another mind13:
So it was, Lord, that you began little by little to work on my heart with your
most gentle and merciful hand, and dispose it to reflect how innumerable were
the things I believed and held to be true, though I had neither seen them nor been
present when they happened. How many truths there were of this kind, such as
events of world history, or facts about places and cities I had never seen; how
many were the statements I believed on the testimony of friends, or physicians,
or various other people; and, indeed, unless we did believe them we should be
unable to do anything in this life. With what unshakable certainty, moreover,
did I hold fast to the belief that I had been born of my particular parents, yet
I could not have known this without believing what I had heard. (Conf. VI v 7;
Boulding’s translation in Rotelle 1997)
Augustine offers a rationalist approach to the problem of believing. To believe
(credere) is not necessarily to be credulous. On the contrary, Augustine (again at
De ut. cred. XI 25) stresses the difference between believing and having an opinion
(opinari). To have an opinion consists of having a belief without the consciousness
that it is a mere belief. To opine entails the ignorance of the difference between the
things we can know (intelligere) and those we can only believe (credibilia); in sum,
opinion is rash belief. But to believe is not wrong per se. We can have true belief,
if we keep in mind that this very belief remains a mere belief, and if we have good
reasons to believe. Hence, the quality of the belief depends on the awareness that
belief is not certain knowledge and on the authority followed for believing.
Thus, just as reason needs faith, so too does faith need reason. More precisely,
faith is the belief in an authority that has—by rational means—been recognized as
being reliable. In the end, this is the very reason that compels us to decide whether
to believe or not:
If, then, it is reasonable that faith precede reason with respect to certain great
truths that cannot yet be grasped, however slight the reason is that persuades us
to this, it undoubtedly also comes before faith. (Letter 120, 3; Teske’s translation
in Ramsey 2004; see also De vera religione XXIV 45)
This rationalistic approach to faith is based on reflection on the authority of the
Bible as a holy book that defines what it is to believe. This view makes it possible to
180 SKEPTICISM
differentiate between Christian faith and credulity: if faith is not blind and irrational
assent, it is necessary to have reasons to assent to facts that are, for the most part,
unbelievable, such as the resurrection of Christ.
For Augustine, the Bible gives an account of God’s manifestations in the world,
which are signs to be interpreted. In this sense, the Bible is like a history book that
requires the evaluation of the reliability of the witnesses and their idiosyncrasies,
etc. Since there can be no direct knowledge of history, we have to rely on accounts
and testimonies, and Augustine tries to offer some sort of method to acquire a
rational approach to the Bible as God’s book, by defending its reliability and overall
coherence, despite the fact that God has chosen different prophets who use a variety
of styles and images to pass on the testimony of his actions on men. The Bible
therefore contains, on the one hand, reasons to believe its contents, and on the
other, truths enabling readers to understand once a method to read and to interpret
it has been provided.14
Hence, there is a connection between Augustine’s reflection on belief and his
thoughts on reading and hermeneutic theory, and this reflection shares some
common features with the problem of skepticism, and more precisely with the
problem of understanding someone else’s thought.15 This feature of Augustine’s
thought has been related to the hermeneutic circle. Augustine’s reflection on belief
shows that, in many things, understanding entails, firstly, making the hypothesis
that there is something to understand (a kind of principle of charity, cf. De ut.
cred. VI 13), and secondly, holding preliminary beliefs in order to be on the path of
understanding (or a pre-understanding principle). This circle comes from Augustine
himself (Conf. I i 1; De Trinitate XV ii 2) and is perhaps one of the most important
theses of Augustinian thought (see, for instance, Pascal Pensées Laf. 919): we cannot
seek God (or truth) without having already found him. Thus, neither the excessive
rationalism of the Manicheans nor any skeptical reading of the Bible is able to
understand the truth of Christianity, because both lack faith.
Augustine conceived this circle through his thinking on the mysteries of the Bible
and the search for a method to read and understand it against the skeptical objections.
As he said in the Confessions, the Bible—and especially the Old Testament—
appeared to him a weak and poor book in comparison with the sophisticated prose
of Cicero (Conf. III vi 9). The comprehension of the full sense—or at least of a true
sense—of the text supposes giving credit to its author and to his intention by way
of some belief.
This circle helps us to understand more precisely Augustine’s stance on faith:
although he emphasized the fact that there are reasons to believe, his stance was not
purely rationalistic. Faith and religious belief can be fed by reason and intellectual
reflection on the meaning of the Bible and, more generally, on God’s manifestations in
the world. But faith does not depend on the sole decision of the will. More precisely,
the will itself is not the result of a purely individual decision, but presupposes the
grace of God without which nobody can really find the path of truth.
This feature takes on great significance given that Augustine was at that point
involved in the debate with the Pelagians, who defended the possibility for men to
save their souls by themselves. For this reason, even if it remains true that faith is
conditioned by rational reasons, the very fact of believing relies on a movement that
originates in God himself. Contrary to what the skeptic seems to believe, assent is
not a completely voluntary act; rather, it is influenced by God’s love and concern
for us.16
4 AUGUSTINE AND THE COGITO

An overview of the relationship between Augustine and skepticism would be
incomplete without a survey of Augustine’s texts that anticipate Descartes’s famous
experience of the cogito. We saw that as early as the Contra Academicos, Augustine
presented arguments that are similar to those employed by Descartes in his First
Meditation, where he showed that some kinds of knowledge (such as mathematical
knowledge) resist the dream argument. But Augustine also deployed some thought
experiments that can be compared to Descartes’s discovery, in the Second Meditation
(AT VII 25), of the indubitable truth of the proposition “ego sum, ego existo.” In
opposition to skepticism, Augustine emphasized the certainty of the existence of
the individual’s own thought as a fundamental experience. It is worth noting that
this very experience is mentioned in the first dialogues (see De vita beata II 7,
Soliloques II 1, De libero arbitrio II iii 7) and appears in later masterpieces like De
Trinitate (e.g., XV xii 21) and De civitate dei (XI, 26).
Putting aside the highly debated question of whether Descartes took his inspiration
from Augustine,17 what is important is determining in what sense Augustine’s stance
is connected to solipsism, which is nowadays deeply connected to skepticism.
In the De civitate dei, the indubitable fact of the existence of the individual’s own
thought takes the form of an anti-skeptical argument—the Academics cannot affirm
that men are always mistaken because there is something indubitable: the very fact
that one does exist, and that one does know that he exists—“for if I am mistaken,
I exist” (si fallor, enim sum, XI 26).18 Accordingly, as in Descartes, doubt cannot
be universal because something within the inner experience of my mind provides
resistance to all skeptical scenarios.
Yet, Augustine’s aim is not to use the skeptical hypothesis in order to reconstruct
knowledge, for one can see that such a methodological doubt does not appear in
Augustine. And, as we saw, as far as knowledge is concerned, he considered rather
the necessity of relying on an external authority. Nevertheless, the discovery of the
inner knowledge of the individual’s own existence as a mind is also fundamental in
Augustine, not in the sense that it could have the function of a first principle, but
rather in the sense that it reveals the distinctive nature of human thought. Not only
does the inner knowledge discovered by Augustine offer a formal criterion of truth
and an instance of an indubitable truth (cf. De Trinitate XV 21), but it also, and
more fundamentally, gives us knowledge of the individual’s own substance:
Now properly speaking a thing cannot in any way be said to be known while its
substance is unknown. Therefore when mind knows itself it knows its substance,
and when it is certain of itself it is certain of its substance. But it is certain of
itself, as everything said above convincingly demonstrates. Nor is it in the least
182 SKEPTICISM
certain whether it is air or fire or any kind of body or anything appertaining

to body. Therefore it is not any of these things. The whole point of its being
commanded to know itself comes to this: it should be certain that it is none
of the things about which it is uncertain, and it should be certain that it is that
alone which alone it is certain that it is. . . . But if it were one of these things
[sc. if mind were a material thing] it would think that thing differently from the
others, not that is to say with a construct of the imagination as absent things
are thought that have been contacted by one of the senses of the body, either
actually themselves or something of the same kind; but with some inner, non-
simulated but true presence (nothing after all is more present to it than itself),
in the same way as it thinks it is living and remembering and understanding
and willing. It knows these things in itself, it does not form images of them
as though it had touched them with the senses outside itself, as it touches any
bodily things. If it refrains from affixing to itself any of these image-bound
objects of its thoughts in such a way as to think it is that sort of thing, then
whatever is left to it of itself, that alone is what it is. (De Trinitate X x 16;
translation by Hill 1991)
Thus, for Augustine, a special knowledge comes from the mind: the mind is present
to itself in a totally different way from that in which other things can be present to
it. Book X of the De Trinitate precisely demonstrates this distinctive characteristic:
by knowing itself, the mind knows that it lives, that it exists—it knows itself entirely
(De Trinitate X iv 6). In this work, references to similar skeptical arguments
probably no longer refer to the skepticism of the New Academy; Augustine has
extracted from this skepticism a new picture of skepticism as an inner experience,19
the experience of solipsism through which he discovered the peculiarity and the
substance of an inner world. Hence, skepticism is not only something that any
rational thinker should refute for the sake of argumentation; it has also become a
gateway to a crucial feature of knowledge in Augustinian philosophy: interiority.
Skeptical scenarios show that knowledge of external reality is not of the same
character as knowledge of the inner self. But even in this context, this experience
cannot be understood as a kind of solipsism that is an experience where one is alone
(solus ipse) and where each thing can be reduced to a subjective representation. It is
for this reason that Augustine perceives this experience revealed in the individual as
the presence of the “interior master”—the presence of God in each of us—thanks
to whom any propositional content can be recognized as a truth.20 The inner self
has the privilege of being an image of God; accordingly, skeptical scenarios are an
efficient means of going inside and finding an inner world. Skepticism remains a
means (however, not the only means) of revealing the inner self and performing
Augustine’s advice: “Do not go outside, come back into yourself. It is in the
inner self that Truth dwells” (De vera religione XXXIX 72, translation by Hill in
Ramsey 2005).
Thus, even though Augustine was not a skeptic and devoted a part of his work
to refuting skepticism, in reality it seems that skepticism played a significant role
in the constitution of his philosophy. And despite the fact that skepticism entered
Augustine’s work through his knowledge of the Academic skeptical tradition, the
vivid discussion he entertained with this philosophical tradition gave rise to new
skeptical problems and to new functions of and uses for skepticism.
NOTES
1. Augustine’s view of skepticism has also had an important influence on the medieval
conception of Skepticism: see Grellard (2011; 2018) and Bouton-Touboulic (2013).
2. On Augustine’s life and education, see Marrou (1939) and Brown (2000).
3. Some scholars have doubted that Augustine earnestly espoused skepticism: see
Alfaric (1918: 352–358), Testard (1958: 129), and Besnier (1993: 91). However,
there is no doubt that Augustine considered skepticism a very serious problem.
4. That hypothesis is already mentioned by Sextus, PH I 234. For a critical approach
to this hypothesis, see Lévy (1978).
5. For a precise analysis of the argumentation in Contra Academicos, see Mosher
(1981), Kirwan (1983), and Curley (1997).
6. For this distinction, see Catapano (2006), who also provides useful references.
7. For an accurate analysis of the function of those analytic counterexamples in
Augustine’s strategy, see Heil (1972: 108–111).
8. All translations of the Contra Academicos are taken from King (1995).
9. See Burnyeat (1982: 28). This paper has been recently challenged by scholars who
denied that Augustine entertained the hypothesis of an external world skepticism:
see Bolyard (2006). One can also consider that the medieval interpretations of
Augustin’s treatment of skepticism were a crucial influence on the modern form of
skepticism: see Grellard (2011: 18; 2018).
10. It is worth noting that Augustine was confronted by accusations of skepticism: see
Contra litteras Petiliani III xxi 24. Cf. Bermon (2009).
11. For the De ut. cred., I follow the translation of Kearney in Ramsey (2005).
12. Quod intellegimus, debemus rationi: quod credimus auctoritati. Cf. also Revisions
XIV 3.
13. Cf. De fide rerum quae non videntur II 4; De ut. cred. XII 26.
14. The definition of such a method is the goal of the second book of De
Doctrina Christiana.
15. According to Stock (1996: 283), the reading theory building upon the experience of
the Bible is an “adaptation of scepticism.”
16. For an accurate analysis of this point, see Koch (2013).
17. Descartes himself was forced to explain the differences between both arguments:
see Bermon (2001: 9–15). The bibliography on the topic is extensive, see Gilson
(1930), Gouhier (1978), Taylor (1989), Matthews (1992), Menn (1998), and
Marion (2008: 87–98).
18. For an accurate explication of this argument, see Matthews (1972; 1992: 29–34).
19. Scholars question whether this new picture should be connected with the Plotinian
treatment of skepticism that can be found in Plotinus’s Ennead V 3 [49].
184 SKEPTICISM
20. On this crucial doctrine, see the De Magistro and De Trinitate XII. An accurate
explanation of the doctrine of illumination can be found in O’Daly (1987: 199–207).
REFERENCES
Alfaric, Prosper. 1918. L’évolution intellectuelle de Saint Augustin: Du manichéisme au
néoplatonisme. Paris: E. Nourry.
Bermon, Emmanuel. 2001. Le cogito dans la pensée de saint Augustin. Histoire des doctrines
de l’Antiquité classique. Paris: Vrin.
Bermon, Emmanuel. 2009. “‘Contra academicos vel de academis’ (retract. I, 1): saint
Augustin et les Academica de Cicéron,” Revue des études anciennes, 111: 75–93.
Besnier, Bernard. 1993. “La Nouvelle Académie selon le point de vue de Philon de Larisse.”
In B. Besnier (ed.), Scepticisme et exégèse: Hommage à Camille Pernot, 85–163.
Fontenay-aux-Roses: ENS Fontenay/Saint-Cloud.
Bouton-Touboulic Anne-Isabelle. 2013. “Scepticism (up to Descartes).” In K. Pollmann
(ed.), The Oxford Guide to the Historical Reception of Augustine, vol. 3, 1694–1697.
Brown, Peter. 2000. Augustine of Hippo: A Biography. London: Faber.
Missed,” Philosophical Review 90: 3–40.
Catapano, Giovanni. 2006. “Quale scetticismo viene criticato da Agostino nel Contra
Academicos?” Quaestio 6: 1–13.
Curley, Augustine. 1997. Augustine’s Critique of Skepticism: A Study of Contra Academicos.
New York: Peter Lang.
Gilson, Étienne. 1930. Études sur le rôle de la pensée médiévale dans la formation du
système cartésien. Paris: Vrin.
Gouhier, Henri. 1978. Cartésianisme et augustinisme au XVIIe siècle. Paris: Vrin.
Grellard Christophe. 2011. “Academicus.” In I. Atucha, D. Calma, C. König-Pralong, and
I. Zavattero (eds.), Mots Médiévaux offerts à Ruedi Imbach, 7–8. Porto: Fédération
internationale des instituts d’études médiévales.
Grellard Christophe. 2018. “Le scepticisme au Moyen Âge, de saint Augustin à Nicolas
d’Autrécourt. Réception et transformation d’un problème philosophique.” Cahiers
Philosophiques.
Heil, John. 1972. “Augustine’s Attack on Skepticism: The Contra Academicos,” The
Harvard Theological Review 65: 99–116.
Hill, Edmund. 1991. Augustine: The Trinity. Volume I/5. New York: New City Press.
King, Peter. 1995. Augustine: Against the Academicians and The Teacher. Indianapolis:
Hackett Publishing.
Kirwan, Christopher. 1983. “Augustine against the Skeptics.” In M. Burnyeat (ed.), The
Skeptical Tradition, 205–223. Berkeley: California University Press.
Koch, Isabelle. 2013. “Représentation et assentiment dans la définition augustinienne de
la croyance.” In L. Jaffro (ed.), Croit-on comme on veut? La controverse classique sur le
rôle de la volonté dans l’assentiment, 47–62. Paris: Vrin.
Lévy, Carlos. 1978. “Scepticisme et dogmatisme dans l’Académie: l’ésotérisme d’Arcésilas,”
Revue des études latines 56: 335–348.
Marion, Jean-Luc. 2008. Au lieu de soi: l’approche de Saint Augustin. Paris: PUF.
Marrou, Henri-Irénée. 1939. Saint Augustin et la fin de la culture antique. Paris: De Broccard.
Matthews, Gareth. 1972. “Si Fallor, Sum.” In G. Matthews (ed.), Augustine: A Collection
of Critical Essays, 151–162. New York: Doubleday.
Matthews, Gareth. 1992. Thought’s Ego in Augustine and Descartes. Ithaca: Cornell
University Press.
Menn, Stephen. 2002. Descartes and Augustine. Cambridge: Cambridge University Press.
Mosher, David L. 1981. “The Argument of St. Augustine’s Contra Academicos,” Augustinian
Studies 12: 89–113.
O’Daly, Gerard James Patrick. 1987. Augustine’s Philosophy of Mind. London: Duckworth.
Ramsey, Boniface (ed.). 2004. Augustine: Letters 100–155, translated by R. Teske. Volume
II/2. Hyde Park and New York: New City Press.
Ramsey, Boniface (ed.). 2005. Augustine: On Christian Belief, translated by E. Hill, R.
Kearney, M. Campbell, and B. Harbert. Volume I/8 of The Works of Saint Augustine:
A Translation for the 21st Century. Hyde Park, NY: New City Press.
Rist, John. 2001. “Faith and Reason.” In N. Kretzmann and E. Stump (eds.), The Cambridge
Companion to Augustine, 26–39. Cambridge: Cambridge University Press.
Rotelle, John (ed.). 1997. The Confessions, translated by M. Boulding. Hyde Park, NY:
New City Press.
Schmitt, Charles Bernard. 1972. Cicero Scepticus: A Study of the Influence of the Academica
in the Renaissance. The Hague: M. Nijhoff.
Stock, Brian. 1996. Augustine the Reader: Meditation, Self-Knowledge, and the Ethics of
Interpretation. Cambridge, MA: Belknap Press.
Taylor, Charles. 1989. Sources of the Self: The Making of the Modern Identity. Cambridge,
MA: Harvard University Press.
Testard, Maurice. 1958. Saint Augustin et Cicéron : Cicéron dans la formation et dans
l’œuvre de Saint Augustin. Paris: Études Augustiniennes.
CHAPTER THIRTEEN
John of
Salisbury’s Skepticism
DAVID BLOCH
1 INTRODUCTION
“The Latin world has had nothing greater than Cicero!” (Orbis nil habuit maius
Cicerone Latinus), said John of Salisbury (c. 1115/20–1180) in his Entheticus
Maior (van Laarhoven 1987: 185). John spent most of his life as a diplomat and
had only limited time for philosophical studies. However, he had studied no
less than twelve years in Paris (1135/36–1147/48) with practically all the great
masters, and he knew much about philosophy, theology, and literature. We may
disagree with him concerning the evaluation of Cicero, but he was certainly
qualified to make the claim. For the history of philosophy, John is also important
as one of the first medieval thinkers who had access to most if not all of Aristotle’s
works on logic. Thus, he had the resources and was naturally inclined to combine
the philosophies of Cicero and Aristotle, a fact that contributed immensely to
his views on skepticism.1 Thus, John was, for better and worse, a peculiar kind
of skeptic: an Academic, Ciceronian skeptic, but (perhaps) not truly so. He set
forth his views in two major works in particular: the Policraticus (on political
theory) and the Metalogicon (on logic and education), both finished in or around
1159. The former relies primarily on examples and quotations from authoritative
authors (although book 7 in particular contains important arguments), whereas
the latter presents arguments in favor of the positions he adopts. I will focus on
the Metalogicon.2
In this work, John makes the claim that the basic texts in education should be
those that constitute Aristotle’s Organon (along with Porphyry’s Isagoge), and the
Stagirite was certainly no skeptic but the founder of modern science and scientific
theory and methodology. At first glance, then, John seems to be in danger of holding
contradictory views. Furthermore, skepticism was not a common position in the
twelfth century.3 In fact, John is apparently the only westerner of that century
who claims to be a skeptic, although the Academic type of skepticism was well
known in the first half of the century, in particular from Augustine’s Against the
Academics (Contra Academicos).4 None of his teachers, who were practically all
JOHN OF SALISBURY’S SKEPTICISM 187
the great masters of his time, were skeptics, and John usually followed their lead
in philosophical matters. But in the Metalogicon he seems to embrace skepticism
contrary to almost all the science and philosophy of his time.
I will argue that John’s desire to be both Ciceronian and Aristotelian makes him
a rather fascinating kind of Aristotelian skeptic. Even if one is sometimes puzzled by
the results of John’s efforts, he deserves our respect for taking such a controversial
and independent stance based on arguments and for his willingness to take both
Cicero and Aristotle seriously.
2 THE THREE KINDS OF ACADEMIC SKEPTICISM

In book 4 of the Metalogicon, John distinguishes three kinds of Academic skeptics:
The Academician is in doubt, and he dares not lay down what is true in each and
every case. But this sect is divided into three groups. The adherents of the first
group claim that they know nothing at all, and as a result of this extreme caution
they have lost the right to be called philosophers. The adherents of the second
group accept as knowledge only things that are necessary and known per se, that
is, things that one cannot fail to know. The third kind comprises those of us who
do not venture to state an opinion in cases that are doubtful to the wise man.
(Metalogicon 4.31, Hall 1991: 168; all translations are mine)5
The first group comprises those who doubt everything, and John quickly dismisses
them. It is, John says elsewhere, ridiculous and absolutely unphilosophical to doubt
everything (Policraticus 7.2, Webb 1909: 96). The second group is surprising in
this context; for without the further premise that they themselves do not think that
“things that are necessary and known per se” are knowable (a premise that is not
made explicit), its adherents are not really skeptics but rather philosophers who
put too much trust in demonstrative science. John’s evaluation of the second group
seems to reflect his own views on knowledge and science.
John himself prefers the third kind of skepticism. The position is noteworthy
for at least two characteristics. First, “holding opinions” is not as such wrong,
but doing so rashly or without proper consideration is problematic. John himself
often held opinions of his own (see below on the problem of genera and species).
Second, in some cases you should hold opinions; it is only when even those who
are best qualified have to give up that one should suspend judgment. This, stated
in positive terms, would seem to mean that John’s kind of Academic skeptic
has every right to state opinions if (1) he examines and considers the matter
carefully beforehand, (2) he is suitably qualified to do so in the relevant area
of knowledge, and (3) the subject matter allows one to hold qualified opinions.
John does not state these requirements explicitly, but they are supported by the
discussions in which he takes his own positions and by his use of Aristotle’s
dialectical method in the Topics. I will illustrate both points in the following
parts of this article.
Probably the most famous of John’s discussions in which he voices and evaluates
many arguments and opinions is the extended treatment of the genus-species
188 SKEPTICISM
problem in book 2, chapters 17 and 20 of the Metalogicon. This and many other
subjects can be discussed with positive results, and skepticism is the best way of
obtaining such results (Policraticus 1.prologus, ed. Keats-Rohan 1993: 25), but there
are also many subjects where rational analysis and argument will provide nothing
useful. In such matters, John says, the only acceptable tool is faith (fides).6 However,
philosophically speaking, another turn in his development is equally interesting,
namely, his Aristotelianism.
3 DIALECTIC AND ARISTOTLE’S TOPICS

It has been argued, then, that John is a rather liberal skeptic. Opinions resulting
from rational analysis and argument are in many cases not only permissible but even
strongly recommendable. On the other hand, he strongly and consistently believes
that one cannot know anything with absolute certainty (Metalogicon 1.prologus, ed.
Hall 1991: 11; Policraticus 7.1, ed. Webb 1909: 93–95). Having learned from, and
adjusted his teaching to, Cicero, what this kind of skeptic needs to do is to find the
best possible source of rational and coherent argument, and John was fortunate. For
the very best had recently become available in full. Parts of Aristotle’s Organon (i.e.,
his works on logic) had been known for some time, and Aristotle was known as the
philosopher. However, it was only in the second quarter of the twelfth century that
the texts of the Organon really entered the schools of Paris (Bloch 2012: 27–44),
and John is one of the first to take advantage of these new sources of learning. Thus,
his overall view naturally becomes the result of his combined reading of Cicero
and Aristotle.
Through Aristotle and later Aristotelian sources (Boethius in particular), the
medievals had learned that there were at least two different kinds of knowledge
and methods of obtaining knowledge: demonstration (demonstrative science) and
dialectic. Aristotle’s Organon had been structured—albeit probably not by Aristotle
himself—in part to reflect this division by including a treatise on deductions (Prior
Analytics), which serves as a tool for both a treatise on demonstrative science
(Posterior Analytics) and one on dialectic (Topics). But, according to Aristotle, the
only real understanding comes from demonstrative science, whereas dialectic is
equally rigid in the deductions but much less strict when it comes to the principles
(i.e., the basic premises) of the deductive process.
So, although the precise nature of the Posterior Analytics is a much disputed
topic in contemporary Aristotelian scholarship, Aristotle’s views on knowledge,
understanding, and science are nowadays usually gathered and deduced primarily
from this text.7 But for John the most important text by far is the Topics. For whereas
the former is concerned with the kinds of knowledge and understanding that are
necessary and based upon clear and distinct principles that cannot be disputed (i.e.,
demonstrative science), the latter is concerned with what is probable and open to
debate (i.e., dialectic). Consequently, when John learns from Cicero that “a mere
human being” (homunculus) should be “striving for the probables” (probabilia
sequens) (Tusculanae disputationes 1.9), he naturally seeks out the Aristotelian work
concerned with probable arguments and knowledge. His introductory account of
the Topics illustrates the importance that John assigns to that work, but it is also a
somewhat surprising passage:
The Topics comprises eight books, and each book brings more coherence and
force to the subject than the preceding ones. The first gives a rough outline, so to
speak, of the subject matter of the other books and establishes some fundamentals
of logic as a whole. For it teaches what a deduction is; what a demonstration is,
and from which sources it is established; what the principles of the arts are,
and the principles of the faith that we gain from the arts; what the dialectical
deduction is, and the contentious . . . . (Metalogicon 3.5, ed. Hall 1991: 120)
Aristotle’s Topics is certainly the natural tool of a man with moderate skeptical
views, and fittingly John also stresses the fact that Aristotle is particularly good at
demolishing other people’s theses and arguments (Metalogicon 3.8, ed. Hall 1991:
125). But John’s account of the purpose and content of the Topics show that he
believes it to be much more than just a useful tool for skeptics. It is rather a text on
logic and how to gather knowledge in general to the extent possible.
Aristotle’s Topics is traditionally divided into eight books. The first is introductory
and establishes the fundamentals of dialectic, books 2–7 treat “accident,” “property,”
“genus,” and “definition,” and book 8 is concerned with rules and procedures for
dialectical debate. John’s use of the Topics goes far beyond logic and dialectic proper
contributing, as it does, to the other kinds of knowledge as well. This comes out
clearly in his careful exposition of the text, which takes up most of book 3 of the
Metalogicon (3.5–10, ed. Hall 1991: 118–139). For instance, in the exposition of
book 3 on accidents, he says:
Since only “accident” admits of comparison, the third book [scil. of the Topics]
explains the power inherent in “comparables” [scil. the value of understanding
what may and may not be done in comparisons]. Pursuing the nature of accidents,
it shows by general rules the reasons for choosing and avoiding, and among the
things to be chosen, it shows which should be preferred over the others, and
among the things to be avoided it shows which should be avoided more than
the others. Consequently, it is evident how much this discipline benefits studies
of natural science and ethics; for this part of the discipline is very effective in
instances where one has to choose and avoid, and generally in every situation
where comparison must be made. Clearly this book [scil. the Topics] has much to
commend it, and earlier scholars were wrong to neglect it, as it is of outstanding
utility, written in a pleasant style and extremely beneficial in the study of both
ethics and natural science. (Metalogicon 3.6, ed. Hall 1991: 123)
Two of the claims made in this passage are particularly important for understanding
John’s position: (1) Aristotle’s Topics was, according to John, not highly regarded in
the twelfth century and thus little used, and (2) the utility of the Topics is great even
in matters of ethics and natural science.
Claim (1) is probably somewhat exaggerated, but it is not completely wrong.8 We
have no extant commentaries on Aristotle’s Topics from the twelfth century, and
scholars were most often perfectly happy with Manlius Boethius’s (c. 480–524/25)
190 SKEPTICISM
version of the Topics. Thus, John was one of the very first thinkers to use the newly
recovered Aristotelian text, and it provided a solid basis for his skepticism. Skepticism
is, according to John, the right approach to things about which we can expect to
learn something by our own faculties, but since extreme skepticism is unreasonable,
we need a positive philosophy as well, which is provided by the dialectical approach
of the Topics. Thus, John is in accordance with Cicero’s description of Carneades,
who is described as a thinker who mastered both proof and demolishing proof (De
oratore 2.38.161).
Claim (2) is based on the view that both natural philosophy and ethics have basic
principles that are by nature probable at most. Nature is defined by movement and
ethics by action, neither of which allows the philosopher the luxury of certainty.
Thus, there will be two possible responses: complete suspension of judgment or
use of a method that is especially designed to treat such cases. This is the dialectical
procedure that is described and analyzed in Aristotle’s Topics. In favor of this view,
John perhaps appealed to the much discussed passage in the Topics, book 1, chapter
2, in which Aristotle can be construed as saying that dialectic is one of the most basic
tools of all philosophical inquiries. Not only does it work on the basis of probable
principles, but it is also the discipline that enables philosophers to examine the
principles of other sciences.
Ethics is particularly important in this context. John is apparently not much
interested in natural science, even though the principles of nature are not necessary,
in the Aristotelian tradition, but only probable like the principles of ethics. Aristotle
himself used several examples from natural philosophy in his Posterior Analytics
to illustrate demonstrative science, and other texts of the twelfth century clearly
show that John’s contemporaries had the option of defending a demonstrative
science of nature despite the apparent conflict between probable principles and a
demonstration.9 However, John focuses on ethics, which he considers by far the
most important discipline (Metalogicon 2.11, 2.13; ed. Hall 1991: 73, 74–76).
Thus, like many of his contemporaries, John thought that ethics was the goal for
which all the sciences were servants, as it were. This view he could have learned
from Cicero himself or from some of his most esteemed teachers, most notably Peter
Abelard (1079–1142),10 and a strongly “ethical” approach to logic is found in the
Latin translations of Arabic scholars.11 The kind of philosophy that interested John
most was the one that would eventually lead to practical results, and such results
could, as Aristotle would also say (Nicomachean Ethics 1.3, 1094b: 12–27), never
attain the degree of precision and necessity required for demonstrative science.
Thus, it seems that John is led both by his regard for Cicero and by his intellectual
inclinations to his mild Ciceronean-Aristotelian skepticism.
4 A SKEPTIC ON DEMONSTRATIVE SCIENCE

Still, since John has such respect for Aristotle and his writings, and since he recognizes
that Aristotle should be the foundation of all proper education, one would have
expected him to take the full step and also embrace Aristotelian demonstration and
demonstrative science. For he is well aware that Aristotelians consider them the
crown jewel of the Philosopher’s theory of knowledge (Metalogicon 4.7, ed. Hall
1991: 145–146).
However, he has, first, his own, very simple—and amusing—skeptical argument
against demonstrative science:
It was long thought that it was impossible to cut diamonds, since their matter did
not yield to sharpened iron or sharpened steel. But when they were eventually cut
using lead and goat’s blood, it became apparent that it was actually quite easy to do
what had previously seemed impossible. (Metalogicon 3.9, ed. Hall 1991: 130)12
Demonstrative science is impossible, he claims, because we can never know for
certain that we actually know what we think we know. The same argument applies
to the more serious case of virgin births (Metalogicon 2.13, ed. Hall 1991: 75).
In fact, this kind of argument could be countered—and was countered by Robert
Grosseteste less than 100 years later—by reference to the Augustinean theory of
divine illumination, a theory which twelfth-century thinkers including John knew
well.13 Understanding and knowledge start by the passive reception of phenomena
through the faculties given by nature, but in order to transform the phenomena into
knowledge human beings need “internal light” (interior lux) provided by divine grace;
without this human beings have no possibility of cognizing the real and universal
truths. John would, however, embrace this description as one that illustrates his
point: human beings have limited cognitive powers when left to themselves, and
divine illumination does not really belong in a theory of demonstration, but in faith.
Second, the state of demonstrative science in the twelfth century was somewhat
unclear. Scholars and philosophers simply did not always know how to use it. John’s
chapter on this issue is important:
The science/knowledge (scientia) found in the Posterior Analytics is a subtle one
and penetrable only for few intellects. There are evidently several reasons for this.
First, the work contains the art of demonstration, and this art is more difficult
than the other systems of argumentation. Second, this art has practically fallen
into disuse as a result of the rarity of practitioners, since the mathematicians are
the only ones to use demonstration, and they too barely use it; even among them,
it is used almost exclusively by the geometers. However, this discipline is not
one that is frequently cultivated among us either, except perhaps in the Iberian
region and in the confines of Africa. For these people, for the sake of astronomy,
practice geometry more than other people. The same is true of the Egyptians
and quite a few of the Arabic peoples. In addition to these observations, the
book in which the demonstrative discipline is transmitted is much more confused
than the others, both because of the transpositions of words and letters and the
out-dated examples that have been borrowed from different disciplines; and
finally—which is not the author’s fault—the work has been so much distorted by
scribal errors that it contains almost as many obstacles, as it contains chapters.
And actually, it is good when there are no more obstacles than chapters. Thus,
many blame the translator for the difficulty and claim that the work has not been
correctly translated. (Metalogicon 4.6, ed. Hall 1991: 145)
192 SKEPTICISM
This rather exotic description shows that the main problems are the limited usefulness
of demonstration and the difficulty of the Posterior Analytics. Demonstrative science
is not completely invalidated; it is simply limited to mathematics. One would think
that John will then comfortably discard any talk of demonstrative science as a valid
method of inquiry, but surprisingly that is not the case.
5 THE METHODOLOGY OF JOHN

OF SALISBURY’S SKEPTICISM
In his only sustained chapter on demonstrative science, John writes:
One must have prior knowledge of the principles, and from these one must gather
the sequence of truths by coherent reasoning based on necessity; and one must
progress by working very hard to place the argument on solid ground, so to speak,
in order that the lack of necessity would not result in the appearance of a gap in
the argument, a gap that would jeopardize the otherwise resulting demonstrative
knowledge. Certainly, not every science/knowledge is demonstrative, but only
those that are based on true, primary, and immediate principles. For just as not
every deduction is a demonstration but every demonstration is a deduction,
so also knowledge/science comprises the demonstrative but not conversely.
(Metalogicon 4.8, ed. Hall 1991: 146)
This remarkable passage is basically a general methodological description that can
be systematized as follows:
1. When one strives to obtain some kind of knowledge, one must recognize the
principles of the knowledge in question as carefully as possible.
2. Conclusions must ultimately be deduced from these principles.
3. The deductions must be necessary, that is, the conclusions must follow
by necessity, every step of the argument must be clear, and there can be
no exceptions.
4. Every effort must be made to ensure that there are no gaps in the
argument. This means that every single step of the argument must be clear
and convincing.
The most surprising thing about this procedure is Step 4. Step 1 is a problem that
Aristotle himself apparently never solved for demonstrative science, and whereas
he would presumably accept that a number of scientific disciplines have strictly
necessary and primary causal principles, John would not (except for mathematics).
Steps 2 and 3 apply to both demonstration and dialectic, but Step 4 is purely
dialectical. The knowledge produced by demonstrative science cannot be disputed,
and thus there can be no reworking of the argument afterwards. If the argument is
truly demonstrative, it is complete. But the skeptic part of John maintains that even
the process of demonstrative science is never complete; there will always—or almost
always—be more gaps to close.
This leaves John with a quasi-demonstrative approach to knowledge. It maintains

that necessity is essential in the deductions from principles to conclusions, and it
also incorporates substantial dialectical elements that have been gathered from
Aristotle’s Topics, not least the fact that the deductions and the arguments can—and
must—always be tested further.
Thus, John has combined methods that would not seem easily compatible. The
reason might be that he believes that all knowledge is somehow connected:
All things have a way of contributing to each other, so that one will become more
proficient in any proposed branch of learning to the extent that he has mastered
neighboring and related branches. (Metalogicon 3.5, ed. Hall 1991: 119; cf. Cf.
Metalogicon 4.1, ed. Hall 1991: 140)
Different sciences do not, according to John, have completely separate domains.
Instead, each somehow contributes to one’s understanding of other sciences. A
scholar needs to be able to use both demonstrative and dialectical methods, and
supposedly “neighboring and related branches” will easily incorporate elements of
both kinds. This further explains how John can say that the Topics also teaches
demonstration (Metalogicon 3.5, ed. Hall 1991: 120, quoted above). The differences
are simply reconcilable, according to John.
For a skeptic, John is strangely obsessed with, and optimistic about, establishing
the proper methods for actually obtaining knowledge.
6 CONCLUSION
In some ways John is then perhaps the least skeptical of all the thinkers discussed in
the present book. He himself sums up his views on how to obtain knowledge, and
to what extent it is possible, as follows:
Divine perfection is to know all things completely, angelic perfection is to avoid
errors in all cases, and human perfection is to have a good perception of most
things. (Metalogicon 3.3, ed. Hall 1991: 110)
This by itself is hardly skepticism. At the very least, it is skepticism with a prominent
Christian superstructure, and thus inherently a rather optimistic kind of skepticism. In
particular, it still leaves room for the intervention of a God who provides the necessary
principles in the Augustinean fashion. But still, if one combines the view that some
subjects are in fact beyond human knowledge and that opinions should not in these
matters be stated as facts with the view that, even when opinions can—and should—be
stated, human beings can never have certain knowledge, one has a respectably skeptical
stance. It is in this sense that John of Salisbury can be considered a skeptic.14
NOTES
1. For general and more thorough information about John of Salisbury’s life and works,
see Liebeschütz (1950), Nederman (2005), and Bloch (2012).
194 SKEPTICISM
2. For analyses of the Policraticus and its use of examples, see von Moos (1988) and
Grellard (2013).
3. For skepticism in the Middle Ages, see Perler (2006), Lagerlund (2010a), Grellard
(2013), and the other articles in Part 2 of the present book.
4. On the lack of knowledge of Cicero’s Academica, see Rouse and Rouse (1978) and
Grellard (2011).
5. Cf. Lagerlund (2010b: 10–11) and, in particular, Grellard (2013: 50–53).
6. Metalogicon 3.10, ed. Hall (1991: 138): “There are many things that do not admit
of discussion, the ones that transcend human reason/argument and are consecrated
only to faith.” Cf. Policraticus 7.1, 7.7, ed. Webb (1909: 95–99, 114–117).
7. See, e.g., Wians (1989) for a summary of the discussion. It is, however, uncertain
whether John really knew this text, and he certainly did not know it well (see
Bloch 2012).
8. See further Metalogicon 4.24 (ed. Hall 1991: 162). For the early tradition of
Aristotle’s Topics, see Stump (1982) and Green-Pedersen (1984).
9. Texts and sources are described and analyzed in Bloch (2012: 27–82).
10. Cicero, Academica 1.9.34 (stating that Strato was no philosopher, because he did
no work on ethics); Abelard, Collationes §§ 2, 67–69 (ed. Marenbon and Orlandi
2001: 82–86).
11. Al-Ghazali, Tractatus de Logica, Prooemium (ed. Lohr 1965: 241–242). It is,
however, unlikely that John knew this particular text.
12. The story is actually much older and found in Pliny the Elder, Augustine, and
Isidore, but lead is not mentioned by any of them.
13. See Policraticus 8.25 (ed. Webb 1909: 418–425). Cf. Entheticus Maior, ed. van
Laarhoven 1987: verses 223–250, 273–278, 629–654, pp. 121, 123, 147–149.
14. I do not here have the space to discuss Grellard (2013), which is now undoubtedly
the most important work on John of Salisbury’s skepticism. I merely note that the
present article is much more focused on John’s Aristotelian inclinations, which
I consider essential to our understanding of John’s skepticism. However, this
approach to John’s work is at the expense of a number of important topics treated
by Grellard.
REFERENCES
Bloch, David. 2012. John of Salisbury on Aristotelian Science. Turnhout: Brepols.
Green-Pedersen, Niels Jørgen 1984. The Tradition of the Topics in the Middle Ages. The
Commentaries on Aristotle’s and Boethius’ “Topics”. München and Wien: Philosophia
Verlag.
Grellard, Christophe. 2011. “La seconde acculturation chrétienne de Cicéron: la réception
des Académiques du IXe au XIIe siècle,” Asterion 11. URL: http://asterion.revues.
org/2350.
Grellard, Christophe. 2013. Jean de Salisbury et la renaissance médiévale du scepticisme.
Hall, John B. (ed.). 1991. Ioannis Saresberiensis Metalogicon. Corpus Christianorum,

Continuatio Mediaevalis 98. Turnhout: Brepols.
Keats-Rohan, K. S. B. (ed.). 1993. Ioannis Saresberiensis Policraticus I-IV. Corpus
Christianorum, Continuatio Mediaevalis 118. Turnhout: Brepols.
Lagerlund, Henrik (ed.). 2010a. Rethinking the History of Skepticism. The Missing Medieval
Background. Leiden and Boston: Brill.
Lagerlund, Henrik. 2010b. “A History of Skepticism in the Middle Ages.” In H. Lagerlund
(ed.), Rethinking the History of Skepticism. The Missing Medieval Background, 1–28.
Liebeschütz, Hans 1950. Mediaeval Humanism in the Life and Writings of John of Salisbury,
Studies of the Warburg Institute 17. London: University of London.
Lohr, Charles H. 1965. “Logica Algazelis. Introduction and Critical Text,” Traditio 21:
223–290.
Marenbon, John and Giovanni Orlandi (eds.). 2001. Peter Abelard: Collationes. Oxford:
Nederman, Cary J. 2005. John of Salisbury, Medieval and Renaissance Texts and Studies
288. Tempe and Arizona: Arizona Center for Medieval and Renaissance Studies.
Perler, Dominik 2006. Zweifel und Gewissheit. Skeptische Debatten im Mittelalter.
Frankfurt am Main: Klostermann.
Rouse, Richard and Mary Rouse. 1978. “The Medieval Circulation of Cicero’s Posterior
Academics and the De finibus bonorum et malorum.” In Malcolm B. Parkes and Andrew
G. Wilson (eds.), Medieval Scribes, Manuscripts and Libraries: Essays Presented to N.R.
Ker, 333–367. London: Scolar Press.
Stump, Eleonore 1982. “Topics: Their Development and Absorption into Consequences.”
In N. Kretzmann, A. Kenny, and J. Pinborg (eds.), The Cambridge History of Later
Medieval Philosophy. From the Rediscovery of Aristotle to the Disintegration of
Scholasticism, 1100–1600, 273–299. Cambridge: Cambridge University Press.
van Laarhoven, Jan (ed.). 1987. John of Salisbury’s Entheticus Maior and Minor, 3 vols.
Leiden: Brill.
von Moos, Peter 1988. Geschichte als Topik. Das rhetorische Exemplum von der Antike
zur Neuzeit und die historiae im Policraticus Johanns von Salisbury. Hildesheim: Georg
Olms Verlag.
Webb, C. C. J. (ed.). 1909. Ioannis Saresberiensis Episcopi Carnotensis Policratici sive De
Nugis Curialium et Vestigiis Philosophorum Libri VIII. Recognovit et prolegomenis,
apparatu critico, commentario, indicibus instruxit, 2 vols. Oxford: Oxford University
Press.
Wians, William. 1989. “Aristotle, Demonstration, and Teaching,” Ancient Philosophy 9:
245–252.
CHAPTER FOURTEEN
Skepticism in Classical
Islam: The Case of Ghazali
PAUL L. HECK
1 INTRODUCTION
The question of skepticism in Islam invariably prompts mention of the name of
Abū Ḥāmid Muḥammad al-Ghazālī (d. 1111 CE, henceforth Ghazali), but it would
be wrong to limit skepticism during Islam’s classical period (ninth to thirteenth
centuries) to single—and apparently anomalous—cases. To be sure, skepticism did
not coalesce into a distinct school (madhhab), although hints of a tendency in that
direction exist in the literature of the ninth and tenth centuries. Also, the sources
of ancient skepticism, notably Cicero and Sextus Empiricus, were not translated
into Arabic, but Islam’s scholars were familiar with the writings of Plato, Aristotle,
and John Philoponus, and writings such as these included material on sophistical
strategies of argumentation. It is therefore best to speak of the phenomenon of
skepticism in classical Islam as part of the overall scholastic dynamic where dogmatic
assertions were always met with reservations and counterarguments, making
contradiction a regular feature of the practice of the art of disputation. The fact that
skepticism in Islam emerged in close relation to the process of defending beliefs may
offer insight for the study of skepticism in other contexts. Are skeptical objections
ever fully separable from contestation over belief? At the same time, Islam’s scholars
used precise language to describe the skeptical outlook.
Nevertheless, while skepticism existed largely within, rather than apart from, the
arena of theological reasoning, it can also be said to have its own distinct features
within that arena. To note some of the specific terminology used in relation to
skepticism: “perplexity” (ḥayra), approximating aporia in ancient skepticism; “the
equivalence of arguments” (takāfu’ al-adilla), approximating isostheneia in ancient
skepticism; and “the inability to know is knowledge” (al-‘ajz ‘an al-idrāk idrāk),
recalling learned ignorance—docta ignorantia—from the Latin West. As will be
seen, it is this last concept (learned ignorance) that best describes Ghazali’s brand
of skepticism.
It is also worth noting that Islam’s scholars distinguished the terminology for
skepticism from that used to designate atheism, which in this context does not mean
SKEPTICISM IN CLASSICAL ISLAM: THE CASE OF GHAZALI 197
denial of God’s existence but rather rejection of Islam’s revelation as a source of truth,
a position that has been traced to only a handful of savants and satirists, dubbed as
“freethinkers” (Stroumsa 1999). Islam’s scholars recognized that to say that one does
not know (lā adrī) or is perplexed (mutaḥayyir) or doubtful (shākk) in the face of
mutually opposing but equally compelling arguments on elusive theological issues is
not the same as denying (inkār) or rejecting (juḥūd) the religion of Islam. Skepticism,
then, had a place within, not apart from, the fold of Islam’s scholastic life. Indeed,
the point in advancing doubts was to establish Islam’s beliefs on a firmer basis, and
yet the process never failed to generate as much confusion as clarity.
A number of problematic questions drove the dynamic of skepticism in classical
Islam. For example, the Qur’an implies that God is corporeal in some fashion,
describing God as having eyes, hands, a face, and as seated on his throne, suggesting
spatial limitation. And prophetic sayings (hadiths) speak of God as beardless, with
curly hair, and clothed in a green garment (Williams 2002). Such revealed texts
stood in contradiction to the conclusions of discursive reasoning (both philosophical
and theological, the two being deeply interwoven during this period) that sought to
preserve the transcendent unity of God from anthropomorphist resemblance to his
creation. In the end, faced with the contradiction that God has a body and that God
does not have a body, one could only admit, “I do not know.” Indeed, according
to the celebrated rationalist scholar of the ninth century, al-Jāḥiẓ (d. 869), those
who adhered to God’s bodily self-description in scripture (anthropomorphism) over
against the demands of rational discourse only encouraged doubts about Islam’s
beliefs, ultimately casting suspicion on the legitimacy of its political mandate. His
opponents, of course, made similar accusations against him for refusing to read
divinely revealed texts in their literal sense. Were humans to question statements
that God had clearly revealed about his corporeal features?
A second factor driving the dynamic of skepticism, in addition to such
contradiction-generating questions, was the very nature of belief in Islam. Was
belief a matter of unquestioning faith—or was it knowledge, that is, compelling and
comprehensible to the mind? The Qur’an, perhaps more so than other scripture,
claims that its message amounts to knowledge (‘ilm), revealed by God to guide
human beings to a favorable outcome on Judgment Day. This would eventually
be understood in a scholastic sense—belief as rationally demonstrable—and thus
raised questions about the status of the masses of believers without skill in scholastic
argumentation by which to show that they actually had knowledge of what they
professed. The issue was particularly troubling within the dominant school of
theology, Ash‘arism, which held that in principle it is not enough for Muslims to
accept their beliefs on hearsay (as handed down on the authority of others, a concept
known as taqlīd) (Frank 1989). If the masses simply believed what scholars told
them to believe, did that qualify as belief? In other words, if they could not offer
rational argumentation for their beliefs, did it constitute genuine belief? However,
the first step in the process was to doubt what one had come to believe by hearsay.
Only after knocking down such “blind belief” could one begin to know it through
rational argumentation. In short, to be a true believer, one had to doubt. To others,
the idea that doubt was essential to belief was absurd. In his critique of Ash‘arism,
198 SKEPTICISM
the great scholar of Andalusia, Ibn Ḥazm (d. 1054) mockingly noted that for that
school, to be a believer, one had to be a skeptic!
A third driver of skepticism was the existence of multiple sects (firaq, sing. firqa)
and religions (Judaism, Christianity, and Zoroastrianism, among others) within the
Abode of Islam. All had their advocates who could defend their doctrines persuasively.
Where did religious truth lie? The plurality of truth claims led diversely affiliated
scholars to pursue skeptical strategies in their debates with one another, seeking to
highlight contradiction (tanāquḍ or ta‘āruḍ) in their opponents’ arguments rather
than common ground upon which all could agree. Such theological wrangling only
suggested that mutually contradictory beliefs could be equally compelling, leading
some to conclude that religious truth is unascertainable. The literary virtuoso of the
eleventh century, Abū Ḥayyān al-Tawḥīdī, inspired by his own deep piety, mocks
the craft of disputation (kalām, lit. speech), citing a number of cases of scholars who
opted for religious skepticism as the logical outcome of theological debate (‘Abbās
1966; Van Ess 1968). More mystically inclined, al-Tawḥīdī is essentially implying
that theological disputation is a waste of time that only produces a cacophony of
contradictory claims, thereby putting religious truth itself in jeopardy.
All of this made it seem that the practice of disputation only ended in the
suspension of belief. Some therefore rejected it as a threat to Islam, opting for
adherence to the literal wordings of scripture. Others, also rejecting it, looked to
philosophy, not as a self-standing method to attain certain knowledge (although
some did make such a claim) but as a means to clean up the theological morass
in Islam. Surely, it was felt, all parties would accept what could be apodictically
demonstrated, and Islam’s scholars, beginning from the ninth and tenth centuries,
sought to formulate the claims of Islam in philosophical categories, a process that
reached a zenith with al-Fārābī (d. 950) and Ibn Sinā (Avicenna, d. 1037) and also
less clearly but no less significantly with al-‘Āmirī (d. 992) and al-Juwaynī (d. 1085).
This, however, only led to new dilemmas. The idea that truth is to be established
by philosophical method rather than by a message from God had the effect of
downplaying scripture at least for the learned elite. The wordings of revelation—
the gardens of paradise and fires of hell—were at best a distant imitation of the
philosophical categories of ultimate happiness and ultimate wretchedness, at worst
merely images useful for controlling the unlearned masses that had no hope of
obtaining certain knowledge philosophically. But philosophical circles were also
not free of taqlīd (Griffel 2005), the practice of accepting beliefs on the authority
of others, the very issue that had encouraged use of the philosophical method in the
first place! Just because Aristotle said something does not mean it is to be uncritically
accepted. How was philosophy to be held to its own standards?
2 GHAZALI, PHILOSOPHICAL SKEPTICISM, AND

THE TURN TO MYSTICAL SCHOLASTICISM
Those conundrums all left their mark on the thought of Ghazali who is best
remembered for the skeptical crisis he faced at the height of his teaching career.
However, it is important to consider carefully the nature of his skepticism, which,
generally speaking, involves reservations about the prevailing scholastic methodology

of his day. Arguing on the basis of the heritage of scholastic philosophy, which he
had mastered, Ghazali made use of a kind of skepticism to counter the claims of
philosophers to have surer knowledge of God than prophets. In other words, he
was not at all opposed to philosophical reasoning but rather used it to expose its
own limits. Several studies in recent years underscore his debt to Avicenna (Frank
1994; Griffel 2009; Treiger 2012). However, it has also been shown that even in his
most philosophical writings, likely addressed to his elite disciples, Ghazali remained
sensitive to the doctrines of Islam: for example, God’s knowledge of particulars in
addition to universal concepts, which philosophers were inclined to deny (Al-Akiti
2009). In the end, Ghazali is keen to show the limits of the philosophical method
even if accepting it as path to certain knowledge. He departs from philosophers, as
will be seen, in his claim that the philosophical method, if properly pursued, ends
in the recognition that the philosophical method itself only begets ignorance of
the reality of God and, in turn, of all things, which, after all, have their existence
in God. In other words, the philosophical method alone is not enough to establish
certain knowledge of the reality (or quiddity) of things (what they truly are rather
than what they seem to be). How is one to know the true nature of things by the
philosophical method if one must admit—by the very same method—ignorance
of the reality (or quiddity) of their creator? The point is that knowledge of the
reality of things requires knowledge of the reality of their creator, but philosophy
has been shown unable to attain certain knowledge of the creator. Another method
is needed, still philosophically sound but now, for Ghazali, enhanced by mystical
insight. This features in the way he views syllogistic reasoning. In The Just Balance
(al-Qisṭās al-Mustaqīm), he shows that Aristotle’s categories of logic are derivable
from the Qur’an, establishing consistency between knowledge attained by syllogistic
reasoning and knowledge revealed by God. However, while generally committed to
Aristotle’s categories of logic, Ghazali has a tendency to conduct his argumentation
on the basis of a mystically informed grasp of the spiritual realm as much as by
syllogistic reasoning. This is not to say that such a pursuit of logic took hold in
scholastic circles. The logic of Avicenna remained dominant. Rather, by grounding
scholasticism in mystical insight, Ghazali was able to dull the challenge of philosophy.
This, however, does not mean that Ghazali’s goal was a simple restoration of
Islam’s revealed message to a place of prominence over philosophy. Ultimately, he
sought to establish a sounder basis for knowledge of the realities of things, a task that
in his view the philosophical method alone failed to accomplish. This is clear in The
Incoherence of the Philosophers (Tahāfut al-Falāsifa), written not in condemnation
of the philosophical method, but rather as a challenge to the supremacy of its claims
to yield certain knowledge, as seen notably in its discussion of causality. There,
Ghazali argues against philosophers who mistakenly assume a quasi-deist view of
the world. His issue with them is not so much the idea of efficient causality, whereby
things happen according to their own natures apart from direct intervention by
God, but rather with their failure to see that such a claim belongs to the realm of
possibility rather than necessity; efficient causality must therefore be qualified as
custom, not as necessity. That is, any causal relation between things—fire making
200 SKEPTICISM
cotton burn—is not due to any necessity. It is simply a custom. Thus, while today
cotton burns because fire is applied to it, tomorrow it might cool as a result. There
is no necessary connection. That would be to limit God’s power over the way things
work. This, of course, was Ghazali’s way of maintaining the sense of miracles,
which would be impossible in a world run by efficient causality rather than by God.
Indeed, as is his wont, Ghazali shows that the arguments of philosophers who argue
for efficient causality fall short of their own method (Marmura 1989). The fact
that philosophical arguments for efficient causality have their flaws is reason to
doubt the philosophical method, casting suspicion on the sufficiency of syllogistic
reasoning. Again, Ghazali is not dismissing philosophy. For him, it is necessary for
the establishment of certainty. Rather, his point is that while necessary, alone it is not
sufficient. Given its flaws, there needs to be a higher authority by which to adjudicate
the claims of philosophy—a higher and thus meta-rational but not contradictory or
anti-rational authority. For this reason, Ghazali turns to mystical insight, not to
replace but rather in order to enhance—more precisely to adjudicate—knowledge
claims reached by the philosophical method. In other words, he does not view
mystical insight as foreign to scholasticism. But what does he mean by mystical
insight? Ghazali was not claiming that individual mystical experience was a source
of knowledge. (After all, a madman could claim such experience.) Rather, he sought
to introduce mystical insight of a kind into the scholastic process in order to buttress
what he saw as its methodological deficiencies. For this reason, we can speak of
Ghazali’s system of thought as a mystical-philosophical scholasticism.
Born toward the middle of the eleventh century in what is today northeastern Iran,
Ghazali was a towering scholar in his own age, writing works in various disciplines,
including philosophy, theology, jurisprudence, and ethics. He was appointed in 1091
to the post of chief teacher at the Niẓāmiyya Madrasa in Baghdad, then the leading
center of learning in the Abode of Islam, where upwards of 300 students at a time
studied under his tutelage. According to his own testimony, Ghazali’s crisis began in
1095 at the height of his professorial career, when he came to the realization that he
was no longer sure of his own scholarly convictions. Excusing himself on the pretext
of making the pilgrimage to Mecca, he departed Baghdad and made his way first to
Damascus, where he would spend the bulk of the next two years in contemplative
retreat. From there, he would visit the holy sites in Jerusalem, Medina, and Mecca.
By 1097, he was back in Baghdad, but he did not remain long. He would resume
teaching, this time in Nishapur, where he had earlier studied under the direction of
al-Juwaynī. He spent his last years there, consolidating his ideas, teaching them to
disciples, and fending off accusations, only to return to his hometown of Tus not
long before his death in the year 1111.
It is in The Deliverer from Error (al-Munqidh min al-Ḍalāl) that Ghazali records
his skeptical crisis, but what exactly is the nature of his skepticism? This question,
which has not gone untreated (Halevi 2002; Kukkonen 2010; Garden 2013), can
only be answered by looking at this quasi-autobiographical work (Menn 2003) in
the context of his overall oeuvre. The Deliverer from Error opens with notice of the
prevailing doctrinal confusion (iḍtirāb al-firaq) and divergence (ikhtilāf) in Islam,
each sect claiming to be the saved sect. Ghazali clears the deck of such discordance,
noting how it had always been his habit to inquire into the beliefs of the varied
schools on his own, apart from the pronouncements of scholarly authorities. In
this way he disassociates himself from taqlīd (i.e., belief on the authority of others),
but he goes further, noting another source of skepticism, namely, the existence of
other religions. His observation that Jewish, Christian, and Muslim youth are all
raised according to family tradition—and tend not to leave it as adults—is meant to
show that his pursuit of knowledge will start from original human nature prior to a
person’s exposure to teachings handed down by parents and teachers (i.e., prior to
taqlīd). Such traditional teachings are not necessarily wrong but cannot be counted
as knowledge if received on the authority of others. Ghazali’s record of his own life
thus reflects the key factors that encouraged doubts about a single religious truth in
the Abode of Islam as outlined in the first section above.
This skeptical posture leads him into crisis. He wants to know things as they
really are, but having begun by rejecting traditional beliefs, he is unsure where to
turn for sources of knowledge. He even doubts the senses. The sun looks small to
the naked eye, the oar bent in the water, but the mind corrects such misperceptions.
If not the senses, then, can he trust the mind and first principles that the mind says
are necessarily true, such as the claim that ten is greater than three or that the same
thing cannot be both true and false? If the senses err, perhaps the mind does, too.
Just as the mind has to correct misperceptions of the senses, perhaps there is a
meta-rational realm where even the conclusions of the mind are judged to be right
or wrong. Dreaming, Ghazali explains, offers a concrete analogy for such a meta-
rational realm; it does not operate contrary to the workings of the senses or the
mind but offers evidence of a cognitive state above and beyond them. Ghazali, it
should be noted, is not looking to dreaming to resolve his dilemma but introduces it
only as evidence for the existence of a cognitive state comparable to mystical insight,
setting the stage for him to reestablish the entire scholastic project on the basis of a
combination of mystical insight and syllogistic reasoning. In the end, as he relates,
he was healed of his skeptical disease, not as a result of a renewed confidence in first
principles (on the basis of which he might build compelling arguments about the
reality of things) but rather by a light (nūr) that God had cast into his breast. This
light, it will be seen, can be taken as intuition of a kind that is at once meta-rational
and the fruit of the philosophical method.
What exactly was this light? It was not a personal or individual mystical
experience. Ghazali was too committed to the scholastic project to rely on personal
or individual experience, which he did not deny but did view as dangerous, since it
offered grounds for individuals to claim to embody God. The light cast into his breast
was, rather, a way of undertaking the scholastic project with mystical insight—that
is, not seeing God (as in a personal or individual mystical experience) but seeing
things in this world only insofar as they exist with God their creator. In other words,
he discovered that rather than limiting himself to the philosophical method in the
traditional sense (i.e., by trying to understand the causal nature of things and building
from there to higher conclusions), he could amplify it by acquiring knowledge of
things not solely in terms of what was rationally comprehensible but also from a
meta-rational perspective. If the scholastic method begins with observation of things
202 SKEPTICISM
and ends with knowledge of their nature (or causality, e.g., the nature of fire being to
cause things to burn), the mystical-philosophical approach involves seeing things—
even the smallest of creatures such as a gnat—not simply as part of a rationally
comprehensible order in which everything is known according to its nature, but
rather first and foremost as part of God’s existence. The philosophical method,
it turns out, is limited, only able to yield knowledge of things according to their
nature, whereas mystical-philosophical scholasticism yields that and more, namely,
knowledge of things as they exist with God—and thus as they truly are. Ghazali saw
this as surer knowledge of the reality of things, following logically from his view that
all things other than God only have existence in a metaphorical sense, that is, insofar
as they receive it from God. In contrast, God alone has existence in reality. Thus, if
God is the only thing with real existence, the surest knowledge of the reality of things
comes not from inquiry into their nature per se but rather by seeing them insofar as
they exist in the singular reality of God. He thus turned to Sufism, not for a personal
mystical experience, but for a cognitive state of a meta-rational kind as referee for
philosophical (rational) claims, just as the mind is referee for sense-based perceptions.
One might argue whether Ghazali is talking about mysticism in a traditional sense
of a direct and ineffable communion with God in which one might see God, but he
is speaking of the meta-rational (what cannot be reached by syllogistic reasoning
alone) and thus, at least in that sense, the mystical. However, since, as we will see, he
reaches this conclusion (the limits of philosophy) by the scholastic method, he never
departs from scholasticism but is, rather, driven to ground it anew not solely in the
scholastic method in the traditional sense but in awareness (albeit indirect) of God
as everywhere present, making it impossible to know things in their true reality (or
quiddity) except insofar as they exist with God as the singular reality, as we will see.
The rest of The Deliverer from Error is largely devoted to explaining the
shortcomings of other approaches to knowledge. Theology (i.e., disputation),
designed for the defense of the faith, has nothing to say to those who do not already
believe; and its practitioners (ahl al-kalām), having left the bounds of apologetics
for more obscure forms of speculation beyond their ken, fail to dispel the doubts
that arise from differences in belief. Philosophy, for its part, has much to offer,
but its leading figures (Ghazali singles out al-Fārābī and Ibn Sinā) are declared
infidels, particularly for their rejection of three doctrines: bodily resurrection, God’s
ignorance of particulars in addition to universal concepts, and the created status of
the world, which they hold to have existed from eternity as a necessary effect of
God’s eternal knowledge. They may or may not be right in these views, but their
method is not sufficient to establish their claims as certain.
Also, in general, Ghazali expresses concerns about the impact of philosophy on
the masses of believers. The assumption that Islam is opposed to philosophy makes
it seem that a believer has to choose between the life of the mind and obscurantist
forms of Islam. As for those who claim to follow the instructions of the infallible
imam (Ismā‘īlism, which makes the imam the infallible source of truth—in contrast
to papal infallibility, which is limited to the preservation of inherited teachings),
Ghazali sees them as the most illogical of all for their claim that the mind is not to be
trusted since people invariably disagree, necessitating the existence of the infallible
imam to resolve all disputes. He repeatedly charges them with perplexity (ḥayra),
that is, with encouragement of obscurantist belief (and thus of skepticism) for their
distrust of syllogistic reasoning, which, Ghazali asserts, can resolve disagreement
among Muslims, obviating the need for the infallible imam, if people would only
take the time to learn it properly!
Finally, he comes to Sufism, which he sees as the method to attain certain
knowledge. He speaks of it in close conjunction with demonstrative proof (burhān),
that is, in terms of the philosophical method, but also as something meta-rational.
His point is that knowledge can be reached apodictically but that certain knowledge
of the reality of things is known only through taste (dhawq), by which he means
mystical insight (based on spiritual consciousness, wijdān, rather than analogical
reasoning, qiyās). In other words, metaphysical realities can be known of through
syllogistic reasoning, but they can only be really known through a mystical seeing
that takes place not with the physical eye but with the mind’s eye. This eye has the
capacity to see metaphysical realities, thereby acquiring certain knowledge of them,
but it needs a light by which to see them clearly, namely, the light that comes from
the niche of prophecy (mishkāt al-nubūwa). Islam’s revealed message—prophecy—
is thus the means to mystical insight into metaphysical realities, allowing for certain
knowledge of the reality of things. In other words, mystical insight as a meta-rational
phenomenon needs a grounding in a meta-rational source (revelation); those who
draw from this source by frequent recollection of the names of God as revealed by
prophecy will be able to cultivate the cognitive state that allows them not only to know
of metaphysical realities with certainty but also to see (know) them directly as the
prophets did. Ghazali is not suggesting that mystically minded scholars can become
prophets, but rather that they can enjoy prophet-like knowledge by cultivating a life
in conformity to that of Muhammad, especially his ritual activity. Here, Ghazali is
influenced by Avicenna, speaking of prophecy not simply as a message from God,
authenticated by a prophet’s performance of miracles, but as an internal cognitive
state that all have the potential to cultivate, allowing them to acquire prophetic
knowledge of the realities of things. And it’s also worth noting that Ghazali is not
claiming that philosophers and prophets disagree. Quite the contrary! What he is
saying is that philosophy, to be true to its own method, requires meta-philosophical
adjudication if it hopes to establish its claims as certain.
It is best to speak of Ghazali’s brand of skepticism as a kind of learned ignorance:
the philosophical method, on its own terms, only yields ignorance of the reality of
God. Ghazali states this clearly in The Highest Goal in Explaining the Most Beautiful
Names of God (al-Maqṣad al-Asnā fī Sharḥ Asmā’ Allāh al-Ḥusnā). There, he shows
that there is no way to reason to the infinite God by analogy to finite creatures (ex
creaturis), making it impossible to acquire certain knowledge of God’s reality, both
essence and attributes (powerful, knowledgeable, living, speaking, seeing, hearing,
etc.), by the philosophical method in the traditional sense. The following passage
captures the gist of his thinking:
If you said: what is the endpoint of the knowledge of the knowers of God the
Exalted [nihāyat maʿrifat al-ʿārifīn bi-llāh taʿālā]? We would say: the endpoint
204 SKEPTICISM
of the knowledge of the knowers is their inability to know [ʿajzuhum ʿan al-
maʿrifa]. Their knowledge in truth is that they do not know him; that it is
completely impossible to know him; that it is impossible that anyone but God the
Mighty and Majestic know God with true knowledge encompassing the essence
[kunh] of the attributes of lordship [ṣifāt al-rubūbiyya]. If that is disclosed to
them by demonstrable proof [inkishāf burhānī], as we noted, they would know it,
that is, they would reach the endpoint that it is possible for creation [humanity]
to know. This is what the great righteous one Abū Bakr, God be pleased with
him, meant when he said: “The inability to grasp comprehension is a kind of
comprehension.” (Ghazali 1971: 54)
The last line of this passage, a quote attributed to one of the leading companions
of Muhammad and his first caliphal successor, indicates that learned ignorance is
not the end of knowledge but yields “a kind of comprehension.” This becomes,
for Ghazali, a new point of departure for scholasticism. This point of departure is
a monistic vision of existence, allowing for knowledge of the reality of things not
simply according to their nature as determined by the philosophical method but as
they exist with God. In short, learned ignorance teaches that God has no definition,
no limit by which to be defined and therefore no opposite. Thus, while God is
not the world, God is also not other than the world. In sum, Ghazali’s skepticism
about the philosophical method—as demonstrated by the philosophical method—
leads him to a monistic vision of existence that requires mystical insight alongside
syllogistic reasoning as proper basis of the scholastic enterprise.
Everything that exists is a light from God’s lights and a trace of God’s pre-eternal
power, and just as one claims to see only the sun when beholding its rays stretching
over mountains, similarly, one can say, “I know only God and I see only God”
(Ghazali 1971: 59). This monistic vision of existence, curious byproduct of his
philosophical skepticism, features across Ghazali’s oeuvre, notably in The Niche of
Lights (Mishkāt al-Anwār) but also in a number of places across his magnum opus,
The Revivification of the Religious Sciences (Iḥyā’ ‘Ulūm al-Dīn). God is now in sight,
at least for mystically minded scholars who see God as manifest without limit (via the
light of the mind’s eye in conjunction with the light of revelation, just as the physical
eye needs the sun to see). Again, mystical insight here is not personal experience.
Indeed, it remains within the realm of scholasticism, enhancing its methodology
with awareness of God’s unbounded existence, everywhere manifest in the effects
of his actions. This yields certain knowledge of the reality of things, since it takes
God’s existence as starting point. In turn, it begets a sound love for God, as seen in
the following passage from The Revivification of the Religious Sciences:
Those with strong insight [baṣīra] and endowment [minna] that is not weakened,
when in a state of inner balance, see only God the Exalted, are not aware of
anything but him, and know there is nothing in existence except God. His acts
are a trace of his power [qudra] and are subject to him. They have no existence
in reality [bi-l-ḥaqīqa] apart from him. Existence belongs to the One, the Truth.
Through him is the existence of all his actions. Those in this state look at actions
only to see in them the agent [God]. . . . The entire world is the composition of
God the Exalted. Those who look at it as God’s action, are aware of it as God’s
action, and love it as God’s action—look only at God, know only God, and love
only God. (Ghazali 2003: vol. 2, 1680–1682, passim)
It is worth noting that this idea of human love for God, which is dependent on
Ghazali’s monistic vision of existence (itself the curious by-product of his skepticism)
and the possibility it affords of meta-rational knowledge obtained through mystical
insight in conjunction with syllogistic reasoning, does not imply any ontological
likeness between creation and creator. The idea of such human-divine affinity did
feature in some circles of Sufism, where spiritual masters claimed to reflect God’s
existence in their very own human being. Rather, for Ghazali, human love for
God—he is more ambiguous about the possibility (and criteria) of divine love of
humans—is the fruit not of ontological affinity but of the knowledge (maʿrifa) of
God’s unbounded existence that results from his system of mystical-philosophical
scholasticism, as seen in the following passage from The Revivification of the
Religious Sciences:
All that exists in relation to the power of God the Exalted is like shade in relation
to the tree, light in relation to the sun; all things are the traces of his power [āthār
qudrathi], and the existence of all belongs to his existence [tābiʿ li-wujūdihi]. . . .
Kindness from people can only be conceived metaphorically, for the one who
is kind is God the Exalted [that is, God is the singular source of all qualities
that are loved]. . . . My goodness! Who can deny that it is really possible to
love God the Exalted [in a way other than simple obedience to his commands
and prohibitions]? . . . Who denies that these descriptions are the descriptions
of beauty, perfection, and goodness . . . and that they describe God? . . . The
closeness of the slave [that is, the human] to his Lord the Mighty and Majestic
is in terms of attributes [not in terms of physical closeness], attributes which
he is commanded to follow and the characteristics of lordship: knowledge,
righteousness, goodness, kindness, mercy, guidance, and so on, which he is
commanded to emulate. . . . All regions of the realm [malakūt] of the heavens
and the earth are the domain of the knower, who takes his place therein as he
wishes. . . . Whoever has awareness [maʿrifa] of God the Praiseworthy will have
disclosed to him the mysteries of the realm [mulk] of God, even a little, and
upon this disclosure, he will be struck in his heart with joy that will almost make
him fly. . . This is something comprehended only by taste [dhawq, the method of
mystical-philosophical scholasticism]. (Ghazali 2003: vol. 2, 1658–1668, passim)
Here, the call to scholarly emulation of God’s attributes shows the ethical climax of
Ghazali’s thought. He had first sought to get the learned elite (i.e., philosophers) to
admit their own ignorance of the reality of things, and that by their own method,
compelling them, if truly learned, to turn to mystical insight—made possible by the
light of prophecy—for certain knowledge of the reality of things. He now seeks
to put them in the position of having to emulate the attributes of God (as known
by revelation and not solely by the philosophical method) in the manner of the
prophets—and to do so because that is what the philosophical method demands
206 SKEPTICISM
in the end! Thus, what began as skepticism of a kind ends with the claim that the
philosophical method, when properly pursued, should convince the learned of
the necessity of living like the prophets! In other words, the call to emulate God’s
attributes is not simply a devotional matter. It actually serves as evidence that those
who claim to have knowledge actually possess it. A scholar who does not live a
prophetic ethics as modeled by Muhammad is therefore scholastically deficient—his
claim to possess knowledge (especially but not only of God) is suspect.
Of course, the emulation of God’s attributes that Ghazali has in mind does not
transgress the ontological divide between creator and creation, as earlier noted.
In The Highest Goal in Explaining the Most Beautiful Names of God, he shows
how it is possible to identify with the attributes of God in appropriately scholastic
fashion. Since God has made these attributes (or names) known by revelation to
humanity, he must have intended them to be comprehensible to humans in some
fashion. In other words, if humans are to identify with them, they have to be able
to grasp their conceptual meaning beyond their literal wordings. Ghazali proceeds
to explain in detail how this is to be done in the case of each of God’s revealed
names. Indeed, for him, it is possible to identify even with the primary name of
God—Allah—through a kind of theosis (ta’alluh), where one immerses one’s heart
and aspiration wholly in God: not seeing other than God, not heeding other than
God, not having hope in—or fear of—other than God. However, in line with
his scholastic approach, which preserves the philosophical method while going
beyond it, Ghazali turns to the philosophical categories of the day to explain the
conceptual meaning of God’s names, thereby allowing human identification with
them. In the end, human emulation of God’s names serves as ethical balance for
the human soul, giving the rational faculty mastery over the soul’s lower appetitive
and irascible faculties. In this way, Ghazali distinguishes between the way in
which the names or attributes are applied to God and the way in which they are
applied to humans, demonstrating the scholastic sense of emulating God’s names
or attributes. For example, when explaining the divine name, “king” (malik), he
begins by describing “king” as one who needs nothing where everything needs him.
In this sense, the name applies exclusively to God. The human being cannot be king
in this fashion but does have a kingdom in his heart and soul where his soldiers are
his appetite, anger, and passion, and his subjects his tongue, eyes, hands, and all
his bodily members. If he rules them and they do not rule him, he attains the rank
of king in his world. Thus, Ghazali’s monistic vision of existence has resulted in
a corresponding monistic vision of ethics whereby humans are able to imitate the
attributes of God in a way that is both grounded in Islam’s revealed message and
philosophically compelling.
To conclude, it is now possible to understand the nature of Ghazali’s skeptical
crisis and its resolution as described in The Deliverer from Error. His skepticism is
skepticism about the knowledge that comes from the philosophical method. He is
not against the philosophical method, but rather seeks to put it in its proper place.
It may yield knowledge of the reality of things, but such knowledge is not supreme,
existing only as fruit of a rationally comprehensible order in which the nature of
things can be determined. Such knowledge requires adjudication by a meta-rational
cognition, just as our sense perceptions need to be adjudicated by the mind. Thus,
for certain knowledge of the reality of things, one must incorporate mystical insight
into one’s scholastic inquiry. This mystical insight, the fruit of learned ignorance,
is the light that Ghazali says God cast in his breast, allowing for the cultivation of
a prophet-like knowledge—that is, certain knowledge—of the reality of things. He
thus forges a close connection between skepticism and (a kind of) mysticism, both of
which remain within the scholastic world to which he belonged. At the same time,
he is led by his own scholastic commitments to raise questions, not of the efficacy
of the philosophical method but of its sufficiency, leading him to doubts, albeit not
outright rejection, of some of the philosopher’s conclusions that are at odds with
Islam’s beliefs.
It is worth noting that Ghazali’s skepticism does not come out of the blue but
builds upon a history of doubt in Islam that questioned the claims of discursive
reasoning (both theological and philosophical) to certain knowledge of the reality
of God (Heck 2014). At the same time, he wrote at a time when such reasoning,
particularly the legacy of Avicenna, was becoming more fully integrated into the
scholarly life of Islam. Ghazali is thus faced with the challenge of both affirming the
philosophical method but also putting it in its place, drawing upon a long heritage
of skeptical stratagem to best philosophy at its own game. In sum, he made the
philosophical method more palatable to Islam’s scholarly milieu by raising doubts
about its supremacy. While helping to consolidate the place of Avicenna in the
scholastic curriculum, he also established grounds for scholars to attack Avicenna
(Shihadeh 2013) and arguably helped set the stage for doubts to be raised about the
sufficiency of the logic of Aristotle as basis of knowledge (Ibrahim 2013).
REFERENCES
‘Abbās, Iḥsān 1966. “Abū Ḥayyān al-Tawḥīdī wa-‘Ilm al-Kalām,” al-Abḥāth 19: 189–207.
Al-Akiti, Afifi M. 2009. “The Good, the Bad, and the Ugly of the Falsafa: Al-Ghazālī’s
Maḍnūn, Tahāfut, and Maqāṣid, with Particular Attention to their Falsafī Treatments of
God’s Knowledge of Temporal Events.” In Y. Tzvi Langermann (ed.), Avicenna and his
Legacy: A Golden Age of Science and Philosophy, 51–100. Turnhout, Belgium: Brepols.
Frank, Richard, M. 1989. “Knowledge and Taqlīd: The Foundations of Religious Beliefs in
Classical Ash‘arism,” Journal of the American Oriental Society 109: 37–62.
Frank, Richard, M. 1994. al-Ghazālī and the Ash‘arite School. Durham: Duke University
Press.
Garden, Kenneth. 2013. The First Islamic Reviver: Abū Ḥāmid al-Ghazālī and his Revival
of the Religious Sciences. Oxford: Oxford University Press.
al-Ghazālī. 1967. al-Munqidh min al-Ḍalāl, eds. Jamīl Salībā and Kāmil ‘Ayyād, 7th edition.
Beirut: Dār al-Andalus.
al-Ghazālī. 1971. al-Maqṣad al-Asnā fī Sharḥ Asmā’ Allāh al-Ḥusnā, ed. Fadlou A. Shehadi.
Beirut: Dār al-Mashriq.
al-Ghazālī. 1993. al-Qisṭās al-Mustaqīm, ed. Maḥmūd Bījū. Damascus: al-Maṭba‘a al-‘Ilmiyya.
al-Ghazālī. 1997. Tahāfut al-Falāsifa, ed. and trans. Michael E. Marmura. Provo, Utah:
Brigham Young University Press.
208 SKEPTICISM
al-Ghazālī. 1998. Mishkāt al-Anwār, ed. and trans. David Buchman. Provo, Utah: Brigham
Young University.
al-Ghazālī. 2003. Iḥyā’ ‘Ulūm al-Dīn, 2 vols. Cairo: Dār al-Salām.
Griffel, Frank. 2005. “Taqlīd of the Philosophers: al-Ghazālī’s Initial Accusations in his
Tahāfut.” In S. Günther (ed.), Ideas, Images, and Methods of Portrayal: Insights into
Classical Arabic Literature and Islam, 273–296. Leiden: Brill.
Griffel, Frank. 2009. Al-Ghazālī’s Philosophical Theology. Oxford: Oxford University
Press.
Halevi, Leor. 2002. “The Theologian’s Doubts: Natural Philosophy and the Skeptical
Games of Ghazālī,” Journal of the History of Ideas 63: 19–29.
Heck, Paul L. 2014. Skepticism in Classical Islam: Moments of Confusion. London:
Routledge.
Ibrahim, Bilal. 2013. “Fahr al-Dīn al-Rāzī, Ibn al-Haytam and Aristotelian Science:
Essentialism Versus Phenomenalism in Post-Classical Islamic Thought,” Oriens: 379–431.
Kukkonen, Taneli. 2010. “Al-Ghazālī’s Skepticism Revisited.” In H. Lagerlund (ed.),
Rethinking the History of Skepticism: The Missing Medieval Background, 2–59. Leiden:
Brill.
Marmura, Michael E. 1989. “al-Ghazālī on Bodily Resurrection and Causality in the
Tahāfut and the Iqtiṣād,” Aligarh Journal of Islamic Thought 2: 46–75.
Menn, Stephen 2003. “The Discourse on the Method and the Tradition of Intellectual
Autobiography.” In J. Miller and B. Inwood (eds.), Hellenistic and Early Modern
Shihadeh, Ayman. 2013. “A post-Ghazālian Critic of Avicenna: Ibn Ghaylān al-Balkhī on
the Materia Medica of the Canon of Medicine,” Journal of Islamic Studies 24: 135–174.
Stroumsa, Sarah. 1999. Freethinkers of Medieval Islam: Ibn al-Rāwandī, Abū Bakr al-Rāzī,
and their Impact on Islamic Thought. Leiden: Brill.
Treiger, Alexander. 2012. Inspired Knowledge in Islamic Thought: Al-Ghazālī’s Theory of
Mystical Cognition and its Avicennian Foundation. London: Routledge.
Van Ess, Josef. 1968. “Skepticism in Islamic Religious Thought,” al-Abḥāth 21: 1–18.
Williams, Wesley. 2002. “Aspects of the Creed of Imam Ahmad Ibn Hanbal: A Study of
Anthropomorphism in Early Islamic Discourse,” International Journal of Middle East
Studies 34: 441–463.
CHAPTER FIFTEEN
Nicholas of Autrecourt
and John Buridan
on Skepticism
CHRISTOPHE GRELLARD
1 INTRODUCTION
After John Duns Scotus’s radical criticism of Henry of Ghent, the skeptical
challenge becomes more and more central to medieval epistemology.1 The
reduction of a theory to skepticism, by showing its skeptical consequences, is
used as a kind of ad hominem argument (see Grellard 2011b). Moreover, the
development of the argument of a deceiving god, in the Scotist and Ockhamist
traditions, urges every philosopher and theologian to answer the skeptic.2 Hence,
a kind of ideal skeptic, who absolutely denied the possibility of knowing anything
with certainty, emerges as an unavoidable adversary. It becomes necessary for any
epistemological system to justify the mere possibility of knowing. This kind of
attitude to the skeptical challenge is key to understanding Nicholas of Autrecourt’s
and John Buridan’s epistemologies. The two philosophers were colleagues at the
Faculty of Arts of the University of Paris, during the 1330s.3 Both tried to refute
skepticism, but their solutions are radically different and even opposed. Indeed,
Autrecourt’s answer to skepticism can be labeled as foundationalist, whereas
Buridan’s can be labeled as externalist. Moreover, Buridan seems to think that
Autrecourt’s foundationalism may lead to a form of skepticism, and he partly
targets him in his general criticism of skepticism. In this chapter, I will first
examine Nicholas’s attempt to answer skepticism and its failure. Such a failure
may explain why Nicholas has been usually considered to be skeptic by most
historians of medieval philosophy. Then, I will turn to Buridan’s solution and
clarify his relation to Autrecourt.
210 SKEPTICISM
2 NICHOLAS OF AUTRECOURT’S
UNWILLING SKEPTICISM
In a letter to the Franciscan friar, Bernard of Arezzo, Nicholas of Autrecourt
explicitly claims that he wants to avoid any form of skepticism:
And, as it seems to me, from your position there follow things that are more
absurd than follow from the position of the Academics. And, therefore, in
order to avoid such absurdities, I have upheld in disputations in the Aula of the
Sorbonne that I am evidently certain of the objects of the five senses and of my
own acts. (Autrecourt 1994: 55–57)
Clearly, to answer skepticism means to preserve the cognitive value of external
(viz., sensible) and internal perception. Therefore, Autrecourt’s solution will consist
in using some kind of perception, along with some intellectual principles, as the
foundations of our knowledge. But before examining his foundationalism, we have
to individuate the kind of skepticism Autrecourt is targeting.
2a The Invention of Modern Skepticism

In the first letter to Bernard of Arezzo, Nicholas is involved in a controversy about
the intuitive knowledge of a non-existent object. In the Scotist tradition to which
Bernard belongs, God, by his absolute power, may destroy an object but preserve a
clear perception of it in the spirit of a man (see Autrecourt 1994: 46–47).4 Nicholas
tries to show how such a position leads to several skeptical absurdities. By this
reductio ad absurdum, Autrecourt wants to disqualify Bernard’s epistemology. But
he also builds a new picture of the skeptic, different in some respects from the
skepticism inherited from Augustine. First, if it is logically possible for the act of
cognition to exist without its object (since we can have a clear idea of an object
that does not exist), we have to conclude that it is impossible to distinguish the true
from the false, since truth consists in the correspondence between an object and an
idea of it. Then, we cannot have evident knowledge of external objects (Autrecourt
1994: 46–47). To this first objection, Bernard replies by distinguishing between the
natural and the supernatural order of knowledge. In the common course of nature,
there is no knowledge of a non-existent object. As we shall see, Buridan defends a
very similar view. So, if we add the clause “God is not intervening in the course of
nature,” such an inference as “a whiteness is seen, therefore a whiteness exists” is
evident. But Nicholas cannot accept such a reply since, according to him, we are
unable to identify the cases in which God is intervening and those in which He is
not (Autrecourt 1994: 50–51). It is therefore impossible to escape a general doubt
about the existence of things outside the mind.
The second step of the reduction to skepticism concerns our mental states.
According to Bernard himself, we don’t have a direct knowledge of our mental
states. Therefore, our perceptions of external things are clearer than the
perception of our inner states. Against the Franciscan, Nicholas uses an a fortiori
argument: if we can’t have an evident perception of external things (as has just
NICHOLAS OF AUTRECOURT AND JOHN BURIDAN ON SKEPTICISM 211
been shown), a fortiori we can’t have an evident perception of things we perceive

less clearly, as is the case with our mental states (Autrecourt 1994: 52–53). From
this, Nicholas concludes that Bernard is a complete solipsist: he does not know
whether external things exist, he does not know whether he himself exists as a
body, and he does not know whether he has mental states, in such a way that
he cannot say that he thinks or has an intellect: “Moreover you will not have
certitude about your own mind either and thus you do not know whether it
exists” (Autrecourt 1994: 54–55).
This reductio ad absurdum clearly shows that, according to Autrecourt,
radical skepticism is strongly linked to the extension of God’s absolute power.
Indeed, such a power leads to a general doubt about the existence of the external
world and to a strong form of solipsism. As Nicholas explains, Bernard’s position
is much more absurd than the Academics’ (see the quotation at the beginning
of Section 2). We probably reach here the first clear distinction between
ancient and modern skepticism. Indeed, in Against the Academicians Augustine
explained, in arguing against Carneades, that the Academics never dared to
question the existence of the external world.5 For this reason, we may consider
the problem of the external world (and the solipsism strongly connected to it)
as a touchstone for distinguishing between ancient and modern skepticism, and
such a touchstone explicitly appears for the first time in Autrecourt’s criticisms
of Bernard of Arezzo.
2b A Foundationalist Answer to Skepticism

In order to avoid skepticism, Nicholas claims, we need to accept some self-evident
basic beliefs as the foundation of all our knowledge. These beliefs are of two kinds:
immediate perceptions and analytic propositions:
The following are evident, properly speaking: sensible objects, and the acts which
we experience in ourselves. These refer to what is incomplex. As regards what is
complex, there are principles which are known from their terms, and conclusions
depending on them. (Autrecourt 1971: 235)
Propositional (or complex, in scholastic terms) evidentness includes the first principle,
namely, the principle of non-contradiction,6 and every analytic proposition that
can be used as a premise in an inference. The first principle plays the role of a
general condition of possibility for every discursive truth, since on the one hand no
truth can be contradictory, and on the other the first principle cannot be falsified
by God’s absolute power. The first principle allows us to secure a first level of
evidentness against the skeptics. The warrant of this first level can be extended to
other propositions by a reduction to the first principle. To reduce means “to show
the foundations” in such a way that the first principle functions as a sort of meta-
principle with regard to analytic propositions, whose main property is the inclusion
of the predicate in the subject (see Autrecourt 1971: 102; Grellard 2005: 69). Since
these propositions may be used in their turn as premises in a syllogism, every proof
contains the first principle virtually, that is, can be reduced to it by a process of
212 SKEPTICISM
regress. Nevertheless, to warrant the transfer of evidentness from the principle to

the conclusion, Nicholas introduces two principles for our inferences:
P1: An inference is evident if and only if the signification of the consequent is
included in, or identical to, the signification of the antecedent.
P2: An inference is evident if and only if it is impossible that the antecedent and
the contradictory of the consequent are true at the same time.7
On these two principles depend two rules of inference:
R1: The inference “if a exists, then b exists” is not evident if and only if (1) “a”
and “b” are two absolute terms; and (2) a is different from b.
R2: The inference “if a exists, then b exists” is evident if and only if (1) “b”
is a term relative to “a”; or (2) if “a” and “b” are two absolute terms, the
signification of “b” is included in the signification of “a.” (see Autrecourt
1994: 64; Grellard 2005: 83–89)
According to R2, we may conclude: “If a house exists, a wall exists,” or “If every
animal is running, every man is running,” but according to R1 we have to exclude
inferences such as: “If an accident exists, a substance exists,” or “If A is produced,
something is producing A” (Autrecourt 1994: 181).8 Therefore, it clearly appears
that, by warranting our knowledge against skepticism, Nicholas excludes from the
realm of science a large part of the Aristotelian natural philosophy and metaphysics.
Such a philosophy, according to Nicholas, is at best only probable, as we shall see.
The second kind of evidentness is perceptual evidence. In order to avoid the problems
of the intuition of non-existent objects, Nicholas claims that a veridical perception
demands a relationship of conformity between the mental act and the external thing.
Such a relation must be necessary (habitudo necessaria) and occurs when the thing
really exists in se and is perceived by the external senses. Such a perception is called
a “complete appearance” (apparentia plena) (Autrecourt 1971: 108; see Perler 2006).
Of course, the problem for Nicholas, if he really wants to answer the skeptic, is to
explain how we may know when we have such an appearance and be sure that the
external thing exists. Facing this objection, Nicholas answers on a twofold level. First,
he claims that there is a qualitative difference between real and fictitious perceptions,
that is, a difference of clarity such that the perception naturally compels the intellect to
assent. Second, we have to assume as probable that everything that appears in a clear
way is true. This probable assumption means that we may generally trust our senses
and accept that our direct perceptions are usually true. With this assumption, I am
allowed to use some clear perceptions to correct the other less clear perceptions. The
complete appearances function as judgments that allow us to correct a posteriori some
misleading perceptions. For example, if I see a broken stick in the water, I may touch it
to correct my vision and confirm my judgment, “The stick is not broken.”
Therefore, Nicholas thinks he may answer skepticism by developing a
foundationalist theory of knowledge, where two kinds of principles, propositional
and perceptual, may warrant the evidentness of some basic beliefs, and from which
we may infer other beliefs. Nevertheless, in some respects, we may think that
Nicholas has failed to reach his goal and is led to an unwilling form of skepticism.
2c Nicholas’s Failure to Avoid Skepticism

Most scholars usually consider Nicholas as the best example of medieval skepticism,
even if he himself claims not to be a skeptic.9 We may identify three reasons that
explain why Autrecourt is often considered a skeptic: his probabilism, his attitude
toward the truth of principles, and the limits of his defense of sensible perception.
Let us consider briefly these three points.
First, as we have seen, Nicholas sharply criticizes Aristotelian philosophy by
claiming that it is at best only probable. Indeed, every philosophy that cannot satisfy
the condition of evidentness can be only probable, that is, rationally grounded. For
example, Nicholas thinks that his own atomistic physics is only probable, but more
probable than Aristotelian natural philosophy, since it can give a better explanation of
natural phenomena.10 A theory is more probable than another if one is more justified
in assenting to it, that is, if one has more and better arguments for it. Autrecourt’s
conception of what is probable is linked both with an evolving conception of
knowledge conceived of as a quest for justification and with a conception of the
history of philosophy as the refutation of previous systems.11 But the consequence is
the weakening of the scholastic model of scientia as evident knowledge. Finally, the
realm of science is strongly limited, and most of our knowledge is not science at all
but only more or less justified opinions. For Nicholas, that is enough, but as we shall
see in the case of Buridan, for most of the scholastics it is not acceptable. For example,
Nicholas’s theory of knowledge leads him to exclude inductive knowledge from the
realm of science. First, induction cannot be evident since causal connections are not.
Indeed, a causal connection cannot satisfy R2, and our experience of causality gives
us nothing more than a mere contiguity. Second, since we cannot enumerate all the
possible causes, we can never exclude absolutely the possibility of counterexamples.
Third, the notion of a cause does not guarantee the constant production of an effect.
We are inclined to accept that similar causes produce similar effects because of a
conjectural habit (habitus conjecturativus), that is, a natural disposition to generalize,
but such a habit gives us no certitude (see Autrecourt 1994: 72; 1971: 119).
The skeptical scope of Nicholas’s probabilism is reinforced by his attitude toward
the principles. For Nicholas, as we have seen, all certitude comes from the first
principle by reduction to it. At the same time, Nicholas claims that, despite its
primacy, the first principle cannot warrant the truth of our other principles or of our
conclusions. Such truths are only useless hypotheses. He first shows that we have an
indirect proof of the truth of evident knowledge, by the following argument:
Accordingly, I shall lay down as a first conclusion that, if the intellect can say
of something “this is true,” there cannot coexist the opposite of what is known
clearly and evidently. Hence, what is clear and evident to the intellect is true—a
statement universally valid and convertible. This conclusion is proven for, when
something is known to be true, if it were possible for its opposite to be true, it
would follow that the intellect could be sure of nothing. But the opposite of this
was maintained in our hypothesis. The reasoning is proven because we have no
certitude concerning first principles of anything else knowable except because we
know them clearly and evidently. (Autrecourt 1971: 116)12
214 SKEPTICISM
Second, he shows that there is no direct proof of the truth of a proposition. If we

want to demonstrate the truth of p by the means of premises q and r, we have to
show that q and r are either true or evident. To show the first point, we need other
premises whose truth has to be demonstrated, etc., and we have an infinite regress.
If we say that p is true because q and r are evident, we have a petitio principii.
Hence, the combination of skeptical modes allows Nicholas to say that truth is only
a hypothesis (Autrecourt 1971: 117). Therefore, there is a primacy of evidentness
over truth. The first point is that we cannot prove the truth of our principles, but
we should assume it as following from their evidentness. But, in the same way, we
have no proof for this evidentness except an indirect one: we need to posit the
first principle if we want to escape the infinite regress argument in our syllogisms.
Undoubtedly, this is a strong concession to the skeptics.
Third, and finally, the strongest concession to skepticism appears in Nicholas’s
solution to the problem of perceptual illusions. Even if Nicholas’s probabilism is very
close to a form of mild skepticism, Nicholas goes even further by acknowledging
some limits to his answer to skepticism. As we explained, against the traditional
skeptical arguments against perception, like the dream argument, Nicholas appeals
to the clarity or self-evidence of complete and veridical perceptions. But we may ask:
how are we able to discriminate complete and partial appearances and to explain
the incompleteness of some appearances? Indeed, in situations involving dreams or
illusions, we do think that we have a complete appearance, which often compels
the will to assent to it as strongly as a complete appearance. The skeptical dream
argument, for example, presupposes a qualitative identity between wakefulness and
sleep. On this point, Nicholas is much more embarrassed. He admits that if our
dreams, and more generally our illusions, were as clear as complete appearances,
we could never make any distinction between them and reality, and we could
not have any certitude at all. We would be reduced to complete skepticism or
absolute relativism:
It is evident that in sleep the appearance is not clear. For, no matter how vividly
it appears to someone in sleep that he has seen a camp, the light of heaven,
etc., nevertheless everyone experiences when awake that the appearance he gets
through sight is clearer and different in kind, and so he is more attracted by this.
For, if they were equally clear, he would have either to say nothing is certain
for him or to admit that in both appearances what appears to be true is true.
(Autrecourt 1971: 107)
Of course, this is an important concession from Nicholas since the skeptical dream
argument makes two assumptions: the qualitative identity of true and false appearances
and a slippery-slope reasoning of the form, “If I am once deceived, I could always be
deceived.” Nicholas tries to escape the problem by claiming that we have to assume
as probable that every clear appearance is true, that is, that our sensations are usually
correct, but, of course, a skeptic would never accept such an assumption.13
The attempt to give a foundationalist answer to skepticism seems to lead
Nicholas of Autrecourt to a dead-end. His position involves too many concessions
to a more or less strong form of skepticism, in such a way that he could not avoid
the accusation to be himself a kind of skeptic. Such an accusation seems to appear in

Buridan’s discussion of Nicholas’s epistemology.
3 JOHN BURIDAN’S ANTI-SKEPTICAL STRATEGY

John Buridan confronts skepticism in several texts. In most of these texts, Nicholas of
Autrecourt is one of his targets, but he is far from being the only one.14 Nevertheless,
due to the radical gap between Autrecourt’s foundationalism and Buridan’s
externalism, the latter clearly uses the process of reduction to skepticism in order
to underline how foundationalism leads to skepticism. For Buridan, skepticism
is mainly a problem for perceptual cognition, and he assumes that an externalist
epistemology can sufficiently answer the skeptical challenge.
3a Who Is Buridan’s Skeptic?

In two questions of his commentaries on Metaphysics and Posterior Analytics, Buridan
directly faces the skeptical challenge by asking: is it possible to know something
with evidence?15 In both questions, the arguments quod non (i.e., the arguments
against the thesis Buridan wants to defend) draw the portrait of the skeptic Buridan
wants to refute. This skeptic mainly uses arguments against perception, which
culminate with the hypothesis of divine deception. In the Questions on Metaphysics,
Buridan introduces ten arguments against perception, most of which are linked to
the relativity of perception. The last one relies on the possibility of illusion, which
is grounded in God’s absolute power. He then introduces seven arguments against
intellectual knowledge. The first four rely on the subordination of intellectual
knowledge to perceptual knowledge. The fallibility of our perception contaminates
intellectual knowledge. Next follows the argument of the Meno,16 the argument of
the infinite regress, Nicholas’s rule of inference (R1) used to criticize causality, and
a “Heraclitean argument” according to which there is no science of contingent and
mutable things. The general structure of the arguments quod non is the same in
the Questions on the Posterior Analytics. Buridan only adds two arguments against
causality and an argument against induction that could be inspired by Autrecourt’s
criticisms. Therefore, for Buridan, the skeptical challenge mainly pertains to
perceptual knowledge and indirectly to intellectual knowledge. It is clear that,
on this point, Autrecourt is not his target. It is highly implausible, contrary to De
Rijk’s (1997) assumption, followed by Zupko (2003) and Klima (2009; 2010), that
Autrecourt would be one of the theologizantes Buridan criticizes in his Summulae de
demonstrationibus (Buridan 2001: 112–113). Indeed, Autrecourt never deals with
the question of the deceiving God, except in his controversy with Bernard, when he
uses Bernard’s own terms and theory to underline their skeptical consequences. For
his part, Bernard seems very close to Franciscan theologians like Walter Chatton
or John of Rodington. Then it is much more plausible that Buridan targets some
aspects of the post-Scotist theology that gave an impulsion to the question of the
deceiving god.17 On the other hand, it cannot be denied that Autrecourt’s criticisms
of our knowledge of substance, along with his criticisms of induction and causality,
216 SKEPTICISM
are mixed with skeptical arguments, which would make it possible to reduce his
epistemology to a form of skepticism. But the problem here is much more the
question of the nature of our principles and of epistemic justification, and only
secondarily, of skepticism.18
To sum up, Buridan’s skeptic is someone who endangers the possibility of
knowing through perception and making inferences from perceptions. Buridan only
partly targets Autrecourt, and the opposition between the two colleagues at the Arts
Faculty rather concerns their respective conceptions of epistemic justification.
3b Buridan’s Externalist Answer to Skepticism

Buridan’s answer to skepticism first relies on a twofold distinction of evidentness,
and second on the idea of a natural inclination to the truth. For Buridan, to avoid
mere fallible opinion, the firmness of our assent must depend not only on subjective
factors (like appearances and dialectical arguments) but also on an objective factor
he calls evidentness (evidentia). This evidentness can be either absolutely infallible
or relatively infallible.19 The first occurs when it can be falsified by no means, either
natural or supernatural. Hence, such evidentness avoids the problem of divine
deception. It is exemplified by the evidentness of mathematics and logic. In this
case, it is almost impossible to refuse to assent to the proposition. For example,
the principle of non-contradiction can be denied verbally (ore), but not sincerely
(corde). The other kind of evidentness is only relatively evident (secundum quid
or ex suppositione), since it depends on the supposition that we are in a strict
natural context (the common course of nature), in which God by definition does
not intervene to modify natural causality. This kind of evidentness is enough for
natural science; that is, we do not need to be certain that just now our knowledge is
preserved from divine deception, but only that, usually, God does not intervene, in
such a way that natural science is true:
In another way, we consider evidentness as relative or hypothetical (accipitur
euidentia secundum quid siue ex suppositione). For example, we claimed before
that we observe it in every being according to the common course of nature,
and in this way it is evident for us that every fire is hot and that the heavens are
moved, even if the contrary is possible by the power of God. Such evidentness
is enough for the principles and conclusions of natural science. (Buridan 1518:
Bk II, q. 1, f. 8vb–9ra)
For Buridan, even if it is always theoretically possible that God may intervene in the
common course of nature and temporarily suspend the causal laws He established,
such a theoretical possibility is nevertheless practically insignificant. In other words,
we may practice natural science as if God would never intervene, but keeping in
mind that He could falsify our knowledge. The scientist can never be absolutely
certain that his natural knowledge is entirely true, since some parameters could
remain unknown, as exemplified by the possibility of divine intervention.
This first level of the answer to skepticism is completed by a second one: the
natural inclination of the intellect toward the truth. For Buridan it is as natural for
the intellect to know the truth as for fire to burn or for wheat to fruit. There is
a natural tendency, due to the biological constitution of these beings, to perform
such an action. Generally speaking, for Buridan the natural power (potentia) of
being, that is an ability to act, will produce its effect if there is no impediment. For
example, a magnet attracts iron, and the body moves toward its proper place. The
ideal action of the intellect is exemplified by the grasping of the first principles.
Such a grasping cannot be erroneous, but there are of course impediments. Since
the intellect is a power determined by many external factors, it is not a strictly
natural power, but a habitual one, which depends on the repetition of the act.
Therefore, it is possible for an intellectual power to be habituated to the contrary
of its natural tendency. By custom, bad education or habits, or by the repetition
of false perceptions, I can be trained to give my assent to the false. Therefore, the
skeptic relies on this bad intellectual education to dismiss all our knowledge. But
Buridan thinks we may correct a posteriori our defective knowledge by identifying
the punctual causes of error. These punctual errors may not be a sufficient reason
for rejecting our natural ability to grasp the truth. In other words, for Buridan,
true knowledge is the rule and error the exception. Indeed, according to him,
our knowledge depends on causal relations that could theoretically fail, but that
succeed in most cases (see Grellard 2014). Usually, I do know the truth, and my
knowledge does not fail.
This natural inclination of the intellect toward the truth is the key to explain
inductive inference. The problem consists in explaining how we can jump from
a partial enumeration of singular cases to a general law subsuming each of these
singulars. A complete enumeration is, of course, impossible. But, according to
Buridan, if we have discovered, after careful examination, no counterexamples,
we are allowed to add the clause et sic de singulis (“and so with each token”).
This clause cannot rely on a principle of the uniformity of nature, otherwise we
would have a petitio principii (the uniformity of nature warrants the inductive
process but is itself known by induction). According to Buridan, we have to
assume that the absence of any counterexamples is sufficient to claim that the
process is reliable enough and depends on our natural inclination to the truth,
even if it is theoretically possible that our inductive principle may be falsified. Of
course, induction provides only a hypothetical or relative evidentness, but it is
nevertheless evidentness.20
The way Buridan answers the skeptical challenge allows him to maintain a
definition of science where evidentness (even of a weaker kind) and truth are still
necessary conditions of knowledge in a strong sense (whereas Autrecourt dismissed
truth as a mere hypothesis and evidentness as the ideal higher level of knowledge).
Indeed, Buridan’s strategy against skepticism is to circumscribe it, to reduce it to a
very exceptional case that very rarely occurs, in order to maintain the high standards
of science. But could Buridan satisfy a skeptic with such an answer? Probably not.
This answer relies on the assumption of some principles the skeptic cannot accept
(e.g., error occurs rarely) and on the possibility of correcting our error a posteriori.
But none of these two assumptions may satisfy the skeptic who asks for an a priori
warrant for the truth of his belief.
218 SKEPTICISM
4 CONCLUSION
What is the general teaching of this medieval debate on skepticism? Two
conclusions may be drawn. First, from a philosophical point of view, the debate
between Autrecourt and Buridan on the best way to avoid skepticism shows that
it is impossible to give a satisfactory answer to the skeptical challenge. Either you
accept the terms of the debate the skeptic calls for, or you try to bypass them by
refusing the skeptical formulation of the problem of knowledge. The first solution is
Nicholas’s choice and leads to a failure since in the end he has to acknowledge some
cases where the skeptic cannot be refuted. The second solution is Buridan’s, but it
cannot satisfy the skeptic since it does not fulfill his demand to meet a high standard
of knowledge that could a priori warrant each instance of our knowledge. In this
case, there is no real dialogue between the dogmatist and the skeptic.
Second, from a historical point of view, the case of Nicholas of Autrecourt
clearly shows that all the necessary conditions for the rise of modern skepticism
are present in the fourteenth century. A mix of institutional and intellectual
factors inhibited the development of such skepticism before the following
centuries. From an institutional point of view, the condemnation of Autrecourt
in 1346 (for reasons alien to skepticism) prevented any positive influence on
late medieval philosophy. From an intellectual point of view, we may assume
that Scotus’s radical criticisms of Henry of Ghent probably made it much more
difficult to explicitly adopt a skeptical position. Nevertheless, the claim that the
translation of Sextus’s writings was the only cause of the development of modern
skepticism should be revised.
NOTES
1. See Perler (2006: 33–115).
2. On this point, see Lagerlund in this book, Gregory (1984), and Perler (2010).
3. Nicholas of Autrecourt was born around 1300 and died in 1369. Due to his
condemnation in 1347, he left the philosophical stage and had little influence on late
medieval Parisian philosophy. John Buridan was also born around 1300 and died
between 1358 and 1361. Throughout his life, he was a teacher at the Arts Faculty,
and he had a wide influence not only in Paris but also in Italy and Central Europe. On
Autrecourt’s life, see Kaluza (1995); on Buridan’s, see Faral (1950). On Autrecourt’s
skepticism, see Thijssen (2000) and Grellard (2010). On Buridan and skepticism, see
Zupko (1993) and Klima (2010).
4. Bernard’s position is very close to Walter Chatton’s. On this point, see Kaluza
(1991).
5. Augustine (1995, 74): “‘How do you know that the world exists,’ replies the
Academician, ‘if the senses are deceptive?’ Your arguments were never able to disown
the power of our senses to the extent of clearly establishing that nothing seems to be
so to us. Nor have you ever ventured to try to do so.”
6. In some texts Nicholas seems to claim that the first principle includes both the principle
of non-contradiction and the principle of identity. On this topic, see Grellard (2005:
69–70), also Krause (2009).
7. On P1, see Autrecourt (1994: 64); on P2, see Autrecourt (1994: 60). P2 is very
common among medieval logicians. P1 is less common and clearly aims to exclude
material implications. For this reason, it can be said that it belongs to the prehistory
of relevance logic.
8. On Autrecourt’s criticisms of the possibility of knowing a substance, see
Robert (2006).
9. As I point out elsewhere (Grellard 2010), Autrecourt’s epistemology is very close to
that proposed in the twelfth century by John of Salisbury, who explicitly claimed
to be a skeptic (see Bloch’s chapter in this book and Grellard 2013a). We may
assume that some historical changes forbade such a claim in the fourteenth century.
A plausible explanation is the following: Scotus’s accusation that Henry of Ghent
was a skeptic made it impossible for anyone to claim to be a skeptic because he took
Henry’s skepticism to be a form of irrationality.
10. On Autrecourt’s atomism, see Grellard (2002; 2009).
11. See Autrecourt (1971: 41): “For although in my opinion they appear far more
probable than what Aristotle said, yet, just as for a long time Aristotle’s statements
seemed to be probable, though now perhaps their probability will be lessened, so
someone will come along and undermine the probability of these statements of
mine.” On Autrecourt’s probabilism, see Grellard (2005: 93–118).
12. The structure of the argument is not very clear, and seems to suffer from some
logical flaws, but we can propose the following reconstruction (I would like to thank
Baron Reed for improving my understanding of the argument):
(1) Kp (assume)
(2) not-p (assume for reductio)
(3) Kp & not-p (conjunction introduction: 1, 2)
(4) if Kp & not-p, then not-Kp (the knowledge of p excludes the truth of not-p)
(5) not-(Kp & not-p) (modus tollens: 1, 4)
(6) not-Kp or p (equivalent to 5)
(7) p (disjunction elimination: 1, 6)
(8) p (reductio ad absurdum: 2, 7)
(9) if Kp, then p (conditional proof: 1, 8)
13. This failure to plainly answer to skepticism may partly explain why Autrecourt
adopts a phenomenalist theory of perception in the second part of his treatise. See
Grellard (2015a).
14. The claim that Buridan is mainly attacking Autrecourt’s so-called skepticism is
found in Zupko (1993; 2003) and Klima (2009; 2010). This claim is not entirely
false, as we shall see, but it should be formulated in a more cautious way.
15. Buridan (1518: Bk II, q. 1, f. 8rb): utrum de rebus sit nobis possibilis comprehensio
veritatis? (Whether it is possible for us to grasp the truth about things?) and Buridan
unpublished, Bk I, q. 2: utrum possibile sit nos aliquid scire? (Whether it is possible
we know something scientifically?).
16. On the status of Meno’s argument in the Middle Ages, see Grellard (2011a).
17. See Lagerlund’s chapter in this book.
18. Buridan also seems to target Autrecourt on induction in the Questions on Ethics,
Bk. VI, q. 11, but without any mention of the skeptical problem. See Grellard
(2005: 270–271).
220 SKEPTICISM
19. I leave aside a third level, the moral evidence that is much closer to mere opinion.
On this point, see Grellard (2013b; 2015 b).
20. On this topic, see Buridan (1518: II, q. 2, f. 9vb; unpublished: I, q. 19 & 20)
Grellard (2005: 264–273), and Economos (2009).
REFERENCES
Augustine of Hippo. 1995. Against the Academicians and The Teacher, translated by
P. King. Indianapolis and Cambridge: Hackett.
Autrecourt, Nicholas of. 1971. The Universal Treatise, translated by L. Kennedy. Milwaukee:
Marquette University Press.
Autrecourt, Nicholas of. 1994. Nicholas of Autrecourt: His Correspondence with Master
Giles and Bernard of Arezzo, critical edition and English translation by L. M. de Rijk.
Leiden, New York and Köln: Brill.
Buridan, John. 1518. Quaestiones in Metaphysicen Aristotelis. Paris: Josse Bade.
Buridan, John. 2001. Summulae de demonstrationibus. Groningen and Haren: Ingenium.
Buridan, John. Unpublished. Quaestiones super Posteriora analyticorum. Typescript by
Hubert Hubien.
De Rijk, Lambert Marie. 1997. “Foi chrétienne et savoir humain. La lutte de Buridan
contre les theologizantes.” In A. Elamrani-Jama, A. Galonnier and A. De Libera (eds.),
Langage et philosophie. Hommage à Jean Jolivet, 393–409. Paris: Vrin.
Economos, Ariane. 2009. Intellectus and Induction: Three Aristotelian Commentators on
the Cognition of First Principles, Including an Original Translation of John Buridan’s
Quaestiones in duos Aristotelis libros Posteriorum Analyticorum. PhD Dissertation.
New York: Fordham.
Faral, Edmond. 1950. Jean Buridan, maître ès arts de l’université de Paris. Paris: Imprimerie
nationale.
Gregory, Tullio. 1984. “La tromperie divine.” In Z. Kaluza and P. Vignaux (eds.), Preuve
et raisons logiques à l’université de Paris. Logique, ontologie et théologie au xive siècle,
187–195. Paris: Vrin.
Grellard, Christophe. 2005. Croire et savoir: Les principes de la connaissance selon Nicolas
d’Autrécourt. Paris. Vrin.
Medieval Epistemology.” In H. Lagerlund (ed.), Rethinking the History of Medieval
Skepticism. The Missing Medieval Background, 119–142. Leiden and Boston: Brill.
Grellard, Christophe. 2011a. “Peut-on connaître quelque chose de nouveau? Variations
médiévales sur l’argument du Ménon,” Revue philosophique de la France et de l’étranger
201: 37–66.
Grellard, Christophe. 2011b. “Academicus.” In I. Atucha, D. Calma, C. König-Pralong, and
I. Zavaterro (eds.), Mots médiévaux offerts à Ruedi Imbach, 5–16. Turnhout: Brepols.
Grellard, Christophe. 2013a. Jean de Salisbury et la renaissance médiévale du scepticisme.
Grellard, Christophe. 2013b. “L’âne et les petites vieilles: psychologie de l’action et logique
de l’assentiment chez Jean Buridan.” In L. Jaffro (ed.), Croit-on comme on veut? Histoire
d’une controverse, 79–101. Paris: Vrin.
Grellard, Christophe. 2014. “How is it Possible to Believe Falsely? John Buridan, the vetula
and the Psychology of Error.” In K. Gosh, D. Denery, R. Copeland, and N. Zeeman
(eds.), Uncertain Knowledge: Interdisciplinary Conversations about Doubt and
Scepticism in the Middle Ages, 91–113. Turnhout: Brepols.
Grellard, Christophe. 2015a. “The Nature of Intentional Objects in Nicholas of Autrecourt’s
Theory of Knowledge.” In G. Klima (ed.), Intentionality, Cognition and Mental
Representation in Medieval Philosophy, 235–250. New York: Fordham University Press.
Grellard, Christophe. 2015b. “Probabilisme et approximation du vrai au xive siècle (Jean
Buridan, Nicole Oresme, Jean Gerson).” In J-Ph. Genet (ed.), La vérité. Vérité et
crédibilité : construire la vérité dans le système de communication de l’Occident (xiiie–
xviie siècle), 65–79. Paris: Publications de la Sorbonne.
Kaluza, Zénon. 1991. “Serbi un sasso il nome: une inscription de San Gimignano et la
rencontre de Bernard d’Arezzo et Nicolas d’Autrécourt.” In B. Mojisch and O. Pluta
(eds.), Historia philosophiae medii aevi. Festschrift für Kurt Flasch zu seinem 60
Geburstag, 437–466. Amsterdam and Philadelphia: Grüner.
Kaluza, Zénon. 1995. Nicolas d’Autrécourt, ami de la vérité. Paris: Imprimerie nationale.
Klima, Gyula. 2009. John Buridan. Oxford: Oxford University Press.
Klima, Gyula. 2010. “The Anti-Skepticism of John Buridan and Thomas Aquinas: Putting
Skeptics in Their Place Versus Stopping Them in Their Tracks.” In H. Lagerlund (ed.),
Rethinking the History of Medieval Skepticism. The Missing Medieval Background, 145–
170. Leiden and Boston: Brill.
Krause, Andrej. 2009. “Nikolaus von Autrecourt über das erste Prinzip und die Gewißheit
von Sätzen,” Vivarium 47: 407–420.
Lagerlund, Henrik (ed.). 2010. Rethinking the History of Medieval Skepticism. The Missing
Medieval Background. Leiden and Boston: Brill.
Perler, Dominik. 2006. “Relations nécessaires ou contingentes? Nicolas d’Autrécourt et la
controverse sur la nature des relations cognitives.” In C. Grellard and S. Caroti (eds.),
Nicolas d’Autrécourt et la Faculté des Arts de Paris (1317–1340), 85–111. Cesena:
Stilgrad editrice.
Perler, Dominik. 2010. “Does God Deceive Us? Skeptical Hypotheses in Late Medieval
Epistemology.” In H. Lagerlund (ed.), Rethinking the History of Medieval Skepticism.
The Missing Medieval Background, 171–192. Leiden and Boston: Brill.
Robert, Aurélien. 2006. “Jamais Aristote n’a eu de connaissance d’une substance: Nicolas
d’Autrécourt en contexte.” In C. Grellard and S. Caroti (eds.), Nicolas d’Autrécourt et la
Faculté des Arts de Paris (1317–1340), 111–51. Cesena: Stilgrad editrice.
Thijssen, Hans. 2000. “The Quest for Certain Knowledge in the Fourteenth Century:
Nicholas of Autrecourt against the Academics.” In J. Sihvola (ed.), Ancient Scepticism
and the Sceptical Tradition (=Acta Philosophica Fennica 66), 199–223. Helsinki:
Philosophical Society of Finland.
Zupko, Jack. 1993. “Buridan on Skepticism.” Journal of the History of Philosophy 31:
191–221.
Zupko, Jack. 2003. John Buridan: Portrait of a Fourteenth Century Arts Master. Notre
Dame: Notre Dame University Press.
CHAPTER SIXTEEN
Divine Deception
HENRIK LAGERLUND
1 INTRODUCTION
Perhaps one of the most fascinating arguments that one encounters as a first-
year philosophy student is the argument for skepticism about the external world.
Entertaining the tantalizing thought that the world as I know it might not exist or
that I am stuck in the Matrix, but unable to take the red pill to find out what the
world is really like, meets the fancy of many undergraduates. The argument itself is
often presented using Descartes’s First Meditation, and the position the meditator
finds herself in at the end of the meditation is captured well by Barry Stroud in The
Significance of Philosophical Scepticism. He writes:
We are confined at best to what Descartes calls “ideas” of things around us,
representations of things or states of affairs which, for all we know, might or
might not have something corresponding to them in reality. We are in a sense
imprisoned within those representations, at least with respect to our knowledge.
Any attempt to go beyond them to try to tell whether the world really is as they
represent it to be can yield only more representations. . . . This can seem to leave
us in the position of finding a barrier between ourselves and the world around
us. There would then be a veil of sensory experience or sensory objects which we
could not penetrate but which would be no reliable guide to the world beyond
the veil. If we were in such a position, I think it is quite clear that we could not
know what is going on beyond the veil. . . . I have described Descartes’s sceptical
conclusion as implying that we are permanently sealed off from a world we can
never reach. (Stroud 1984: 32–33)
Now, Descartes himself is not a skeptic, of course, since we are only in the first
meditation and by the end of the third, when he has proved the existence of a
benevolent God, he is in a position to claim that our ideas reach beyond the veil and
hook up with things in the world.
The arguments Descartes uses to cast doubt on our ideas of the external world
are famous, but certainly the most spectacular is that of an evil demon deceiving
us. Descartes uses a demon because he does not think God could deceive us;
in fact, he thinks it goes against God’s nature to do so, and once the existence
DIVINE DECEPTION 223
of an all-powerful God is established, there is no reason for the original doubt,

and the methodological skepticism is overcome. Even though Descartes finds it
problematic to think of God as a deceiver, it was in the early fourteenth century
commonly thought that he could deceive human beings, and it generated similar
worries about the existence or nature of the external world as it did for Descartes
300 years later.
The late medieval debate about divine deception ranges from the early fourteenth
to the seventeenth century. Descartes himself was most likely aware of it, particularly
since he was reminded of it by Mersenne in his objections to the Meditations (AT
VII 125; CSM I 89–90). It was a debate with very high stakes, which had an impact
on all parts of philosophy and theology. It represented a new wave of skeptical
thought that went well beyond anything in the earlier history of philosophy.
In this chapter, I will look closer at the origin of the idea of God as a deceiver and
the debates about skepticism that it led to. I will also outline some of the reasons
why it never led philosophy to global skepticism.
2 A NEW MODAL THEORY

In his very instructive article “Does God Deceive us?” Dominik Perler starts by
looking at thirteenth-century views about God as a deceiver. Thomas Aquinas’s
view was, for example, very similar to Descartes’s. Aquinas argued that you have
to have a bad intention in order to deceive someone and that this is contrary to
God’s nature or rather to his benevolence. Demons can deceive, of course, but
their power is limited, and although they can pull tricks on our senses, they are
not powerful enough to fiddle with our intellectual capacities. Aquinas developed
a very particular view of perception according to which we can, despite possibly
being deceived by demons, grasp the essences and gain knowledge (Perler 2010:
173–178). It is foremost through the metaphysical underpinnings of his theory
of cognition that any skepticism due to deceiving demons is prevented. This
metaphysics, as well as the metaphysics of modality, went through fundamental
changes in the late thirteenth and early fourteenth centuries, which changed
philosophy fundamentally, and as a result new epistemological problems also
emerged (see Lagerlund 2010b).
The starting point for a discussion of divine deception has to be a discussion of
God’s omnipotence. The changes that took place in the fourteenth century began
with the development of a new modal theory underlying a new conception of divine
omnipotence. The background for these developments was, as so often, set by
Aquinas. In Summa Theologiae I, q. 25, art. 3, he writes:
God is said to be omnipotent, because he can do everything that is absolutely
possible. . . . It must be said that something absolutely possible or impossible
derives from the nature of the terms [in a proposition], in the way that something
is possible if the predicate [term] is not repugnant to the subject [term], like in
“Socrates is sitting.” Something is, however, impossible, because the predicate
[term] is repugnant to the subject [term], like in “A human being is a donkey.”
224 SKEPTICISM
This view of modality takes as fundamental what is or what is not repugnant to the
essence to which the subject term refers. It assumes the essence to be and then asks
what is or is not repugnant to it. It is, for example, not repugnant to the existing
essence of Socrates that he sits or stands, hence it is possible for Socrates to sit and
the corresponding proposition: “Socrates is sitting” is possible. It is repugnant to
the essence of Socrates, and indeed every other human being, that he be a donkey.
Such a genus-species arrangement cannot occur. Likewise, it is not possible, that
is, it is impossible, that Socrates is not rational, since it is repugnant to the essence
of Socrates that he not be rational. He would simply not be what he is, namely, a
human being, if that were the case. Hence, on Aquinas’s view of omnipotence, God
can do anything that is absolutely possible, but he does not have the power to bring
about something that is absolutely impossible. This view of possibility is obviously
limited in the sense that it starts from what is actual.
John Duns Scotus, who is mainly responsible for the radically new view of modality
that emerged in the early fourteenth century, retains some of the terminology used
by Aquinas. He even accepts that God’s omnipotence should be explained in terms
of what is absolutely possible. He, however, changes the notion of what is absolutely
possible. For him, it must instead be equated with what he calls “logical possibility”
(possibile logicum). Scotus is the first to use this terminology (see Knuuttila 1993,
2012; Gelber 2004).
Scotus explains logical possibility by a thought experiment. If we assume that
neither God nor the world exists, but that the proposition “The world will be”
exists in some intellect, then this proposition is possibly true, as is the proposition
“The world is possible”; that is, they are logically possible or true (Scotus, Lectura I,
d. 7, n. 32). This is also what he means when he modifies the so-called positio rule
in obligations logic and explains what he means by contingency. He writes:
I do not call something contingent because it is not always or necessarily the case,
but because the opposite of it could be actual at the very moment when it occurs.
(Ordinatio I, d. 2, p.1, q. 1–2, n. 86)
What both these examples illustrate is that for Scotus possibility is not tied to, or
have anything to do with, what actually exists. The subject term, to follow Aquinas’s
example, is not assumed to pick out an actual being anymore, but it is instead assumed
to range over all possible beings. Hence, what is possible might not ever be actual.
The best way to understand this is in terms of possible worlds, even though it is a
bit anachronistic to borrow terminology from the seventeenth century. On Scotus’s
view, there are an infinite number of possible worlds or possible combinations of
things. The actual world is merely one among these.
On this view, a proposition like “Every human being is an animal” is not
necessarily true, since there are possible worlds where there are no humans and
hence the proposition is false there. Instead, it is contingent. Scotus thinks that a
proposition like “God exists” is necessarily true, since God exists in all possible
worlds. He is thus a necessary being, unlike humans, who are contingent beings. A
proposition like “A Chimera exists” is impossible, because there is no possible world
where Chimeras exist.
As mentioned, Scotus continues to associate God’s omnipotence with absolute

possibility, but on his view what is absolutely possible is a much larger class than
what Aquinas had in mind. God’s omnipotence is associated with what is logically
possible, and as a result God’s power has vastly increased since what he cannot do
is now only what is logically impossible; for example, he cannot make Chimeras
exist and he cannot make the principle of non-contradiction false, but otherwise his
power has very few limitations. To fully see how medieval philosophers expanded
this view, we must have a look at the famous distinction between God’s absolute
power and his ordained power.
3 ABSOLUTE VERSUS ORDAINED POWER

The way medieval philosophers, at least after William Ockham in the early fourteenth
century, explained God’s power was through a distinction between his absolute and
his ordained power. A representative example of how this distinction was used in
the late fourteenth century can be found in Peter of Ailly’s Sentences commentary
(for more about him see Oakley 1964). He explains:
What God can do is understood in two ways. In one way, in accordance with his
absolute power, and in another in accordance with his ordained power. It is not
the case that there are two powers in God, one absolute and one ordained, but
that God can do with his absolute power (absolute potest) what he simply and
absolutely can do. And in this way we have to understand God’s absolute power.
(Commentary on the Sentences I, q. 13, art. 1)
The use of “absolute potest” above is, of course, a reference to the idea of absolute
possibility as used by Aquinas and modified by Scotus to mean logical possibility. The
notion of ordained power is explained in a twofold way by Ailly in the continuation
of the passage just quoted:
What God can do in accordance with his ordained power can be understood
in two ways. In one way and strictly speaking, it is what God can do when he
holds to what he ordained, which he wills eternally to be such and such, and
then he can only do what he ordained to be the case. In another way, this can be
understood more widely so that it means what God can do when he holds to the
true law and the divine scripture. And as such it is possible to say that God can do
with his ordained power what is absolutely possible but only as long as he does
not take away some already ordained true law or sacred scripture.
Once God had created the world in a certain way or, to follow the spirit of Scotus’s
view, once God had chosen to make one possible world actual, he has committed
himself to it and does not change his mind. This act of God’s will binds him, and
whatever is in line with this initial choice or does not contradict it is within God’s
ordained power, but he still can act contrary to this through his absolute power. This
is a straightforward way of explaining miracles, for example, which are contrary to
the normal causal interactions in the created world. When fire does not have any
effect on the bush or the bodies of the children, God is intervening in the ordained
226 SKEPTICISM
order through his absolute power. From the way Ailly outlines the distinction, it
follows a distinction between absolute or logical possibility and ordained or natural
possibility. Natural possibility or necessity is what is possible or necessary given the
actually created world, while logical possibility or necessity is what is possible or
necessary over all possible worlds.
4 DIVINE DECEPTION
It is this view of God’s power put in the particular context of the new modal theory
that generates the skeptical hypothesis of divine deception. God is so powerful that
the mere possibility of deception threatens to cast philosophy into global skepticism
and undermine any knowledge of the external world. John Buridan, an important
figure in this debate a generation before Ailly,1 puts it succinctly in his commentary
on Aristotle’s Metaphysics:
As is commonly said, the senses can be deluded, and it is certain that the species of
sensible things can be preserved in the organs of sense in the absence of sensibles,
as is mentioned in De somno et vigilia. And then we judge about what does not
exist as if it existed, and so we err through the senses. And the difficulty is greatly
augmented by the fact that we believe on faith that God can form sensible species
in our senses without the sensible things themselves, and preserve them for a long
time. In that case, we judge as if there were sensible things present. Furthermore,
since God can do this and greater things, you do not know whether God intends
to do this, and so you have no certitude and evidentness [regarding the question]
whether there are men before you while you are awake or while you are asleep,
since in your sleep God could make a sensible species as clear as—indeed, a
hundred times clearer than—what sensible objects could produce. And so you
would then judge formally that there are sensible objects before you, just as you
do now. Therefore, since you know nothing about the will of God, you cannot
be certain about anything. (Buridan, Quaestiones in Aristotelis Metaphysicam, II,
q. 1, fol. 8rb-va)
Buridan is very clear in his response to the skeptical argument that our intellects
have the ability to correct for sensory illusion and that our senses are overall reliable.
Under any natural circumstance we can thus trust our senses and realize when they
are deceiving us, but in the case of divine deception this is obviously not possible, as
he explains. If God deceives us, we have no way of knowing it, and hence we cannot
be certain about anything.
The kind of argument presented by Buridan swept over fourteenth-century
philosophy and theology, and we find all major thinkers of the time commenting on
it. Even though the doctrine of divine deception had the potential to cast medieval
philosophy into complete skepticism, it did not do so. With only one exception,
there were no skeptics in the fourteenth century, but it did force thinkers to change
their views of the conditions under which we can achieve knowledge, and it also
ultimately changed their view of knowledge.
Ockham, Robert Holkot, and Adam Wodeham were all very influential
philosophers and theologians before Buridan and Ailly in the early to mid-
fourteenth century who all agreed that God could be deceiving us. Holkot writes in
his Sentences commentary that
God can deceive a rational creature, both in an immediate way, namely by himself,
and in a mediate way, through good human beings and angels. (Commentary on
the Sentences III, q. 1, BBB; see also Perler 2010: 179)
Wodeham adds to this the following observation about the consequences of the
doctrine of possible divine deception: “We cannot know of any external thing—
more precisely, of any thing other than our own mind—that it exists” (Karger
2004: 229). There is hence no doubt that they were all aware of the implications of
that doctrine. It has even been argued that Ockham’s famous doctrine that we can
have intuitive cognition of things that do not exist is an attempt to get away from
the implications of divine deception (see Karger 2004 and the response to this in
Panaccio and Piché 2010), though it seems very unlikely that this was what it was
intended for.
The only thinker of the fourteenth century who ends up defending a kind of
skepticism was Nicholas of Autrecourt.2 He develops his position in an attempt to
refute the skeptical views of the otherwise unknown Bernard of Arezzo, an early-
fourteenth-century follower of Scotus. The issue at hand was whether we can have
an intuitive cognition of non-existent things. Nicholas argues that Bernard’s position
implies Academic skepticism since one can have intuitive cognitions without any
object that they are about. If this were the case, it would therefore be impossible
to be certain about the existence of external objects. Bernard responded to this
objection by introducing the distinction between God’s absolute and ordained
power, and then claimed that there is no intuitive cognition of non-existent things
without divine intervention. Nicholas then replies that, if we must qualify all our
knowledge claims by saying “unless God is intervening,” then we can never be sure
about anything, since we cannot know whether God is deceiving us or not. On
this view, we cannot be certain about the existence of sensible things, he argues.
In fact, he argues further that Bernard’s position implies what we now would call
solipsism, and claims that Bernard cannot even be certain about the existence of
his own mind.
Based on these arguments against Bernard, Nicholas presents himself as an anti-
skeptic, but, as Grellard (2010) shows, his views actually imply a new form of
skepticism. Nicholas proposes a kind of foundationalism as an answer to Bernard.
The one thing that cannot be doubted in the face of divine deception, he argues, is
the principle of non-contradiction, and hence anything based on it can be said to
be certain. His position is somewhat similar to Leibniz’s in the sense that all certain
knowledge is reducible to the principle of non-contradiction. All other knowledge is
merely probable, according to Nicholas. This view was an anomaly in the fourteenth
century, but it motivated powerful reactions and led ultimately to a reformulation of
the concept of knowledge.
228 SKEPTICISM
5 ARGUMENTS AGAINST DIVINE DECEPTION

As already mentioned, the doctrine of divine deception did not lead the fourteenth
century into skepticism. One might ask, why? There are at least three reasons for
this. First of all, some thinkers argued that divine deception is simply not possible,
since it would contradict God’s benevolence. Second, although God can deceive, he
will not do so because of a covenant between God and humankind laid down at the
moment of creation. The third and the most philosophically interesting view consists
in revising the concept of knowledge and relativizing the concept of evidence.
Gregory of Rimini, another early-fourteenth-century philosopher, argued that
God cannot be a deceiver since this would go against his benevolence. Gabriel Biel
in the fifteenth century and Francisco Suárez in the sixteenth century defended a
similar view. This seems to be the view that Descartes falls back on in the Meditations
and the Objections and Replies (Rimini and Biel are mentioned by Mersenne).
Rimini’s argument is that God cannot deceive, since it would imply saying or making
someone believe something false. He writes: “God cannot say something false to
someone, willing thereby that the person, to whom he says it, assents to what is
said” (Lectura I, d. 42–44, q. 2; see Perler 2010: 182 for this quote). As Perler
(2012: 183) analyzes Rimini’s position, he holds that God cannot say something
false or make someone else say something false, if that means that what is said is
said in a meaningful way, without irony, and in an assertive mode. Saying something
false in this way amounts to lying, argues Rimini, and God cannot lie even given his
absolute power. Rimini is thus limiting God’s omnipotence by saying that, given his
benevolence, he cannot be a deceiver.
A second reason for why divine deception did not have the implications for
philosophy it could have had was the appeal to a contract or covenant between
God and humankind. This was a very common view, particularly among certain
theologians following Ockham, and all the main thinkers who were concerned with
divine deception adhered to this view. We find it in Ockham, Holkot, Rimini, Ailly,
and Biel. The notion of a covenant governs both the natural and the moral order
established by God in the creation of the world. It takes its starting point in the idea
that what is, is determined by God’s will, and whatever natural or moral laws there
are, are also determined by an act of God’s will. These laws are hence not absolute
but decided on or chosen by God. God had infinite possibilities from which to
choose in creating this world, but he chose to establish this particular moral law and
this particular causal order, which, of course, make them special and necessary in
the relative or natural sense discussed above.
As someone like Ailly understands the notion of a covenant, which we could
already see indicated in the above discussion of ordained power, there are two
covenants, corresponding to the two kinds of ordained power. The first covers the
world in general and is made with all humans. The second is made with the church
(see Courtenay 1971). According to the first covenant, God has promised to uphold
his creation and all the laws that govern it. According to his absolute power, God
could act in whatever way he wanted toward his creation: change it, deceive us,
etc. However, according to his ordained power, he will not and instead he acts in
accordance with the established laws. This view does not limit God’s omnipotence
in any way, since God is not limited by the laws he has put down, but given the
covenant he will abide by them. As humans, we can hence trust that God will not
make changes and that he is not deceiving us. The doctrine of a covenant plays an
important role in late medieval and early modern thought, but perhaps the most
important aspect of the theory is that it did not allow divine deception to take hold
of philosophy.
The third way to deal with divine deception is philosophically the most interesting.
Buridan is generally credited with coming up with this approach, but after him it was
repeated by almost everyone well into the sixteenth century (see Karger 2010). I will
here present the view as defended by Albert of Saxony, a slightly later contemporary
of Buridan, and Ailly. It takes as its starting point the assumption that, in the face
of possible divine deception, knowledge can no longer be assumed to be infallible.
Given God’s omnipotence, as we saw Buridan explain above, we can no longer be
certain of anything and hence, if we hold on to the idea that knowledge is infallible,
we will have to accept skepticism. Buridan’s way out is to say that knowledge is
fallible and that we must relativize the concept of evidence to the domain of inquiry
we are concerned with. Albert of Saxony explains this very clearly. He writes:
It must be noted that evidence is used in two ways, namely maximal and natural
evidence. Maximal evidence is used for propositions in accordance with which
the intellect by its nature cognizes with assent to the proposition, and cannot
dissent from it. In this way, we say that the principle of non-contradiction is
evident, and, similarly, it is by this evidence that it is evident that I know that I
exist. Natural evidence, however, is such evidence that nothing opposite appears
through human reasoning unless by sophistry. And in this way natural principles
and conclusions are said to be evident. It must be noted that [such principles] are
not strictly speaking evident, since the intellect that holds this to be evident can
be deceived by a supernatural power.3
Albert goes on to explain that God may have changed the appearances so that
Socrates who appears human to me is in fact a donkey or that fire is hot when it in
fact is cold, and hence it is not maximally evident that “Socrates is a human being”
or that “Fire is hot.” These are only naturally evident, which means that there is
no natural counterevidence to them. It is true given that we are not being deceived
by God.
Ailly explains this distinction in a very similar way. He, however, couches it in
terms of absolute versus conditional evidence. For the second he also uses secundum
quid and ex suppositione. Natural evidence on Albert’s view is conditional in Ailly’s
terminology, since it assumes that God is not deceiving us, that is, it puts God
aside so to speak. He notes as well that whatever is absolutely evident is infallible
(infallibilis), and the examples he gives are the principle of non-contradiction and
that I know that I am alive. He also adds to this list truths like that one cannot
be made cold by warmth (Sentences I, q. 1, art. 1, O, fol. 48r). Ailly also adds
a distinction between what is conditionally evident and what is merely probable.
Something is probable if it is possible to conclude the opposite, but if something is
230 SKEPTICISM
evident, on the other hand, then there is no counterargument to it, save, of course,
divine deception (Sentences I, q. 3, art. 3 DD, col. 78r).
Mainly through Ailly this concept of knowledge and the new account of evidence
that followed from it had an enormous influence on later philosophy. We can see its
influence well into the sixteenth century (see Karger 2012). It very elegantly short-
circuited the global skeptical threat of divine deception and, although it ultimately
moved from a view of knowledge as infallible to a view of it as fallible, it allowed
philosophy and natural science to develop without the threat of skepticism.
NOTES
1. On Buridan, see further Christophe Grellard’s chapter in this book.
2. On Autrecourt, see further Christophe Grellard’s chapter in this book.
3. Albert of Saxony, Quaestiones subtilissime super libros Posteriorum Analyticorum
Aristotelis, I, q. 3, fol. 4rb–va, transcribed in Fitzgerald 2002: 348–349.
REFERENCES
Ailly, Peter of. 1968. Quaestiones super libros sententiarum cum quibusdam in fine
adjunctis. Frankfurt: Minerva.
Aquinas, Thomas. 1952. Summa Theologiae, edited by P. Caramello. Turin and Rome:
Marietti.
Buridan, John. 1518. Quaestiones in Aristotelis Metaphysicam: Kommentar zur
Aristotelischen Metaphysik. Paris.
Courtenay, W. 1971. “Covenant and Causality in Pierre d’Ailly,” Speculum 46: 94–119.
Descartes, René. 1984–1991. Oeuvres, edited by P. Tannery and Ch. Adam. Paris: Vrin.
Descartes, René. 1984–1985. The Philosophical Writings of Descartes, translated by J.
Cottingham, R. Stoothoff, and D. Murdoch. Cambridge: Cambridge University Press.
Duns Scotus, John. 1950. Ordinatio I, d. 1-2. Opera Omnia 2. Vatican City: Typis
Poliglottis Vaticanis.
Duns Scotus, John. 1960. Lectura I, prologus, d. 1-7. Opera Omnia 16. Vatican City: Typis
Poliglottis Vaticanis.
Duns Scotus, John. 2010. “A Buridanian Response to a Fourteenth Century Skeptical
Argument and Its Rebuttal by a New Argument in the Early Sixteenth Century.” In H.
Lagerlund (ed.), Rethinking the History of Skepticism: The Missing Medieval Background,
215–232. Leiden: Brill.
Fitzgerald, M. 2002. Albert of Saxony’s Twenty-Five Disputed Questions of Logic, Leiden:
Brill
Gelber, Hester. 2004. It Could Have Been Otherwise: Contingency and Necessity in
Dominican Theology at Oxford, 1300–1350. Leiden: Brill.
Medieval Epistemology.” In H. Lagerlund (ed.), Rethinking the History of Skepticism:
The Missing Medieval Background 119–144. Leiden: Brill.
Holkot, Robert. 1967. In quatuor libros Sententiarum quaestiones. Frankfurt: Minerva.

First published in 1518.
Karger, Elizabeth. 2004. “Ockham and Wodeham on Divine Deception as a Skeptical
Hypothesis,” Vivarium 42: 225–236.
Knuuttila, Simo. 1993. Modalities in Medieval Philosophy. London: Routledge.
Knuuttila, Simo. 2012. “Modality.” In J. Marenbon (ed.), The Oxford Handbook of
Medieval Philosophy, 312–341. Oxford: Oxford University Press.
Lagerlund, Henrik (ed.). 2010a. Rethinking the History of Skepticism: The Missing
Medieval Background. Leiden: Brill.
Lagerlund, Henrik. 2010b. “Skeptical Issues in Commentaries on Aristotle’s Posterior
Analytics: John Buridan and Albert of Saxony.” In J. Marenbon (ed.), The Oxford
Handbook of Medieval Philosophy, 193–214. Oxford: Oxford University Press.
Oakley, Francis. 1964. The Political Thought of Pierre d’Ailly: The Voluntarist Tradition.
New Haven: Yale University Press.
Panaccio, Claude and David Piché. 2010. “Ockham’s Reliabilism and the Intuition of Non-
Existents.” In H. Lagerlund (ed.), Rethinking the History of Skepticism: The Missing
Perler, Dominik. 2010. “Does God Deceive Us? Skeptical Hypotheses in Late Medieval
Epistemology.” In H. Lagerlund (ed.), Rethinking the History of Skepticism: The Missing
Rimini, Gregory of. 1979–1984. Lectura super primum et secundum Sententiarum, edited
by A. D. Trapp and V. Marcolino. Berlin: de Gruyter.
Stroud, Barry. 1984. The Significance of Philosophical Scepticism. Oxford: Oxford
University Press.
CHAPTER SEVENTEEN
Michel de Montaigne
GIANNI PAGANINI
1 LIFE AND WORKS

Michel de Montaigne was the first European intellectual to fully realize the strong
impact of the rebirth of ancient Pyrrhonism on early modern philosophy. Widely
known as a skeptical student of philosophy and as a painter of himself as well as of
mankind, he used a new literary form, the essay, in which he tends to digress into
anecdotes and personal meditations, according to his motto: “I am myself the matter
of my book.” At the same time, the apparently loose form of the essay permitted him
to adopt an approach of free doubt and inquisition toward any matter: the essayist
is the modern incarnation of the “seeker” (zētētikos) or the “inquirer” (skeptikos) as
it was depicted by ancient Pyrrhonism, some sort of “researcher” who practices the
skeptical art of zētēsis (free research).
Montaigne was born on February 28, 1533, in the Château de Montaigne, close
to Bordeaux.1 His father, Pierre Eyquem, Seigneur of Montaigne, was a French
Catholic soldier in Italy for a time and had also been the mayor of Bordeaux. His
mother, Antoinette López de Villanueva, converted to Protestantism, whereas
his maternal grandfather came from a wealthy Spanish-Jewish family who had
converted to Catholicism. The oldest of eight surviving children, Montaigne had a
brother and a sister who became Protestants. It is very probable that the presence
of these different religious allegiances in the family history stimulated Montaigne
to be open-minded with regard to problems of faith and pushed him to adopt an
approach of wide religious toleration. To this effect, the experiences deriving from
the fierce religious war in France between Catholics and Calvinists that raged during
the sixteenth century also contributed in a strong way.
Montaigne’s education followed a pedagogical plan sketched out by his father
and refined by the latter’s humanist friends. The objective was for Latin to become
his first language, as it in fact was. He also learned Greek, though without becoming
as fluent in it as in Latin. Around the year 1539, Montaigne was sent to study at a
prestigious boarding school in Bordeaux, the Collège de Guyenne, then under the
direction of the greatest Latin scholar of the era, George Buchanan. Afterward,
the young Montaigne studied law in Toulouse; then, he entered a career in the
local legal system, becoming a counselor of the Court des Aides of Périgueux and,
in 1557, counselor of the Parlement in Bordeaux. From 1561 to 1563, he was a
MICHEL DE MONTAIGNE 233
courtier at the court of King Charles IX, obtaining the highest honor of the French
nobility, the collar of the Order of St. Michael. During this period at the Parlement
in Bordeaux, he became very close friends with the humanist poet Étienne de la
Boétie, by whose death in 1563 he was deeply affected. It has been argued that after
losing in Étienne his best conversation partner, Montaigne sought for a different
interlocutor who could take the place of the dead friend. It was thus that he started
writing the Essais (Essays) as his means of communication and expression: this time
he would aim not at a friend but at a reader. In 1565, Montaigne married Françoise
de la Cassaigne, basically under pressure from his family, who encouraged him to
accomplish marriage as a sort of social duty. The couple had six daughters, though
only the second-born, Léonor, survived childhood.
Following the recommendation of his father, Montaigne started to work on the
first translation of Raymond Sebond’s Theologia naturalis, which he published a
year after his father’s death in 1568. Then he inherited the family’s estate, the
Château de Montaigne, to which he moved back in 1570, thus becoming the Lord
of Montaigne. Another literary accomplishment was Montaigne’s posthumous
edition of his friend Boétie’s works, among which Discours de la servitude volontaire
was prominent.
In 1571, Montaigne retired from public life to the tower of his castle, where
he almost totally isolated himself. Locked up in his library, he had several Greek
maxims (some of skeptical intonation) sculpted on the beams of the ceiling and
finally began to work on his Essays.
During the Wars of Religion, being still a Roman Catholic, he acted as a
moderating authority and obtained the esteem of both the Catholic King Henry III
and the Protestant Henry of Navarre (Henry IV). From 1580 to 1581, he traveled
in France, Germany, Austria, Switzerland, and Italy, partly in search of a cure for
the painful kidney stones from which he was suffering. To this effect, he established
himself at Bagni di Lucca, a place famous for the healing power of its waters.
During this travel Montaigne kept a fascinating Journal, which was published only
posthumously, in 1774.
While traveling through Rome in 1581, he experienced the strong pressure of
the Inquisition. Actually, he was detained in order to have the text of his Essays
examined by Sisto Fabri who served as Master of the Sacred Palace under Pope
Gregory XIII. After Fabri’s close scrutiny, the Essays were returned to its author on
March 20, 1581. The main objections concerned references to the pagan notion of
“fortuna” and praise for the emperor Julian the Apostate and some heretical poets.
Eventually, the author was released with a strong recommendation to make some
emendations to the text (which he followed only partly).
Being elected mayor of Bordeaux, Montaigne returned to France and took up
office. He was reelected in 1583 and served until 1585, again trying to moderate
between Catholics and Protestants. In 1586, he left his castle for two years, because
of the plague that had broken out and the Wars of Religion that were raging.
Montaigne continued to extend, revise, and oversee the publication of the
Essays. The first two books were published in 1580 and included 94 chapters.
The earliest were short, impersonal, more akin to a collection of anecdotes with
234 SKEPTICISM
brief conclusions. The following chapters were deeply influenced by the Stoic ideal
of resisting the ills of life by means of reason and will. However, the chapters of
1574–1575 already show a stronger discontent with Stoical solutions, until the
chapter that is the longest by far, the Apology for Raymond Sebond, results in a
strong attack on Stoicism, and more generally on any sort of dogmatism, both
philosophical and theological. In this chapter, Montaigne drew heavily upon
Greek skepticism, especially that of Pyrrho of Elis as presented in the works of
Sextus Empiricus and in the detailed Life of Pyrrho contained in the philosophical
biographies of Diogenes Laertius.
In 1588, Montaigne supervised the fifth edition of the Essays, the first to
contain the thirteen substantial chapters of Book III. He also met the writer Marie
de Gournay, who became his literary executrix. He attended the Estates-General
in Blois, where Henry III had Henri de Guise (the Catholic leader of the League
and the instigator of the revolt against the king) assassinated. After King Henry
III’s murder (1589), Montaigne helped to keep the city of Bordeaux loyal to the
Huguenot Henry of Navarre, who would go on to become King Henry IV.2
Montaigne died of quinsy at the age of fifty-nine, on September 13, 1592, at
the Château de Montaigne. Remaining in possession of all his other faculties, he
requested mass and died during the celebration of that mass. Since 1588 Montaigne
had not written new books or chapters of his Essays; however, he kept adding a
thousand passages that reveal a bolder and bolder awareness of his independent
thought. Deeply considering the harsh experience of religious wars, he sharpened
his views on the state of religion in France, blaming both Catholics and Protestants
for their fanatical and intolerant behavior. He became extremely conscious of the
necessity of man’s moral autonomy; thus, he criticized any conception of morality
that would rely on external authorities, such as religion and law. Furthermore,
Montaigne wrote indignantly against the cruelty and hypocrisy of the civil wars,
condemning torture and inquisition as means that do not fit any sense of humanity;
finally, he developed his plan to portray himself as a way to penetrate the universal
human form. His rejection of the Stoical rigor became sharper in the late Essays,
whereas his view of death changed completely. At the twilight of his life, he
considered death as the end, and not the goal, of the entire existence.
2 THE INTRODUCTION OF SKEPTICISM

INTO MODERNITY
As we have already said, Montaigne introduced the genuine Pyrrhonian philosophy
into early modern philosophy by giving it wide circulation. Before him it was
mostly a topic of study for philologists or was used to give support to apologists.
The immediate background of Montaigne’s position was the revival of skepticism
caused by the recovery of the works of Sextus in Renaissance European culture.3
The rediscovery, the publication, and finally the Latin translation of Sextus’s works
allowed Montaigne two great achievements that influenced the aftermath of the
skeptical trend during the seventeenth century. First of all, by reading Sextus’s
Pyrrhoniae Hyptyposes, Montaigne became able to grasp the profound originality

of Pyrrhonism in comparison not only to the other ancient philosophies, but also
to its “false friends,” such as Academic skepticism. Secondly, by having access to
Pyrrhonian sources, Montaigne could appreciate the full philosophical substance
of skepticism and grasp its principal aims, epokhē and ataraxia (suspension of
judgment and imperturbability or peace of mind). Even though he realized the
possible religious utility of skepticism as an introduction to naked faith, nevertheless
he avoided identifying it with a pure apologetic instrument, as was the case in the
Renaissance interpretations.
Pyrrhonism was the last ancient philosophy rediscovered by humanists, a
resurrection that occurred nearly at the end of the Renaissance. While Academic
sources—mainly Cicero’s Academica and the polemical work by Augustine, Contra
Academicos—were available during the Western Middle Ages, the few manuscripts
of Sextus’s works that were still extant in European libraries apparently passed
unnoticed before the sixteenth century. In early humanism, Florence and Rome
were the main centers of the renascent interest in those works. Giovanni Pico
(1463–1494) used it for his polemic against the astrologers (Disputationes adversus
asrologiam divinatricem) while his nephew, Gianfrancesco Pico della Mirandola
(1469–1533), was the first to make extensive use of Sextus’s philosophical works in
order to support his own fideistic apologia for Christianity. In his Examen vanitatis
doctrinae gentium (1520), Gianfrancesco Pico presented Pyrrhonian skepticism as
the best preparation to religion, which was completely alien to the true ancient
sources (cf. Cao 2007). His main target was to demolish the “pagan” dogmatism
of the Aristotelians and the other philosophical schools, not in order to achieve
the sheer suspension of judgment like a true Pyrrhonist, but in order to reach
the superior and transcendent certitude of religion. The heavy recourse to the
Pyrrhonian tropes (amply displayed and paraphrased from Sextus’s Hypotyposes)
was in Pico closely connected to the idea that Christian revelation would be better
set up on the ruins of philosophical reason. According to Pico, it is up to skepticism
to display its corrosive effect on the body of all dogmatic philosophies, so as to
open the way to the transcendent authority of religion and church. From then on
an understood alliance was consolidated in European culture between skepticism
and apologetics. In the seventeenth century, this typically modern blending of
skepticism and strong religious belief took the name of “Christian Pyrrhonism,”
finding a fervent supporter in Blaise Pascal in the second part of the century, and a
more controversial representative in Pierre Bayle, at the end of the century. In the
same vein one can read the long invective contained in Henricus Cornelius Agrippa’s
De incertitudine et vanitate scientiarum (1530), which blatantly counterpoised the
ignorance of simple believers with the pride of knowledge displayed by dogmatists,
which was considered by Agrippa as a “veritable plague” to be fled.
Apart from the circulation of manuscripts, which became significant at the end of
the fifteenth century, a wider first-hand knowledge of Sextus’s work was first made
possible thanks to the printing of Henri Estienne’s Pyrrhoniae Hypotyposes (1562)
and Gentian Hervet’s Adversus mathematicos (1569) Latin translations. Another
primary source for Pyrrho’s life and doctrines was available in an edition of Diogenes
236 SKEPTICISM
Laertius’s Lives of philosophers (1570) but one had to wait until 1621 to have the
editio princeps of Sextus’s Greek text printed (cf. Floridi 2002). The importance
of this hard philological work should not be underestimated; yet, it was not just
an episode inside the studia humanitatis tradition. Once rediscovered, ancient
skepticism was immediately exposed to a subtle enterprise of reinterpretation, in
which fidelity to the original meanings was the least of concerns.
We will give just two conspicuous examples to show that even the first editions
provided the occasion for free reinterpretations. The first editors, the Catholic
Gentian Hervet and the Calvinist Henri Estienne, were both worried about giving
to the modern reader the right key to understand the works they had edited and
translated. In their respective prefaces, the two humanists highlighted the anti-
dogmatic character of Sextus’s arguments, even while emphasizing the religious
utility that could possibly derive from skepticism. Of course, their theologies
were different and the religious advantages were accordingly different, too. The
Huguenot Estienne presented the usefulness of skeptical teaching in terms of moral
reform and the improvement of knowledge, whereas the Counter-Reformist Hervet
presented Sextus’s tropes and his devastating attack on reason as a confirmation
of his own objections to Protestant positions, which he had interpreted as a fresh
example of renewed dogmatism.
It is with Montaigne’s reappraisal that ancient skepticism in general comes to
be appreciated anew in its own original features and as a true philosophical point
of view, not only as an apologetic weapon. It is nearly impossible to underestimate
Montaigne’s importance as an expounder of this “new” skepticism. Besides drawing
on a whole new corpus of Pyrrhonian texts, he clearly marked the difference between
the real “skeptics”—namely, the Pyrrhonians—on the one side, and on the other
both “dogmatists” and negative dogmatists—these latter being neo-Academics who
positively and conclusively affirmed the impossibility of knowing, instead of merely
the need to suspend judgment. Furthermore, the author of the Essays rethought
and popularized fundamental notions such as phenomenon, criterion, epochē,
equipollence between contrary arguments (isostheneia), ataraxia or apathy, vicious
circle (diallelos), and “infinite regress.” For the first time, Montaigne translated
the entire Pyrrhonian technical terminology into a modern language and, what is
more, redirected the ancient notions to suit modern needs. With this reworking,
Pyrrhonian skepticism passed from the register of philology (or, in certain cases,
apologetics) to that of philosophy.
3 PHENOMENA AND APPEARANCES

In order to achieve this result, that is, to reinstate skepticism as one of the major
trends of early modern philosophy, the first step had to be to resurrect the basic
notion of phenomenon, which was at the very heart of Pyrrhonian philosophy. In
reality, Montaigne never used the word “phenomenon,” which was perceived at the
time as a Greek word and not accepted either as a Latin or a French term until the
age of Descartes and Leibniz (the latter was the first to use this word extensively
both in his Latin and in his French writings). If not the word, however, the meaning
was already present in Montaigne. Instead of the term “phenomenon,” Montaigne

usually spoke of “appearance” (apparence), which is a calque of the Latin apparentia
(plural) used by Henri Estienne in his translation and commentary of Sextus’s
Hypotyposes. In Estienne’s work apparentia and phainomena are strictly equivalent.4
This is not the only influence of Estienne’s philological work on Montaigne’s
vocabulary. Another, no less important, concerns the notion of phantasia, which
Estienne considered as equivalent to phenomenon. Estienne’s Latin interpretation
actually renders phantasia as “persuasionem et coactam passionem” (“assent and
involuntary affection”), which largely overlaps the definition of phenomenon. The
two influences (the one regarding phainomenon, the other connected to phantasia)
tightly intertwine. In a crucial passage of the Apology for Raymond Sebond (in
fact, a synthesis of PH I 72–78), Montaigne explains what a phenomenon is by
appealing to the appearance, at the same time connecting it closely to the notion
of fantasie (in French), which means to him “imagination,” or more precisely
“sensitive representation.”5 Both choices are not merely lexical; in reality, they
deeply characterize Montaigne’s understanding of Pyrrhonian skepticism.
Montaigne adopts an epistemologically weak version of phantasia or fantasie,
which is no longer a phantasia katalēptikē (the apprehensive representation that,
according to the Stoics, can provide a sure criterion for distinguishing between false
and reliable representations of reality). Montaigne’s skeptical fantasie is not able to
fulfill this function, and that situation clears the way for the study of the subjective
states of the mind in themselves and disconnected from reality. These subjective
states of the mind are now rather connected to the organic equipment of the senses
and are relative to circumstances, conditions, locations, frequency or rarity, habits,
etc. In the Apology Montaigne closely follows the famous Ten Modes of Aenesidemus
that had been expounded by Sextus. However, unlike the relativistic readings that
betrayed the authentic skeptical meaning of these tropes,6 the author of the Essays
really grasps their original content, which was calling into question the paradigm of
normality on which Aristotelian epistemology was founded. According to Aristotle
and the Aristotelians, any man, put in a “normal” situation, with senses and intellect
well-tuned according to nature, is able to get an adequate and true representation
of external reality. The traffic of species, which come from the object to the subject
bearing forms, is supposed to provide a reliable bridge between representation and
what is represented. On the contrary, in the Essays not only are species (both sensible
and intelligible) banished7—the words species and espèce literally disappear from
Montaigne’s lexicon—but the entire complex Aristotelian psychology, which is the
basis of his fundamentally realistic theory of knowledge, is abandoned. In a crucial
passage of the Apology, Montaigne compares dreaming and being awake, being
sober and being drunk, youth and old age, jaundiced and correct perception, health
and illness (Montaigne 1999: 595 ff.). Far from being a sheer rhetorical device,
this comparison plays an important role in Montaigne’s argumentation. The real
meaning of these comparisons consists in showing that, from the epistemological or
ontological points of view, normal and abnormal representations are on the same
foot: despite all of their differences, they are all simply phenomena or appearances
mediated by the internal constitution of the subject, the particular nature of the
238 SKEPTICISM
medium, the complex of circumstances in which both the subject and the object are
placed. There is neither direct nor privileged access to reality in itself. Considering
their causes and features, “normal” and “abnormal” representations depend on the
same natural processes that produce both. Unlike the Aristotelian species, skeptic’s
“images” or fantasies are not copies that contain the same forms as their patterns,
but just indirect, remote, and complicated effects of distant causes.
Starting from these principles, Montaigne can now include all subjective states
within the skeptical category of phenomena or appearances. In this light one should
read his apparently paradoxical claim: “We are always with some sickness” (Montaigne
1999: 569). For him, to perceive an object always means to be in a subjective state
that reflects the quality of the subject rather than the nature of reality itself; or, better,
any state of mind reflects a peculiar mixture of both. When perceiving a thing, says
Montaigne, we accommodate it to ourselves, transforming the thing according to
the self. This process of assimilation is at the same time a process of falsification that
prevents us from judging what the thing is in and of itself:
Now, because our state accommodates things and turns them according to itself,
we do not know what the actual things are any more; as nothing comes to us but
falsified and altered by our senses. (Montaigne 1999: 600; cf. 562, 587, 592,
599, 603)
In reality, the person “who judges by appearances judges on the basis of something
different from the thing itself” (Montaigne 1999: 601), relying on phenomena that
belong to the subject and its states—and not strictly to the object. These appearances
or phenomena are not part of the object, not even in a formal sense. As a consequence,
for a Pyrrhonist such as Montaigne is at least in the Apology, perceptions and
any sort of mental representations are not warranted either epistemologically or
ontologically, as was the case in the Aristotelian model of “normal knowledge.”
4 MONTAIGNE’S INNOVATIONS: SKEPTICISM AS

PHENOMENALISM AND PHILOSOPHY OF DOUBT
Montaigne’s interpretation had a strong impact in leading early modern philosophers
to seeing Pyrrhonian skepticism as a sort of phenomenalism.8 The foundations of
this reading can be summarized in two points. (a) The main scene of skepticism
becomes the dichotomy between appearance and reality. The first is knowable. The
second—which includes both formal essences and real substances—is unknowable.9
(b) About reality we know only the way in which the external object perceptually
appears. “Phenomenon” and “perception” become synonymous expressions in
Montaigne and in his followers, whereas, according to Sextus, there was also a
kind of phenomena that were disclosed only to the intellect; they did not have
sensible features but were rather like noumena (intelligible objects). Nowadays, we
might wonder whether Montaigne’s interpretation is philologically and exegetically
correct, in comparison with the ancient Pyrrhonian sources. We might also ask
whether this kind of dichotomy between appearance and reality does not end up
reinstating some sort of negative dogmatism, proclaiming that the latter is impossible
to know and imagining a mysterious metaphysical substructure behind phenomena.
This would be an outcome absolutely at odds with the sheer anti-dogmatic effort
displayed by Pyrrhonists. Our judgment, however, must be changed when we place
ourselves in the perspective either of Rezeptionsgeschichte or of Wirkungsgeschichte.
Both of them include the history of misunderstandings and unfaithful traditions,
sometimes much more than the sheer history of faithful transmission. Very often the
former results in sterile repetitions, whereas the latter produces interesting novelties.
In this connection, it is historically true that Montaigne’s phenomenalistic version of
ancient skepticism became the prevailing reading in early modern philosophy and
deeply influenced the entire epistemological debate.
The adoption of the skeptical pattern based on this phenomenalistic approach
permitted Montaigne to break the link between reality and appearances and to
narrow all of our knowledge to the latter. The clearest statement of this break can
be read in the so-called Socrates’s portrait dilemma: how can we be certain that the
portrait is a picture of Socrates if we only have access to his representation, namely,
to appearances (phenomena), and not to the original model? The Pyrrhonian
problem of the criterion is tightly connected to this new scenario, which is the
coherent result of abandoning the Aristotelian notion of species (the bearer of the
form) as a link between reality and its sensible or intellectual representation. Since
appearance and reality are now split, it is necessary to appeal to a “third” term (the
criterion) that could guarantee the relation between the two. However, this criterion
should in turn require a new one, so raising a sort of infinite regress in the search for
further criteria that can guarantee the previous ones. The attempts made in order to
stop this aporia fall into two possible traps: the diallelos, or vicious circle, and the
dogmatic pretense to allegedly objective and uncontroversial evidence. In reality,
both of these aporias result in the impossibility of overcoming the basic dilemma of
the criterion (Montaigne 1999: 600–601).
Besides these epistemological considerations, which brought all the strength of
Pyrrhonian skepticism into the European culture, already weakened by the crisis of
Aristotelian scholasticism, the Apology for Raymond Sebond also contains the critical
examination of a wide range of dogmatic conceptions belonging to the special fields of
metaphysics, theology, politics,10 morals, and even religion.11 Criticism of perceptual
representations, contradictions between philosophical schools on their great topics
of knowledge (God, soul, immortality, ontological foundations of reality, laws of
nature, the divide between the so-called wild peoples and the civilized ones, etc.): all
these are the themes of Montaigne’s skepticism, which rejuvenates old Pyrrhonian
topics in close contact with modern problems, thus marking the beginning of a
rather lively version of the “life without dogmatic beliefs.” Montaigne elaborates
also a new concept of belief, emphasizing its emotional and practical character12
and giving prominence in the field of religion to that particular instinct from above
that is divine grace. This was also for him the way to distinguish between “true”
religion, which is mysterious and supernatural, ultimately an indisputable gift
coming from God, and the real religions, which are purely human since they are
historical and sociological phenomena. Unlike the former (transcendent revelation),
240 SKEPTICISM
these latter (historical phenomena) can be fully submitted to critical analysis, and
this close examination takes in Montaigne the form of tough disenchantment. In
spite of his overtures to grace and supernatural revelation,13 the author of the Essays
present in other chapters a realistic depiction of the all-too-human forms taken by
religions when they are transformed into blind belief, absurd superstition, intolerant
fanaticism, or simply traditional conformism.
Another major novelty brought about by Montaigne concerns his transforming a
philosophy based on epochē and ataraxia into a philosophy of doubt. In the Greek
tradition, ataraxia or peace of mind flowed from suspension of judgment—namely,
from epochē—rather than from knowledge and judgment about things. Montaigne
maintains the basic avowal of ignorance concerning moral laws, no matter whether
they are conceived as natural laws or as prescriptions connected to the particular
settlement of every society. However, he eventually ends up by making doubt,
instead of epochē and ataraxia, the climax of skeptical approach. This is not without
consequences for the entire conception of Pyrrhonism, all the more so since doubt is
perceived by Montaigne as a state of restlessness and discomfort rather than of calm
and moderation of emotions.14 In this connection, Montaigne’s position is defined
by a double move: first, he identifies skepticism with the activity of doubting; second,
he stresses the discomfort involved in the attitude of doubting. Thus, skepticism
and doubt become co-extensional and synonymous, which was not the case in the
ancient sources.
Once again, the contribution made by Estienne’s translation to this shift is
considerable. Following Cicero, Estienne rendered the Greek verb aporein, with
its multiple meanings, with the much more univocal Latin verb dubitare (cf. Mates
1996: 30). Under this influence, Montaigne is the first modern writer to stick
to the idea that the main activity of the skeptic should be doubting, in the exact
meaning that this term takes in the theory of knowledge: an activity fundamentally
concerning thoughts whose truth one evaluates. The main activity of the modern
skeptic turns out to be a mental exercise, whereas the ancient Pyrrhonist does
a discursive practice: the former brings up doubts in his mind, the latter utters
opposite statements, balancing each other until he reaches isosthenia. In the Greek
Pyrrhonian tradition, doubting or wondering was not the main activity of the
inquirer. In fact, the ancient skeptic used to compare conflicting propositions, so as
to show that they were no more (ou mallon) asserted than denied. The proper stance
of the Pyrrhonist was the suspension of judgment, rather than a fluctuating state of
mind, like incertitude and hesitation. After Montaigne and under his influence, these
mental and psychological aspects of skepticism, which are represented mostly by
the activity of doubting, decidedly prevailed over the original Pyrrhonian linguistic
formulations, and it contributed to render extremely problematic the ideal of living
comfortably following appearances, according to the original Pyrrhonian project.
This relation of synonymy established by Montaigne between the activity of
doubting and skepticism is found especially in a crucial passage of the Apology,
where the primacy of doubt overlaps a wider panoply of ancient formulations.
At first, the author correctly mentions all of the main phonai (phrases) typical of
Pyrrhonism, showing how they intertwine, converging into epochē (Montaigne
1999: 505). However, when he needs afterwards to narrow the outlook and focus
on a few main points, Montaigne drastically reduces the wider range of skeptical
prescriptions, limiting them to two basic propositions: he stresses the formula of
Socratic ignorance (J’ignore), which actually is neither properly Pyrrhonian nor
exclusively skeptical, and he puts forward the very activity of doubting (Je doubte)
as the main activity of skepticism. “I doubt” becomes thus the true motto of the
skeptic (Montaigne 1999: 527). Needless to say, both formulas betray the genuine
meaning of epochē, since they imply a state of mind quite different from the mere
suspension of judgment. “I ignore” and “I doubt” substitute the balanced neutrality
of the suspension of judgment with either a dogmatic denial (“I do not know”) or a
statement of wavering perplexity (“I doubt”). Despite this true misunderstanding of
the original skeptical approach, Montaigne’s interpretation had a strong impact on
his posterity and deeply influenced seventeenth-century presentations of Pyrrhonism.
It is only in the second edition of the Essays that Montaigne reformulates his
approach to skepticism as a kind of interrogation (Que sçay-je?, i.e., “What do I
know?”) rather than of a positive or negative assertion.15 He thus intends to avoid
the danger of falling into the logical inconsistency represented by the liar paradox,
according to which a proposition like “I do not know anything” or “I doubt
everything” should apply to itself, with the paradoxical and unacceptable result of
being at the same time true and false.
5 THE SKEPTICAL SELF

Montaigne has constantly been celebrated as the first inventor of the modern self,
or as the father of modern subjectivity. Self-knowledge and self-control play a
great role in the Essays, indeed. Montaigne studied himself more than any other
subject; he was convinced that the attempt to depict the most faithful portrait of
an individual life could lead also to a better knowledge of humanity (Montaigne
1999: 379, 665, 806). He clearly stated this very conviction: “Each man bears the
entire form of man’s estate. Authors communicate themselves to the people by some
peculiar and strange mark; I am the first to do that by my universal being, as Michel
de Montaigne, and not as a grammarian or a poet or a lawyer” (Montaigne 1999:
805). Self-study was for him a school of human nature. This is also the reason
why he shared with Socrates full allegiance to the Delphic precept: “Know thyself”
(Montaigne 1999: 1001). This is in accordance with the notion of a phenomenon as
an internal state of mind, to which the subject has direct access. Furthermore, even
when he seems to look at others, in reality it is just to better understand himself: “I
tell the others, but I do that in order to recount more myself”; “I watch my life in
that of the others” (Montaigne 1999: 1076).
Montaigne was in a good position to develop this art of self-portrayal. Self-study
led him to practice the art of self-description. Nearly all of the Essays practice this
kind of technique, in style as well as content. Furthermore, the portrait must follow
the variability, the contingency of its subject; thus, the topic of the essay changes as
the subject changes, according to its multiplicity and variability (Montaigne 1999:
378, 805, 818, 964, 1095). The closing of the Apology argues that man cannot
242 SKEPTICISM
have firm knowledge of being, because he is rather a creature of flux, according

to the Heraclitean interpretation of the skeptical diatribe; therefore, man cannot
participate in full being, which must be stable. Thus, the only real being that
does not become is God. The very conclusion of the Apology is against Stoicism,
considered as the supreme form of dogmatism, inasmuch as it holds that man can
achieve by himself the highest degree of wisdom and natural perfection. According
to Montaigne, it is rather the reverse: God alone can rescue man from the perpetual
flux of becoming and raise him above humanity (Montaigne 1999: 601–603).
The depiction of the self is evident and extremely enlivened in the Essays. For
all the appreciation that deserves, we must nevertheless distinguish between, on
the one hand, the artistic practice of subjectivity, of which Montaigne is a master,
reaching absolutely outstanding results, and on the other the philosophical theory
of the subject, for which he still lacks the main conceptual premises and in which
he is not really interested. No trace of a proper philosophical theory of the subject
can be found in the Essays, despite all his concern about knowledge of his internal
and external states. In this regard, the Pyrrhonian model turned out to be more an
obstacle than a real help to any theoretical formulation of subjectivity: the more
it helped him to an experimental description of the self, the less it contributed to
elaborating a theory of the subject.
It is true that, according to Montaigne, the “sacramental formula” of the skeptics
is epechō (“I suspend [judgment]”), which is a first-person point of view. In reality,
this does not imply that Pyrrhonism relies on a deep reflection on the role of the
subject. Actually, Pyrrhonism can be defined more as an impersonal philosophy than
as a first-person view. Thus, skeptical phenomena are usually described by Sextus
using the impersonal form of phainetai, that is, “it seems,” “it appears.” Even though
phenomena are connected in practice to the peculiarities of the perceiving subject,
from the theoretical point of view there is no need to make explicit the role played
by this perceiver; or rather, the perceiver is put at the same level as the perceived
and the natural medium that connect each other. Even in Aenesidemus’s Ten Modes,
in which the common topic is the relation connecting subject and object, the tone
of the treatment is openly naturalistic and very far from any hint of the primacy of
the human subject. This latter is mostly considered as a piece of nature plunged into
a material world. The knowledge process itself is explained as a “combination” of
strongly naturalistic elements (sense organs, humors, particles, and membranes).
If we have a look at the psychological section of Montaigne’s Apology, we can
immediately realize that the approach of this modern skeptic is also extremely
objectivistic. He examines the main different schools of philosophical psychology,
showing their hopeless clash, according to the skeptical method of diaphonia
(disagreement). Treating this topic is like being in the middle of the “Babel tower”
(Montaigne 1999: 553). In his extensive doxology, Montaigne depicts the subject
and the soul as parts of nature. Far from grounding the esprit’s prerogative, centrality,
and activity in a theory of the autonomy of the soul, Montaigne draws especially
on Stoic and Epicurean crude materialism, while considering Platonism, with its
complex doctrine of the psychē, to be a fairy tale and poetic philosophy. About
Aristotle, he thinks that he was intentionally obscure with regard to his psychology,
willing to hide his attitude toward the thesis of the intellect’s mortality (Montaigne
1999: 552). At the end of this harsh examination of all the psychological theories,
the intellect itself is described as “an instable, dangerous, and daring instrument”
(“un outil vagabond, dangereux et temeraire”), upon which one should not rely
too much. It would be better to assign to it “the tightest possible constraints” (“les
barrieres le plus contraintes qu’on peut”) (Montaigne 1999: 559),16 so as to control
its extreme fickleness. The full awareness of man’s mutability, inconstancy, and
inconsistency was another cause that prevented Montaigne from outlining a strong
doctrine of subjectivity. It was up to Descartes to resume from Montaigne the new
characterization of skepticism and, at the same time, to overcome the aporias in
which the author of the Essays remained trapped.17
NOTES
1. On Montaigne’s life, see Frame (1965).
2. On Montaigne’s political involvements, see Desan (2006).
3. Popkin (2003) is still a useful work about modern skepticism. For a more in-depth and
up-to-date analysis, see Paganini (2003; 2008b). On Renaissance skepticism in general,
see Popkin and Maia Neto (2004) and Paganini and Maia Neto (2008). For a wider
look at recent research on early modern skepticism, see Moreau (2001), Charles and
Bernier (2005), and Maia Neto, Paganini, and Laursen (2009).
4. For this interpretation of Montaigne’s skepticism and its drawing on Estienne’s
translation of Sextus, see Paganini (2008b: 15–60). On the difference with Sanches’s
case, cf. Paganini (2004). For example, when translating Sextus’s main chapter about
epochē’s limitations (“Whether skeptics reject phenomena,” PH I 19–20), Estienne
felt it necessary to interpolate the text to explain better the sense of the Greek word
phainomena. To this aim, he used all the variations on the verb “to appear” in the
lexicon. In the next chapter (PH I 21–23), Sextus established an equivalence between
the phenomenon—Estienne’s apparentia—and the phantasia.
5. As was first pointed out by Dumont (1972: 44–45), the entire long passage from
the Apology for Raimond Sebond (Essays, II, 12) is a brilliant paraphrase of Sextus
Empiricus’s text. Cf. Montaigne (1999: 600–601).
6. The reasons why skepticism is not relativism have been correctly illustrated by Annas
and Barnes (1985: 96–98, 144): instead of suspending judgment, relativism leads to
the affirmation that the object is neither this nor that, or more precisely, it is one thing
when it stands in one relation, another when it is in another relation.
7. For a sharp criticism of the Aristotelian pattern of knowledge by similarity, which is
at the basis of the notion of species, see Montaigne (1999: 601).
8. For this impact, especially on Hobbes, see Paganini (2003: 3–35).
9. In some famous passages, Montaigne contrasts the perception of the quality of a
thing (often varying among subjects) with the thing itself (Montaigne 1999: 587).
In other places, he contrasts observable qualities with the essence of the object,
which is unknowable (Montaigne 1999: 598–599). Sometimes Montaigne attributes
“l’ignorance de la vraye essence de telles choses” to the lack of faculties apt to grasp
the “hidden” properties of things (Montaigne 1999: 590).
244 SKEPTICISM
10. On Montaigne’s skeptical political thought, see Laursen (1992).

11. On this last topic, see Panichi (2010: 45–96, 119–152).
12. This aspect is at the heart of the recent French interpretation: see Brahami (2001:
167–234); Giocanti (2001: 91–140).
13. See, for example, the end of the Apology, where the absolute transcendence of the
divine is openly proclaimed (Montaigne 1999: 601–604; cf. 927–928). On these
topics, see Carraud and Marion (2004), Desan (2008).
14. To characterize modern skepticism, most scholars have focused on the abandonment
of the conception of happiness as ataraxia. This is an important feature in itself, but
it is secondary compared to what we are dealing with here. For more details, see
Paganini (2008a: 166–169).
15. Montaigne (1999: 527): “Cette fantasie est plus seurement conceüe par interrogation:
Que sçai-je? comme je la porte à la devise d’une balance.” Clearly, it is an allusion to
the medal Montaigne had made to engrave with this symbol, and, at the same time,
a reference to the skeptical theme of the equal balance of two opposite assertions,
or isostheneia.
16. On the problems raised by Montaigne’s naturalism, see Gontier (1998).
17. On Montaigne’s influence, besides the classical study by Brunschvicg (1942), see
Marchi (1995), Melehy (1997), Gessmann (1997), Paganini (2008b: ch. 1), Panichi
(2009), Hartle (2013), and Panichi and Spallanzani (2013).
REFERENCES
Annas Julia and Jonathan Barnes. 1985. The Modes of Scepticism: Ancient Texts and
Modern Interpretations. Cambridge: Cambridge University Press.
Brahami, Frédéric. 2001. Le travail du scepticisme: Montaigne, Bayle, Hume. Paris: Presses
Universitaires de France.
Brunschvicg, Léon. 1942. Descartes et Pascal lecteurs de Montaigne. Neuchâtel: La
Baconnière.
Cao, Gian Mario. 2007. Scepticism and Orthodoxy: Gianfrancesco Pico as a Reader of
Sextus Empiricus. Pisa and Roma: Fabrizio Serra Editore.
Carraud, Vincent and Jean-Luc Marion (eds.). 2004. Montaigne: scepticisme, métaphysique,
théologie. Paris: Presses Universitaires de France.
Charles, Sébastien and Marc André Bernier (eds). 2005. Scepticisme et modernité. Saint-
Étienne: Publications de l’Université de Saint-Étienne.
Desan, Philippe (ed.). 2006. Montaigne politique. Paris: Honoré Champion.
Desan, Philippe. 2008. Montaigne: Les formes du monde et de l’esprit. Paris: PUPS.
Desan, Philippe (ed.). 2008. Montaigne et la théologie: “Dieu à nostre commerce de société”.
Genève: Droz.
Dumont, Jean-Paul. 1972. Le Scepticisme et le phénomène: Essai sur la signification et les
origines du pyrrhonisme. Paris: Vrin.
Eva, Luiz. 2004. Montaigne contra a Vaidade: Um estudio sobre o ceticismo na Apologia de
Raimond Sebond. São Paulo: Humanitas.
Faye, Emmanuel. 1998. Philosophie et perfection de l’homme: De la Renaissance à Descartes.

Paris: Vrin.
Frame, Donald. 1965. M. Montaigne: A Biography. London: Hamilton.
Gessmann, Martin. 1997. Montaigne und die Moderne. Zu den philosophischen Grundlagen
einer Epochenwende. Hamburg: Felix Meiner Verlag.
Giocanti, Sylvia. 2001. Penser l’irrésolution: Montaigne, Pascal, La Mothe Le Vayer. Trois
Gontier, Thierry. 1998. De l’homme à l’animal (Montaigne et Descartes). Paris: Vrin.
Green, Felicity. 2012. Montaigne and the Life of Freedom. Cambridge: Cambridge
University Press.
Hartle, Ann. 2013. Montaigne and the Origins of Modern Philosophy. Evanston:
Northwestern University Press.
Laursen, John Christian. 1992. The Politics of Skepticism in the Ancients, Montaigne,
Hume, and Kant. Leiden, New York and Köln: Brill.
Maia Neto, José R., Gianni Paganini, and John Christian Laursen (eds.). 2009. Skepticism
in the Modern Age: Building on the Work of Richard Popkin. Leiden: Brill.
Marchi, Dudley. 1995. Montaigne Among the Moderns: Receptions of the Essais. Providence:
Berghahn.
Mates, Benson. 1996. The Skeptic Way: Sextus Empiricus’s Outlines of Pyrrhonism. New
York: Oxford University Press.
Melehy, Hassan. 1997. Writing Cogito: Montaigne, Descartes, and the Institution of the
Modern Subject. New York: State University of New York Press.
Montaigne, Michel de. 1999. Essais, edited by P. Villey. Paris: Presses Universitaires de
France.
Moreau, Pierre-François (ed.). 2001. Le scepticisme au XVIe et au XVIIe siècle. Paris: Albin
Michel.
Paganini, Gianni. 2003. “Hobbes Among Ancient and Modern Sceptics: Phenomena and
Bodies.” In G. Paganini (ed.), The Return of Scepticism: From Hobbes and Descartes to
Bayle, 3–35. Dordrecht, Boston and London: Kluwer.
Paganini, Gianni. 2004. “Montaigne, Sanches et la connaissance par phénomènes. Les
usages modernes d’un paradigme ancien.” In V. Carraud and Jean-Luc Marion (eds.),
Montaigne: scepticisme, métaphysique, théologie, 107–136. Paris: Presses Universitaires
de France.
Paganini, Gianni. 2008a. “The Quarrel over Ancient and Modern Scepticism: Some
Reflections on Descartes and His Context.” In J. Popkin (ed.), The Legacies of Richard
Popkin, 173–194. Dordrecht: Springer.
Paganini, Gianni. 2008b. Skepsis: Le débat des modernes sur le scepticisme. 2nd edition,
2017. Paris: Vrin.
Paganini, Gianni and José R. Maia Neto. 2008. Renaissance Scepticisms. Dordrecht:
Springer.
Panichi, Nicola. 2004. I vincoli del disinganno. Per una nuova interpretazione di Montaigne.
Firenze: Olschki.
Panichi, Nicola. 2010. Montaigne. Roma: Carocci.
246 SKEPTICISM
Panichi, Nicola (ed.). 2009. Montaigne et les philosophes, Montaigne Studies 21.
Panichi, Nicola and Maria Franca Spallanzani. 2013. Montaigne and Descartes:
A Philosophical Genealogy. Special issue of Montaigne Studies 25.
Popkin, Richard. 2003. The History of Scepticism from Savonarola to Bayle. Oxford:
Popkin, Richard and José R. Maia Neto (eds.). 2004. Skepticism in Renaissance and Post-
Renaissance Thought. New Interpretations. Amherst: Humanity Books.
Sève, Bernard. 2008. Montaigne: Des règles pour l’esprit. Paris: Presses Universitaires de
France.
CHAPTER EIGHTEEN
Pierre Charron
GIANNI PAGANINI
1 LIFE AND WORKS

Pierre Charron was born in Paris in 1541, one of the twenty-five children of a
bookseller. After studying law, he took his degrees in Montpellier (1571) and
afterward practiced as an advocate, for a few years, before deciding to study
theology, still in Montpellier, where he resided from 1565 to 1566 and again from
1570 to 1571. He then entered the church and soon was a popular priest, rising to
become a canon.
He moved to the southwest of France, invited by Arnaud de Pontac, Bishop
of Bazas. He was appointed priest in ordinary to Marguerite de Valois, wife of
Henry IV of Navarre. In about 1588, Charron decided to become a monk, but,
being rejected by both the Carthusians and the Celestines, he returned to his old
profession. He delivered a course of sermons in Angers, and the next year he moved
to Bordeaux where he formed a famous friendship with Michel de Motaigne. On
July 2, 1586, Montaigne presented him with a copy of Catechism by Bernardino
Ochino (the famous Italian monk who fled his native country and religion to
adopt Protestant ideas), with the annotation liber prohibitus (“forbidden book”).
According to La Rochemaillet, Charron, residing in Bordeaux from 1589 to 1593,
had many opportunities to visit Montaigne during the last four years of the latter’s
life. On Montaigne’s death, in 1592, Charron was requested in the will to bear the
Montaigne arms.1
In many circumstances of his life and during his activity as a preacher, Charron
got involved in the events of the French religious wars, first in the struggles between
the Catholic League and the Huguenots, and afterwards in the difficult procedures
aiming to enforce the Edict of Toleration (Nantes Edict). Indeed, Charron directly
experienced the fierce troubles provoked by religious intolerance: in Montpellier
he was imprisoned and taken hostage by the Protestants (1565–1566). In 1588, in
Angers, he was reached by the news of the assassination of the leader of the League,
the Duke of Guise, by order of King Henry III, against whom he had rebelled. In
the Discours chrestien, qu’il n’est permis au subjet pour quelque cause et raison que
ce soit, de se liguer, bander et rebeller contre son Roy (written in 1588 and published
in 1606), Charron remembers to have been caught then by a “strong fit of anger
248 SKEPTICISM
and disdain,” but to have been quickly “cooled,” definitely abandoning his previous
attachment to the League. This Discours contains a complete refutation of the ideas
held by the monarchomachs, preaching instead absolute fidelity to the sovereign,
even when he is “bad.”
Charron spent from 1594 to 1600 under the protection of Antoine Hérbrard de
Saint-Sulpice, Bishop of Cahors, who appointed him grand vicar and theological
canon. In 1595, he was chosen deputy to the general assembly of the clergy, for
which he became chief secretary. For the occasion he preached also in the church of
Saint Eustache, one of the most populous parish churches in Paris.
In 1601, Charron published in Bordeaux his main work, De la Sagesse, a system
of moral philosophy that develops some ideas from Montaigne, realizing a peculiar
synthesis of skeptical wisdom and a strong Stoic sense of human autonomy. In this
work, Charron also connected Montaigne’s skepticism to the anti-rational strand
in Christianity, underlining the narrow limits that constrain human knowledge
in theological matters. He worked afterwards on a new revised edition but died
suddenly of a stroke on November 16, 1603. This new edition appeared in 1603, just
after the author’s death. It contains numerous changes in comparison with the first
one; in particular, Charron modified the plan of Book I: he added six chapters and
modified several passages that were considered scandalous by theologians.2 Despite
these changes, the work was condemned by the Faculty of Theology in Paris (1603)
and was after included in the Index of Forbidden Books (1605). De la Sagesse was
also attacked by the Jesuit François Garasse (1585–1631), who described Charron
as an atheist; yet the work was defended by Duverger de Hauranne, the abbot of
Saint Cyran, one of the leading personalities within the Jansenist movement.
A summary and defense of the book, written shortly before his death, appeared
in 1606 (Traicté de Sagesse). In 1604, Charron’s friend Michel de la Rochemaillet
prefixed a “Life” to an edition of De la sagesse, which depicts Charron as an
amiable man of good character. His complete works, with this contribution by de la
Rochemaillet, were published in 1635.
2 SKEPTICISM AND THE RISE OF

MODERN SUBJECTIVITY
In his treatise De la Sagesse, Charron tried to summarize and rearrange Montaigne’s
thought, bringing forth that handbook of wisdom that was already implicit in the
Essays but still needed to be explicitly and systematically expounded.3 In so doing,
Charron made of De la Sagesse a cornerstone for the new culture of the honnête
homme that spread all over the seventeenth century, influencing both the skeptical-
libertine milieux and Cartesian morals. In Charron’s book various philosophical
trends converge and cooperate: the Stoic reevaluation of the autonomy of virtue
goes together with an argument against superstition and subordination of morality
to faith. At the same time, the drive to an entirely mundane preud’homie (“human
wisdom”) appears to be full of Epicurean suggestions and contains daring references
to an unscrupulous political prudence.
PIERRE CHARRON 249
From the point of view of the history of skepticism, Charron’s work is a landmark
that separates modern forms from the ancient ones. De la Sagesse uncovers a new
face of skepticism, which had deep impact on the common image of this movement
during the seventeenth century and after. In Charron’s work, the skeptical sage is
not only an active and important character, but, what is more, also is someone who
breaks away from the classical Pyrrhonian pattern as well as from its more recent
reenacting in Montaigne’s Essays. As regards the ancient pattern, the attainment
of epochē was basically the result of a process of “detachment from the self,” the
fruit of an “accentuated passivity” of sensations and thoughts, “a paralysis of
reason by itself,” as Myles Burnyeat has put it (1983: 133). According to Jonathan
Barnes, epochē was a pathos, some sort of passivity or passion, that happened
to inquirers (skeptikoi) at the end of their investigations. In other words, for the
ancient Pyrrhonians, even the suspension of judgment was less a voluntary effort
than the product of a causal sequence impacting on them (the sequence included:
investigation (zētēsis)—disagreement (diaphōnia)—equipollence (isostheneia)—and
finally suspension of judgment (epochē) (Barnes 1983: 7)). Nearly the same can be
said about Montaigne, except that this latter added a sense of pessimism and human
frailty that was unknown to the ancient skeptics. When one reads Montaigne, it
seems that skepticism inflicts a heavy blow not only to human arrogance, but also to
the value of the intellect and the senses themselves. Not by chance, reason is defined
in the Essays as “an instrument of lead and wax, which one can lengthen, bend, and
adapt to any direction and extent” (Montaigne 1999: II xii 565).
From this point of view, Charron’s skepticism is quite different. First of all, the
author of De la Sagesse does not hesitate to stress the strength of reason, emphasizing
at the same time the positive feature of skeptical examination. It is true that reason,
and mostly skeptical reason, does not have “jurisdiction” on all subject matters:
for example, Charron proclaims that skepticism makes no metaphysical claims and
that it relies on revelation in theological matters. Nevertheless, in the mundane
domain of “human wisdom,” reason preserves its complete right “to judge about
everything” and to “restrain assent” when faced with inadequate reasons (Charron
1986: 399). In his own field, the sage shows full “universality of spirit”; he is not
narrowed by the “municipal law” (1986: 406) that rules closed systems of beliefs,
whether these beliefs be the elementary, sociological beliefs of the village or the
more sophisticated beliefs of nations, philosophical schools, religions (i.e., actual
religious rites and behaviors), and churches.
For the first time in the modern age, Charron’s skeptical sage is someone who
fights against an entire corpus of beliefs, someone who decides to doubt and seeks
for arguments to this effect. This is the main Charronian achievement: epochē or
suspension of judgment becomes now a vigorous and voluntary liberating move
from a system of beliefs rather than an imponderable point of balance between
different opinions, as it was before in the ancient idea of equipollence. Far from
being the passive effect produced by the unresolved diaphōnia of appearances
that are simply registered by the observer, Charron’s art of doubting requires a
conscious and voluntary decision from the subject. Skepticism demands a whole
set of epistemic virtues that preserve the wise from precipitancy, prejudices, and
250 SKEPTICISM
mistakes. Modern skepticism, as it is reworked by the author of De la Sagesse, needs

a discipline of the intellect and its judgments. After Charron, the modern skeptic
is someone who wants to cast into doubt almost everything whereas the ancient
Pyrrhonian was cast in doubt by different phenomena. With this work, one can
see the author initiate the typical modern attitude of negative suspicion against
any belief whose sole justification is habit, authority, or tradition. From passive,
skepticism becomes active; it is no longer the position of an impartial observer but
the pose of a conscious protagonist who does not want to be enslaved to prejudices
or opinions devoid of any rational justification.4
This is also the reason why Charron’s philosophy, probably more than Montaigne’s,
and certainly more than ancient Pyrrhonism, is one of the birthplaces of modern
subjectivity (cf. Paganini 2008). One major contribution to this achievement is the
sharp separation between externality and interiority in the attitude of the sage.
Whereas external appearances can be given up to social conformity, so that the
skeptic would not challenge laws, common opinions, and public offices, interiority,
that is, private judgment, becomes in Charron the domain of free “jurisdiction”
that acknowledges no authority but reason. The external sphere is the field in
which particularity and municipality rule, whereas the internal becomes the reign
of “universality of spirit.” The “free spirit” or the esprit fort, to which Charron
recognizes the right to judge about everything, is the antecedent of the modern
intellectual to whom the right to criticize established values and conventions is
attributed, without any other commitment, except to reason and morality. All these
innovations, along with the clear influence of Stoicism, converge in Charron’s
philosophy into the discovery of man’s autonomy: autonomy that means firstly self-
government according to the lead of reason, and secondly disentanglement toward
any external, traditional, social, and political constraint.
This profound split between internality and externality runs along the line dividing
the private and the public spheres. After Charron, modern skepticism develops in
the sage’s interior region, while actions are confined to exterior conformism. This
is a major innovation in comparison to the ancient Greek sources that ignored this
dramatic divide. In a much clearer way in Charron than before in Montaigne, this
radical split is not only a matter of personal reservation; indeed, it stretches to the
extremes of simulation and dissimulation. Faced with the constraints imposed by
external requirements, the modern Charronian sage feels free to recur to pretense
and reticence in order to keep his own interiority safe and free. We should not
understand this attitude as a kind of overt rebellion; on the contrary, the skeptic does
not shirk his own duties and responsibilities, in line with the Pyrrhonian precept that
recommends following the rules, customs, and laws of one’s country. Nevertheless,
and despite this apparent conformism, the modern skeptic becomes the bearer of a
new sense of intellectual emancipation that was unknown to the ancients. Actually,
the relationship between the sage and society has completely changed with respect
to antiquity. The moderns are confronted with schools of thought, political and
religious bodies of beliefs that are organized, and compelling systems requiring from
the subject the performance of precise behaviors, but also requiring him to take sides
in the course of the ideological struggles that tear the commonwealth, the church,
PIERRE CHARRON 251
and their institutions. Thus, the modern skeptic has to experience the strength and
the coercion of ideological systems that were almost unknown to antiquity.
In order to face this new and troubling situation, the modern skeptical sage,
according to Charron, should adopt a cold and detached conception of acting that
is tightly connected to the distinction between the external and the internal. The
skeptic must give to the institutions just the minimum of his external commitment,
the minimum that is absolutely necessary for the needs of praxis and for the
functioning of the social body. By contrast, he ought to preserve for the internal and
individual sphere his intimate beliefs. The modern skeptic can feel free to criticize in
his internal sphere the common wisdom and the values of the city, which were not
cast into doubt by the ancient skeptic. The main target of the criticism of the latter
were not so much the beliefs and the rites of the city, but only the moral and political
theories elaborated by the dogmatic philosophers (cf. Charron 1986: 411–415).
Furthermore, and despite all of Charron’s caution in keeping different fields
and their respective prerogatives sharply separated, one must admit that with this
modern skeptical scrutiny the foundations for an uncompromised assent to the
public institutions are at least implicitly undermined. The least one can say is that, in
comparison to those of ordinary people, the skeptic’s choices become more pragmatic
and independent of any partisan dogmatism. Even though he conforms to the beliefs
and rules that are in force and displays his profession of conformity, whenever he can
withdraw into “the forum of consciousness,” the skeptical sage refuses to abdicate
the freedom of his own judgment; his external obedience is completely disentangled
from any internal orthodoxy. According to Charron, abiding by a law does not
necessarily involve the duty of considering it true, right, or rational. All this has an
evident impact on the holding of the skeptic’s beliefs. By practicing the skeptical
scrutiny of any dogma, including those that constitute morals and politics, the sage
is now able to denounce the “mystical ground” of laws and customs: “mystical”
in the sense of being obscure, mysterious, and ultimately arbitrary or unjustified.
Echoing one of Montaigne’s famous statements, Charron declares that “laws and
customs keep their credit not because they are right and good, but for being laws
and customs: this is the ground of their authority” (1986: 498).5 This sentence
overtly expresses skepticism regarding the official motivations that are given to
support ethics and politics. This split between the internal and the external can
sometimes result in a plain contradiction, as Charron himself recognizes: “enjoying
his own right to judge and examine everything, it will often happen to the sage that
his judgment and his hand, the spirit and the body contradict themselves, and he
will act outside in one way and judge inside in another way” (1986: 393). This is the
effect of the new protagonist pose that the sage takes, at least in his internal sphere,
toward the whole body of public opinions. Whereas the ancient Pyrrhonians usually
stressed the perfect conformity of the skeptic to the customs and beliefs that make
up not only ordinary life but also political life, Charron emphasizes the complete
autonomy of the sage, who conceives himself as an independent subject, even while
seeking refuge in the sphere of private consciousness.
We can understand now why the value of isostheneia is no longer central in
this new perspective. By challenging the cautious advice of “general indifference,”
252 SKEPTICISM
Charron sticks much more to the rational basis of skepticism. He declares that “the
search for truth” is the real and explicit aim of any skeptical investigation, and that
its instrument (outil) is reason. This free use of reason, which is not only allowed but
also recommended by skepticism, becomes the new standard for the conception of
being truly human. Human dignity is now based on using reason in order to “judge”
about everything:
The real prerogative of humans, their proper and more natural practice, their
worthier activity, is that of judging. Why are humans beings that speak, reason,
and intend? Because perhaps they have minds in order to build castles in the air
or to fill to bursting with silliness and trifles, like most people? Who ever had
eyes for darkness? Certainly, humans are made to see, to understand, to judge all
things to such a point that they are called the guardian, supervisor, controller of
nature, world, God’s works. To deprive them of this right is the same as willing
them not to be anymore humans but beasts. (Charron 1986: 389–390)
This recourse to the prerogative of judging is all the more crucial, since humans
are immerged in that condition of uncertainty that the skeptical crisis has revealed.
This situation of doubt demands from the sage the most attentive practice of
rational discernment:
Since there is only one truth among a thousand lies and of a thousand opinions
about the same subject only one is the true one, why shouldn’t I investigate which
one is best, the most reasonable, good, useful, and convenient using reason? Is it
possible that among so many laws, habits, beliefs, different and opposite customs,
only those we practice are the good ones? (Charron 1986: 387)
Charron’s conclusion is univocal: “The sage will judge everything. Nothing will
escape him that he lays on the table and on the scale to investigate” (Charron
1986: 392).
It is essential to notice that this kind of “judging” is neither dogmatic nor
Pyrrhonian. As to the former, Charron explains that the sage does not “judge” in the
sense of “establishing, affirming, determining,” but in the sense of a continuous quest
that has no end (“to examine, evaluate, weigh and balance reasons on one side and
on the other, and thus search for truth”). As to the latter, Charron admits that the
sage can opine (contrary to the Pyrrhonian injunction against neo-Academics), even
though “without any determination, decision, affirmation.” This is a kind of having
“weak” opinions, according to Academic probabilism (Charron 1986: 386–387).
It must be added, however, that this right of doubting and judging, which is at the
heart of Charron’s active interpretation of skepticism, is not the business of everyone
but just of those who are well equipped for it, namely, “spirits well born, strong, and
vigorous” (“esprits bien nez, forts et vigoureux”). These peculiarities depend on the
prerogative of reason, and do not derive from any privilege of nobility. According to
Charron, an esprit fort is someone who is capable of using reason and conducting his
own life according to this kind of rational examination, which is the basic procedure
of skepticism. The Charronian sage is well aware that social, political, and religious
institutions rely on passivity, and yet that they are absolutely necessary for ordinary
PIERRE CHARRON 253
people. What is more, Charron is overly pessimistic about the possibility that most
people could act according to his ideal of self-determination. The large majority is
made of “not too well talented men”; these latter strongly support the character that
laws and conventions impose upon them.
Unlike ancient skepticism, which did not explicitly make any difference
between various kinds of persons, modern skepticism, as it was shaped by Charron
in the wake of Montaigne, is intrinsically hierarchical. Charron divides human
beings according to intellectual capacity and correlates with this their ability to
suspend judgment, not giving their assent to what is doubtful or simply plausible.
Most people are on the lowest degree of this tripartite hierarchy (“weak spirits,
of low and little ability, born to obey, to serve, and to be conducted”); in the
middle are “pedants,” who are perfect models of the “dogmatist”; on top are the
few who “would rather doubt and suspend their judgment” than enter the rush
of dogmatic and unjustified affirmations. This tripartite classification of spirits
was already present in Montaigne, but it is only with Charron that it enters as
a constituent point into the construction of a skeptical anthropology (Charron
1986: 291–292).6
On the whole, it can be said that the main aim of Charron’s skepticism is no longer
the Pyrrhonian ataraxia. The importance of this rather passive value is superseded
now by the exercise of a triad made up of much more active practices: “judging of
everything” (juger de tout), “retaining the assent” (sursoir la determination), keeping
“the universality of spirit” (universalité d’esprit) (Charron 1986: 406). Even though
he still draws on Montaigne for illustration of the details in which this practice of
doubting ultimately consists,7 it is only in Charron that this practice takes the form
of an emancipatory program supported by a deliberate education and exercise of
the spirit.8
3 CHARRON’S SKEPTICISM COMES BACK FROM

PYRRHONISM TO NEO-ACADEMICISM
Since the theory and praxis of epochē are largely common to the two main skeptical
traditions, Pyrrhonism and neo-Academic philosophy, Charron feels authorized to
draw on both of them. He often puts Pyrrhonism and Academicism on the same
level (Charron 1986: 405), equating the ataraxia of the former to the “neutrality
and indifference” of the latter (Charron 1986: 410). In so doing, he inaugurates
the typical modern koinē of skepticism, which is some sort of eclecticism in the
art of doubting. Blending together these two trends, Charron goes much further
than Montaigne, who had rather stressed the novelty of Pyrrhonism and rejected
probabilism. The author of De la Sagesse opens the way to considering both
traditions as belonging nearly to the same body of philosophy, which meant, for
some “dogmatic” opponents of skepticism like Descartes, rejecting Pyrrhonism
and Academicism altogether, and for some sympathizers of skepticism like Hume,
allowing them to shift from the one form to the other, always keeping allegiance to
the skeptical approach.
254 SKEPTICISM
From the strictly epistemological point of view, Charron does not take up the
precise characterization of Pyrrhonism that was contained in Montaigne’s Apology
for Raymond Sebond. This marks a major shift in his own description of skepticism,
which turns out to follow in the wake of the neo-Academic revival that preceded
Montaigne’s turn toward Sextus’s texts.
Among the neo-Academic features that are recognizable in De la Sagesse,9 there
is first the emphasis on the ideal of intellectual integrity, which comes from Cicero
(Acad. II. 7–9), and the insistence on the necessity of studying and neutralizing the
internal and the external causes of any strong attachment to uncertain, false, and
precipitant opinions. “Retaining the assent” (surseance) is the first condition for
keeping the freedom of one’s spirit (Charron 1986: 404). Avoiding precipitancy and
error ends up being more important than attaining truth: whereas this latter usually
is extremely difficult to reach, the former always stays within our range.
The second major Academic influence on Charron consists in an overt
profession of Socratic ignorance. The device of the wise man, which figures in the
frontispiece of De la Sagesse, is the famous Socratic motto “je ne sais” (“I do not
know”). The affirmation of akatalēpsia was blamed by Pyrrhonians as an example
of negative dogmatism. By contrast, it is taken up by Charron, despite all of its
logical inconsistencies (one could object that at least the sage knows one thing, that
he knows nothing). Nor is the author of De la Sagesse interested in Montaigne’s
reform of skepticism, which had transformed the profession of ignorance into an
interrogation (“que sçais-je?”). Charron seems inclined to give to his sage a stable
intellectual and moral position (which is made of intellectual integrity, rejection of
errors, refusal to hasten his judgments, emendation of prejudices, etc.). To this effect,
he does not like the kind of Pyrrhonian irresolution that was practiced by his master
Montaigne. The assured position of the neo-Academic sage seems to him steadier
than the passive and indifferent pose typical of the Pyrrhonian philosopher.10
The third Academic feature is the emphasis put on the value of self-knowledge.
Moving in the wake of his beloved predecessors, Socrates and Montaigne, the
author of De la Sagesse gives a systematic form to what in the Essays was much
more a matter of practice than of theory. In Charron, self-observation and self-
inquisition develop in a true anthropological doctrine, without losing all of his
skeptical approach: in reality, this whole corpus of knowledge that starts from the
self and ends with humanity is basically aimed at fighting dogmatic presumption
and ignorance; skeptical anthropology is mainly a doctrine about the moral and
epistemic limits of anyone’s self (Charron 1986: 28, 33, 37, 44–46, 376).
Finally, the fourth positive influence of the Academic tradition on Charron makes
him more open than Montaigne to Carneades’s notion of probability as a guide
to the complex world of opinions, most of all practical opinions.11 In this regard,
Charron clearly breaks off from both Sextus, who denied any significant difference in
opinions concerning probability, and Montaigne, who saw in probabilism a challenge
not only to Pyrrhonian indifferentism but also to his Catholic traditionalism. By
contrast, Charron thought that there was a way of taking a doctrine as probable
that was compatible with a true skeptical approach, provided that one adopted
Carneades’s and Cicero’s point of view.
PIERRE CHARRON 255
4 THE SKEPTICAL IMPACT ON MORALS

AND RELIGION
The consequences for morals and religion of this new conception of wisdom, in which
skeptical themes play a crucial role along with a Stoic inheritance, are extremely
important. The comparison with Montaigne is still useful in order to appreciate the
novelty of Charron’s position. Whereas the former blended the Pyrrhonian fourfold
rule of life with modern Catholic traditionalism, which prevented him from
adhering to the Reformation, Charron, even while remaining apparently faithful
to his religious confession, used skepticism to go beyond the limits of Montaigne’s
cautious conformism, even in the field of morals and religion. As to the former
(morals), Charron took up the Platonic and Stoic topic of the “inborn seeds,”12 willing
to emphasize the autonomy of morals from any external constraints, such as those
of religions and laws. The origin of this topic is still classic, coming basically from
Cicero and Seneca; however, Charron’s reinterpretation is new in that he stresses its
possible polemical meaning, which is now clearly directed against the overwhelming
burden of religious eternal norms (see Kambouchner 2008; Gaiu 2010). This is a
radical turn in comparison to the traditional ethics that relied on a transcendent
foundation and not on a “natural” tendency of man, appealing much more to the
threat of natural rewards and the promise of supernatural punishments than to the
intrinsic value of virtue.13 In contrast with an entire millennial religious tradition,
Charron wants men to be good even “without hell and paradise” (1986: 464). Nor
do original sin and theological grace have any importance in the philosophical ethics
of De la Sagesse, even though the author will be compelled in the second edition
to mitigate his views, adding (in Book III, Ch. III) a paragraph about “grace as a
perfection of nature,” which would be taken up again in the Traicté de sagesse with
the title: “warning about the necessity of the divine grace” (1986: 846–847; cf.
Faye 1998: 270–271). However, and despite this softening of the original thesis, it
is still true that in the first version of De la Sagesse, human wisdom (preud’homie)
is considered by the author as antecedent and independent with regard to religion.
Charron continually emphasizes the difference between “probity” on the one side
and “piety” on the other, claiming that “religion is posterior, a special and particular
virtue, distinguished from the other virtues” (1986: 464).14 At the same time, the
traditional conception of wisdom related to metaphysics and speculative theology is
completely abandoned, whereas the emphasis is put on its practical and humanistic
import. Charron stresses the narrow borders of human reason: while restraining
philosophy to what lies within the natural reach of the human mind, he polemically
reacts to the excessive pretensions of the metaphysicians who aim at delivering the
first principles and causes. The object that is naturally adequate to human reason is
not knowledge of essences or substances, or dogmatic physics, but human morals
and life (Charron 1986: 25–32).
In the expanded preface to the second edition of De la Sagesse, Charron
explicitly distinguishes between divine wisdom and human wisdom, leaving to the
theologians the former and adopting the latter as his own subject. In so doing,
he appeals to his qualification as a philosopher and not as a theologian. It is on
256 SKEPTICISM
religion that this skeptical-stoical sort of wisdom had the greatest and the most
influential impact. In principle, true religion and true wisdom should converge,
since—says Charron—when someone is really a sage, he “acts according to God,
according to himself, to law, to nature” (1986: 424). In reality, when one examines
religions in their historical and actual substance, leaving out of consideration the
supernatural pretensions that they unanimously claim, rightly or wrongly, it is easy
to recognize that all of them counterpoise true wisdom. Consequently, one basic
distinction that De la Sagesse draws is that between “true piety” and “superstition.”
The superstitious man is dominated by fear and makes God to be like men, thereby
running into “blasphemy.” He reduces the relation with the divinity into a utilitarian
trade, some sort of “trafficking” in which the expectation of rewards and the fear
of punishments end up being more important than reason and virtue. Noticing
how much superstition is natural among men, Charron draws from Montaigne
a whole web of considerations that develop, in the text of De la Sagesse, into a
true philosophical anthropology of religious behaviors (see Kogel 1972: 90–104).
In Book II Ch. V, Charron opposes “vraye pieté” to actual religions: these latter,
however various and strange, have many features in common, such as sacrifices,
humiliation of the spirit, credulity, multiplication of miracles, and some sort of
illusory reversal in the human relations to the divinity. While aiming at propitiating
gods, these religions end by annihilating human nature. “What overturning of sense,
when one thinks to appease the divinity by inhumanity, to gain divine goodness by
our affliction, to please his justice by cruelty?” (1986: 449). Even though Charron
would like to carry out “true wisdom and true piety,” joining them and “blending
together,” it is still true that, even in the most favorable hypothesis of a mutual
concord, one should keep them sharply distinguished: “[I want it to be so] that each
of them subsists and stands alone, without the help of the other, and acts by its own
strength” (Charron 1986: 466). Wisdom must be “already born in you, with you,
planted by nature”; by contrast, religion cannot but be “posterior,” “learnt” through
revelation and instruction. In the relation between religion and wisdom, the latter
comes first and is the cause of the former, not the reverse (1986: 466–467).
Because of his official commitments in ecclesiastical institutions as a preacher
and a teacher,15 Charron had the opportunity to write many theological works,
some of which preceded De la Sagesse. The first was Les trois vérités contre les
athées, idolatres et juifs (1593), an apologetic treatise aimed at demonstrating the
existence of God and the truth of Christian religion. It is interesting to see how
these apologetic aims developed in such a way as to connect with skeptical themes
that would be treated at a later point in De la Sagesse. The main points in common
with this latter work concern the polemic against superstition and the adherence to
the themes of negative theology, on the basis of the idea that “God is impossible
to know” (1635: II 12). Combining the skeptical thesis about the inadequacy of
human knowledge with the theses of mystical and apophatic theology grounded
on God’s ineffability, Charron goes so far as to write statements that only seldom
could be found in any apologetic work. The “Discourse on God’s knowledge” is
particularly explicit, stating that “every divinity, which is proven and established
by reason, is a false divinity, not a true one,” since “there is no passage from
PIERRE CHARRON 257
the finite to the infinite” (1635: II 13). In this same Discourse, Charron declares
every theological representation to be absolutely relative and therefore devoid of
any effective content. After describing with mystical overtones the soul that rises
“abstracting itself from everything” and “merging into a vague and infinite void,”
Charron concludes that every “image” or representation of the divinity is “always
false, that is faulty and defective,” even though man cannot but try to fabricate one.
Despite the tone filled with compunction, the ultimate result is a strong skepticism
about the value of any discourse about God. Charron states so: “The last and
highest degree to which everyone can rise and reach by the extreme effort of his
imagination is for him his own God” (1635: II 19–20). In reality, every definition,
every positive statement about God is at the same time some sort of restriction
of the infinite, which is unacceptable. This apophatic method (according to which
we can say of God what he is not but not what he really is) also serves Charron’s
strategy against atheism because he sustains that even a negation of God, being as
dogmatic as the affirmation, would intend to set a limit to divine nature; therefore,
atheism also falls into the defect of restricting the infinite to the finite. Like the
dogmatic theologian, the atheist is guilty of an “abuse,” since both try to submit
the infinite to the limits of finite human reason (1635: II 13). This propensity of
Charron toward negative theology, as opposed to “dogmatic” theology, is strongly
emphasized in the Discours chrestiens (1601). In this work, he appeals to the label of
“learnt ignorance” and declares that his work aims at arousing “admiration” rather
than offering “knowledge” about God. Religious discourse should be simply a way
of “honoring” God (“without any affirmation, determination, prescription”), and
not an effort to describe Him (1635: I 12–13). When it is applied to theological
discourse, skepticism encourages some sort of transvaluation of the attributes of
God. In the wake of this tendency, Charron does not just emphasize the absolute
transcendence of the divinity; he even goes much further than supporting the
traditional distinction between ordained and absolute power in God, stating that He
can do everything that is not contrary to Himself, that is, to being and truth (1635: I
23). The most daring statement by Charron in this direction is that, even though we
cannot “conceive goodness, wisdom, power except in a human way,” nevertheless
one should remove all the bounds from these notions, going as far as to imagine a
kind of “superwisdom” (sursagesse) in God (1635: I 8).
NOTES
1. The main source for Charron’s life is still La Rochemaillet in Charron (1635),
reprinted in Desan (2009: 273–286).
2. A bibliography of Charron’s works can be found in Charron (1960: 147–151) and in
Dini and Taranto (1987: 419–435).
3. Referring to his sources in general, Charron claims that his work is the product of his
readings; however, he says, “the form and the order are mine” (1986: 34).
4. On this peculiar feature of Charron’s skepticism, see Popkin (1954; 2003: 57–63,
68–70, 100–107) and Paganini (1991: 201–214).
258 SKEPTICISM
5. This is an echo of a famous passage by Montaigne (1999: III xiii 1072).

6. See Montaigne (1999: II xvii 657). For this tripartition, see Paganini (1987).
7. See e.g. Charron (1986: 406–411) for the considerations on barbarism, universality
of nature, variety of religions, etc.
8. Another important source of this emancipatory program can be traced to the
secularization of some ideas related to the movement of “free spirit” or “libertinism,”
typical of the sixteenth century and condemned by John Calvin. From this heritage
probably comes—even though from far—the continuous hints made by Charron to
the Pauline motto “spiritualis omnia iudicat et a nemine iudicatur” (1 Cor. 2: 15),
by which he tries to legitimate his appeal to freedom of judgment and universality
of spirit (Charron 1986: 41, 389).
9. For the emphasis on neo-Academic features in Charron’s work, see José Maia Neto’s
interpretation (Maia Neto 2003, 2014).
10. It is significant that, when it is a matter of describing the grounds of epochē, Charron
recurs much more often to neo-Academic formulations that to the truly Pyrrhonian
ones (Charron 1986: 399–400). Also when characterizing the result of epochē,
Charron emphasizes active more than passive features of the spirit (see Charron
1986: 404).
11. On the issue of whether Carneades’s stance is a sort of probabilism, see Harald
Thorsrud’s contribution in Part One of this book.
12. On this crucial topic, see Horowitz (1971; 1974; 1998).
13. For the novelty represented by Charron in this field, see Gregory (1986: 71–113),
translated in Gregory (2000: 115–155).
14. For the originality of Charron’s notion of “wisdom,” see Rice (1958: 178–207) and
Abel (1987: 187–208).
15. On this activity, see Belin (1995: 7–205).
REFERENCES
Abel, Günher. 1978. Stoizismus und Neuzeit. Berlin: De Gruyter.
Barnes, Jonathan. 1983. “The Beliefs of a Pyrrhonist,” Elenchos 4: 5–43.
Belin, Christian. 1995. L’Oeuvre de Pierre Charron (1541–1603): littérature et théologie de
Montaigne à Port-Royal. Paris: Honoré Champion.
Burnyeat, Myles. 1983. “Can the Skeptic Live His Skepticism?” In M. Burnyeat (ed.), The
Charron, Pierre. 1593. Les trois vérités contre les athées, idolatres, juifs, mahumétans,
hérétiques, schismatiques par Charron. Bordeaux: S. Millanges.
Charron, Pierre. 1601. Discours chrestiens. Seconde partie. Bordeaux: S. Millanges.
Charron, Pierre. 1635. Toutes les Œuvres de Pierre Charron…. Paris: Jacques Villery.
(Reprinted in 2 vols. Genève: Slatkine Reprints, 1970.)
Charron, Pierre. 1986. De la sagesse. Paris: Fayard.
Desan, Philippe (ed.). 2009. Pierre Charron. Special issue of Corpus 55.
Dini, Vittorio and Domenico Taranto (eds). 1987. La saggezza moderna. Temi e problemi
dell’opera di Pierre Charron. Napoli: Edizioni Scientifiche Italiane.
PIERRE CHARRON 259
Faye, Emmanuel. 1998. Philosophie et perfection de l’homme. De la Renaissance à Descartes.

Paris: Vrin.
Gaiu, Claudiu. 2010. La Prudence de l’homme d’esprit. L’éthique de Pierre Charron.
Bucuresti: Zeta Books.
Gregory, Tullio. 1986. Etica e religione nella critica libertina. Napoli: Guida.
Gregory, Tullio. 2000. Genèse de la raison classique de Charron à Descartes. Paris: Presses
Universitaires de France.
Horowitz, Maryanne. 1971. “Pierre Charron’s View of the Source of Wisdom,” Journal of
the History of Philosophy 9: 443–457.
Horowitz, Maryanne. 1974. “Natural Law as Foundation for an Autonomous Ethics:
Pierre Charron’s De la Sagesse,” Studies in the Renaissance 21: 204–227.
Horowitz, Maryanne. 1998. Seeds of Virtue and Knowledge. Princeton: Princeton University
Press.
Kambouchner, Denis. 2008. “Descartes et Charron: prud’homie, générosité, charité,”
Corpus 55: 193–208.
Kogel, Renée. 1972. Pierre Charron. Genève: Droz.
Maia Neto, José. 2003. “Charron’s Epoché and Descartes’ Cogito: The Sceptical Base of
Descartes’ Refutation of Scepticism.” In G. Paganini (ed.), The Return of Scepticism:
From Hobbes and Descartes to Bayle, 81–131. Dordrecht, Boston and London: Kluwer.
Maia Neto, José. 2014. Academic Skepticism in Seventeenth-Century French Philosophy:
Montaigne, Michel de. 1999. Les Essais, edited by Pierre Villey. Paris: PUF.
Paganini, Gianni. 1987. “‘Sages,’ ‘spirituels,’ ‘esprits forts.’ Filosofia dell’ ‘esprit’ e tipologia
umana nell’opera di Pierre Charron.” In V. Dini and D. Taranto (eds.), La saggezza
moderna. Temi e problemi dell’opera di Pierre Charron, 113–156. Napoli: Edizioni
Scientifiche Italiane.
Paganini, Gianni. 1991. Scepsi moderna. Interpretazioni dello scetticismo da Charron a
Hume. Cosenza: Busento.
Paganini, Gianni. 2008. Skepsis: Le débat des modernes sur le scepticisme. 2nd edition,
2017. Paris: Vrin.
Popkin, Richard. 1954. “Charron and Descartes: the Fruits of Systematic Doubt,” Journal
of Philosophy 51: 831–837.
Rice, Eugene. 1958. The Renaissance Idea of Wisdom. Cambridge, MA: Harvard University
Press.
CHAPTER NINETEEN
Francisco Sanchez:
A Renaissance
Pyrrhonist against
Aristotelian Dogmatism
DAMIAN CALUORI
1 INTRODUCTION
Francisco Sanchez (1550/51–1623) was a well-known skeptic in the sixteenth and
seventeenth centuries. His one skeptical work, That Nothing is Known (Quod nihil
scitur), first published in 1581, was widely read and went to four editions.1 Pierre
Bayle praised him in his Dictionnaire historique et critique as “a great Pyrrhonist”
(Bayle 1740: vol. IV, 133), and Gabriel Naudé recommended in 1644 that Sanchez’s
skeptical work should be part of every learned library—next to that of Sextus
Empiricus (see Jolly 1990).2 Even as late as 1702, Leibniz declared in a letter to the
French mathematician and physicist Pierre de Varignon that he had often thought a
reply to Sanchez would be a very useful thing indeed (Gerhardt 1859: 94–95). After
this, Sanchez’s influence waned. He no longer seems to have been discussed much
in the eighteenth century.
In what follows, I will first discuss what kind of skepticism Sanchez adhered to:
whether he was an Academic or a Pyrrhonian skeptic. Then I will consider Sanchez’s
critique of the scholastic Aristotelians, the dogmatists of his day. Finally, I will ask
where Sanchez stands in relation to modern Cartesian skepticism and whether his
skepticism is purely theoretical or whether it also has implications for the practical
affairs of life.
2 SANCHEZ, THE PYRRHONIST

There is a traditional view of skepticism—a view found, for example, in Hegel’s
influential Lectures on the History of Philosophy—according to which the skeptic
FRANCISCO SANCHEZ 261
claims that nothing can be known or that nothing can be known for certain (Hegel
1979: 253). Sometimes this version of skepticism is called “dogmatic” skepticism
since it dogmatically claims that nothing can be known (or that nothing can be known
for certain). Already in antiquity, Academic skepticism was sometimes interpreted
in this way.3 Thus, Sextus Empiricus criticizes Academic skepticism for asserting
that nothing can be known and considers this a mark distinguishing Academic from
Pyrrhonian skepticism: while the Academic skeptic dogmatically states that nothing
can be known, the Pyrrhonist continues his search (PH I 3, 226).4
It is easy to see how the title of Sanchez’s work, That Nothing is Known,
can be understood as defending a dogmatic skepticism. In order to counter this
misunderstanding right from the start, Sanchez begins his work with the following
words: “Not even this do I know, that I know nothing. I suspect, however, that
neither I nor anyone else knows anything”5 (QNS 1, LT 172).6 Further elaborating
on this, Sanchez denies that he can prove the truth of the claim that he knows
nothing. This incapability of proving his ignorance, he thinks, confirms his suspicion
that he does not know that he knows nothing. If, however, he had proof of his total
ignorance, then he would know that he knew nothing. He could then assert that
he knew nothing. Pursuing this hypothetical alternative a little further, Sanchez has
an imaginary opponent state: “If you know how to prove that you know nothing,
the contrary follows, namely, that you already know something” (QNS 1, LT 172).
Sanchez, not being a dogmatic skeptic, can easily reply that he knows of no such
proof and since he has no such proof, he does not commit himself to any claim to
knowledge. Rather, like an ancient Pyrrhonist, Sanchez does not know that nothing
can be known. He is just not aware of any knowledge that has so far been established.
Thus, it seems to me that Sanchez’s own words rule out an interpretation such as
that of Richard Popkin who states in his History of Scepticism: “Sanchez’s totally
negative conclusion is not the position of Pyrrhonian skepticism, the suspense of
judgment as to whether anything can be known, but rather the more full-fledged
negative dogmatism of the Academics” (Popkin 2003: 41).
That Sanchez was not a skeptic of the sort Popkin believes him to be does not as
such imply that he was a Pyrrhonian skeptic. There seems to be at least one further
possibility. Elaine Limbrick, for example, shares Popkin’s view that Sanchez was
an Academic skeptic. Yet she understands Academic skepticism as advocating, not
a negative dogmatism, but “a system of methodological doubt which would free
the human mind of the fetters of the past and enable the skeptical enquirer to re-
establish knowledge on a firmer empirical basis” (Limbrick 1988: 88). According
to her interpretation, the knowledge that Sanchez seeks to establish is not firm or
certain knowledge (and thus not the sort of knowledge that the dogmatists believe
themselves to have achieved). Rather, it is revisable knowledge based on the best
evidence we have. Should new evidence show that our knowledge is inadequate—
and it is perfectly conceivable that such new evidence can be found—our knowledge
should be revised. Sanchez’s skepticism becomes, according to this interpretation,
a methodical skepticism in the sense that, although we assume that knowledge is
achievable, we never assume it to be definitely achieved but always continue to
question and test it. In allowing for revisable knowledge, Sanchez’s skepticism would
262 SKEPTICISM
be distinct from the radical Pyrrhonian skepticism that doubts that knowledge is
achievable at all. If this interpretation is correct, Sanchez becomes, thus understood,
a profoundly modern thinker.
What makes this view nevertheless Academic, in Limbrick’s interpretation, is the
idea that we base our knowledge claims on the Academic notion of what is persuasive
or probable (probabile): we are justified in believing ourselves to know something if we
have better reasons to believe it than not to believe it.7 On the basis of this justification,
the skeptic of this ilk believes that the truth of her beliefs is more likely (probabilius)
than their falsehood. Thus, while a skeptic of this sort does not commit herself to
any firm or certain knowledge, she nevertheless believes that it is possible to achieve
knowledge in the sense of a set or system of justified beliefs that remains open to reform.
While this is doubtless a picture attractive to us moderns, it seems to me that
we have no good reason to attribute it to Sanchez.8 Let us first consider two pieces
of evidence adduced by Limbrick (1988: 78) for her claim that Sanchez was an
Academic skeptic: he signs off a letter to the famous Renaissance mathematician
Clavius with “Carneades philosophus,” using the name of one of the most famous
Academics; and in his commentary on Galen’s De pulsibus ad tyrones, he claims
to withhold judgment “Academicorum more.” For this to count as evidence for
a Sanchezian Academic skepticism, we would have to assume that Renaissance
philosophers clearly distinguished between Pyrrhonian and Academic skepticism.
However, as Charles Schmitt (1972: 7, 14) has shown and as the studies of
Emmanuel Naya (2009: 24) have confirmed, they often used the words “skeptic,”
“Academic,” and “Pyrrhonian” more or less synonymously. Thus, these two pieces
of evidence show only that Sanchez considered himself a skeptic. Yet they do not
show that he considered himself an Academic, as opposed to Pyrrhonian, skeptic.
In That Nothing is Known Sanchez uses the term “persuasive” or “probable”
(probabile) only three times. As I have argued elsewhere (Caluori 2007: 34–36),
he nowhere uses it unambiguously in the epistemic sense relevant to Limbrick’s
interpretation. Moreover, if Sanchez wanted to establish a science based on the
notion of what is persuasive or probable, we would expect him to explain this
notion and how it forms the foundation for his new science. Yet none of this can be
found in his writings. So it seems to me that we have no good reason to believe that
the notion of the probabile played a key role in Sanchez’s skepticism.
Finally, we may be inclined to believe that Sanchez is about to establish a
foundation for a new (revisable) science because he refers four times to a method of
knowing (methodus sciendi) as if he was dealing with it in another book.9 At the end
of That Nothing is Known, for example, he states:
If you know anything, teach me. For I shall be very grateful to you. In the
meantime, I prepare myself for the examination of things and I shall put forward
in another book whether anything is known and [if anything is known] how, a
book in which I shall expound, as far as human frailty allows, the method of
knowing. Farewell. (QNS 245, LT 290)
No such book is extant nor do we have any evidence that Sanchez ever wrote it. It
is thus impossible to prove that such a book would (or would not) have contained
a method of knowing that would allow us to build a new and revisable science. The
best information we possess about Sanchez’s project is contained in the passage just
quoted. But here Sanchez does not commit himself to the view that anything can be
known. His promised exposition of the (or a) method of knowing presupposes the
settling of the question of whether anything can be known. For if nothing can be
known, developing a method of knowing is quite pointless.
This passage tells us, however, that Sanchez is not giving up his pursuit of truth.
He promises to continue his search for it by preparing himself for the examination
of things. In the same vein, he states in the address to the reader:
Yet I will not at all promise you the truth because I am ignorant of it—as of
everything else. Nevertheless, I will pursue it with all my power and you for your
part shall pursue it. . . . Yet do not expect ever to capture it. . . . Let the chase
suffice for you, as it does for me. For this is my aim and my end—an end you too
must aim for. (QNS xiii, LT 170)
Now compare this passage to how Diogenes Laertius describes the Pyrrhonists: “All
these were called Pyrrhonists after the name of their master, but also . . . seekers
because they were ever seeking the truth; skeptics [i.e. inquirers] because they were
always looking for a solution and never finding one” (DL IX 69–70). Sanchez’s
description of his search for the truth is clearly modeled on the basis of what he
knows about ancient Pyrrhonism.10 Sanchez is seeking the truth without ever having
found it, and while he does not claim that it is impossible to find the truth, he is not
certain that it will ever be found and cautions expectations accordingly.
3 AGAINST ARISTOTELIAN DOGMATISTS

Aristotelians dominated Western universities throughout the sixteenth century
(see Schmitt 1983). In fact, as Charles Lohr has shown, the number of Latin
commentaries on Aristotle composed between 1500 and 1650 “exceeds that of
the entire millennium from Boethius to Pomponazzi” (1974: 228). Aristotelianism
was not, of course, a unified school or doctrine at the time (nor at any other), and
there were many different strands, schools, and traditions.11 For our purposes,
however, it will be sufficient to consider the epistemological consequences of
two competing metaphysical views that were popular among the Aristotelians of
Sanchez’s time12: first, the nominalist view in the tradition of Ockham. Nominalists
claim that only particulars (as opposed to universals such as species) possess real
existence; second, the realist so-called species theory that goes back to Thomas
Aquinas and regained popularity with the revival of Thomism in the sixteenth
century.13 Sanchez argues against them separately. His conclusion, however, is in
both cases the same: if the corresponding metaphysical view is true, nothing can
be known.
Before discussing these two views, the following should be noted: when
Aristotelians talked about knowledge (scientia), they had in mind scientific
knowledge—the sort of knowledge Aristotle discusses in his Analytica Posteriora.
Accordingly, Sanchez’s target is scientific and not ordinary knowledge. I will come
264 SKEPTICISM
back below to the relation of these two kinds of knowledge and what this means for
Sanchez as a skeptic.
Let us now look at Sanchez’s critique of the knowledge claims of the nominalists.
In order to show some of the difficulties into which nominalists run, Sanchez adopts,
for the sake of argument, a nominalist position. He states:
Do you expect even further proof of our ignorance? I will provide it. You have
already seen the difficulties with species. But you will concede that there is no
knowledge of particulars as they are infinite. But there are no species—or, in any
case, they belong only to the imagination. (QNS 80, LT 212–213)
Sanchez uses a problem (aporia) already stated in Aristotle’s Metaphysics and turns
it into a skeptical argument: we can have scientific knowledge (scientia) only of
universals. But there are no universals. Therefore, there is no scientific knowledge
(see Met. 999a: 24–b3).14 Now, a nominalist may be unimpressed by this argument
because she does not deny the existence of universals tout court but only the real
existence of universals. So she would reply to Sanchez that there are universals,
namely, species, but that these species have no real existence. Instead, they exist
only as mental entities. To which Sanchez replies:
You will say that you are not studying particulars, as they are not objects of
scientific knowledge (scientia), but rather universals such as human being,
horse, and so on. This is indeed the case and I have already said it before: your
knowledge is knowledge, not of real human beings, but of a fiction that you
invent. Hence, you know nothing. (QNS 46, LT 196)15
According to the nominalist position understood thus, universals are mental entities
with no real existence. But if we understand them in this way, then according to
this argument, mental entities, rather than really existing things, become the objects
of knowledge. But if so, then we are incapable of acquiring any knowledge about
things in the external world. Nothing is known if the nominalists are right.16
Now, in order to avoid the conclusion that species are purely mental (and
perhaps even fictitious) entities, we may want to give up nominalism and claim that
species are in some way part of extramental reality. The so-called species theory
was perhaps the most sophisticated version of this view in the sixteenth century.
According to Aquinas, with whom the species theory originates, all our knowledge
derives from the senses. Yet even though all sensory objects are particulars, we are
capable of arriving at the cognition of universals—and this in roughly the following
way. When I see a particular human being, the sensible part of my soul receives
through my eyes an impression of this particular human being. What the soul thus
receives is a sensible form. Now my active intellect abstracts the universal form
(i.e., the intelligible species) from the sensible form. As a result of this process of
abstraction, the soul possesses an intelligible species, which is the essence of the
object: in our example the essence human being.17 What is crucial for our purpose
is the view that the species (the universal) is not simply a representation of the
external object: it is not an internal object somehow representing the essence of the
external object. Rather, it is formally identical with it. This makes it possible for the
intellect (or so defenders of this view claimed) to grasp the essence of the external
object itself. In this way, the species theory is a realist theory and distinct from the
nominalist view considered above.
Sanchez’s first argument against the species theory is a classic Pyrrhonist attack
on the reliability of the senses (through which all knowledge is acquired according
to the Aristotelians). When examining the reliability of sense perception, Sanchez
quite reasonably assumes that, in order to have normal sense perception (i.e., sense
perception presenting the things the way they are), certain criteria must be fulfilled.
One of them concerns the medium through which we perceive the object. For
depending on the medium, things appear differently to us. So the medium must be
such as to allow for normal sense perception. But how can the mind be certain that
the medium is such as to allow for normal sense perception? Well, certain criteria
will need to be fulfilled so that the medium is such as to allow for normal sense
perception. How does the mind know that these criteria are fulfilled? Sanchez sums
up: “There is no conclusion, only endless doubt” (QNS 172, LT 254).
Even if we assume for the sake of argument that sense perception works properly
and that we possess normal sense perception, Sanchez sees further problems with the
species theory. I will restrict the discussion to one of them. Defenders of the species
theory claim that all knowledge is based on sense perception. What our senses
perceive, however, are only accidental features of particulars—not their essence.
Accordingly, the sensible form must exclusively consist of accidental features. But
if this is the case, the essence cannot get abstracted from the sensible form. For the
essence is neither part of the sensible form nor can it be inferred from a thing’s
accidental features:
The whole thing would be bearable, if we had images from the senses of all things
we wish to know. Yet the contrary is the case: we do not possess images of the
most important things but only of accidental features which do not, as they say,
contribute anything to the essence of a thing. But the essence is the source of true
knowledge; and accidental features are the most useless of all things. (QNS 127,
LT 237)
Sanchez here assumes that species are only images of extramental objects and
implicitly denies that species are formally identical with the objects whose species they
are.18 Now, even if the species theorist rejects this representationalist interpretation,
the main argument still stands. The active intellect is incapable of deriving the
essence of a thing from sensible forms. For sensible forms, being the only material
from which the active intellect can abstract intelligible forms, exclusively consist of
accidental features. And no matter how many accidents of a thing we are aware of,
by no process of abstraction can an essence be distilled out of them. Hence, Sanchez
concludes, if the species theory is true, we will not be capable of cognizing essences.
And since essences, according to the Aristotelians, are the only objects of scientific
knowledge of the world, nothing can be known if the realists are right.
I have only discussed two of the arguments Sanchez presents against the dominant
Aristotelian positions of his day. But I hope Sanchez’s strategy has become clear.
He shows that both nominalists and realists fail to ground their knowledge claims
266 SKEPTICISM
within their own respective dogmatic frameworks. In his arguments, Sanchez only
uses premises either granted by his opponents or derived from claims granted by
them. Thus, in line with both forms of ancient skepticism, he commits himself
neither to nominalism nor to realism nor to any other dogmatic framework.
4 SANCHEZ AND CARTESIAN SKEPTICISM

In a classic paper, Myles Burnyeat shows in what ways Cartesian and ancient
skepticism are distinct. Ancient skeptics doubted that we possess knowledge about
the world as it really is. No ancient philosopher, however—and thus no ancient
skeptic—knows “the problem of proving in a general way the existence of an
external world” (Burnyeat 1982: 19). Descartes’s hyperbolic doubt is more radical
in that it presents, by means of the evil demon, precisely this problem. As Burnyeat
shows, this radicalization of skepticism leads to its refutation by means of the cogito.
Moreover, Burnyeat follows Bernard Williams (1978) in arguing that Descartes’s
project is one of pure enquiry. Ancient skepticism, by contrast, is a way of life and
thus an eminently practical affair. Sextus tells us that suspension of judgment leads
to tranquility and thus to a sort of life that appears desirable. Hence, the ancient
skeptic offers what his dogmatic opponents promise but cannot deliver: a good and
happy life. Cartesian skepticism, on the other hand, is purely theoretical. Descartes
does not show us a way to a good and happy life achievable through skepticism.
Burnyeat thinks that these two points (i.e., doubting the existence of the external
world and seeing skepticism as a purely theoretical affair) are related. Ancient
skepticism did not question, Burnyeat claims, that one can walk around in the
world “because it was in fact entirely serious about carrying skepticism into the
practical affairs of life” (Burnyeat 1982: 40). Descartes, by contrast, is in a position
to question the existence of the external world because his skepticism is methodical
and divorced from practical concerns.
We have seen in Section 3 that the sort of knowledge whose attainability Sanchez
doubts is scientific knowledge (scientia). So it may seem that his skepticism is also
of a purely theoretical kind, detached from one’s life. But Sanchez’s doubt is not
methodical and not part of a larger project of pure enquiry. It rather is a genuinely
skeptical doubt as I hope to have shown in Section 2. What is more: Sanchez was
a medical practitioner at the Hôtel Dieu, the hospital of Toulouse, and, later in
his life, became a professor at the faculty of medicine at the town’s university. So
we need an account of Sanchez’s attitude toward practical issues. Since his doubt
was not methodical, we have to ask how he could be a practitioner and teacher of
medicine. For we may be inclined to assume that these activities presuppose the
acknowledgment of a body of scientific knowledge.
The answer to this question can be found if we heed the fact that Sanchez’s
skepticism was influenced by the ancient medical school of empiricism (see Caluori
2007).19 Ancient empiricists practice medicine, but they do so without reference to
any theoretical framework (see Frede 1987). They reject medical theories, the use
of formal inference, and the postulation of entities that (according to their dogmatic
opponents) can be grasped only by reason. Instead, they rely on experience for their
practice. A medical empiricist acquires experience in relying on what she herself
perceives, on the carefully checked reports of other people, and on the transition
from cases she is familiar with to cases that are similar to them. In keeping with their
skeptical attitude, they do not believe themselves, however, to possess a rational
justification for their procedures. Rather, it appears to them that, if they proceed in
this way, they can practice medicine quite successfully. Moreover, the procedures
they use seem to them to be the same as those that people use in their ordinary
lives. Ordinary people live their lives without relying on theories and without the
postulation of the existence of abstract entities.
Sanchez makes very similar claims. Like the ancient empiricists, Sanchez does not
doubt that ordinary people possess experiences and lead their lives on the basis of
them. For example, he does not doubt that the farmer can distinguish his donkey
from the neighbor’s ox (QNS 171, LT 254). And like the ancient empiricists he
doubts the existence of theoretical entities: “For who could understand non-existent
things? Thence come Democritus’s atoms, Plato’s Forms, Pythagoras’s numbers, and
Aristotle’s universals. . . . In this way they [i.e. the dogmatists] entrap the unwary,
claiming to have discovered unknown things and the secrets of Nature” (QNS vii,
LT 168). In another passage he states: “Much experience, then, makes a person
both learned and wise. This is why old people are more learned, at least in terms
of experience, and thus better adapted to the conduct of human affairs than the
young” (QNS 225, LT 281). We trust experts in such fields as agriculture, navigation,
commerce, and medicine because of their experience. Experience appears to be a
tremendously successful guide to the conduct of human affairs.
Sanchez was well acquainted with ancient empiricism. One main source for this
view is Galen—in the sixteenth century the medical authority par excellence. Galen’s
work includes a number of treatises and passages discussing medical empiricism.20
It is true that Renaissance Galenism was rather dogmatic (Gilbert 1960; Schmitt
1985). But this does not exclude the possibility that a Renaissance skeptic could
find in Galen a view opposed to dogmatic Galenism—a view that would allow him
not only to live a skeptical life but also to practice and teach medicine without
believing in the truth of any dogmatic theory.21 It seems to me that Sanchez was such
a Renaissance skeptic, a skeptic living a skeptical life and practicing and teaching
medicine in a non-dogmatic way.
Sanchez thus shares with the ancient Pyrrhonists the view that a skeptic can live
his skepticism and even practice an art without committing himself to any theory.
Given the practical side of his skepticism, it is perhaps also unsurprising that he
does not doubt the existence of the external world, the world we live and act in.
However, there is one crucial difference between Sanchez and ancient Pyrrhonism.
Sanchez nowhere promises that his skepticism may lead to tranquility or happiness.
His view on human life is less hopeful, and he often complains about the misery
of human life and laments about our epistemic situation. What motivates him
nevertheless to pursue his theoretical enquiries is not the hope of tranquility but
rather the desire to understand.
268 SKEPTICISM
NOTES
1. Lyon 1581, Frankfurt 1618, Toulouse 1636, and Rotterdam 1649.
2. Also Delassus, Sanchez’s first biographer and editor of the Toulouse edition of That
Nothing is Known, calls Sanchez a Pyrrhonist. See Limbrick (1988: 68 n. 6).
3. This interpretation has since been refuted, as far as the major Academic skeptics of
the Hellenistic period are concerned. See Frede (1997).
4. See also Photius, Bibl. 212, 169b. Another influential source for this view was
Augustine, who was widely read in the Renaissance. See Augustine, C. Acad. 2.11.
5. A very similar statement is attributed to Arcesilaus by Cicero at Acad. I 12, 45.
6. When quoting Sanchez, I refer to the paragraph number of the latest critical edition
of the Latin text by Howald, Caluori, and Mariev (2007) and to the page number
of the English translation by Thomson in Limbrick and Thomson (1988). All
translations are my own.
7. It should be noted that Limbrick’s understanding of Academic skepticism is
compatible only with the Metrodorian-Philonian line and stands in contrast to the
classical skepticism of Clitomachus and Carneades. For more details, see Frede
(1997) and Brittain (2001).
8. I discuss this in detail in Caluori (2007).
9. In addition to the passage quoted below, see QNS 26 (margin), 215 (margin), and
221 (LT 278). The edition and translation of Limbrick and Thomson does not
render Sanchez’s marginal notes.
10. As far as his sources are concerned, we do not know whether Sanchez knew Sextus’s
writings, but he explicitly refers to Diogenes Laertius and to Plutarch’s works
Adversus Colotem and Lucullus (see Sanchez’s marginal notes to QNS 24, 34, 130,
and 164). In addition, Sanchez knew Galen, for which see below.
11. For the variety of Renaissance Aristotelianisms, see Kristeller (1974: 30–49) and
Schmitt (1983: 10–33).
12. In this part, I follow Howald, Caluori, and Mariev (2007: cxxiv–cxlviii).
13. For the renewed interest in Thomas Aquinas in the sixteenth century, see Spruit
(1995: 93–158).
14. In Aristotle it is, of course, an aporia to be resolved.
15. Sanchez’s argument goes back to Duns Scotus (Commentary on Aristotle’s
Metaphysics, lib. 7, Quaest. 13).
16. Sanchez also argues against an alternative modification of the original argument a
nominalist could present. For she could claim that there is knowledge of particulars.
For a discussion of Sanchez’s argument against this, see Howald, Caluori, and
Mariev (2007: cxxxii–cxxxvi).
17. For more details on Aquinas’s theory, see Kretzmann (1993: 138–146).
18. This argument had already been developed by Petrus Johannis Olivi and William
Crathorn in the thirteenth and fourteenth centuries. See Perler (2003).
19. Ancient empiricism was related to Pyrrhonism. See James Allen’s chapter in Part I of
this book.
20. See Galen’s On the Sects for Beginners; On Medical Experience; An Outline of
Empiricism (all in Frede 1985); and On Therapeutic Method (in Hankinson 1991).
21. For textual parallels between Sanchez and Galen on empiricism, see Caluori (2007).
REFERENCES
Bayle, Pierre. 1740. Dictionnaire Historique et Critique. Amsterdam: P. Brunel.
University Press.
Missed,” The Philosophical Review 91: 3–40.
Caluori, Damian. 2007. “The Skepticism of Francisco Sanchez,” Archiv für Geschichte der
Philosophie 89: 30–46.
Frede, Michael (ed.). 1985. Galen: Three Treatises on the Nature of Science. Indianapolis:
Hackett.
Frede, Michael (ed.). 1987. “The Ancient Empiricists.” In his Essays in Ancient Philosophy,
Frede, Michael (ed.). 1997. “The Sceptic’s Two Kinds of Assent.” In M. Burnyeat and
M. Frede (eds.), The Original Sceptics: A Controversy, 127–151. Indianapolis: Hackett.
Gerhardt, Carl Immanuel. 1859. Leibnizens Mathematische Schriften, vol. 4. Halle: Verlag
H. W. Schmidt.
Gilbert, Neal. 1960. Renaissance Concepts of Method. New York and London: Columbia
University Press.
Hankinson, R. J. (ed.). 1991. Galen on the Therapeutic Method. Oxford: Oxford University
Press.
Hegel, G. W. F. 1979. Vorlesungen über Geschichte der Philosophie. Werke in zwanzig
Bänden. Auf der Grundlage der Werke von 1832–1845. Neu edierte Ausgabe. Redaktion
Eva Moldenhauer und Karl Markus Michel, Band 19. Frankfurt am Main: Suhrkamp.
Howald, Kaspar, Damian Caluori, and Sergei Mariev (eds.). 2007. Franciscus Sanchez.
Quod nihil scitur. Dass nichts gewusst wird. Introduction, Latin text, and German
Translation. Hamburg: Meiner Verlag.
Jolly, Claude (ed.). 1990. Gabriel Naudé: Advis pour dresser une bibliothèque. Reproduction
de l’édition de 1644. Paris: Alain Baudry et Cie.
Kretzmann, Norman. 1993. “Philosophy of Mind.” In N. Kretzmann and E. Stump (eds.),
The Cambridge Companion to Aquinas, 128–159. Cambridge: Cambridge University
Press.
Kristeller, Paul Oskar. 1974. Humanismus und Renaissance I: Die antiken und
mittelalterlichen Quellen. Munich: Fink.
Limbrick, Elaine and Douglas Thomson (eds.). 1988. Francisco Sanches: That Nothing is
Known. Cambridge: Cambridge University Press.
Limbrick, Elaine. 1988. “Introduction.” In E. Limbrick and D. Thomson (eds.), Francisco
Sanches: That Nothing is Known, 1–88. Cambridge: Cambridge University Press.
270 SKEPTICISM
Lohr, Charles. 1974. “Renaissance Latin Aristotle Commentaries: Authors A-B,” Studies in
the Renaissance 21: 228–289.
Naya, Emmanuel. 2009. “Renaissance Pyrrhonism: A Relative Phenomenon.” In G. Paganini
and J. R. M. Neto (eds.), Renaissance Skepticisms, 13–32. Dordrecht: Springer.
Perler, Dominik. 2003. “Wie ist globaler Zweifel möglich? Zu den Voraussetzungen des
frühneuzeitlichen Aussenwelt-Skeptizismus,” Zeitschrift für philosophische Forschung
57: 481–512.
Schmitt, Charles. 1972. Cicero Scepticus: A Study of the Academica in the Renaissance. The
Hague: Martinus Nijhoff.
Schmitt, Charles. 1983. Aristotle and the Renaissance. Cambridge, MA: Harvard University
Press.
Schmitt, Charles. 1985. “Aristotle among the Physicians.” In A. Wear, R. K. French, and
I. M. Lonie (eds.), The Medical Renaissance of the Sixteenth Century, 1–15. Cambridge:
Spruit, Leen. 1995. Species Intelligibilis. From Perception to Knowledge. Vol. 2: Renaissance
Controversies, Later Scholasticism, and the Elimination of the Intelligible Species in
Modern Philosophy. Leiden: Brill.
Williams, Bernard. 1978. Descartes: The Project of Pure Enquiry. London: Penguin Books.
PART THREE
Modern Skepticism
Introduction
Modern Skepticism
BARON REED
The beginning of the second great flowering in the skeptical tradition coincides with
two of the major transformations in European history. First, the development of a
globalized economy brought the exchange across vast distances of not only material
goods but also people and ideas. Many Europeans reacted with the sort of self-
interested chauvinism that has always tended to shape human interactions, but the
most perceptive recognized the intellectual, ethical, and spiritual challenges posed
by these very different forms of human life. Michel de Montaigne, for example, was
prompted by his encounter with a group of Tupinambá people, who had traveled
from Brazil to France, to reconceive what it means to be barbaric or civilized and, a
century later, Pierre Bayle punctured the almost universally accepted notion that an
atheistic society couldn’t be moral by pointing to the ancient civilization of China.1
The increasingly close contact with the wider world served as an implicit critique of
European beliefs, values, and customs.
Second, the Reformation—together with the resistance it provoked—led to a
profound change in the way moral and intellectual authority functioned in society.
Religious disagreements became far more widespread, and it was no longer possible
to simply treat unorthodox ideas as impermissible heresies. These disagreements
also became deeper from an epistemological point of view, as Catholics and
Protestants no longer shared a commitment to the same way of thinking about how
these differences could be resolved. These theological conflicts, along with the rise
of a new way of engaging in natural philosophy, prompted a widespread concern
with intellectual methodology: what are the proper sources of human knowledge?
Martin Luther, Francis Bacon, René Descartes, and John Locke, among many
others, gave widely diverging answers to the question.
The ancient skeptical texts were among the intellectual resources mustered to
come to terms with these transformations.2 Arguments drawn from the ancient
skeptics, particularly those having to do with the problem of the criterion—that
is, determining which criterion is the correct one to resolve a disagreement and to
judge what is true—were used by both sides in the Reformation dispute, and they
also shaped the development of epistemology in early modern philosophy.3 More
broadly, a skeptical outlook became a prominent thread in European culture, visible
not only in the essays of Montaigne and the philosophical writings of Descartes but
also, for example, in the plays of Shakespeare and the fiction of Cervantes.4
Modern skepticism is often thought to differ from ancient skepticism in a
number of ways. Some scholars have argued that modern skepticism targets
274 SKEPTICISM
knowledge, while the ancients directed their arguments against belief. This picture
is perhaps best exemplified in Bernard Williams’s characterization of Descartes’s
methodological use of doubt as a crucial first step in a project of “Pure Enquiry,”
intended to culminate in “a systematic vindication of knowledge” (1978: 20).5
Before beginning that project, Descartes adopts a “provisional moral code,” so that
he will not be indecisive in his actions while his judgments are in doubt.6 He then
advances skeptical arguments, “not copying the sceptics, who doubt only for the
sake of doubting and pretend to be always undecided,” but only “to reach certainty”
(CSM I 125; AT VI 29).7 Because these skeptical considerations are being raised
only when they will not threaten the ordinary course of life, and because Descartes
thinks he has a decisive response to them, the doubts he is willing to entertain can
become as radical as he likes.8 What had been a way of life in the ancient world was
now merely an intellectual problem to be solved.
Something like this attitude, though in a more pessimistic form, is also visible
in some of Hume’s remarks about skepticism. In his famous characterization of
Berkeley’s arguments as skeptical, he says that this can be seen from the fact “that
they admit of no answer and produce no conviction. Their only effect is to cause
that momentary amazement and irresolution and confusion, which is the result
of skepticism” (Hume EHU XII 122 n. 1). So, far from having any relevance to
our behavior, Hume instead holds that “the great subverter of Pyrrhonism or the
excessive principles of scepticism is action, and employment, and the occupations of
common life” (EHU XII 126). To rid himself of his “philosophical melancholy and
delirium,” he says,
I dine, I play a game of back-gammon, I converse, and am merry with my
friends; and when after three or four hour’s amusement, I wou’d return to these
speculations, they appear so cold, and strain’d, and ridiculous, that I cannot find
in my heart to enter into them any farther. (T 1.4.7 9)
Although there is an element of truth in this picture, it leaves out much of what
is most interesting in modern skepticism. It overlooks the way in which Descartes
is concerned not only to build a body of certain knowledge but also to reform
the minds of those who follow his method.9 Descartes would not have recognized
a clear distinction between the theoretical and the practical, as he thought the
experience of certainty itself would have a profoundly transformative effect upon
those who follow his course of meditation. The evil demon hypothesis is designed
to position his readers precisely so that the first-hand experience of certainty will
have the greatest impact. For this reason, the meditator doesn’t simply note the
possibility that he is being deceived; rather, he “will suppose” or “pretend” that
this is actually happening, so that “the weight of preconceived opinion is counter-
balanced” (CSM II 15; AT VII 22).
This picture also ignores Hume’s efforts to articulate a way of proceeding in
philosophy despite his skeptical worries. They are not abandoned, as idle and practically
irrelevant; rather, he says, “A true sceptic will be diffident of his philosophical doubts,
as well as of his philosophical conviction” (T 1.4.7 14). We would do well, he thinks,
to adopt “a more mitigated scepticism or academical philosophy,” which brings with
BARON REED 275
it “a degree of doubt, and caution, and modesty, which, in all kinds of scrutiny and
decision, ought for ever to accompany a just reasoner” (EHU XII 129). Skepticism,
understood in this way, is not subverted by “the occupations of common life” but
rather ought to be inextricably bound up with them.
And, finally, this picture misses the wonderful variety of skeptical arguments and
perspectives found in other modern philosophers. Although many of them were
deeply influenced by the Academics and the Pyrrhonists, there turned out to be no
single way of being a modern skeptic. Modern skepticism did not comprise a way of
life as Pyrrhonism did; there was no modern Sextus, laying out a common goal and
a shared sense of what the implications of the skeptical arguments might be.10 What
we have instead is an endlessly fascinating, many-sided dispute by philosophers who
were intrigued, worried, and sometimes delighted by human cognitive limitations.
This part of the book, then, reflects the full diversity of purpose and perspective
in modern skepticism, including essays on both famous and less well-known
philosophers of the modern era.
If ancient and modern skepticism cannot be distinguished from each other in
virtue of their scope or their practical relevance, perhaps the best way to understand
their differences is to look at the changing circumstances in which modern skepticism
developed. We have already seen two vastly important transformations—the move
toward a fully globalized world and the Reformation—but it is also worth noting
another broad difference from the ancient world. Skepticism changed in the modern
period, in large part, because dogmatism changed. With the rise of the major
monotheistic religions, different dogmatic worldviews came to be incompatible
rivals with each other. Within Christianity, in particular, salvation was thought to
depend on having the correct theological beliefs. This had enormous implications
not only for the individual believer but also for the shape of society at large. The
consequences of dying outside the faith were terrible enough, Augustine thought, to
motivate the use of coercion in matters of religion, which could be justified by Jesus’s
use of the parable of the banquet, in which the master of a house tells his servant,
“Go out into the highways and hedges and compel them to come in that my house
may be filled” (Luke 14:23). Bayle’s Philosophical Commentary on Augustine’s view
is the most direct attempt to push back against this sort of aggressively intolerant
dogmatism.11 Bayle also used skeptical arguments to reconsider, in broader terms,
the relation between religion and morality. Where we should be able to expect
perfection in the moral order, given the nature of God, we find instead depravity
and suffering. And where we should be able to expect virtue in the actions of the
faithful, and vice in the behavior of the faithless, we instead find virtuous atheists
and vicious Christians. Bayle concludes that “it is a mistake to believe that the moral
practice of a religion corresponds to the doctrines of its confession of faith.”12
Skeptical arguments were also brought to bear on religion through a consideration
of “superstition,” usually in the guise of the ancient pagan religions. One of the chief
difficulties in grasping the nature of modern skepticism is that much of this work is
intentionally ambiguous. Philosophers, with varying degrees of sincerity, articulated
criticism of the irrational nature of paganism and “idolatry,” before remarking
that this sort of criticism of course did not apply to “the true religion” (Bayle
276 SKEPTICISM
1965: 401). Where skeptical arguments do make contact with Christianity, Bayle
says that “one learns of the excellence of faith” and “of the necessity of mistrusting
reason and of having recourse to grace” (1965: 435). This apparent move toward
fideism can be found in other early modern thinkers as well, including François de
La Mothe Le Vayer, Pierre-Daniel Huet, and Johann Georg Hamann.13 Scholars are
divided on how to understand Bayle’s attitude toward religion, with some taking the
fideist remarks at face value and others arguing that those are intended to deflect
unwelcome attention from “libertines” and freethinkers.14 Contemporaries of the
early modern skeptics tended to view their fideism with suspicion; as one of them
remarked, “That Method of attacking, when one seems to defend, is not new.”15 As
the grip of orthodoxy relaxed later in the modern period, the need for a fideistic
shield began to lessen.16 Hume’s criticisms of superstition, for example, are less
ambiguous than Bayle’s, though even Hume couches some of his arguments in the
form of an imagined dialogue set in ancient Athens.17
At the beginning of the early modern period, orthodoxy in religion was intimately
bound up with orthodoxy in philosophy.18 Skeptical arguments proved to be a
useful tool for undermining the latter as well as the former. Both Gassendi and
Descartes used skepticism as a way of circumventing the well-entrenched Aristotelian
Scholasticism, though Gassendi had a greater sympathy for skepticism in its own
right.19 The usual Scholastic practice was to recount what had been said by earlier
philosophers on a particular question before presenting one’s own view. Raising
skeptical concerns from the outset allowed Descartes to avoid being needlessly
involved in a thousand controversial matters before getting to a fundamentally new
way of doing philosophy. For this reason, he begins the Meditations by attacking
the Scholastics’ empiricist assumption that “whatever I have up till now accepted
as most true I have acquired either from the senses or through the senses” (CSM II
12; AT VII 18). In doing so, he was able to set aside the entire Scholastic tradition
while also posing an epistemological challenge for any potential rival view: if only
the Cartesian theory can overcome skepticism, then Cartesianism must be the best
of all philosophies. Other philosophers, of course, did not think that Descartes was
ultimately successful in meeting this challenge. Gassendi, Huet, Foucher, and Bayle
all used skeptical arguments to attack various aspects of Cartesianism.20
Skeptical considerations played a crucial role in modern philosophy, not only
negatively by shaping criticism of Scholasticism, but also positively through the
development of a new account of perception. Reflection on the Ten Modes of
Pyrrhonism convinced many modern philosophers that material objects were not
really as they seemed to our senses. Sensible qualities like colors and odors were
thought to be mental items that represented objects without resembling them, while
other qualities we could perceive, like motion and spatial extension, existed just as
they appeared to us. This view, espoused by Galileo, Descartes, Boyle, Locke, and
others, proved to be unstable. As Bayle noted, “If the objects of our senses appear
colored, hot, cold, odoriferous, and yet they are not so, why can they not appear
extended and shaped, in rest and in motion, though they are not so?”21 In this
way, the modern account of perception, which Reid called the “common theory of
ideas,”22 gave rise both to a physics that fruitfully narrowed its focus to the qualities
BARON REED 277
that most easily permit mathematical description and to the philosophical problem
of skepticism about the external world. Berkeley and Reid tried to solve that problem
by denying, in different ways, that perception is deceptive; where Berkeley rejected
the distinction between our ideas and the objects they are purportedly of, Reid
argued that we should abandon the theory of ideas altogether and take ourselves
to have a direct awareness of mind-independent objects.23 Kant, by contrast, was
willing to grant that our perceptions may not give us an unproblematic access to the
nature of mind-independent objects as they are in themselves, but he also argued
that our awareness of ourselves as determined in time is possible only if we are
aware of something relatively permanent in our perceptual experiences, which can
serve as a frame of reference for the temporal order of those experiences.24
Skepticism had a significant impact, not only on the work of philosophers who
tried to refute it, but also on the development of some of the major philosophical
systems of the modern period. Spinoza followed Descartes in placing God at the
center of his epistemology, though not by invoking a divine guarantee for our
cognitive faculties; rather, the certainty of which we are capable arises only in the
relatively rare cases when our ideas can be properly deduced from God’s essence.25
Hume’s efforts at systematic philosophy turn toward naturalism as the only possible
way of moving forward, in light of the unanswered skeptical challenges he has
raised. Both aspects of his philosophy, the skepticism and the naturalism, manifest
his inheritance from ancient skepticism.26 Both Hegel and Nietzsche also took
skeptical modes of thought to be essential to philosophical theorizing; where Hegel
thought skepticism was itself an integral part of rationality, it is for Nietzsche bound
up with the “revaluation of all values.”27
Skepticism continued to be of central importance in the twentieth century.
Bertrand Russell, early in his career, attempted to solve the problem of the
external world by showing how everyday objects, like trees and books, are logically
constructed out of sense-data. Our knowledge of the former is supposed to be
secured in virtue of our unproblematic knowledge of the latter.28 Although G. E.
Moore also incorporated sense-data into his epistemology, he did not regard that as
necessary for a refutation of skepticism. His famous “proof of an external world”
presented the skeptical challenge in the starkest possible terms: either you reject the
possibility of any sort of knowledge in light of your inability to tell whether you are
dreaming, or you acknowledge that you know you have hands and know you are
awake. No philosophical argument sufficient to establish the skeptical conclusion
could be more certain than your knowledge that you have hands, Moore contends,
and so we can confidently reject skepticism.29 Ludwig Wittgenstein thinks that
Moore’s way of dealing with the skeptic is implausible because it is unnatural—
Moore’s proof fails because he has taken the skeptic’s doubts seriously. But doubts
can be raised only in the scope of a “language-game,” where some things are taken
for granted. If doubting one thing means accepting another, though, it is impossible
for the sort of general skepticism Descartes raises to get a grip on us.30
As should be clear from the rich variety of ways in which it has been defended
and rejected, used as an instrument and embraced as an end, skepticism is one of
the most important facets of intellectual life in the modern period. Its influence
278 SKEPTICISM
can be traced in both the rise of secularism and in some of the religious reactions
to rationalist critique. Skeptical arguments have played a role in the development
of modern science, and skepticism has profoundly shaped every branch of
philosophy, from epistemology and metaphysics to ethics and political philosophy.
Encompassing everything from Descartes’s evil demon to Hume’s “just reasoner,”
modern skepticism provides an indispensable framework for understanding the past
several centuries and the world we inhabit today.
NOTES
1. See “Of Cannibals” in Montaigne (1958); for Bayle’s view of classical Chinese
philosophy, see Israel (2006: 644 and 2007: 7–8). See also Hazard (2013: ch. 1).
2. Cicero’s Academica was published with a commentary by Omer Talon in 1547; see
Schmitt (1972: 81). Sextus Empiricus’s Outlines of Pyrrhonism were printed by Henri
Estienne in 1562; see Popkin (2003: 18). Popkin (2003: 6) also notes that Savonarola
encouraged the production of a Latin edition of Sextus’s works in 1494 while in
the midst of his disagreement with the Pope, though Sextus did not become widely
known until Estienne’s publication of the Outlines. More generally, Popkin (1960,
1979, 1980, and 2003) argues that the revival of skepticism in the early modern
period is largely due to the rediscovery of Sextus. See Maia Neto’s chapter on Popkin
in this book for an account of the profound impact his work has had on the study of
modern skepticism. See Maia Neto (1997 and 2014) and Michael Hickson’s chapter
on Huet and Foucher in this book for an argument that the influence of Academic
skepticism was greater than has been generally recognized.
3. On the uses of the criterion argument in religious disagreement, see Popkin (2003),
ch. 1, 4. Descartes says, in the Second Replies, that he has “reheated this cabbage,”
namely, the arguments of the ancient skeptics; see Descartes (CSM II 94). See Fine
(2000) on the relation between Descartes’s presentation of skepticism and ancient
skepticism. Arnauld famously raised the problem of the “Cartesian Circle,” which is
an instance of the problem of the criterion, in the Fourth Objections to Descartes’s
Meditations (CSM II 150; AT VII 214).
4. See, e.g., Bell (2002) and Rescher (1993) on Shakespeare and Cervantes, respectively.
5. Cf. Williams (1986: 118). See also Gaukroger (1995: 310–321): “Epistemological
doubt questions whether one’s beliefs amount to knowledge, whereas doxastic
doubt questions whether one is entitled to hold the beliefs one has” (311).
Gaukroger takes doxastic doubt to be a distinctive feature of Pyrrhonism (311),
while epistemological doubt is “something new” in Montaigne’s work (315) and
characteristic of Cartesian skepticism. See Fine (2000: 195–196) for a survey of
others who draw this sort of contrast between ancient and modern skepticism and
for a critical response to it.
6. Discourse on the Method, Part Three (CSM I 122; AT VI 22). In the Meditations, he
begins the project only after he has arranged for himself “a clear stretch of free time”
(CSM II 12; AT VII 18).
7. Diderot expresses a similar thought in his article on the “Pyrrhonic or Skeptical
Philosophy” in the Encyclopédie: “Doubt is the first step toward science or truth”
(1965: 294). (This passage was censored in the original version.)
BARON REED 279
8. Descartes describes the doubts as “exaggerated” (CSM II 61; AT VII 89), “very
slight and, so to speak, metaphysical” (CSM II 25; AT VII 36), and ones that “no
sane person” would ever have (CSM II 11; AT VII 16).
9. See Rorty (1986) and Hatfield (1986), who both emphasize the “meditational”
aspect of Descartes’s work.
10. It is important to keep in mind that ancient skepticism, taken as a whole, was not a
single way of life. The revival of Pyrrhonism occurred precisely because there was
widespread disagreement in the Academy over the legacy of Carneades and the
best way of conducting inquiry; see Brittain’s introduction to Cicero (2006). One
of the unstated aims of Sextus’s Outlines may well be to create a consensus about
skepticism that did not already exist.
11. See Bayle (2005), Robinson (1931: 72), and John Christian Laursen’s chapter on
Bayle in this book. For the intersection of skepticism and political thought in the
early modern period more generally, see Laursen and Paganini (2015).
12. “Jupiter,” remark D, in the Historical and Critical Dictionary (1965: 109). Cf.
“David,” where Bayle recounts the morally questionable exploits of David, “a
man after God’s own heart” (1965: 45), and asks, “is it not better to consider the
interests of morality rather than the glory of a particular person?” (1965: 63). For
Bayle’s reflections on the problem of evil, see “Manichaeans,” “Paulicians,” and the
“Third Clarification” in the Dictionary (1965: 144–153, 166–193, and 399–408).
For Bayle’s consideration of atheism, see his Various Thoughts on the Occasion of a
Comet (2000) and the “First Clarification” in the Dictionary (1965: 399–408). See
also Laursen’s article on Bayle in this book. As in other ways, Hume pushed to a
further extreme an attitude found in Bayle—in this case, the extrication of morality
from religion; as Don Garrett (2015: 309) notes, Hume’s “account of the basis of
morality makes no substantive appeal to religion.” An earlier manifestation of this
impulse can be found in the work of Pierre Charron; see Gianni Paganini’s chapter
on Charron in Part II of this book.
13. See the chapters in this book by Sylvia Giocanti on La Mothe Le Vayer, by Hickson
on Huet, and by Gwen Griffith-Dickson on Hamann.
14. See van der Lugt (2016: 37) for an overview of this dispute in scholarship on Bayle.
15. Durand (1730: 19). Durand says that it was employed by Carneades and Cicero,
among others, and he hints that Bayle may also be a practitioner of the method.
Cf. Allen (1964: 11), who says that, according to Voetius—the orthodox Dutch
theologian who pushed the University of Utrecht to condemn Descartes’s work—
“One of their cleverer tricks is to attack an atheist, fully expound his Lucianic
remarks, and then, after a tepid confrontation, depart, ‘leaving certain subtle
curiosities in the midst of things.’”
16. It should be noted, however, that the skeptical undermining of reason in defense
of religion continued to find adherents in later thinkers, including Hamann and
Kierkegaard. See Griffith-Dickson’s chapter on Hamann in this book and “Hume
and Kierkegaard” in Popkin (1980: 227–236). Popkin (1980: 236) says that
Kierkegaard’s religious outlook was significantly influenced by Hume via Hamann’s
reflections on Hume’s chapter on Miracles (EHU X).
17. EHU XI 104110. According to Garrett, “Hume saw his age as a battleground
between what he called ‘philosophy,’ on the one hand, and ‘superstition,’ on the
other” (1997: 7).
280 SKEPTICISM
18. Stillman Drake (2001) argues that Galileo’s troubles with the Inquisition were
caused by conflict with Aristotelian philosophers, rather than with theologians. See
also Voetius’s dispute with Descartes over the challenge Cartesianism poses to “the
traditional philosophy”; Descartes recounts this conflict in the “Letter to Father
Dinet” (CSM II 393–397; AT VII 596–603).
19. See especially Gassendi’s first published work, Exercitationes paradoxicae adversus
Aristoteleos, Maia Neto (2014: ch. 3), and Antonia LoLordo’s chapter on Gassendi
in this book. On Descartes, see Rorty (1986), Hatfield (1986), and David Cunning’s
chapter in this book.
20. See Watson (1987), LoLordo’s chapter in this book on Gassendi, Hickson’s chapter
on Huet and Foucher, and Laursen’s chapter on Bayle. See also Bayle’s entry on
“Pyrrho,” remark B, in the Dictionary (1965: 194–199).
21. See “Pyrrho,” remark B, in the Dictionary (1965: 197). Bayle attributes the argument
to Foucher (fn. 11). See also Watson (1987) and Reed (2015).
22. See Reid’s Essays on the Intellectual Powers of Man II.14 (1872: 298).
23. See Margaret Atherton’s chapter on Berkeley and René Van Woudenberg’s chapter
on Reid, both in this book. See also Van Cleve (2015: ch. 3) on Reid’s rejection of
the “way of ideas.”
24. See Georges Dicker’s chapter on Kant in this book.
25. See Alison Peterman’s chapter on Spinoza in this book.
26. See Donald Baxter’s chapter in this book for an account of how Hume adapted the
sort of passive acquiescence found in Pyrrhonism to his own system. See Garrett
(2015: ch. 7) for an interpretation of Hume as embodying a mitigated form of
skepticism. This mitigated skepticism, which Hume calls “academical” (EHU XII
129), may derive to some extent from the ancient Academics, given his use of
probability as a guide to belief. See Harald Thorsrud’s chapter on Carneades in Part
I of this book for an account of how probability may have functioned in a mitigated
Academic skepticism.
27. See Dietmar Heidemann’s chapter on Hegel and Andreas Urs Sommer’s chapter
on Nietzsche, both in this book. See also Berry (2011) on Nietzsche’s relation
to Pyrrhonism.
28. See Peter Graham’s chapter on Russell in this book.
29. See Annalisa Coliva’s chapter on Moore in this book. See also Reed (2015: 74–75).
30. See Michael Williams’s chapter on Wittgenstein in this book. See Sluga (2004) for an
interpretation that takes Wittgenstein to be sympathetic in some ways to Pyrrhonism.
REFERENCES
Allen, Don Cameron. 1964. Doubt’s Boundless Sea: Skepticism and Faith in the Renaissance.
Baltimore, MD: Johns Hopkins Press.
Bayle, Pierre. 1965. Historical and Critical Dictionary, translated by R. Popkin. Indianapolis,
IN: Bobbs-Merrill.
Bayle, Pierre. 2000. Various Thoughts on the Occasion of a Comet, edited by R. Bartlett.
Albany, NY: State University of New York Press.
BARON REED 281
Bayle, Pierre. 2005. A Philosophical Commentary on the Words “Compel Them to Come
In, That My House May Be Full,” edited by J. Kilcullen and C. Kukathas. Indianapolis,
IN: Liberty Fund.
Bell, Millicent. 2002. Shakespeare’s Tragic Skepticism. New Haven, CT: Yale University
Press.
Berry, Jessica. 2011. Nietzsche and the Ancient Skeptical Tradition. Oxford: Oxford
University Press.
Cicero. 2006. On Academic Scepticism, translated by C. Brittain. Indianapolis, IN: Hackett.
Descartes, René. 1984. The Philosophical Writings of Descartes, vol. II, translated by
J. Cottingham, R. Stoothoff, and D. Murdoch. Cambridge: Cambridge University Press.
Descartes, René. 1985. The Philosophical Writings of Descartes, vol. I, translated by
Diderot, Denis, et al. 1965. Encyclopedia, translated by N. Hoyt and T. Cassirer.
Indianapolis, IN: Bobbs-Merrill.
Drake, Stillman. 2001. Galileo: A Very Short Introduction. Oxford: Oxford University
Press.
Durand, Denis. 1730. The Life of Lucilio Vanini, Burnt for Atheism at Thoulouse. London:
W. Meadows.
Fine, Gail. 2000. “Descartes and Ancient Skepticism: Reheated Cabbage?” Philosophical
Review 109: 195–234.
Garrett, Don. 1997. Cognition and Commitment in Hume’s Philosophy. Oxford: Oxford
University Press.
Garrett, Don. 2015. Hume. London: Routledge.
Gaukroger, Stephen. 1995. Descartes: An Intellectual Biography. Oxford: Clarendon Press.
Hatfield, Gary. 1986. “The Senses and the Fleshless Eye: The Meditations as Cognitive
Exercises.” In A. Rorty (ed.), Essays on Descartes’ Meditations, 45–79. Berkeley, CA:
Hazard, Paul. 2013. The Crisis of the European Mind, translated by J. May. New York:
New York Review Books.
Hume, David. 1975. Enquiries Concerning Human Understanding and Concerning the
Principles of Morals, edited by P. H. Nidditch. Oxford: Clarendon Press.
Hume, David. 2000. A Treatise of Human Nature, edited by D. Norton and M. Norton.
Israel, Jonathan. 2006. Enlightenment Contested. Oxford: Oxford University Press.
Israel, Jonathan. 2007. “Admiration of China and Classical Chinese Thought in the Radical
Enlightenment (1685-1740),” Taiwan Journal of East Asian Studies 4: 1–25.
Laursen, John Christian, and Gianni Paganini (eds.). 2015. Skepticism and Political Thought
in the Seventeenth and Eighteenth Centuries. Toronto: University of Toronto Press.
Maia Neto, José R. 1997. “Academic Skepticism in Early Modern Philosophy,” Journal of
the History of Ideas 58: 199–220.
Maia Neto, José R. 2014. Academic Skepticism in Seventeenth-Century French Philosophy:
Montaigne, Michel de. 1958. The Complete Essays of Montaigne, translated by D. Frame.
Stanford, CA: Stanford University Press.
282 SKEPTICISM
Popkin, Richard. 1960. The History of Scepticism: From Erasmus to Descartes. Assen: Van
Gorcum.
Popkin, Richard. 1979. The History of Scepticism: From Erasmus to Spinoza. Berkeley:
Popkin, Richard. 1980. The High Road to Pyrrhonism. Papers collected by R. Watson and
J. Force. San Diego: Austin Hill.
Popkin, Richard. 2003. The History of Scepticism: From Savonarola to Bayle. Oxford:
Reed, Baron. 2015. “Skepticism and Perception.” In M. Matthen (ed.), The Oxford
Handbook of Philosophy of Perception, 66–80. Oxford: Oxford University Press.
Reid, Thomas. 1872. The Works of Thomas Reid, edited by W. Hamilton. Edinburgh:
Maclachlan and Stewart.
Rescher, Nicholas. 1993. “Quixotic Scepticism,” American Philosophical Quarterly 30:
183–184.
Robinson, Howard. 1931. Bayle the Sceptic. New York, NY: Columbia University Press.
Rorty, Amélie Oksenberg. 1986. “The Structure of Descartes’ Meditations.” In A. Rorty
(ed.), Essays on Descartes’ Meditations, 1–20. Berkeley, CA: University of California
Press.
Schmitt, Charles. 1972. Cicero Scepticus. The Hague: Nijhoff.
Sluga, Hans. 2004. “Wittgenstein and Pyrrhonism.” In W. Sinnott-Armstrong (ed.),
Pyrrhonian Skepticism, 99–117. Oxford: Oxford University Press.
Van Cleve, James. 2015. Problems from Reid. Oxford: Oxford University Press.
van der Lugt, Mara. 2016. Bayle, Jurieu, and the Dictionnaire Historique et Critique.
Watson, Richard. 1987. The Breakdown of Cartesian Metaphysics. Atlantic Highlands, NJ:
Humanities Press International.
Williams, Bernard. 1978. Descartes: The Project of Pure Understanding. London: Routledge.
Williams, Michael. 1986. “Descartes and the Metaphysics of Doubt.” In A. Rorty (ed.),
Essays on Descartes’ Meditations, 117–139. Berkeley, CA: University of California Press.
CHAPTER TWENTY
François de La
Mothe Le Vayer1
SYLVIA GIOCANTI
1 INTRODUCTION
François de La Mothe Le Vayer (1588–1672) was no doubt a skeptic, since he
explicitly claimed to belong to the skeptical philosophical tradition (both Pyrrhonian
and Academic) and made frequent use of the adjective “skeptical” in the title of
many of his opuscules.2 It is nonetheless legitimate to ask: in what respect was
he skeptical?
Indeed, the reception of his oeuvre as “libertine”3 raises doubts about the
sincerity of his participation in the skeptical philosophy, and particularly about
the sincerity of his adherence to fideism, thanks to which skepticism could, despite
the uncertainty of reason, prepare for the Christian faith. On the hypothesis that
skeptical argumentation is nothing but a rhetorical tool designed to dissimulate
subversive theses, skepticism would be only a cover for his libertinage. But in either
case (fideism or libertinage), the skepticism of La Mothe Le Vayer would not exist
in itself and by itself, but merely as a means to other ends.
Thus, to wonder in what respect La Mothe Le Vayer may be skeptical is at the
same time to reflect upon the possibility of the autonomy of skepticism within
philosophy, as a way of thought that could have value in itself, despite the fact that
it needs to feed on dogmatism either to combat it or to serve it.4 In other words, one
can recognize with Pascal that “nothing fortifies Pyrrhonism more than that there
are some who are not Pyrrhonian,” without thereby subscribing to what he adds
immediately afterward: “If all were so, they would be wrong” (Pascal 1962: frag.
33). Reading La Mothe Le Vayer invites us to pose anew the question of the value
of skepticism, independently of its polemical anchorage.
In Section 2, I will show that reason, which La Mothe Le Vayer reduces to opinion
(δόξα), is sanitized in its intellectual activity by the practice of paradox, substituted
for the practice of equipollence (ἰσοσθένεια). In Section 3, I will argue that this
critical activity introduces a novel relationship between skepticism and orthodoxy,
which explains why La Mothe Le Vayer may be considered both a libertine and a
skeptic, to the extent that his discourse is used to enjoy the transgression of norms.
284 SKEPTICISM
In Section 4, I will claim that the artifices employed to transgress the norms are not
however offered as recipes that work mechanically, but rather as expedients that are
skeptically utilized in order to calm the soul.
2 THE SKEPTICAL DISORDER OF REASONS

La Mothe Le Vayer seems to run counter to the idea peculiar to the Enlightenment
that skepticism is intellectually untenable, that doubt can only be provisional and
is to be surpassed by knowledge, since he builds up his skepticism on the basis
of the insurmountable irresolution of reason. Even if it turned out that La Mothe
Le Vayer utilized the skeptical argumentation to dissimulate the audacity of his
thought, it would still be clear that his skepticism cannot be entirely feigned, since
his philosophy rests upon one of the characteristics of modern skepticism, namely:
the observation of the irregularity of the way in which each person reasons (La
Mothe Le Vayer [LMLV] 1970e: 744), which is nothing but the effect of the
irregularity of reason itself.
La Mothe Le Vayer radicalizes Montaigne’s view of the malleability of reason
(Essays II, 12, 565),5 for he claims that, as a “shameless courtesan who, hidden under
the mask of appearance, shamefully abandons herself to every sort of candidate”
(LMLV 1970g: 262), reason is incapable of providing rules. To make a skeptical use
of reason consists in using it in accordance with what it is, namely, “a multipurpose
toy” (un jouet à toutes mains) (1970g: 262). One must bang reasons together as in
a game of balls, not in order to ridicule reason in its dogmatic use—as is the case
in Montaigne—but in order to exhibit its inherent uncertainty. Within the context
of this “Saturnalia of opinion,” in which everyone amuses himself by changing his
mind, the disorder of reasons is a real spectacle, or even a “rhapsody.”6 In this
game, the discursive discontinuity is irreducible, since the game is not unified, as
in Montaigne, by the project of portraying oneself as gradually constructing the
subject’s identity by means of a public reflection on oneself. The author’s position
cannot be determined—and in this it is clearly skeptical—for the aim of the text
cannot be related to the coherence of his thought, since his ideas remain uncertain.
Thus, one would be tempted to say regarding La Mothe Le Vayer what Bayle
reports about Carneades in the article of his Dictionary devoted to the latter:
[Clitomachus] was never able to discover what were the opinions [sentiments]
that Carneades approved of. That was not the result of the obscurity of his
expressions, but of the skepticism of this philosopher. He did not find anything
certain, he successively held and refuted the same doctrines: that is why one
could not discern whether he approved of any of them. (Remark O)
In a similar way, La Mothe Le Vayer’s skeptical argumentation does not let the
author’s point of view show through, because it is fleetingly discernible, blurred
by other points of view adopted successively, before a final retraction. Does his
skepticism really make him incapable of making up his mind one way or the other?
Or does he avoid explicitly saying what he thinks, out of prudence, like the libertines
of his time (on whom see Cavaillé 2002 and Gouverneur 2005)?
FRANÇOIS DE LA MOTHE LE VAYER 285
La Mothe Le Vayer may well employ skepticism to protect himself from

the accusation of impiety or atheism, thus eluding religious persecution. He will
nonetheless remain someone who is both committed to skepticism and a strong spirit
(esprit fort), insofar as his disavowing what he writes is coherently based on the
discredit of reason, and a fortiori on the discredit of the rationalism that reigns in
Cartesianism. Indeed, to accuse reason of poisoning the mind at the very moment in
which Cartesianism prevails does not have the same significance as in Montaigne’s
time (see Giocanti 2013). Being skeptical for La Mothe Le Vayer is to take the
opposing view to Descartes’s most famous claims expressing the confidence one must
accord to reason, as the guarantee of the progressive acquisition of all the perfections
that man is capable of. It is against the Cartesians that La Mothe Le Vayer writes his
Little Skeptical Treatise on This Common Way of Speaking “To Lack Common Sense”
(LMLV 1970g), where he declares that reason (or common sense) is not the thing
best shared in the world, that reason is not a source of order and the construction of
knowledge, but of the dissolution and corruption of the mind in fantasy and madness.
In order to find a remedy for rational obstinacy, this plague of the soul that
keeps reappearing in reaction to the spontaneous disorder of reasons, the skeptic
suggests ceaselessly counterbalancing the claims of reason by means of the practice
of equipollence, henceforth conceived of as the practice of paradox that purifies the
mind. This means that so-called reason, which acts as good sense or common sense,
is on the side of δόξα, and that for the skeptic it is therefore necessary to apply
oneself to struggling against it, systematically taking the opposing view to what is
affirmed. The claim of independence and the quest for verisimilitude peculiar to
the skeptical Academy are thus reused in the context of a combat, deemed equally
Herculean, both against the overly strong adherence to dominant opinions and
against the fervor of beliefs.7
3 THE ACCOMPLISHMENT OF
SKEPTICISM IN LIBERTINAGE
By the adoption of this rebellious attitude toward the common way, and hence
toward the authority of tradition—of everything that is handed down by moral,
intellectual, and religious customs—the skepticism of La Mothe Le Vayer clearly
differs from that of Montaigne. It is by such an attitude that his skepticism leads
La Mothe Le Vayer to libertinage, in a challenge to orthodoxy that constitutes
skepticism’s natural development.
3a The Libertine Practice of Paradox Substituted for the

Skeptical Practice of Equipollence
The affirmation of reason’s helplessness to hold on firmly and definitively to well-
founded certainties goes hand in hand with the affirmation of its power to challenge
unfounded certainties. It is this power that La Mothe Le Vayer deploys when calling
into question the attitude of respect toward laws and customs and the authority of
tradition in general.
286 SKEPTICISM
Such is the main difference between Montaigne and La Mothe Le Vayer in

their redeployment of the Tenth Mode of Aenesidemus as preserved by Sextus
Empiricus, and particularly in the practical consequences that they both draw from
it. Montaigne, favorable to anything that can contribute to preventing reason from
ranting, concludes his Essays by stressing the value of the aptitude for falling in with
the common model.8 It is never a matter of struggling against the obstinacy of those
who strive to remain in the defeated paths of the common way, but rather of using
the model of animality and simplicity to correct the excesses of speculative reason.
By contrast, for La Mothe Le Vayer the skeptic has the distinctive feature of taking
account of opinions other than those handed down by δόξα, that is to say, by the
religious and moral orthodoxy. Qua strong spirit who is against the vulgarity of
common sense, the skeptic opposes to it a contrary opinion, not in order to readjust
such vulgarity with the aim of disciplining himself, but in order to grant full license
to his thought, making it take, like the goat, “the isolated paths.” The practice of
paradox, substituted for the practice of equipollence, does not therefore have the
function of neutralizing opinions via the equipollence of arguments, as is the case in
ancient Pyrrhonism or in Montaigne’s skepticism. Its function is rather to promote
the reverse of ordinary opinion, for “the healthiest opinions are perhaps—supposing
that we possess some health in this respect—the most paradoxical” (LMLV 1970g:
271).9 Thus, whenever La Mothe Le Vayer compares pagan rites (which in his time
have become extraordinary and illicit) with Christian rites (customary and licit),
he does so in favor of the former, if only because the latter are considered without
regard for their differences in a way that makes them lose their exclusiveness.
Because of this, it is not only completely questionable, as Maia Neto (1995: 126)
has remarked, to claim that La Mothe Le Vayer is the representative of a Christian
Pyrrhonism, but it is also possible to advance the view that he covertly devotes
himself to contesting the validity of Christianity, despite the fideist remarks that, as
Popkin (1995: 137) has noted, permeate all of his works.
3b A Singular Conjunction between Skeptical Fideism and

Libertine Disrespect of the Common Way
Such claims of obedience to Christianity do not actually have any credibility. But the
reason is not only that, having abandoned every rational criterion, it is not possible
to place Christianity beyond the scope of doubt. This argument is too weak to call
into question the sincerity of La Mothe Le Vayer’s skepticism, for, as Popkin (1995:
142–143) has argued, given that one could raise the suspicion of hypocrisy against
any skeptic whatsoever, there is no more reason for suspecting La Mothe Le Vayer
than any other skeptic. However, as Paganini (2008: 93–94, 99) has rightly remarked,
in the second part of On the Virtue of Pagans devoted to “Pyrrho and the skeptical
sect” (1970f: 194–196), La Mothe Le Vayer himself denies any faith implicit in
Pyrrho and condemns him to damnation on the grounds that, unlike the other pagan
philosophers, this leader of the skeptical school has placed doubt at the heart of his
philosophy. On this specific point, La Mothe Le Vayer adopts the same position as
Bayle, according to which “faith does not accept the Academic spirit: it wants that
we deny, or that we affirm” (Historical and Critical Dictionary, art. “Chrysippus,”

remark G).10 Religion requires an invariability that is impossible to attain without
grace and that is incompatible with doubt. Now, in La Mothe Le Vayer’s skeptical
anthropology, doubt is inevitable on account of the human inconstancy that is the
spontaneous manifestation of the unbelief of the mind (see especially LMLV 1970d:
605). One should then conclude that skepticism leads to irreligion, and not, as Popkin
believes, that it constitutes an excellent preparation for Christianity. Admittedly, La
Mothe Le Vayer explicitly makes this latter claim when he presents his skepticism
throughout his work. But one may wonder—taking as an indication of an attitude of
dissimulation the abovementioned text on Pyrrho—if the link he establishes between
skepticism and preparation for faith is sincere, if the skeptical position itself is not
merely a posture that makes it possible to attack religion in an indirect way. La
Mothe Le Vayer would then be neither skeptical nor Christian, but libertine.
But one will retort that La Mothe Le Vayer strives to make an exception for
his own form of skepticism insofar as it is “circumcised,” that is, purified because
Christianized. The question is then once again: what is the credibility of this pious
exception? There is a criterion that makes it possible to resolve the dispute about the
credibility of La Mothe Le Vayer’s fideism in comparison with that of Montaigne,
and a fortiori that of Pascal: whether one finds in him an attitude of respect toward
the authority of tradition, an injunction to obey the laws and customs of one’s
country. The skepticism of La Mothe Le Vayer is, in a manner quite singular in
the skeptical tradition, inseparable from a monolithic criticism of customs, which
are denounced as tyrannical.11 It is from this criticism that results a disrespectful
treatment of religious authority. For instance, in the dialogue On Divinity, La Mothe
Le Vayer defends, in the tradition of Bacon, atheism’s superiority over superstition.
And this is not in favor of the Christian religion, since he adds on the same page
(LMLV 1988e: 339), following Charron, that “all religions are foreign and horrible
to common sense,”12 which seems to warrant the equation of all religions with the
phenomenon of superstition (see Paganini 2008: 81). In this text, Christianity does
not therefore seem to be an exception nor to be susceptible, as orthodoxy requires,
of being in agreement with common sense, which is nonetheless fundamental in the
adherence to Christianity as a preamble to faith (Maia Neto 1995: 131–132).
Certainly, the claims against religion are caught in the skeptical equipollence,
which, as already noted, is transformed in La Mothe Le Vayer into the practice of
paradox. The reason for such a transformation is to better display the other side
of the orthodox position, with a subversive aim, since by valuing the reverse of
the opinion one produces a disproportion between the orthodox theses (weakened
because agreed) and the heterodox ones (whose critical impact is much more
striking). La Mothe Le Vayer takes pleasure in the criticism of opinions received with
reverence and systematically devalues the attitude of submission that is inseparable
from devotion and piety. He praises in blasphemous terms what he calls “La Divine
Sceptique,” that is, the divine skeptical philosophy, in that it leads, following the
“decalogue” (i.e., the Ten Modes) of “Divine Sextus” (LMLV 1988c: 83), to the
ecstatic rapture of the mind of the character of Télamon, suddenly converted to
beatitude by means of doubt (LMLV 1988d: 301–302).
288 SKEPTICISM
What is at issue here is not a lack of differentiation between religions that are
distinct (viz., the pagan and the Christian), but rather a competition between the
Christian religion and skepticism in favor of the latter. For several times La Mothe
Le Vayer cedes exclusively to the skeptical philosophy the merits and satisfactions
normally provided by the Christian religion (see LMLV 1988b: 61–62).13
His skepticism, insofar as it accomplishes the desire to free oneself from the rules
that are authoritative, becomes an enterprise of subversion of values, an enterprise
that is incompatible not only with Christian apologetics (skeptical fideism), but with
a sincere profession of faith, by which he tells us one must not be duped.14 But how
could one be, since far from making a virtue of obedience to the laws and customs
of his country, La Mothe Le Vayer lays down a challenge to the tyrannical authority
of the latter, in the context of his skeptical escapades whose unpredictable and
capricious character he defends?
3c The Skeptical Subversion of Norms as a Condition

of the Libertine’s Enjoyment
Contrary to what Popkin (1995: 138) maintains, La Mothe Le Vayer is not an
“insipid Montaigne.” Rather, he would be a “scandalous Montaigne,” who realizes
the libertine (deviant) potentialities of skepticism, by means of an unprecedented
redeployment of its content, out of reach from what he calls, in the same terms as
Charron (1986: 261), “the universal contagion” or “spiritual contamination” of the
crowd (LMLV 1988e: 303). Even if obedience to the Christian tradition may have
a value for the herd, it is not convenient to respect it in one’s heart, but rather to
contest its validity and to conform to it only to “keep up appearances,” the function
of religion being “to offer a rational explanation of moral phenomena” (LMLV
1988e: 330–331).
The flexibility required in La Mothe Le Vayer’s skepticism, qua ethical virtue that
makes it possible to adapt to the diversity of phenomena (LMLV 1988f: 365), does
not therefore have the same sense as in Montaigne, who praises the sociability of a
“soul of diverse stages that knows both to tense up and to dismantle itself” (Essays
III, 3, 821). The flexibility of the soul in La Mothe Le Vayer is conceived of in the
context of “skeptical wisdom”—which is opposed to divine wisdom—and is defined
as “wandering prudence” that “is proud of changing at any moment and of turning
the sail according to the wind” (1970e: 735). To be flexible means, then, to gently
accommodate oneself to all the customs introduced by mores (LMLV 1988d: 273),
which also means to try one’s hand at all practices, including the most debauched,
after having freed oneself from every scruple. It is not a matter, as in Montaigne,
of accepting one’s natural irresolution and displaying a plasticity of both mind and
body in the face of social constraints that require one to adapt oneself. It is rather a
matter of freeing oneself from these constraints, of enjoying the transgression of the
established order, and of spreading a perfume of scandal, initiating oneself without
shame into those practices that are judged to be the most extravagant and the most
shameful, on the pretext that one finds examples of them in nature (see LMLV
1988c and 1997).15
La Mothe Le Vayer relates then the enjoyment promised by the skeptical discourse
to a diversification of mores that contravenes that which is licit. He thus extends
doubt to cultural norms—not only to their intrinsic value but also to the respect
that is due to them—whereas in Montaigne doubt bore only on the natural limits.
By so doing, not only does he subvert social norms and incite the transgression of
their prohibitions, but he also establishes the transgression of social conformism
as a condition of personal flourishing. The game of reason—“shameless courtesan
who, hidden under the mask of appearance, shamefully abandons herself to every
sort of candidate”—in which La Mothe Le Vayer indulges proves then to be more
audacious than that of the “crooked, lame and wiggling” (Essays II, 12, 565) reason
of Montaigne (and this despite the image associated with sexual relations with
cripples at Essays III, 11), to the extent that it accomplishes skepticism in libertinage.
If it is always a matter of learning how to use reason according to itself, of acquiring
“the self-sufficiency to know how to circumvent it,” as Montaigne said (Essays II,
12, 565), it is remarkable that the attitude of intellectual disobedience, related to a
moral anti-conformism, supposes a sophisticated and even cunning use of reason.
It is now necessary to clarify in what respect La Mothe Le Vayer’s skepticism, in its
strictly discursive dimension, consists in according ethical and therapeutic virtues to
a sophistical use of reason.
4 THE THERAPEUTIC VIRTUES OF THE

SKEPTICAL DISCOURSE
Contrary to the point of view of apologists who take the sole aim of libertine
discourse to be justifying a life of debauchery motivated by the passions alone,
La Mothe Le Vayer’s skeptical point of view on mores shows the pleasure one
finds in transgressing the proprieties within discourse above all else. This linguistic
enjoyment rests upon a “poietic” (productive) conception of reason: reason is a
protean faculty that molds its discourses in a hundred different ways, following its
imaginative representations, as the potter models his work.16
Reason so conceived is not a faculty that naturally leads us to knowledge of the
truth, but an instrument that has “more than one trick up its sleeve” as regards
fantasizing and artifice, so as to “smoke out the mind.” And the first to be interested
in this conjuring, in the moral domain, is the skeptic himself, who deploys his
ingenuity to deceive himself (LMLV 1970c: 534). For him, to actualize the mind’s
potentialities, following a penchant that leads us to lying, means to employ artifice
(LMLV 1970i: 538). Like every man, the skeptic is a “philomyth” animal, that is,
a lover of fables (LMLV 1970c: 539). In this respect, it should be noted that the
Dialogues Made in Imitation of the Ancients are presented as fables (LMLV 1988a:
14). Unlike the other men, the skeptic dares prefer lying over truth (LMLV 1970b:
478–479) and allows himself to practice a fabulating art in the hope of experiencing
pleasure or at least of overcoming anxiety.
The human condition is, according to La Mothe Le Vayer, so calamitous that,
without this potion of error and ignorance that nature makes us drink at birth, no
290 SKEPTICISM
one would resist the temptation to commit suicide (LMLV 1970b: 393, 409). But
this observation of human misery, apparently very Augustinian, is not an invitation
to search for the truth, without which we could not live at rest. For in La Mothe
Le Vayer the restriction of the skeptical investigation to what is verisimilar has no
other function than to guarantee the possibility of recanting. To abide by what is
verisimilar does not mean to confine oneself to an adequate vision of what is true,
but to adopt a point of view that would always make it possible to adopt another
one that is more joyful (LMLV 1970g: 268). This is why, following an orientation
diametrically opposed to that of Pascal, La Mothe Le Vayer authorizes us to lie to
ourselves, trying to approach in the mode of mockery that which would afflict us
if we looked at it close up. To laugh with Democritus at the miserable condition
of man transfigures this condition in a way that may prove to be “salutary” for the
mind’s health, for its rest. Thus, far from fighting lying and error, as did Montaigne
(who values telling the truth as sincerity that authenticates his own discourse), La
Mothe Le Vayer incites us to deceive ourselves and to develop the artifices that make
it possible. Admittedly, Montaigne too makes use of diversion and of the powers of
fantasizing of the mind (Essays III, 4), taking advantage of a deceptive reason that
indistinctively tells the truth and lies (Essays II, 12, 565). But he does not go so far
as to claim that reason must be employed to deceive ourselves insofar as it has more
grace in lying than in telling the truth, as La Mothe Le Vayer (1970g: 262) maintains.
In this respect, La Mothe Le Vayer’s skeptical ethics is not similar to the libertine
naturalism of Vanini or Charron.17 Artifice is at the heart of the devices used with
therapeutic aims, in order to cure the soul of the grief that often has no attributable
cause, human life appearing as a chronic illness in the eyes of a mind that is compared
to a “debauched stomach” (LMLV 1970c: 530). This is why it does not suffice, as
in the Pyrrhonian tradition, to get rid of dogmatic opinions in order to, through the
suspension of assent, recover from the quartan fever that accompanies melancholy
and attain the state of undisturbedness (ἀταραξία). This purge of opinions that
had such a beneficial effect on Henri Estienne, the translator of Sextus, would
leave the desolate mind of someone like La Mothe Le Vayer without something to
occupy itself. Such a purge would abandon him to the vacuity of other thoughts,
which could turn out to be just as corrosive, since the mind is capable of turning
honey into venom, of converting pleasure into sadness, under the influence of
its representations.
It is then the creative freedom of fantasy, productive of reasonings likened to
ingenious fables, that becomes the only possible source of consolation in the context
of the irrational (because imaginative) working of the mind. Because the notions of
good and evil are psychological, that is to say, relative to the mind that examines
them, this fantastical treatment of the emotions—within which one learns to content
oneself with imaginary objects—seems more suitable than their rationalist treatment.
Hence, the discourses apparently borrowed from the Stoic and Epicurean traditions
are often less taken over by La Mothe Le Vayer than skeptically distorted.
One can thus provide part of a reply to the question posed at the outset of
this chapter: La Mothe Le Vayer’s skepticism enjoys a certain autonomy within
philosophy, if only because it claims to have come up with remedies capable of

relieving the misfortunes of the human soul—specific remedies that draw on the
inventive potentialities of the mind. In this regard, Paganini (2008: 77) is right in
insisting on the “positive potentialities” of La Mothe Le Vayer’s skepticism. The
development of the Tenth Mode as expounded in Sextus represents the opportunity
to divert the mind from its grief, to extract it from the upsetting disparity of its
reasonings, so as to make it admire the diversity of the world (even of the worlds)
through the spectacle of the ἀνωμαλία (irregularity) of phenomena, thus arousing
enthusiasm. When he manages to delight in the wonders of nature or of human
societies, La Mothe Le Vayer does not hide the pleasure he experiences in opening
his mind to a multitude of ways of thinking, far from common prejudices (Paganini
2008: 70).
But this way of looking at things results from an effort of diversion. La Mothe Le
Vayer’s skeptic is also a bilious spirit, who suffers from a wandering freedom that
leads him to take a disenchanted and pessimistic view of men (whose obstinacy and
injustice are everywhere criticized), of himself (who does not tolerate the disparity
of his reasonings and his moods), of the world (the stage where the pitfalls of fortune
are performed), and of human life (which he finds it difficult not to curse).
La Mothe Le Vayer’s skepticism has then both positive and negative potentialities.
And one should not be surprised: insofar as things are such as we represent them,
they are also at the mercy of these representations.18 According to an image that La
Mothe Le Vayer borrows from Montaigne (Essays II, 12, 581), each thing is like
a pot with two handles, which one can take by either the right or the left (LMLV
1670: 43). From a logical point of view, that refers to the structure of equipollence,
and from an empirical point of view, that refers to the combination of pleasure and
displeasure in all things.
It follows that, despite the therapeutic use of reason (which is likened to the
imagination) in La Mothe Le Vayer’s skeptical discourse, ἀταραξία, which remains
the pursued aim, is not within easy reach and is by no means guaranteed by the
adoption of the skeptical stance, nor by the reading of texts written from the
skeptical perspective. The skeptic, more than any other philosopher, admits that his
thought is subject to conditions he does not control and that it is always possible for
him to succumb to the irrationality of the working of his mind. He therefore accepts
that he is not able to really enjoy the rest associated with the adoption of a given
position. Because of this, La Mothe Le Vayer’s writings stand out by an awkward
mode of enunciation, which nonetheless manages, more or less constantly, to put to
the test the intellectual, moral, and religious orthodoxy of his time. This certainly
provokes perplexity in the reader.
Thus, even if it frees us from the troubles caused by opinion, leading us to
libertinage, La Mothe Le Vayer’s skepticism remains a genuine, non-feigned
skepticism. It is so not because it leads to ἀταραξία, but paradoxically (with
respect to the Pyrrhonian tradition) because it expresses, by means of its manner of
exposition, an intellectual and moral discomfort that reveals a way of thinking that
takes no position, a discomfort that affects both the author and the reader.
292 SKEPTICISM
NOTES
1. This chapter has been translated from the French by Diego Machuca.
2. These opuscules include On the Skeptical Philosophy, The Skeptical Banquet,
Academic Homilies, Discourse to Show that the Doubts of the Skeptical Philosophy
Are of Great Use in the Sciences, Skeptical Discourse on Music, Little Skeptical
Treatise on This Common Way of Speaking “To Lack Common Sense,” Skeptical
Problems, Skeptical Doubt, and Skeptical Soliloquies.
3. One did not have to wait until the twentieth century (Pintard 1983), for Pierre
Bayle attests to this reception in the article “Le Vayer” of his Historical and Critical
Dictionary (Bayle 1740) when he writes: “One suspects that he did not have any
religion” (Remark B).
4. In La Mothe Le Vayer (1970g: 269–270), Samson, a “perfect figure of the skeptical
philosophy,” after having killed the lion, feeds on the honey he finds in its mouth.
The skeptic, then, feeds on the dogmatism he combats. But he is at the service of
dogmatism when he takes up a discourse in line with the apostolic message of Saint
Paul (LMLV 1988e: 308).
5. I cite Montaigne’s Essays by book, chapter, and page number in the standard Villey-
Saulnier edition (Montaigne 1992).
6. On the “Saturnalia of opinion,” see La Mothe Le Vayer (1988d: 365); and on
rhapsody, see the “Author’s Letter” in La Mothe Le Vayer (1988a: 201).
7. The claim of independence of the skeptical Academy (Acad. II 8) and the importance
of verisimilitude in the conduct of life are pointed out by La Mothe Le Vayer in his
general presentation of Pyrrhonism (1970f: 192–194) and taken over on his own
account especially in La Mothe Le Vayer (1970g: 268). He explicitly refers to the
Herculean effort made by Carneades to combat the tendency to judge rashly (Acad.
II 108) in La Mothe Le Vayer (1970h: 326).
8. Essays III, 13, 1116: “The most beautiful lives are, in my opinion, those that fall
in with the common and human model, with order, but also without miracle and
without extravagance.” Cf. Essays II, 12, 558–559: “All the extravagant ways anger
me. … And there is no beast to which blinders are more justly to be given to keep its
vision submissive and restrained in front of its steps, and to guard it from wandering
here and there.”
9. In the “Author’s Letter” in La Mothe Le Vayer (1988a: 14), he announces that in
this work one will find nearly nothing but paradoxes.
10. For a more general comparison between Bayle and La Mothe Le Vayer, see Giocanti
(2004: 243–263).
11. In the “Author’s Letter” in La Mothe Le Vayer (1988a: 14–15), he denounces “the
tyrannical authority of time, and of the customs that the [opinions of the laypeople]
have established.” Unlike Montaigne (Essays III, 13, 1080), he does not praise
the metamorphic power of custom, which is compared to the beverage of Circe.
La Mothe Le Vayer regards one’s flexibility toward custom as an escape from the
authority of the dominant customs.
12. In relation to this reference to common sense, it should be noted that, although in
On Divinity La Mothe Le Vayer takes common sense as a criterion of evaluation
in order to attack religion, he criticizes common sense itself in La Mothe Le
Vayer (1970g). On his ambiguous attitude toward common sense, see Giocanti
(1996: 27–51).
13. In La Mothe Le Vayer (1970g: 265), the skeptical philosophy is presented,
adapting an Augustinian slogan, as “the sole port of salvation for a mind that
loves calm.”
14. In La Mothe Le Vayer (1970h: 328) one reads: “That he [the reader] not believe me
unchanged by the opinions I may have, or pretend to have.”
15. As Cavaillé (1995: 135 n.320) remarks, this naturalistic argumentation, used for
instance in La Mothe Le Vayer (1988c: 102–103) in favor of incest, might find its
source of inspiration in Vanini (1616: 326–327).
16. On the mind as “another Proteus,” see La Mothe La Vayer (1970d: 666); and on the
mind as “potter,” see La Mothe Le Vayer (1970i: 504).
17. See Gouverneur (2005: 276–290) on “fabulation as imaginary evasion” in La Mothe
Le Vayer, “far from Charron’s naturalism.”
18. “Man is the measure of all things, which become such as he represents them to
himself” (LMLV 1970c: 530).
REFERENCES
Bayle, Pierre. 1740. Dictionnaire historique et critique. 4 vols. Amsterdam: P. Brunel et al.
Cavaillé, Jean-Pierre. 2002. Dis/simulations. Cesare Vanini, François de La Mothe Le Vayer,
Gabriel Naudé, Louis Machon, Torquato Acetto. Religion, morale et politique au XVIIe
siècle. Paris: Champion.
Charron, Pierre. 1986. De la sagesse. Paris: Fayard.
Giocanti, Sylvia. 1996. “La perte du sens commun dans l’œuvre de La Mothe Le Vayer.” In
A. McKenna and P. F. Moreau (eds.), Libertinage et philosophie au XVIIe siècle, 27–51.
Saint-Etienne: Publications de l’Université de Saint-Etienne.
Giocanti, Sylvia. 2001. Penser l’irrésolution. Montaigne, Pascal, La Mothe Le Vayer. Trois
itinéraires sceptiques. Paris: Champion.
Giocanti, Sylvia. 2004. “Bayle et La Mothe Le Vayer.” In A. McKenna and G. Paganini
(eds.), Pierre Bayle dans la République des Lettres, Philosophie, religion, critique, 243–
263. Paris: Champion.
Giocanti, Sylvia. 2010. “Le scepticisme instrument de transgression du licite. Le cas de
La Mothe Le Vayer.” In A. McKenna and P. F. Moreau (ed.), Le libertinage est-il une
catégorie philosophique?, 83–100. Saint-Etienne: Publications de l’Université de Saint-
Etienne.
Giocanti, Sylvia. 2013. “La Mothe Le Vayer: un anti-cartésien?” In D. Kolesnik-Antoine
(ed.), Qu’est-ce qu’être cartésien?, 77–94. Lyon: ENS Editions.
Gouverneur, Sophie. 2005. Prudence et subversion libertine. La critique de la raison d’Etat
chez François de La Mothe Le Vayer, Gabriel Naudé et Samuel Sorbière. Paris: Champion
La Mothe Le Vayer, François de. 1670. Soliloques sceptiques. Paris: Billaine.
La Mothe Le Vayer, François de. 1970a. Œuvres, 2 vols. Genève: Slatkine Reprints. (This
edition corresponds to the edition of Louis Billaine [Paris, 1669] as it was republished
by Michel Groell [Dresde, 1756].)
294 SKEPTICISM
La Mothe Le Vayer, François de. 1970b. “Opuscules ou Petits Traités.” In François de La

Mothe Le Vayer, Œuvres, vol. I, 316–492. Genève: Slatkine Reprints.
La Mothe Le Vayer, “Prose Chagrine.” In François de La Mothe Le Vayer, Œuvres, vol. I,
505–542. Genève: Slatkine Reprints.
La Mothe Le Vayer, François de. 1970d. “Discours ou homilies académiques.” In François
de La Mothe Le Vayer, Œuvres, vol. I, 572–681. Genève: Slatkine Reprints.
La Mothe Le Vayer, François de. 1970e. “La promenade en neuf dialogues.” In François de
La Mothe Le Vayer, Œuvres, vol. I, 690–758. Genève: Slatkine Reprints.
La Mothe Le Vayer, François de. 1970f. “De la vertu des payens.” In François de La Mothe
Le Vayer, Œuvres, vol. II, 119–219. Genève: Slatkine Reprints.
La Mothe Le Vayer, François de. 1970g. “Petit traité sceptique sur cette commune façon de
penser ‘n’avoir pas le sens commun’.” In François de La Mothe Le Vayer, Œuvres, vol.
II, 251–272. Genève: Slatkine Reprints.
La Mothe Le Vayer, François de. 1970h. “Doute sceptique.” In François de La Mothe Le
Vayer, Œuvres, vol. II, 306–328. Genève: Slatkine Reprints.
La Mothe Le Vayer, François de. 1970i. “Petits traités en forme de Lettres écrites à diverses
personnes studieuses.” In François de La Mothe Le Vayer, Œuvres, vol. II, 341–749.
Genève: Slatkine Reprints.
La Mothe Le Vayer, François de. 1988a. Dialogues faits à l’imitation des Anciens. Paris:
Fayard.
La Mothe Le Vayer, François de. 1988b. “De la philosophie sceptique.” In François de La
Mothe Le Vayer, Dialogues faits à l’imitation des Anciens, 17–62. Paris: Fayard.
La Mothe Le Vayer, François de. 1988c. “Le Banquet sceptique.” In François de La Mothe
Le Vayer, Dialogues faits à l’imitation des Anciens, 63–113. Paris: Fayard.
La Mothe Le Vayer, François de. 1988d. “De l’ignorance louable.” In François de La Mothe
Le Vayer, Dialogues faits à l’imitation des Anciens, 209–302. Paris: Fayard.
La Mothe Le Vayer, François de. 1988e. “De la divinité.” In François de La Mothe Le
Vayer, Dialogues faits à l’imitation des Anciens, 303–352. Paris: Fayard.
La Mothe Le Vayer, François de. 1988f. “De l’opiniâtreté.” In François de La Mothe Le
Vayer, Dialogues faits à l’imitation des Anciens, 353–386. Paris: Fayard.
La Mothe Le Vayer, François de. 1997. L’Hexameron rustique. Paris: Zanzibar.
Maia Neto, José. 1995. The Christianization of Pyrrhonism. Scepticism and faith in Pascal,
Kierkegaard, and Shestov. Dordrecht: Kluwer.
Montaigne, Michel de. 1992. Essais. Paris: PUF.
Paganini, Gianni. 2008. Skepsis. Le débat des modernes sur le scepticisme: Montaigne, Le
Vayer, Campanella, Hobbes, Descartes, Bayle. Paris: Vrin.
Paganini, Gianni, Miguel Benitez, and James Dybikowski. (eds.). 2002. Scepticism,
Clandestinity and Free-Thinking. Paris: Champion.
Pascal, Blaise. 1962. Pensées, ed. Lafuma. Paris: Du Seuil.
Pintard, René. 1983. Le libertinage érudit dans la première moitié du XVIIe siècle. Génève:
Slatkine. (First published in 1943.)
Popkin, Richard. 1995. Histoire du scepticisme d’Erasme à Spinoza. Paris: PUF.
Vanini, Giulio Cesare. 1616. Les Arcanes admirables de la Nature, Reine et Déesse des
mortels. Paris: Adrien Périer.
CHAPTER TWENTY-ONE
Gassendi on Skepticism
ANTONIA LOLORDO
1 INTRODUCTION
In note B of his Dictionary article on Pyrrho, Pierre Bayle remarked that
before Gassendi,
one hardly knew the name of Sextus Empiricus in our schools. The methods he
had proposed so subtly for bringing about suspense of judgment were not less
known than the Terra Australis, when Gassendi gave us an abridgement of it,
which opened our eyes. . . . Now no good philosopher any longer doubts that the
skeptics were right to maintain that the qualities of bodies that strike our senses
are only appearances. (Popkin 1991: 197)
This is—to put it charitably—an exaggeration.1 However, Gassendi did do a great
deal to make skepticism better known. One factor was his use of skeptical techniques
to attack rival systems, as in his Exercitationes paradoxicae adversus Aristoteleos
and his Objections and Counter-Objections to Descartes’s Meditations (jointly, the
Disquisitio Metaphysica). Another is the detailed account of ancient skepticism
provided in his magnum opus, the posthumous Syntagma Philosophicum.2
In the Exercitationes, his first work, Gassendi portrays himself as a skeptic.
But by the time of the Syntagma, he disavows that label, counting himself instead
among those who hold that some truth can be found and that “a certain middle
way between the skeptics . . . and the dogmatists should be followed” (1.79b). This
via media is sometimes called “mitigated skepticism” or “constructive skepticism,”
following Popkin 1979, but the term should not mislead us. The via media of the
Syntagma is not a form of skepticism at all, but an epistemology that insists on three
things. First, the scope of our knowledge is limited: it cannot reach all the way to the
innermost nature of things. Second, knowledge does not require absolute certainty
but only probability or verisimilitude. And third, knowledge does not require a
justification that is accessible to the knower. When there is no real disagreement,
some things count as knowledge without any such justification.
Gassendi’s interest in ancient skepticism is epistemological and methodological.
He is aware that skepticism originally had an ethical dimension. But this is not
something he cares much about. His long discussion of ancient skepticism in the
296 SKEPTICISM
Syntagma concludes with the brief remark that “how the skeptics have attained
their end—ataraxia, an imperturbable state of mind—by this withholding of
assent can be passed over at this point” (1.76b).3 Nowhere else in his work, as
far as I know, does the ethical function of suspension of judgment receive any
real discussion.4
2 GASSENDI’S HISTORY OF SKEPTICISM

Gassendi was familiar with the Pyrrhonian tradition from early on. One of his earliest
surviving letters explains that in his youth, his “mind was not yet . . . thoroughly
Pyrrhonist” but that “only one thing, experience, was lacking to make [him] an
imitator of Sextus Empiricus” (to Henri du Faur de Pibrac, April 8, 1621: 6.1b).
The sources he mentions are Cicero, Charron, and Montaigne. In the Preface to
his first published work, the 1624 Exercitationes paradoxicae adversus Aristoteleos,5
Gassendi explains that although he studied the views of all the various sects of
philosophers, “none of them ever pleased [him] like the akatalepsia praised by the
Academics and Pyrrhoneans” (3.99). Thus, he explains,
During the six years I devoted to the task of teaching philosophy, especially
Aristotelean philosophy, at the University of Aix, I always made sure that my
students would be able to defend Aristotle properly. But as an appendix to the
course I also taught the claims that totally undercut Aristotelean dogmas. . . . In
this way my students . . . saw that no proposition or opinion is so thoroughly
accepted or plausible that its contrary cannot be shown to be equally probable or
even more probable. (3.100)
Most of the sources he lists are modern: Juan Luis Vives, Gianfrancesco Pico della
Mirandola, Petrus Ramus, Charron again (3.99), and Cicero (3.100).6 But it’s
clear, especially in Book II, that by this point Gassendi was also familiar with two
important ancient sources: Sextus Empiricus’s Outlines of Pyrrhonism and Book IX
of Diogenes Laertius’s Lives of the Philosophers, the Life of Pyrrho.
In both the Exercitationes and the Syntagma, Gassendi’s presentation of the Ten
Modes of suspension of judgment follows Outlines of Pyrrhonism closely.7 In the
Syntagma—where Gassendi is unusually careful (by seventeenth-century standards)
about citing his sources—he also refers to the Life of Pyrrho repeatedly. But he
cites a number of other ancient sources as well: Sextus’s Against the Logicians and
Against the Grammarians; Cicero, especially his Academica; Augustine’s Contra
Academicos; Seneca’s Letters; and many others.
The Syntagma’s Logic begins with a history of logic that is designed to make the
need for Gassendi’s own system clear. Because the goal of logic is the truth, Gassendi
centers his history around disagreements over the criterion of truth. Much of this
history deals with the philosophers who deny the existence of a criterion and insist
that nothing is known (1.69b). Gassendi emphasizes that the criterion in question
is supposed to be one “which is subject to no falsity and which thus gives rise to
certain and infallible knowledge” (1.70a). This is parasitic on the way the skeptics’
opponents think of the criterion. If Descartes or the Aristotelians, for instance, set
GASSENDI ON SKEPTICISM 297
a very high bar for knowledge—absolute certainty—then the skeptic need only cast
doubt on our ability to achieve absolute certainty.
One surprising feature of Gassendi’s history of skepticism is the sheer number
of people it includes. The obvious suspects are there: Pyrrho and his followers, and
those “who [Sextus] Empiricus thought hardly differed from the Skeptics, namely
the Academics” (1.72b).8 But it also includes Democritus, Empedocles, Euripides,
Heraclitus, Hippocrates, Homer, Philo, Socrates, Xenophanes, Zeno of Elea, and
more. One reason Gassendi includes such a dizzying list of figures is that he is always
concerned to emphasize the diversity of non-Aristotelian traditions. But a second,
more important one is that he has simply taken over much of the list from the Life
of Pyrrho (1.70a–1.71b) and Against the Logicians (1.71b–1.72b).
3 THE TEN MODES

Gassendi’s skeptical arguments rely heavily on the Ten Modes. In the Exercitationes,
he uses familiar examples to show that “the judgments of different men concerning
the things that are perceived by the senses are very different” (3.197b) and hence
that “men do not know the inner natures of things” (3.203a). In the Syntagma,
he explains that “one and the same thing can appear in different ways to different
animals and different men, and even to one and the same man according to his
various senses and his various affections (which are the first four Modes)” (1.84a).
Moreover, Gassendi argues, this disagreement cannot be resolved.
Consider an example. The same wine tastes sweet to some men and bitter to others
(3.198a ff). How can we decide whose taste constitutes the standard? We cannot say that
the way wine tastes to a healthy man is the way it really is, because healthy men disagree.
We cannot say that the way wine tastes to most people is the way it really is, because
there’s no clear majority verdict and no reason a mere majority should constitute the
standard. And so on (3.198ff). If one considers such cases, Gassendi thinks, one will find
oneself withholding assent from any opinion about the wine’s inner nature.
This, Gassendi makes clear, is compatible with continuing to assent to
the appearances:
When [the skeptics] say that nothing can be known with certainty or that no
Criterion exists, they are not talking about what things appear to be and what
is manifested through the senses . . . but about what things are in themselves.
(1.70a)
For:
If it is asked whether snow is white, honey sweet, fire hot, [the skeptics] will
respond that it appears so to them, and they do not say that their vision perceives
another color than white in snow or that their tongue perceives another taste than
sweet in honey or that their hand perceives another quality than heat in fire. But
because these are only external adjuncts and things that strike the senses, if it’s asked
whether the things themselves, secundum se and from their innermost nature, are
really the way they appear to be . . . this is what they will stick at. (1.13b)
298 SKEPTICISM
The disagreement between various men about the taste of wine is not a dispute
over the contents of their minds. Gassendian appearances are far closer to what the
ancient skeptics had in mind than to Cartesian knowledge of one’s own subjective
mental states.
Knowledge of the appearances includes not only claims like the claim that I
am now seated rather than standing or that wine appears sweet to me, but also
mathematical knowledge:
Whatever certainty and evidentness there is in the discipline of mathematics,
it pertains to appearances and in no way to the true causes or inner natures of
things. However, I add that through mathematics, I can become more certain
that, for example, the earth is round. For this can be apparent through the
eclipses of the moon and the variation in the height of the poles. But why is
the earth round? What is its true nature? . . . Why does it rest motionless in the
centre, or if it moves, what power moves it?9 As soon as you go beyond those
things which are apparent and fall under sense and experience, to inquire into
things further within, both the discipline of mathematics and all other disciplines
are surrounded entirely by darkness. (3.209a)
Moreover, the appearances include not only what is apparent to us now but also
what may become apparent later. For even in the Exercitationes Gassendi is willing
to allow that we can infer to things beyond the appearances. Such inferences will not
get us all the way to the true natures of things:
If you say that the intellect can infer from the things that happen in experience
or appear to the senses to other, much more hidden things, I will reply that
nevertheless, it can proceed no farther by reasoning than to things which may again
be experienced or things of which some appearance can be exhibited. But what we
deny is that we can penetrate all the way to the inner natures of things. (3.207a–b)
However, although Gassendi allows more knowledge than one might initially
suspect, the knowledge he allows is still far more limited than Aristotelian scientia.
It differs in two other ways as well.10 It lacks the absolute certainty characteristic
of Aristotelian scientia, as Gassendi explains in an attempt to avoid the charge of
self-refutation. The kind of knowledge the skeptic claims to have
is not entirely certain or evident and does not derive from some Aristotelian
cause or demonstration. But nevertheless, it does not lack its own certainty and
evidentness, and hence probability, that relies on very apparent inferences and
reasons that can restrain the intellect from assenting to the proposition that there
is Aristotelian knowledge. So use the word “opinion” or any other term you
like, this doesn’t really matter. For we say “to have an opinion” and “to know”
indifferently, as the manner of speech of the vulgar shows. And if you attend
properly, scientia and opinion can be considered synonyms. And one can say
“certain scientia” and “certain opinion” as well as “weak [imbecillam] scientia,”
and “weak opinion.” (3.206a)11
The certainty in question here is simultaneously psychological and epistemological.
Finally, the justificatory structure of Gassendian probability differs from that

of Aristotelian scientia or, for that matter, Cartesian clear and distinct perception.
Gassendian knowledge need not be deduced from first principles, and it need not
derive from a source whose epistemic bona fides we have established. It is enough,
as we’ll see below, if there is no actual disagreement to provide a reason for doubt.
4 THE AGRIPPAN TRILEMMA

Gassendi does recognize the existence of skeptical tropes beyond the Ten Modes.
Near the end of his history of skepticism, he remarks:
When Pyrrhonism was revived, some other Modes were added, but it is more
than enough to recount the three principal ones. The first is what can be called
regressus in infinitum, where what has been put forward to confirm something
is said to require confirmation by another thing, and that again by another, so
that no way out can be found. The second is diallelus . . . where it is shown that
someone who is giving a proof asserts one thing in order to establish a second,
and asserts the second in order to establish the first. . . . The last is hypothetical,
or from supposition, where, after something has been supposed, someone asserts
that he is permitted to make the contrary supposition. (1.75b)
Using these modes again allows the skeptics to say that there is no criterion (1.75b).
Gassendi’s interest in the Agrippan trilemma is slight in comparison to his interest
in the Ten Modes. This fact requires some explanation. It’s easy to see why the Ten
Modes are dominant in the Exercitationes, which attacks knowledge conceived of as
Aristotelian scientia: “certain and evident cognition of some thing that comes from
knowledge of a necessary cause or a demonstration” (3.192b).12 For
since it is established that all of our knowledge either is sensation or proceeds
from the senses, it also seems to be established that we cannot make a judgment
about anything unless the senses have first given testimony of it. (3.192b)
Thus, if Gassendi can eliminate the beliefs about the natures of things that are
derived from the senses, he has destroyed the foundation of Aristotelian scientia.
One might expect the Agrippan trilemma to play a greater role in Gassendi’s
objections to Descartes. However, it does not. To see why, let’s look at three
different strands in these objections.
The first strand concerns Descartes’s use of skepticism. Gassendi argues, at
some length, that Descartes misunderstands two crucial distinctions: the distinction
between the inner natures of things and their appearances, and the distinction
between the actions of life and the inquiry into truth. Failure to recognize these
distinctions leads Descartes to apply skeptical techniques where they are not
appropriate. Gassendi explains that the skeptics rejected indifference entirely where
the actions of everyday life are concerned:
But for the investigation of truth, they drew a further distinction . . . between
ta phainomena, “the things which appear to the senses,” like the heat of fire,
300 SKEPTICISM
the sweetness of honey, and others; and ta noumena, “the things which are
understood by the mind,” like . . . internal and proximate natures and causes,
and the properties of those same things. And then, although they considered the
latter kind of things uncertain, they still allowed the first kind except when their
state of mind made it pleasant to take on the insolence of the dogmatics. (3.286a)
This is consistent with Gassendi’s treatment of skepticism in the Exercitationes. But
it does not amount to much as a critique of Descartes. Here it seems that Gassendi
has simply misunderstood the purpose of Cartesian skepticism.13
A second line of objection, which concerns Descartes’s reliance on clear and
distinct perception, is more interesting. Gassendi expresses the familiar worry that
Descartes has argued in a circle by using clear and distinct perception to argue that
God exists and then using God’s existence to argue that whatever is clearly and
distinctly perceived is true. But he also worries about whether clear and distinct
perception can serve as the criterion, posing a dilemma for Descartes. On the one
hand, if clear and distinct perception is picked out by its psychological aspect alone,
there will be conflicting clear and distinct perceptions. Different men—all of whom
are honest, all of whom have thought about the matter carefully—perceive different,
incompatible things with clarity and distinctness (3.315a). Moreover, Gassendi
continues, he himself has clearly and distinctly perceived different, incompatible
things at different times. As a youth, for instance, he clearly and distinctly perceived
that two lines continually approaching each other more closely must meet. Later he
learned about asymptotes and came to perceive, with equal clarity and distinctness,
that they need not meet (3.314b). Hence, clear and distinct perception cannot serve
as the criterion of truth.
Descartes would object, of course, that not all of these conflicting perceptions
are genuinely clear and distinct (AT 7.361). Gassendi is somewhat impatient with
this response, replying that “the fact that men go to meet death for the sake of
some opinion seems to be a perspicuous argument that they perceive it clearly
and distinctly” (3.317a). But impatience aside, this leads to the other horn of the
dilemma. If we build truth into the nature of clear and distinct perception, then we
open up a gap between genuine clear and distinct perception and merely apparent
clear and distinct perception. And once this gap is open, we cannot use clear and
distinct perception as the criterion. Thus, nothing as sophisticated as the Agrippan
trilemma is needed.
Another reason for Gassendi to de-emphasize the Agrippan trilemma emerges
when we look at the third line of objection: that the First Meditation skeptical
hypotheses simply fail to produce doubt. For the fact that we make mistakes in certain
circumstances does not threaten the knowledge we have in other circumstances:
When we examine a nearby tower and touch it, we are certain that it is square,
even though when we were further from it we had reason . . . to doubt whether it
was square or round or some other figure. . . . In the same way, the sense of pain
which still appears to be in a foot or hand after the limbs have been amputated
can sometimes deceive . . . but those who are whole are so certain that they
feel pain in the foot or hand which they see pricked that they cannot doubt it.
In the same way, because while we live we are alternately awake and asleep,
we may be deceived through a dream . . . but nevertheless we are not always
asleep and when we are really awake we cannot doubt whether we are awake or
asleep. (3.388a–b)
The dream example is the most interesting one. Gassendi grants that we are
sometimes deceived in dreams and that we sometimes believe we are awake when
we are really dreaming. Nevertheless, he insists, when we are actually awake, we
know that we are awake, and the falsity of beliefs acquired in dreams does not
affect the status of beliefs acquired when awake. Our knowledge that we are awake,
when we are awake, is not mere supposition and does not require further proof: it
is simply obvious to all men.
Gassendi returns to the Agrippan trilemma in the Syntagma, where he makes
essentially the same point:
The three modes added to the ten pertain to the general form of arguing and
objecting. But . . .whenever someone objects that there . . . is no criterion . . .
because anyone who says that there is does so either without a demonstration
or with one, but that either way, etc., it can be said, first, that we have some
demonstration. Although it’s not an Aristotelian demonstration or a demonstration
which requires a precise investigation of some previous sign or criterion or the
like, it’s the sort of demonstration that men who are prudent, intelligent, and
furnished with good sense will generally accept as a proper reason and which
cannot be contradicted except out of sheer contrariety [contradicendi animo].
For it seems that we should stand firm on this and not become frightened of
being drawn into an infinite regress or circularity. (1.85b)
He goes on to explain that
demonstration . . . is not necessary when things are so evident that they just need
to be proposed. Such things include not only particular things which appear to
the senses and are approved by experience, but also general propositions against
which no objections can be brought, such as . . . if equals are taken from equals,
the remainders are equal . . . And although it is countered that someone who
simply makes a declaration and does not prove it should not be believed, and that
the opposite can be asserted hypothetically and claimed as true by anyone . . .
wishing to maintain the other side, it is clear that this can indeed be done in
doubtful matters, where neither experience nor some convincing and reasonable
argument comes to our support, but it cannot be done without folly in other
cases. (1.86a)
The key difference between this and the way Gassendi treats the Ten Modes is
that in the Ten Modes we start from actual disagreement about particular facts.
But in the examples of the Agrippan trilemma Gassendi is considering here, actual
disagreement plays no role. No one is actually denying that if equals are taken from
equals, the remainders are equal. And hence, Gassendi thinks, there is no reason to
start looking for proof in the first place. A similar point is made in the Disquisitio
302 SKEPTICISM
Metaphysica, where Gassendi argues that Descartes’s skeptical hypotheses fail to

raise genuine doubt in cases where “the thing always appears in the same way, as the
sun always appears round and shining, or as the crossing of two straight lines always
creates two right angles or angles equal to two right angles” (3.279b).
5 GASSENDI’S VIA MEDIA

By the time of the Syntagma, Gassendi thinks of skepticism as nothing more than a
useful check on the pretensions of the sciences. Explaining the need for a via media
between the skeptics and the dogmatists, he writes that
the dogmatists do not really know all the things they suppose they know, and
they do not have an appropriate criterion for judging those things. However, it
does not seem that all the things which are thrown into dispute by the skeptics
are so unknown that we cannot have some criterion for judging them. And
because the majority of the things which the dogmatists suppose they know are
really unknown, too often in physics the occasion arises for declaring that we
are fortunate when we arrive not at what is true but at what is probable. (1.79b)
This claim is itself, of course, merely probable (1.79b).
Gassendi’s ultimate view is that there are two criteria of truth. First, the senses,
which serve as the criterion for the appearances. Since even the skeptics allow
knowledge of the appearances, Gassendi does not spend much time on this. The
more interesting criterion is reason, by which we can know the “hidden” truth that
is “lurking under the appearances” (1.80b). How exactly this works depends on why
the truth in question is hidden. Here, again following Sextus, he distinguishes three
categories of hidden truths.
Some things are only circumstantially hidden (1.69a), and these can be known by
memory. When we see smoke but no fire, we form the belief that there is a temporarily
hidden fire because we remember smoke being produced by fire and nothing else.
A second category of things are “entirely hidden, so that they can in no way fall
under the scope of the understanding” (1.68b). These things cannot be known by
any criterion.
The third category contains things that are hidden by nature. These are things
that cannot be evident per seipsas but can come to be known through “indicative
signs,” to use the term of Sextus, who rejected their use (1.68b). Gassendi gives a
number of examples. Motion signifies the void, because we observe motion and infer
that it can exist only if there is empty space for moving bodies to move into (1.81a).
Certain actions that could not be performed without a soul signify the existence of
a soul (1.82b). The fact that iron is attracted by a magnet signifies a power in the
magnet (1.82b). The orderliness of the universe signifies that God exists and created
the universe (1.82b). Refraction signifies that light is a body (1.82a). And sweat
signifies that there are pores in the skin (1.68b).
Gassendi’s use of indicative signs is an important point of difference from the
ancient skeptical tradition, and marks a clear way in which he is not a skeptic by
the time of the Syntagma. The last example—the pores in the skin—is particularly
interesting in this connection because pores belonged to the third category of hidden
things before the invention of the microscope, and now belong to the first category.
Gassendi gives other examples of this as well. We used to infer that mites have feet
from the way they move, but now we can see their feet. We used to infer that the
Milky Way is made up of stars, and can now see the individual stars that compose it
with a telescope (1.82a). Such examples, Gassendi argues, confirm the legitimacy of
inferences from indicative signs. And for all we know,
many of the things which are now hidden from us and which we until now
perceive only by understanding will one day, by some instrument thought up by
our descendants, also become things that are perceived by the senses. (1.82a)
At this point, Gassendi’s via media may seem closer to dogmatism than skepticism. I
think this is right. However, he insists that there’s a big difference between claiming
to know the existence of things and claiming to know their nature. We know that
there is a soul and a God and a power in the magnet because we know there must be
some cause of the things we perceive. What we do not—and cannot—know are the
nature of the soul, of God, and of the magnet’s power (1.82b). We know that there
is some cause without knowing what the cause is like in itself.
This difference is crucial in Gassendi’s explanation of how to escape the
Ten Modes:
It seems that one and the same thing can appear in different ways to different
animals and different men, and even to one and the same man according to
his various senses and his various affections. . . . But although so many various
Phantasiae or appearances are created, nevertheless it’s clear that there is in
the thing or object some general cause which suffices for all the things which
are manifest. And so, to whatever extent the effects are not like each other,
nevertheless there are two things which are certain and can be proven. . . . One
is that there is a cause in the thing itself, or the object, and the other is that there
is a different disposition in the faculties that encounter it. (1.84a)
Consider the action of the sun, which melts wax but hardens clay. We know that
there is something in the sun that enables it to cause both the melting of the wax
and the hardening of the clay. And we know that there is some difference between
the wax and the clay that explains why the sun melts the first but hardens the
second. But we can only characterize these features of the sun, the wax, and the clay
relationally. Gassendi’s claim that we can know appearances but not inner natures,
then, is the claim that we can know the relational properties of things but not the
intrinsic properties that ground them.
NOTES
1. See Popkin (1979: chapter 2) and Floridi (2002). See also the chapters on Montaigne
and Charron in this book.
2. I cite all Gassendi’s works by book, page, and (where relevant) column number in
Gassendi (1964). The Syntagma occupies all of volumes 1 and 2.
304 SKEPTICISM
3. All translations are my own, but I have benefited greatly from consulting translations
by Craig Brush (Gassendi 1972), whose Selected Writings contains a great deal of
material about skepticism; Bernard Rochot (Gassendi 1959; Gassendi 1962); and
Sylvie Taussig (Gassendi 2004).
4. Perhaps surprisingly, there is no real discussion of the skeptics in the Syntagma’s
Ethics 1.1, “Quae felicitatem faciunt, meditanda.” (There’s a brief mention later,
at 2.696b.)
5. Book I was published in 1624. Book II was written around the same time but not
published until after Gassendi’s death.
6. All these authors are mentioned in various places in the body as well. For details, see
Bernard Rochot’s extremely useful footnotes in Gassendi (1959).
7. See Exercitationes 2.6.2, 3.192b ff and, in the Syntagma, Logic 2.3, 1.72b ff.
8. This genealogy doesn’t put a lot of emphasis on the difference between Pyrrhonian
and Academic skepticism, but Gassendi is aware of it. He notes, for instance, that
“when Carneades founded what is called the Third Academy . . . he revived the
criterion . . . not for the true but only for the probable [verisimile]” (1.72b). And in the
Preface to the Syntagma, he tells us that he will follow Sextus’s taxonomy, dividing
the Greeks into three categories: the Dogmatics; Arcesilaus and the Acataleptics, so
called “because they reckon that all things are akatalepsa, incomprehensible”; and
Pyrrho and the Skeptics (1.13b).
9. Gassendi was in fact a committed Copernican. In the preface to the Exercitationes,
he says that in Book IV, “rest will be brought to the fixed stars and the sun: but
motion will be given to the earth, as if it’s one of the planets” (3.102, no columns).
See LoLordo (2014) for more on this.
10. These are also ways in which it differs from Cartesian clear and distinct perception.
11. In later works, Gassendi remains concerned to show how the charge of self-
refutation is avoided. But he makes clear that the skeptic need not be claiming to
have a different form of knowledge than the form she is attacking. Rather, he can
make skeptical suggestions without claiming any kind of knowledge at all. Gassendi’s
explanation here follows Sextus: if someone objects that the skeptic refutes himself
by claiming that nothing is known, the skeptic “responds that while phrases or
propositions of this sort deny others, they also deny themselves, like medications
which are taken into the body to purge the humors and expel themselves along with
them” (1.76b).
12. Gassendi restricts himself to Aristotelian scientia for several reasons. As we’ve
already seen, he wants to allow knowledge of the appearances and to avoid self-
refutation. But he also wants to exclude matters of faith—something which is
grounded in both genuine personal conviction and in pragmatic considerations.
13. It also seems that he has misunderstood the Pyrrhonian notion of ta phainomena,
since in Sextus appearances need not be “things that appear to the senses” but can
be intellectual as well as perceptual.
REFERENCES
Descartes, René. 1996. Oeuvres de Descartes. 11 volumes, edited by C. Adam and P.
Tannery. Paris: Vrin.
Gassendi, Pierre. 1959. Dissertations en forme de paradoxes contre les Aristotéliciens
(Exercitationes paradoxicae adversus Aristoteleos) Livres I et II, edited and translated by
Bernard Rochot. Paris: J. Vrin.
Gassendi, Pierre. 1962. Recherches métaphysiques; ou, Doutes et instances contre la
Métaphysique de R. Descartes et ses réponses (Disquisitio metaphysica; seu, Dubitationes
et instantiae adversus Renati Cartesii Metaphysicam et responsa), edited and translated
by B. Rochot. Paris: Vrin.
Gassendi, Pierre. 1964. Petri Gassendi Opera Omnia in sex tomos divisa. 6 volumes. Lyon:
Laurent Anisson and Jean-Baptiste Devenet, 1658. Reprinted in facsimile with an
introduction by T. Gregory. Stuttgart-Bad Canstatt: Friedrich Frohmann.
Gassendi, Pierre. 1972. The Selected Works of Pierre Gassendi, edited and translated by
C. G. Brush. New York: Johnson Reprint.
Gassendi, Pierre. 2004. Pierre Gassendi (1592–1655): Lettres latines. 2 volumes, translated
by S. Taussig. Turnhout: Brepols.
LoLordo, Antonia. 2014. “Copernicus, Epicurus, Galileo, and Gassendi,” Studies in the
History and Philosophy of Science 51: 82–88.
Popkin, Richard. 1979. The History of Scepticism: From Erasmus to Descartes. Berkeley:
Popkin, Richard (ed.). 1991. Pierre Bayle: Historical and Critical Dictionary: Selections.
Indianapolis and Cambridge: Hackett.
CHAPTER TWENTY-TWO
Descartes and the Force

of Skepticism
DAVID CUNNING
1 INTRODUCTION
In Meditations on First Philosophy, René Descartes (1596–1650) considers a number
of skeptical arguments with the aim of locating beliefs that are immune from all
doubt whatsoever. He writes:
Some years ago I was struck by the large number of falsehoods that I had accepted
as true in my childhood, and by the highly doubtful nature of the whole edifice
that I had subsequently based on them. I realized that it was necessary, once in
the course of my life, to demolish everything completely and start again right
from the foundations if I wanted to establish anything at all in the sciences that
was stable and likely to last. . . . So, for the purposes of rejecting all my opinions,
it will be enough if I find in each of them at least some reason for doubt. (First
Meditation, AT VII 17–18, CSM II 12)1
Descartes does not propose to abandon all of the beliefs that the skeptical arguments
call into question. He will recover those that he can demonstrate to follow from
new and improved grounds. This chapter will treat the skeptical arguments that
Descartes unveils in the First Meditation and their eventual relation to the goal
of securing results that are “stable and likely to last.” Section 2 considers both the
argument from occasional sensory deception and the dream argument. Section 3
treats Descartes’s refrain that the main purpose of the skeptical arguments is to
help us to detach from our senses and then recognize the truth of non-sensory (or
what we today might call a priori) philosophical axioms. Section 4 highlights that
in the final analysis Descartes is not just skeptical about the veridicality of waking
perceptions; he thinks that there are transparent non-sensory philosophical axioms
that entail that the senses hardly provide an accurate representation of reality at all.
Section 5 is a discussion of the hyperbolic doubt that appears in the latter half of the
First Meditation and a discussion of how Descartes thought he had the resources to
overcome it.
DESCARTES AND THE FORCE OF SKEPTICISM 307
2 SKEPTICISM AND THE SENSES

The initial skeptical arguments of the First Meditation call into question our beliefs
about the everyday sensible objects that we presume to surround us. Descartes begins:
Whatever I have up till now accepted as most true I have acquired either from the
senses or through the senses. But from time to time I have found that the senses
deceive, and it is prudent never to trust completely those who have deceived us
even once. (AT VII 18, CSM II 12)
Here he is appealing to a common experience in which a person has a sensory
experience of an object but subsequently concludes that the experience was misleading
or deceptive. As Descartes rehashes the argument later in the Sixth Meditation:
Sometimes towers which had looked round from a distance appeared square
from close up; and enormous statues standing on their pediments did not seem
large when observed from the ground. In these and countless other such cases,
I found that the judgments of the external senses were mistaken. (AT VII 76,
CSM II 53)
The initial conclusion that Descartes draws from this argument from occasional
sensory deception is that we should never put our complete trust in any of our
beliefs about the bodies that surround us.
Descartes retreats from this far-reaching conclusion immediately, however,
just as most of us would do if we were working through the argument along with
him. Presumably, the reason that we would ever call into question a belief about a
distant tower or any other non-ideally situated object is that we have had conflicting
observations of the object in more optimal circumstances—in circumstances that we
fully trust. If so, the definitive conclusion that our senses in some cases mislead us
presupposes that they sometimes (and indeed often) do not. Descartes continues:
Yet although the senses occasionally deceive us with respect to objects which are
very small or in the distance, there are many other beliefs about which doubt is
quite impossible, even though they are derived from the senses—for example,
that I am here, sitting by the fire, wearing a winter dressing-gown, holding this
piece of paper in my hands, and so on. Again, how could it be denied that these
hands or this whole body are mine? Unless perhaps I were to liken myself to
madmen, whose brains are so damaged by persistent vapours of melancholia that
they firmly maintain they are kings when they are paupers, or say they are dressed
in purple when they are naked, or that their heads are made of earthenware, or
that they are pumpkins, or made of glass. But such people are insane, and I would
be thought equally mad if I took anything from them as a model for myself. (AT
VII 18–19, CSM II 12–13)
This seems to be the right thing to say in response to the contention that all of our
sense-based beliefs are suspect if our senses sometimes mislead us: the way that
we tell that such a belief is misleading is by relying on sensory perceptions that we
trust—sensory perceptions “about which doubt is quite impossible.”
308 SKEPTICISM
We tend to think that our beliefs about objects in our immediate vicinity are
unimpeachable, but Descartes actually thinks that our confidence is unfounded even
here. He continues the Meditation by remarking that we sometimes have dreams
that are so vivid that, while we are having them, we cannot differentiate them from
waking experiences. After making clear that he is not a madman, he writes:
A brilliant piece of reasoning! As if I were not a man who sleeps at night, and
regularly has all the same experiences while asleep as madmen do when awake—
indeed sometimes even more probable ones. How often, asleep at night, am I
convinced of just such familiar events—that I am here in my dressing gown, sitting
by the fire—when in fact I am lying undressed in bed! (AT VII 19, CSM II13)
Presumably most of us have had the experience of thinking that a dream was real,
and we might have even had the experience of asking ourselves at a given moment
whether or not we are in a dream and then concluding that we are not, but then we
wake up. Descartes continues:
Yet at the moment my eyes are certainly wide awake when I look at this piece of
paper; I shake my head and it is not asleep; as I stretch out and feel my hand I
do so deliberately, and I know what I am doing. All this would not happen with
such distinctness to someone asleep. Indeed! As if I did not remember other
occasions when I have been tricked by exactly similar thoughts while asleep! As
I think about this more carefully, I see plainly that there are never any sure signs
by means of which being awake can be distinguished from being asleep. (AT VII
19, CSM II 13)
Here Descartes focuses on the vividness of dream experiences to secure the result
that sometimes they are just as real-seeming as experiences that we have while
awake. If so, then for any belief that we take ourselves to have formed in ideal
viewing conditions, that belief might have for its evidence a perception that is not
reflective of how things are. Descartes would appear to have found himself again at
the conclusion that “it is prudent never to trust [the senses] completely.”
As he did in the case of the argument from occasional sensory deception,
Descartes quickly puts the reins on the conclusion that he will draw from the dream
argument. We may not be able to tell the difference between waking experience and
dreaming, he supposes, but for any sensory perception that we have there must exist
the basic elements that are necessary for sensory perception to occur at all:
Suppose then that I am dreaming, and that these particulars—that my eyes are
open, that I am moving my head and stretching out my hands—are not true.
Perhaps, indeed, I do not even have such hands or such a body at all. Nonetheless,
it must surely be admitted that the visions which come in sleep are like paintings,
which must have been fashioned in the likeness of things that are real, and hence
that at least these general kinds of things—eyes, head, hands and the body as a
whole—are things which are not imaginary but are real and exist. For even when
painters try to create sirens and satyrs with the most extraordinary bodies, they
cannot give them natures which are new in all respects; they simply jumble up the
limbs of different animals. Or if perhaps they manage to think up something so

new that nothing remotely similar has ever been seen before—something which
is therefore completely fictitious and unreal—at least the colours used in the
composition must be real. (AT VII 19–20, CSM II 13)
So presumably there exist hands and faces, or at the very least things like size and
shape. We might be wrong to trust any particular sensory perception, but at least we
can trust that generally speaking the external world contains items that are familiar.2
Descartes does not say where it is that he is getting the assumption that we could
not have sensory perceptions unless there existed basic elements out of which those
perceptions are fashioned. The assumption certainly has not been demonstrated at
this point in the Meditations, but one thing to note for now is that Descartes allows a
number of unexamined assumptions in the First Meditation. This is to be expected,
actually. The anticipated outcome of working through the skeptical arguments is that
we will once and for all locate claims that are immune from any possible doubt, but
that means that we are not in possession of any such claims for use in the skeptical
arguments themselves (see also Curley 1978: 50–51; Larmore 2014: 55–59). One of
Descartes’s objectors (Pierre Bourdin) raised this concern to Descartes explicitly—
how he can trust the premises of the skeptical arguments of the First Meditation if
those arguments are meant to finally locate beliefs that are completely certain and if
his beliefs are otherwise unexamined (Seventh Objections and Replies; AT VII 527–
530, CSM II 358–361). Descartes’s response to the concern is dismissive: he points
out that in the construction of a building, the tools that are used to remove the prior
dwelling and any rubble do not need to be of the same quality as the materials that
compose the new foundation (AT VII 538, CSM II 367; AT VII 544, CSM II 371;
AT VII 542, CSM II 370). He adds that in the First Meditation he was “not yet
concerned with establishing any truths” (AT VII 523, CSM II 355) and that “there
is nothing at all that I asserted ‘with confidence’ in the First Meditation: it is full of
doubt throughout” (AT VII 474, CSM II 319). As we have already seen, Descartes
allows a number of unexamined assumptions in the First Meditation: for example,
that what we know we know through the senses; and that the distinction between
waking perceptions and dreams is a distinction between perceptions that are veridical
and perceptions that are not.3 Now he is allowing the assumption that we could not
have sensory perceptions at all unless there existed basic elements that constitute
them. The first result of the Meditations that is worth its salt is “I am, I exist” (AT
VII 25, CSM II 17), and the First Meditation skeptical arguments do a lot of work to
help us to arrive at it.4 As we shall see, they help us to recognize the transparency and
perspicuity of claims that are not known through the senses.
3 SKEPTICISM ABOUT THE A PRIORI

The dream argument supposes that waking experience provides an accurate
representation of reality, that dream experience does not, and that we are in
possession of no clear sign to distinguish the two. What is striking is that in the
final analysis Descartes holds that the external world is not as we know it through
310 SKEPTICISM
the senses, but is a plenum that contains no sensory qualities—no color, no sound,
no taste, no light. Below we will consider his arguments for the view that sensory
qualities are not literally in physical objects—not in anything like the way that
we experience them—but first we must consider his methodological assumption
that non-sensory axioms are the appropriate sort of premise for a philosophical
demonstration. In Section 4, we will see how he attempts to appeal to such axioms
to generate the result that the external world is not as we sense it.
Descartes holds that, upon reflection, the clearest and most obvious results are not
known through the senses. In contrast to the level of certainty that is afforded to the
senses in the early moments of the First Meditation, Descartes in fact thinks that we
have another cognitive faculty to which the senses pale in comparison. He writes:
It is clear that we do not have [perfect] certainty in cases where our perception is
even the slightest bit obscure or confused; for such obscurity, whatever its degree,
is quite sufficient to make us have doubts in such cases. Again, we do not have
the required kind of certainty with regard to matters which we perceive solely by
means of the senses, however clear such perception may be. . . . Accordingly, if
there is any certainty to be had, the only remaining alternative is that it occurs in
the clear perceptions of the intellect and nowhere else. (Second Replies, AT VII
145, CSM II 105–106)
Descartes then provides a list of the sorts of truth that he thinks are undeniable upon
reflection—so long as we are sufficiently detached from the senses. He says:
Now some of these perceptions are so transparently clear and at the same time so
simple that we cannot ever think of them without believing them to be true. The
fact that I exist so long as I am thinking, or that what is done cannot be undone,
are examples of truths in respect of which we possess this kind of certainty.
(Second Replies, AT VII 145, CSM II 106)
The following are examples of [common notions or axioms]: It is impossible for
the same thing to be and not to be at the same time; What is done cannot be
undone; He who thinks cannot but exist while he thinks; and countless others. It
would not be easy to draw up a list of all of them; but nonetheless we cannot fail to
know them when the occasion for thinking about them arises, provided that we are
not blinded by preconceived opinions. (Principles I 49, AT VIIIA 24, CSM I 209)
Descartes thinks that there are axioms that are the most perspicuous that the
human mind can ever confront, and he calls them alternately “primary notions”
or “common notions or axioms.” Given how clear and obvious they are, they are
preferable to claims or contentions that are much less perspicuous, or to contentions
that the axioms entail are false. Still, we can fail to register the force of a primary
notion if we think too much in terms of sensory pictures and images and are not
practiced at thinking abstractly (Principles I 73, AT VIIIA 37, CSM I 220), or if we
are wedded to opinions that entail that the primary notion is implausible or false
(Principles I 72, AT VIIIA 36, CSM I 219–220). Descartes is contending with just
such a meditator at the start of the First Meditation, someone who has not yet
examined his opinions and who is committed to views like that what is known best
is known through the senses. The skeptical arguments are then meant to help us to
recognize that the senses are not the only game in town:
In the First Meditation reasons are provided which give us possible grounds for
doubt about all things, especially material things. . . . Although the usefulness
of such extensive doubt is not apparent at first sight, its greatest benefit lies in
freeing us from all our preconceived opinions, and providing the easiest route by
which the mind may be led away from the senses. (“Synopsis of the following six
Meditations,” AT VII 12, CSM II 9)5
If we detach from our senses and become able to recognize the clarity and force of
the deliverances of the intellect, we will notice that these are much more perspicuous
than the claims that we have asserted heretofore. We will revise and update our
beliefs accordingly, and so long as we stay on the wagon, our former beliefs will be
kept at bay:
In order to philosophize seriously and search out the truth about all the things
that are capable of being known, we must first of all lay aside all our preconceived
opinions, or at least we must take the greatest care not to put our trust in any
of the opinions accepted by us in the past until we have first scrutinized them
afresh and confirmed their truth. Next, we must give our attention in an orderly
way to the notions that we have within us, and we must judge to be true all and
only those whose truth we clearly and distinctly recognize when we attend to
them in this way. When we do this we shall realize, first of all, that we exist in
so far as our nature consists in thinking; and we shall simultaneously realize both
that there is a God, and that we depend on him, and also that a consideration of
his attributes enables us to investigate the truth of other things, since he is their
cause. Finally, we will see that besides the notions of God and of our mind, we
have within us knowledge of many propositions which are eternally true such as
“Nothing comes from nothing.” We shall also find that we have knowledge both
of a corporeal or extended nature which is divisible, moveable, and so on, and
also of certain sensations which affect us, such as the sensations of pain, colours,
tastes and so on. . . . When we contrast all this knowledge with the confused
thoughts we had before, we will acquire the habit of forming clear and distinct
concepts of all the things that can be known. These few instructions seem to me
to contain the most important principles of human knowledge. (Principles I 75,
AT VIIIA 38–39, CSM I 221)
This all sounds fair enough—we should locate axioms that are so obvious as to be
undeniable, and then derive whatever it is that follows from them. We should also
reject whatever our new results entail is false.
4 SKEPTICISM ABOUT THE SENSES IS

WARRANTED UPON REFLECTION
In the final analysis, Descartes does not think that waking perceptions are veridical.
He thinks that our non-sensory cognitive faculties are a better guide to truth, and
312 SKEPTICISM
he appeals to primary notions to generate the result that the external world is very
different from how we sense it. First, we consider a statement of the view:
Now we understand very well how the different size, shape and motion of the
particles of one body can produce various local motions in another body. But
there is no way of understanding how these same attributes (size, shape and
motion) can produce something else whose nature is quite different from their
own—like the substantial forms and real qualities which many <philosophers>
suppose to inhere in things. . . . Not only is all this unintelligible, but we
know that the nature of our soul is such that different local motions are quite
sufficient to produce all the sensations in the soul. What is more, we actually
experience the various sensations as they are produced in the soul, and we do
not find that anything reaches the brain from the external sense organs except
for motions of this kind. In view of all this we have every reason to conclude
that the properties in external objects to which we apply the terms light, colour,
smell, taste, sound, heat and cold—as well as the other tactile qualities . . .—
are, so far as we can see, simply various dispositions in those objects which
make them able to set up various kinds of motions in our nerves <which are
required to produce all the various sensations in our soul>. (Principles IV 198,
AT VIIIA 322, CSM I 285)
Here we have a statement of the view that sensory qualities exist as sensations in
our mind and can be said to exist mind-independently only in the sense that bodies
exist that are able to trigger sensations in us. The passage also includes a gesture at
Descartes’s argument.
First, he makes the claim that there is no way to understand how bodies with
features like size, shape, and motion can produce qualities like color and taste and
sound. Here he appears to be assuming without argument that bodies with features
like size and shape and motion do not already have qualities like color and taste and
sound. If they did, it would not be so difficult to understand how they could produce
color and taste and sound yet again. In fact, Descartes is supposing a background
argument—an argument that piggybacks on one of his arguments for substance
dualism. In that argument, he reflects on the nature of substance and infers that
every substance has a principal attribute in terms of which all of its modifications
can be explained or understood:
Each substance has one principal property which constitutes its nature and
essence, and to which all its other properties are referred. Thus extension in
length, breadth and depth constitutes the nature of corporeal substance; and
thought constitutes the nature of thinking substance. Everything else which
can be attributed to body presupposes extension, and is merely a mode of an
extended thing; and similarly whatever we find in the mind is simply one of
the various modes of thinking. For example, shape is unintelligible except in an
extended thing; and motion is unintelligible except as motion in extended space;
while imagination, sensation and will are intelligible only in a thinking thing.
(Principles I 53, AT VIIIA 25, CSM I 210–211)
To generate the result that minds are immaterial substances, Descartes supposes
that there are thinking modifications that are modifications of a substance and that
any substance also has a principal attribute in terms of which its modifications can
be understood. Descartes also supposes that substances have only one principal
attribute. This is because we recognize (he thinks) that a substance and a principal
attribute constitute a complete thing—that is, they together constitute a thing in
its own right that does not depend on any other creature for its existence (Fourth
Replies, AT VII 223, CSM II 157). A substance with the principal attribute of
thought is a substance that has only modifications that can be understood in terms
of thought; the substance does not have any other kinds of modification because
the substance starts and stops with the attribute of thought, and modifications of a
different kind would be understood in terms of a different principal attribute, which
would be the principal attribute of a different substance (see also Rozemond 1998:
ch. 1). Descartes is then making a similar argumentative move in concluding that
bodies do not have sensory qualities. He thinks that if we have successfully detached
from our senses, we are in a position to think of three-dimensional extension as
colorless, tasteless, soundless, etc., and we recognize it to be a complete thing. If
so, the only modifications of an extended substance are modifications that can be
understood or explained in terms of colorless, tasteless, and soundless extension—
features like size, shape, and motion.
Descartes performs the thought experiment more vividly in the Second
Meditation when he thinks of a piece of wax as not having qualities like color or
taste or sound. He says,
Let us concentrate, take away everything which does not belong to the wax, and
see what is left: merely something extended, flexible, changeable. . . . I would not
be making a correct judgment about the nature of wax unless I believed it capable
of being extended in many more different ways than I will ever encompass in my
imagination. (AT VII 31, CSM II 21)
He appears to hold that, as a matter of fact, when we think of something extended
and flexible in isolation from any sensory qualities, we are thinking of something
that on its own constitutes a substance. Since sensory qualities cannot be explained
or understood in terms of that, they are not modifications of bodies. But we do have
sensations, and these are modifications of our mind (see also Principles I 66–68).
In the end, Descartes offers a pragmatic account of the role of the senses. They do
not inform us of the details of the bodies that surround us, but instead they help us
to keep track of the features of our environment that are relevant to our well-being
and survival. The function of sensory perceptions is not to be veridical—and they
are not veridical—but they do help us to organize our activities and to live our lives:
For the proper use of the sensory perceptions given me by nature is simply to
inform the mind of what is beneficial or harmful for the composite of which the
mind is a part; and to this extent they are sufficiently clear and distinct. But I
misuse them by treating them as reliable touchstones for immediate judgements
about the essential nature of the bodies located outside us; yet this is an area
314 SKEPTICISM
where they provide only obscure information. (Sixth Meditation, AT VII 83,
CSM II 57–58)
It will be enough, for the present, to note that sensory perceptions are related
exclusively to this combination of the human body and mind. They normally tell
us of the benefit or harm that external bodies may do to this combination, and do
not, except occasionally and accidentally, show us what external bodies are like
in themselves. (Principles II 3, AT VIIIA 41–42, CSM I 224; see also (Principles
IV 204, AT VIIIA 327, CSM I 289
Here Descartes is clear that even the most vivid of sensory perceptions is not an
indicator of the details of the bodies that surround us. Instead, sensory perceptions
serve as a guide to help us to preserve our mind-body union (see also Simmons
2014). This setup is presumably very wise, for we might not pay attention to those
aspects of our environment that are relevant to our survival if the signals were not
gripping and loud.
Descartes thinks that our sensory experience of the external world is non-veridical
in another respect as well. Not only does the physical universe have no color, taste,
sound, or light—again, not in anything like the way that we experience them—but it
is a fully dense plenum, and the “empty” spaces that we take to be separate objects in
fact are brimming with being and activity. Once again the reasoning that Descartes
offers here is (and presumably had better be) non-sensory: if there was nothing
between two objects, they would be touching, and so any objects that appear to be
separated by empty space are in fact in contact with bodies that form a continuous
stretch in between (Principles II 18, AT VIIIA 50, CSM I 231). We, however, tend to
think that things are real to the extent that they are sensible, and to the extent that
they are relevant to our needs and well-being, so when we do not sense anything in
the “space” between objects, we conclude that there is nothing there (Principles I 71,
AT VIIIA 35–36, CSM I 218–219).
It cannot be overemphasized just how foreign is the world of external bodies, if
Descartes’s non-sensory argumentation is right. The universe is a colorless plenum,
etc., and there is a lot more that is going on in our local (and our faraway) environment
than we would be prepared to admit. Descartes does not highlight the drama of the
result here, but a passage from one of his contemporaries does the job nicely:
Though it was not the intention of God or nature to abuse us herein, but a
most wise contrivance thus to beautify and adorn the visible and material world,
to add lustre or embellishment to it, that it might have charms, relishes, and
allurements in it, to gratify our appetites. Whereas otherwise reality in itself,
the whole corporeal world in its naked hue, is nothing else but a heap of dust or
atoms, of several figures and magnitudes, variously agitated up and down. So that
these things which we look upon as such real things without us, are not properly
the modifications of bodies themselves, but several modifications, passions, and
affections of our own souls. (Cudworth 1996: 148–149)6
For Descartes, the distinction between waking and dreaming is not a distinction
between perceptions that are veridical and perceptions that are not. If we converge
on primary notions that are not known through the senses, we can conclude on
the firmest grounds possible that the world is a plenum that is devoid of sensory
qualities and that the world that appears to us is an incomplete but attention-getting
snapshot by which we can notice and register what is pertinent to our survival.
Descartes thus has a view of sensory faculties that is similar to the view that we
find in Plato. Descartes does not just think that the senses deliver less probability or
certainty than the intellect—where sensory perceptions might be exactly veridical,
but we just cannot be sure. Instead, he holds with Plato that when it comes to
metaphysical truth sensory perception is off kilter:
Then what about the actual acquiring of knowledge? Is the body an obstacle
when one associates with it in the search for knowledge? I mean, for example, do
men find any truth in sight or hearing, or are not even the poets forever telling us
that we do not see or hear anything accurately, and surely if those two physical
senses are not clear or precise, our other senses can hardly be accurate, as they
are all inferior to these. . . .Then he will [grasp knowledge] most perfectly who
approaches the object with thought alone, without associating any sight with his
thought, or dragging in any sense perception with his reasoning, but who, using
pure thought alone, tries to track down each reality pure and by itself. (Plato,
Phaedo 65a–66b; I use the translation in Plato 2002)
Descartes agrees that the picture of reality that is secured through the senses is hardly
accurate at all. He is using the skeptical arguments of the First Meditation, not to
show that our beliefs about the external world are less probable than we thought,
but to get us to practice a non-sensory kind of thinking that (on his view) reveals
with full confidence that the external world is almost nothing like we experience
it. He recognizes that the First Meditation skeptical arguments are not especially
original (see also Fine 2000). What is original, he thinks, is how he repurposes them
in order to help us to detach from our senses and arrive at non-sensory results that
are the real guide to how things are.
5 HYPERBOLIC DOUBT IS UNINTELLIGIBLE

UPON REFLECTION
We still need to consider the remaining skeptical arguments that Descartes advances.
After offering the dream argument, he introduces an even more radical worry—that
perhaps we are wrong about matters that we find to be completely evident. As he
says in his Sixth Meditation restatement of the worry:
Since I did not know the author of my being (or at least was pretending not to), I
saw nothing to rule out the possibility that my natural constitution made me prone
to error even in matters which seemed to me most true. (AT VII 77, CSM II 53)
In the First Meditation, he offers three versions of the worry that perhaps our minds
are deceived about matters that we find completely evident. In the first, he considers
that since God is omnipotent, He certainly has the power to create our minds to
316 SKEPTICISM
be defective, and since our own occasional errors are consistent with His goodness,
for all we know He has reason for creating our rational faculties to be deceived
all the time. Descartes does not fill in the details as to why this rationale might be
compelling, but we might speculate—perhaps God wants us to recognize that our
rational faculties are worthless so that we focus more of our energies on faith.7
In the second version, Descartes concedes that for all we know God does not
exist, and that, if so, our minds are the product of a creator of a different kind. He
says that perhaps
I have arrived at my present state by fate or chance or a continuous chain of
events, or by some other means; yet since deception and error seem to be
imperfections, the less powerful they make my original cause, the more likely it
is that I am so imperfect as to be deceived all the time. (AT VII 21, CSM II 14)
As in the first version of the argument, Descartes is making an assertion to the effect
that something is possible, and then he is concluding from that assertion that it is
possible that our minds are mistaken about matters that are completely evident to us.
In the third version, Descartes returns to the question of the motives of a fully
benevolent God and allows that presumably such a being would not have created us
with defective minds, but still there might exist an evil demon that is constantly at
work to deceive us (AT VII 22–23, CSM II 15). Descartes supposes that this being
might be so powerful and malicious that “it is not in my power to know any truth”
(AT VII 22–23, CSM II 15, emphasis added).8 We recognize that two and three add
to five, but perhaps a demon is tricking us into thinking that they add to five when
in fact they add to something else.
At this point we might want to ask whether or not Descartes is entitled to assume
the various possibilities that undergird the three versions of the First Meditation
skeptical argument about the reliability of our minds. The reason that we want to
ask is that there are passages later in the Meditations and also outside the Meditations
in which Descartes says without prevarication that God cannot deceive and is a
necessary existent and the author of all reality—our minds included. For example,
in the Third Meditation he concludes that God exists, is not a deceiver, and is the
creator of all things in such a way that nothing is independent of Him (AT VII 45,
CSM II 31; AT VII 51–52, CSM II 350). In the Fifth Meditation he argues that God
exists necessarily and that deception is inconsistent with His nature (AT VII 65–71,
CSM II 45–49). Outside of the Meditations, to one of his correspondents, he writes,
“God is thought of as a deceiver.” This is foolish. Although in my First Meditation
I did speak of a supremely powerful deceiver, the conception there was in no way
of the true God, since, as he himself says, it is impossible that the true God should
be a deceiver. But if he is asked how he knows this is impossible, he must answer
that he knows it from the fact that it implies a conceptual contradiction—that is,
it cannot be conceived. (Letter to Voetius, May 1643, AT VIII B 60, CSMK 222)9
Descartes thinks that it follows from non-sensory axioms that God is a necessary
existent and that there does not exist the possibility that we are deceived about
matters that are completely evident to us. That is, he thinks that it follows from
non-sensory axioms that there does not exist the possibility that God created us with
defective minds, or the possibility that God did not create us, or the possibility that
God created an evil demon that is constantly deceiving us. If Descartes is right that
all of these follow from undeniable primary notions, then he would conclude that
none of the First Meditation possibilities is a possibility at all. They are unexamined;
they conflict with results that are transparent and obvious; and as per Principles I 75
they should be dismissed.
Descartes does not finally arrive at an Archimedean point until early in the
Second Meditation, and the skeptical results of the First Meditation help him to
do that. The first two skeptical arguments work to neutralize our allegiance to the
senses, with the goal that we might recognize non-sensory axioms as a higher caliber
point of departure. The remaining skeptical arguments turn us yet further away
from the senses and begin to focus our attention on our minds and their activities
and contents. We are not yet ready to conclude “I am, I exist” with the highest level
of clarity, however (see also Cunning 2007). Descartes assumes that because we
are so practiced at thinking in terms of sensory pictures and images, we enter the
Meditations with ideas of mind and God that are highly adulterated—we think of
God as a bearded man on a cloud, for example, and of mind as something that can
be pictured and touched and felt.10 The hyperbolic arguments in the second half
of the First Meditation help us to focus our attention on mind, and then when we
finally think of mind in a way that does not involve the senses, we arrive at the sort
of non-sensory result that Descartes thinks is the bread and butter of philosophical
demonstration and that makes other results pale in comparison. At the start of the
Second Meditation, we consider whether there is anything that is still true even if
we have no senses and our minds are deceived. Descartes writes:
Is there not a God, or whatever I may call him, who puts into me the thoughts
I am now having? But why do I think this, since I myself may be the author of
these thoughts? In that case am not I, at least, something? But I have just said
that I have no senses and no body? This is the sticking point: what follows from
this? . . . But I have convinced myself that there is absolutely nothing in the
world, no sky, no earth, no minds, no bodies. Does it not follow that I too do not
exist? No: if I convinced myself of something then I certainly existed. But there
is a deceiver of supreme power and cunning who is deliberately and constantly
deceiving me. In that case too I undoubtedly exist, if he is deceiving me; and let
him deceive me as much as he can, he will never bring it about that I am nothing
so long as I think that I am something. (AT VII 25, CSM II 16–17)
Here Descartes is using the skeptical arguments to zero in on a result that can
survive them. He proceeds to converge on other non-sensory results as the
Meditations unfolds: that body is essentially extended, changeable, and flexible
(and in ways that neither the senses nor imagination are able to capture); that
something cannot come from nothing (or, as Descartes puts it in the case of ideas
in the Third Meditation, that the objective reality of an idea requires a cause with
at least as much formal reality)11; and that the idea of God has infinite objective
reality. When we recognize the truth of these claims and others like them, then so
318 SKEPTICISM
long as we are sufficiently disciplined to trust what is clear upon reflection over
what is clear at first glance, we will appreciate that the contentions of the skeptic
have to be weighed against non-sensory primary notions, and Descartes thinks that
the former cannot compete.
Descartes did not take skeptical arguments to be threatening at all, except to
beliefs that deserve to be cast aside. His own skeptical arguments might upend
some of our beliefs, however. If we think that the distinction between waking and
dreaming is a distinction between perceptual experience that tracks what the outside
world is like and perceptual experience that does not, or if we think that there really
exists the possibility that we were created by chance, or the possibility that an evil
demon is deceiving us, then we have a lot of work to do. Descartes did not think
that these possibilities existed. He did not take skeptical arguments to be serious.
He had a much higher opinion of the non-sensory axioms that come in the Second
Meditation and beyond.
NOTES
1. Translations are from Descartes (1984/5, 1991). References also specify the standard
pagination in Descartes (1996).
2. Bernard Williams (1978: 54) famously remarks that the dream argument is highlighting
the universal possibility of illusion, but not the possibility of universal illusion.
3. Descartes takes all of these to be false. For example, he argues at the end of the
Second Meditation that a central component of any belief that we have about a body
is an idea of extension that is not acquired through the senses (AT VII 30–32, CSM II
20–21). See also Cunning (2010: ch. 2).
4. See also the discussion in Broughton (2008: 97–98). Broughton also makes a number
of nice connections between the skeptical arguments of the First Meditation and the
argumentation in Descartes’s dialogue, The Search for Truth.
5. See also “Dedicatory Letter to the Sorbonne,” AT VII 4, CSM II 5; Second Replies,
AT VII 130–131, CSM II 94; and Third Replies, AT VII 171–172, CSM II 121.
6. See also Berkeley: “It is thought strangely absurd that upon closing my eyelids, all
the visible objects round me should be reduced to nothing; and yet is not this what
philosophers commonly acknowledge, when they agree on all hands, that light
and colours, which alone are the proper and immediate objects of sight, are mere
sensations that exist no longer than they are perceived?” (Berkeley 1975: I 46, 105–
106). See also Cavendish 1655: 174, and Cavendish 1653: 44–45.
7. See, for example, the discussion of seventeenth-century views of faith versus reason
in Popkin (1979: ch. 7).
8. See also the Second Meditation: “I will make an effort and once more attempt the
same path I started on yesterday. Anything which admits of the slightest doubt I will
set aside just as if I had found it to be wholly false; and I will proceed in this way until
I recognize something certain, or, if nothing else, until I at least recognize for certain
that there is no certainty” (AT VII 24, CSM II 16).
9. See also Principles I 14; AT VIIIA 8, CSM I 197–198.
10. Second Replies, AT VII 130–131, CSM II 94; Fifth Replies, AT VII 365, CSM II
252; Principles I 71–73, AT VIIIA 35–37, CSM I 218–220.
11. See the Third Meditation, AT VII 40, CSM II 28; and Second Replies, AT VII 135,
CSM II 97.
REFERENCES
Berkeley, George. 1975. Principles of Human Knowledge. In M. Ayers (ed.), Philosophical
Works. London: J. M. Dent.
Broughton, Janet. 2008. Descartes’s Method of Doubt. Princeton: Princeton University
Press.
Cavendish, Margaret. 1653. Poems and Fancies. London: Printed by T. R. for J. Martin
and J. Allestrye.
Cavendish, Margaret. 1655. The Worlds Olio. London: Printed for J. Martin and J. Allestrye.
Cudworth, Ralph. 1996. A Treatise Concerning Eternal and Immutable Morality, edited by
S. Hutton. Cambridge: Cambridge University Press.
Cunning, David. 2007. “Descartes on the Dubitability of the Existence of Self,” Philosophy
and Phenomenological Research 74: 111–131.
Cunning, David. 2010. Argument and Persuasion in Descartes’ Meditations. New York:
Cunning, David (ed.). 2014. The Cambridge Companion to Descartes’ Meditations.
Curley, E. M. 1978. Descartes against the Skeptics. Cambridge, MA: Harvard University
Press.
Descartes, René. 1984/5, 1991. The Philosophical Writings of Descartes, 3 vols, edited
by J. Cottingham, R. Stoothoff, D. Murdoch, and A. Kenny. Cambridge: Cambridge
University Press.
Descartes, René. 1996. Œuvres de Descartes, 11 vols, edited by C. Adam and P. Tannery.
Paris: Vrin.
Fine, Gail. 2000. “Descartes and Ancient Skepticism: Reheated Cabbage?,” The
Philosophical Review 109: 195–234.
Larmore, Charles. 2014. “The First Meditation: Skeptical Doubt and Certainty.” In
D. Cunning (ed.), The Cambridge Companion to Descartes’ Meditations, 48–67.
Plato. 2002. Phaedo. In Plato: Five Dialogues, 2nd edn., translated by G. M. A. Grube.
Popkin, Richard. 1979. The History of Skepticism, from Erasmus to Spinoza. Berkeley:
California University Press.
Rozemond, Marleen. 1998. Descartes’s Dualism. Cambridge, MA: Harvard University Press.
Simmons, Alison. 2014. “Sensory Perception of Bodies: Meditation 6.5.” In D. Cunning
(ed.), The Cambridge Companion to Descartes’ Meditations, 263–274. Cambridge:
Williams, Bernard. 1978. Descartes: The Project of Pure Enquiry. London: Penguin.
CHAPTER TWENTY-THREE
Varieties of Modern
Academic Skepticism:
Pierre-Daniel Huet and
Simon Foucher
MICHAEL W. HICKSON
1 INTRODUCTION
Thanks in large part to decades of research by Richard H. Popkin, it is now beyond
doubt that there was a revival of ancient skepticism in the early modern period.
However, current historians have begun to doubt Popkin’s claim that the skeptical
problem in early modernity was “more aptly described as a crise pyrrhonienne than
as a crise academicienne” (Popkin 2003: xx). Following Popkin, the literature on the
history of skepticism in the early modern period was for decades almost exclusively
focused on the revival of ancient Pyrrhonism as it is outlined by Sextus Empiricus.
Whether ancient Academic skepticism—that of Arcesilaus, Carneades, and Cicero
for example—influenced the writings of seventeenth- and eighteenth-century
skeptics was mostly ignored. While Charles B. Schmitt (1972) demonstrated that
Cicero’s Academica was the focus of scholarly interest up to the Renaissance, it was
not until the work of José R. Maia Neto (1997) that historians of skepticism took
seriously the idea that early modern skepticism was much indebted to the ancient
Academics. There may have been a crise academicienne after all.
In a recent book, Maia Neto (2014) offers a broad interpretation of early
seventeenth-century philosophy that, like Popkin’s work (2003), focuses on the
importance of skepticism, but this time Academic skepticism. The unifying thread
is Pierre Charron’s 1601 work, De la sagesse, which lays out a skeptical account of
wisdom, which Charron describes as the highest excellence achievable by human
beings, but one which “does not presuppose the attainment of certain knowledge as
the dogmatist would claim, but only full accomplishment of its integrity” (Maia Neto
2014: 26). Charron understands integrity in the way expressed in the Academica,
VARIETIES OF MODERN ACADEMIC SKEPTICISM 321
where Cicero suggests that Academics “are more free and untrammeled in that [they]
possess [their] power of judgment uncurtailed [integra nobis est iudicandi potestas]”
(Cicero 1979: 475; Maia Neto 2014: 25). Maia Neto argues that Charronian
skeptical wisdom was a central concern of philosophers in the early seventeenth
century, some of whom defended the concept (La Mothe Le Vayer and Gassendi),
others of whom attempted to go beyond mere integrity to the possession of certain
knowledge (Pascal and Descartes). Maia Neto’s groundbreaking article continues
this interpretation into the late seventeenth century, when “Academic skepticism
had a spectacular comeback . . . with some of the major anti-Cartesians of the
period” (1997: 203). Other scholars (Lennon 2003a, 2008; Charles 2013a, b) have
added support to Maia Neto’s claims that Pierre-Daniel Huet and Simon Foucher,
in particular, were greatly influenced by Academic skepticism. Even Popkin, in the
last edition of his magnum opus, concedes that Foucher is better understood as an
Academic than as a Pyrrhonist (Popkin 2003: 275).
Classifying past philosophers with labels such as “skeptic,” “Pyrrhonist,”
or “Academic” is little more than a game, however, unless the labels highlight
similarities between those philosophers’ texts and ultimately help us to understand
the texts better. So in this chapter I ask whether labeling Huet and Foucher Academic
skeptics helps us to identify important similarities shared by these philosophers.
In good skeptical fashion, I will argue “yes and no.” Both Huet and Foucher can
meaningfully be called Academic skeptics. The label “Academic” applies to Huet
and Foucher because both wrote histories of the ancient Academics and both tried
to revive core aspects of Academic skepticism. Moreover, the common label focuses
our attention on a common goal of the philosophers: to undermine dogmatic
aspects of Cartesianism by means of self-styled Academic arguments. However,
matters quickly become complicated because Huet and Foucher had radically
different conceptions of Academic philosophy, and both explicitly rejected the
label of “Academic skepticism” for their works. Moreover, while both Huet and
Foucher principally attacked Descartes, Huet’s attack is complete and unrelenting,
undermining Cartesian epistemology and metaphysics, while Foucher’s attack
is limited to Cartesian metaphysics and, ironically, is based firmly on Descartes’s
epistemology, which Foucher even sought to defend against the radical sort of
skepticism advanced by Huet.
In what follows, I give overviews of the lives and works of Huet and Foucher, and
I present and analyze each philosopher’s interpretation of Academic philosophy.
The theme of skepticism in the writings of these two figures is too broad to treat
exhaustively in this chapter, so I limit my attention to the skeptical attacks of each
thinker on the philosophy of Descartes, and in particular to the views of Huet and
Foucher on Descartes’s criterion of truth. These are the topics that best expose the
similarities, but also the differences, between these Academic skeptics.
2 PIERRE-DANIEL HUET: LIFE AND WORKS

Pierre-Daniel Huet (1630–1721) was a quintessential citizen of the Republic
of Letters—erudite and eloquent, well-connected and well-traveled, polymath,
322 SKEPTICISM
polyglot, and prolific.1 From his pen we have an edition and Latin translation of
Origen’s biblical commentaries; a spin-off treatise on translation; scores of Latin
and French poems on subjects as diverse as the reign of Louis XIV and the chemical
nature of salt; a refutation of Descartes’s philosophy (Huet 1689, 1694/2003); a
satirical story about the Cartesians; a novel, as well as a history of the genre of
novels; a treatise on the location of the Garden of Eden; and a work that could have
been called Summa skeptica but was instead entitled A Philosophical Treatise on the
Weakness of the Human Mind (Huet 1723; henceforth, Treatise).2
Huet was born in the village of Caen, France, to noble parents, both of whom he
lost by the age of six. The gloomy tone of Huet’s Memoirs becomes much cheerier
once Huet begins his study of belles lettres and philosophy at the Jesuit College in
Caen, which he later fondly remembered as his “second parent and foster mother”
(Huet 1810: vol. 1, 57). A scholarly career was launched that would quickly gain
the attention of Europe’s greatest minds, as well as its most renowned monarchs.
Queen Christina beckoned Huet numerous times to join her in Sweden, and King
Louis XIV made Huet, along with Bossuet, one of the Dauphin’s preceptors. Huet’s
prestige increased in 1674 with his election to the Académie française, and then
again in 1692 when he was consecrated bishop of Avranches.
Precisely how, when, and why Huet became a fierce opponent of Cartesianism
and a skeptic is debated (see Lennon 2008: 6–29). From Huet’s Memoirs we learn
the following relevant details. Shortly after completing his study of philosophy
with the Jesuits in Caen (in the mid-1640s), Huet reports that he learned about
Descartes’s new system of philosophy, and that he “could not rest till I had procured
and thoroughly perused his book; and I cannot easily express the admiration which
this new mode of philosophizing excited in my young mind, when, from the simplest
and plainest principles, I saw so many dazzling wonders brought forth” (Huet 1810:
vol. 1, 29). But Huet would not remain Cartesian forever:
I was for many years closely engaged in the study of Cartesianism . . . and I long
wandered in the mazes of this reasoning delirium, till mature years, and a full
examination of the system from its foundations, compelled me to renounce it, as
I obtained demonstrative proof that it was a baseless structure, and tottered from
the very ground. (Huet 1810: vol. 1, 30)
To understand why Huet ultimately rejected Descartes, we need another piece of
Huet’s intellectual biography. Around the time he discovered Descartes, Huet also
discovered the study of ancient “Oriental” languages and literatures through his
friendship with Samuel Bochart (see Shelford 2007: 35–39). The war of influence
between Bochart’s antiquarianism and Descartes’s modernism in Huet’s young
mind was unstable and ultimately forced Huet to take a side. Huet’s rejection of
Cartesian philosophy likely dates to 1674, when he read Malebranche’s Search after
Truth (1674/1997), parts of which attack the kind of erudite learning that Bochart
helped Huet to appreciate and cultivate (see Lennon 2008: 21–29). Huet sided with
antiquity.
Not surprisingly, Huet later found himself, as a member of the Academie
française, in the midst of the Battle of the Ancients and Moderns (officially begun
in 1687). Huet discusses the Battle several times in his Memoirs and indicates that
those like Marets and Perrault who championed the moderns did so because of
arrogance, ignorance, lack of taste, insanity, and to their “perpetual disgrace” (see
Huet 1810: vol. 2, 2, 186, 189). Arrogance and ignorance are again the cardinal
vices Huet identifies in Malebranche’s Search after Truth (see Lennon 2008: 23).
The Cartesians are obsessed with appearing novel, yet—and this is what irks Huet
more than anything else—they rarely advance beyond the ancient philosophers,
and they plagiarize their best arguments (Huet 1694/2003: 214–229). Fifteen years
elapsed between Huet’s reading of Malebranche and the publication of his first anti-
Cartesian work. The impetus finally to make his objections to Descartes and the
Cartesians public may have been Foucher, whose anti-Cartesian works discussed
below demonstrated to Huet that Cartesianism was not only misguided but also
dangerous to Christianity (see Lennon 2003a: 122–128; 2008: 32–41; Rapetti
2003: 111–142).
The weapon ultimately employed by Huet against the Cartesians was skepticism,
which, ironically, Huet first encountered in the writings of Descartes (see Maia
Neto 2008b for the Cartesian roots of Huet’s skepticism). Huet knew only the
name of Sextus Empiricus until around 1660, when he met Louis de Cormis, a
former president of the Parliament of Aix, who became Huet’s close friend and who
spoke of little else besides the ancient philosophers. It was Cormis who first urged
Huet to peruse the works of Sextus. Later, in the 1680s, Huet recounts that he
immersed himself again in the study of ancient philosophy, this time with Diogenes
Laertius as his guide, along with the commentary on Diogenes by Huet’s friend,
Gilles Menage. As we learn from the posthumous Treatise, Huet found through his
careful study of the ancients proof after proof that “the truth cannot be known by
human understanding, by means of reason, with perfect and complete certainty”
(Huet 1723: xxix).
Huet’s most skeptical writings—the Treatise3 (Huet 1723), Censura philosophiae
cartesianae (Huet 1689; Huet 1694/2003), and Alnetanae Quaestiones (Huet 1690),
all composed between 1680 and 1690—were meant to be parts of a single treatise
with the sections corresponding roughly to the works just given in the order just
given (see Maia Neto 2008a: 167–172). Huet’s system would establish the limits
and the correct use of reason in supporting the Catholic faith. For a variety of
reasons (see Maia Neto 2008a: 164), Huet abandoned the larger, unified project
and the works were published separately. Together, these works elaborate a claim
that Huet made at the outset of his earlier Demonstratio evangelica (Huet 1679),
namely, that skepticism, which renders doubtful and uncertain everything that we
know by reason and the senses, is not opposed to religion (Huet 1679: 5).
Still, Huet did not wish to urge an irrational faith upon his readers. It was his
contention that the careful study of antiquity—the languages, culture, times, and
places when the Hebrew and Christian Revelations were given—was the best
preparation for the faith, though this preparation could only ever achieve probability.
However, since the time of Descartes, philosophers and even some theologians were
pleased with the use of reason in the preambles of faith only if reason attained
the highest degree of certainty—clarity and distinctness—as it allegedly did in
324 SKEPTICISM
Descartes’s proofs for the existence of God. But in Huet’s view, reason at its best
never attains such a pitch of certainty, so if the bar is set this high, reason’s real,
but humble, achievements in support of faith will be disdained. So Huet had to
correct his readers’ expectations and show the value of the correct antiquarian
path that reason paves toward the faith (see Quantin 1994). The Treatise deals with
reason along with its limits, weaknesses, and need for the faith. The Censura is an
application of the Treatise, refuting the Cartesian view of reason while attempting
to repair the damage done by it to the faith. The Alnetanae Quaestiones finally treat
the proper accord of faith and reason, while attempting to prove that the study of
antiquity leads to the faith. There is much to say about these three works, but in
what follows I limit myself to important aspects of the Treatise and Censura that
clarify Huet’s relation to Academic skepticism and to Foucher.
3 HUET’S INTERPRETATION AND PRAISE

OF ACADEMIC SKEPTICISM
Book one of Huet’s Treatise is devoted to proving that the “truth cannot be known by
the human understanding, by means of reason, with perfect and complete certainty”
(Huet 1723: xxix). Huet opens the book with twelve philosophical arguments for
this conclusion, derived mainly from classical skeptical tropes, such as the lack
of a criterion of truth, the unreliability of the senses, and the changing nature of
things. The thirteenth argument is historical and takes up the rest of book one. It
involves a sweeping history of philosophy, from Pythagoras to Maimonides, which
demonstrates that the inaccessibility of the truth by human reason has always been
acknowledged to varying extents by the best writers. The bulk of this thirteenth
argument is a lengthy treatment of the various ancient Academies, their evolution,
and their relation to Pyrrhonian skepticism. Huet’s conclusion is that only the
Academic and Pyrrhonian skeptics, who constitute a single sect in Huet’s view, fully
appreciated the inaccessibility of absolute certainty in the human and divine sciences.
From the outset of his history of the Academy, it is clear that Huet associates the
school primarily with doubt and the profession of ignorance: “From [Socrates] there
emerged several sects of philosophers, of which the most famous, the Academy,
followed the wise method of doubting everything, and even augmented and brought
this method to its fullest perfection” (Huet 1723: 107). The Academy evolved over
time, but each instantiation was just another branch of a single tree, whose trunk
was “this first principle, posited by Socrates, that man knows nothing” (Huet 1723:
132). After surveying historical rival attempts to enumerate ancient Academies,
Huet settles on the following division: “We must agree, therefore, that properly
speaking there were two Academies: the old [l’ancien] Academy, that of Socrates
and Antiochus, and the new Academy, that of Arcesilaus, Carneades, and Philo; and
I maintain that the new Academy was nothing other than Pyrrhonian philosophy”
(Huet 1723: 138).
The method of the old Academy involved doubt and the constant refutation of
the views of others. It led, however, to a positive assertion that brought consolation,
namely, that a very limited wisdom is available through the recognition that “all
that I can know is that I know nothing.” The new Academy followed but greatly
extended the method of doubt and disagreement of the old Academy, even to the
point of renouncing Socrates’s famous adage just quoted: we cannot even know
that we know nothing. It was Arcesilaus who brought the Academy to this skeptical
extreme: “We must consider [Arcesilaus], therefore, as not only the restorer, but
also as the reformer of the doctrine of Socrates and of the old Academy. He is the
one who gave birth to the new Academy, which is based on a more solid foundation
than the old” (Huet 1723: 111). Arcesilaus was converted to his way of thinking by
Pyrrho, and though he retained the name of the Academy, Arcesilaus turned it into a
Pyrrhonian school. Although the views of the next head of the Academy, Carneades,
differed from those of Arcesilaus in the particulars, fundamentally the two were
of like mind because they denied the existence of any criterion of scientific truth.
Huet’s only remark about the next head, Philo, is that he too taught that we cannot
understand anything completely by means of reason.
The new Academics were therefore all Pyrrhonian skeptics, who, according
to Huet, were characterized by the facts that “they acknowledged no truth rule,
no reasoning, no mark for recognizing the truth; they affirmed nothing, defined
nothing, judged nothing; they did not even believe that a thing was more this than
that” (Huet 1723: 125). Like the Pyrrhonists, the Academic skeptics were zetetics;
the work of their philosophy involved searching for the truth, but never settling
on anything in particular as The Truth, not even the assertion that the truth is
unattainable. Huet considers numerous possible differences between the Academic
and Pyrrhonian skeptics, but he refutes all of them, reducing the differences to mere
language (Huet 1723: 139–150).
Huet devotes so much attention to Academic skepticism in his Treatise that the
reader is led to believe that the author of this work must himself be an Academic
skeptic. But in book two of the Treatise, which is an exposition of the “surest
and most legitimate way of philosophizing,” Huet denies very explicitly that he
belongs to the sect of Academic skepticism, or to any sect for that matter: “Let us be
especially careful never to become attached to the opinions of any particular author,
or to become members of any sect” (Huet 1723: 213). True philosophers must
instead follow their own philosophies, taking whatever seems best from all sects
and all authors. The way to determine what is best among the competing views is to
“weigh all things in the balance of our mind, reserving at all times a complete liberty
to think and to speak on every philosophical subject” (Huet 1723: 213).
Yet this disavowal of every sect is in fact a strong argument for interpreting
Huet as an Academic skeptic, and for recognizing a distinction in Huet’s mind
(malgré lui) between Academic and Pyrrhonian skepticism. According to Cicero
(Acad. II 8, in Cicero 1979: 475), it is part of the nature of Academic skepticism to
uphold the freedom of one’s judgment and consequently to refrain from settling
in any camp and from accepting any label for one’s views: Academic skeptics must
always deny that they are members of any school, including that of Academic
skepticism. Sextus Empiricus, on the other hand, permitted at least a tentative
commitment to the school of Pyrrhonism (PH I.16-17; Sextus Empiricus 2000: 7).
326 SKEPTICISM
After refusing to pledge allegiance to any philosophy, Huet tellingly justifies this
practice by citing a number of authors who were likewise fiercely independent
thinkers: Arcesilaus, Carneades, Philo, and Antiochus (Huet 1723: 213–214)—
all Academics!
That said, it is important always to recall that Huet’s revival of Academic
skepticism is in many ways indistinguishable from a revival of Pyrrhonian skepticism.
We will see this most clearly in the next section when we consider Huet’s skeptical
objections to Descartes which are largely based on Pyrrhonian skeptical tropes.
However, Huet’s great interest in Academic skepticism, evidenced by his lengthy
history of it that was given pride of place in his Treatise; his emphasis on the
Academic goal of complete freedom of judgment (noted in the paragraph above);
his clear preference for the title “Academic” rather than “Pyrrhonist” based on the
long list of distinguished philosophers who held the former (Huet 1723: 150); and
his very explicit rejection of the goal of ataraxia for his philosophy (Huet 1723:
215), give us reasons to follow Maia Neto in classing Huet among the Academic,
rather than Pyrrhonian, skeptics. Moreover, as we will see below, one of Huet’s
contemporaries, Foucher, considered Huet part of a revival of the ancient Academy.
Nevertheless, the ambiguity between Academic and Pyrrhonian skepticism in Huet’s
works is one of the chief reasons for my own skepticism, noted in the Introduction
above, about the use of the term “Academic” to denote the spirit of the French anti-
Cartesians of the late seventeenth century.
4 HUET’S OBJECTIONS TO DESCARTES’S

CRITERIA OF TRUTH
Huet put his skepticism to work against the Cartesians, who claimed to possess not
only some part of wisdom, but more importantly the very foundation of all wisdom:
the criteria of truth. All eighteen sections of the second chapter of the Censura are
devoted to undermining Descartes’s criteria of truth—criteria, not criterion, since
Huet begins by observing that there are at least three criteria that Descartes employs
in his philosophical works: the natural light, clarity and distinctness, and evidence
[évidence]4 (Huet 1694/2003: 120). Huet observes that some Cartesians treat I
think, therefore, I am as if it were itself a criterion, but Huet considers this view
absurd since propositions, their truth, and the mark of their truth must be distinct
things (Huet 1694/2003: 115). Thus the skeptical objections to Descartes’s criteria
will focus on the natural light, clarity and distinctness, and evidence. Objections
to criteria of truth are central to Huet’s Academic skepticism, for in his view “it
is one of the main points of skeptical teaching that nothing is evident” (Huet
1694/2003: 111).
Huet attempts to give some order to the three separate criteria before
undermining them:
[Descartes] plainly distinguishes [the natural light and clarity-and-distinctness]
when he asserts that the natural light does not attain anything that is not true,
in so far as it is attained by the natural light, that is, in so far as it is clearly and
distinctly perceived [see Descartes, Principles I 30; AT VIIIA 16; CSM I 203]. It
thus follows that the natural light is what attains the object of perception, and
that a clear and distinct perception is the action with which a thing is attained
by the natural light. It is as if the natural light were the criterion through which
[per quod], and clear and distinct perception the criterion according to which
[secundum quid]. (Huet 1694/2003: 120–121)
Huet’s interpretation of Principles I 30 is plausible, since Descartes refers to the
natural light there as “the faculty of knowledge” itself; so it is the thing that actually
grasps objects of perception, while clear and distinct perception is that faculty’s
activity by means of which those objects are perceived. Huet then interprets
Cartesian evidence to be a more general criterion:
[Descartes] seems to take evidence to be something general that ought to be in
everything we perceive that deserves to be regarded as true, whether it is perceived
through the natural light, or through clear and distinct perception, or through
reasoning, or through the senses, or in some other way. (Huet 1694/2003: 121)
Huet’s objection to the natural light is that it is too general a criterion:
Descartes defines the natural light as “the faculty of knowing given to us by
God” [Descartes, Principles I 30; AT VIIIA 16; CSM I 203]. From this it follows
that whatever we know through a faculty of knowing given to us by God is
known to us by the natural light. But whatever we know is something that we
know through a faculty of knowing that is given to us by God. Whatever we
know, therefore, is known by the natural light, which is clearly absurd. (Huet
1694/2003: 121)
As Huet sees it, the problem that a criterion of truth is supposed to resolve is how
to distinguish genuine knowledge from merely apparent knowledge. Our faculty of
knowledge is therefore being called into question. To appeal to that very faculty of
knowledge to resolve this issue—which is what we do by appealing to the natural
light—is to beg the question, not to resolve it.
Huet employs a battery of arguments against the criterion of clear and distinct
perception, but his strongest objections are his applications of the skeptical modes
of Agrippa outlined by Sextus. Huet observes that clarity and distinctness is a
relative concept (the third mode of Agrippa): one and the same idea can appear
clear and distinct to one person, but not to another. This phenomenon gives rise to
interminable rational disagreements (the first mode of Agrippa): “The Cartesians
also disagree among themselves and, using the same standard of truth, maintain
opposite and contradictory views” (Huet 1694/2003: 127). Either these disputing
Cartesians will simply assert that their own perceptions (and not their opponents’)
are clear and distinct, in which case they will fall prey to the Agrippan mode of
hypothesis (the fourth mode); or they will bring in something other than clarity and
distinctness to decide these disputes, which demonstrates that clarity and distinctness
cannot be the ultimate criterion of truth. But whatever additional criterion they
appeal to in order to decide the dispute will then be questioned, and the dispute
328 SKEPTICISM
will either go on ad infinitum (the second mode), with each further criterion being
bolstered by yet another criterion, or the dispute will be circular (the fifth mode),
since the disputants will support the further criterion by means of criteria posited
earlier in the dispute (Huet 1694/2003: 128–129).
In the Censura, the criterion of evidence gets little further attention, since Huet
considers the forgoing arguments sufficient for refuting this criterion:
If the senses, the mind, and the operations of the mind, the clear and distinct
perceptions of things, and the natural light are all deceptive means of knowing
the truth, however great the evidence attached to them, then they will not in the
least be certain. (Huet 1694/2003: 129)
Once the faculty of reason itself, the natural light, has been called into question,
all additional criteria of truth are undermined, including evidence. However, in
the Treatise Huet offers several pages of additional objections to evidence, most of
which are based on Pyrrhonian skeptical tropes. One novel objection, however, is
Huet’s neurological objection to evidence. All ideas and images in the understanding,
Huet claims earlier (Huet 1723: 49–50), are formed by impulses on the brain and
movements of nerve fibers and animal spirits. Since evidence is just a quality or
modification of ideas and images, it follows that the causes of evidence are the
same as the causes of ideas and images—brain impulses and various other internal
motions. Once reduced to brain and other bodily activity, it is possible that evidence
is accidentally caused by some physical aspect of the knower, rather than by the
truth of the matter in question.
In Huet’s mind, these rejections of Descartes’s criteria of truth are an indispensable
preamble to receiving the faith, since they humble reason and demonstrate its
need for faith. They are Huet’s most fundamental Academic skeptical exercise.
I have dwelt so long on these objections because they distinguish Huet’s Academic
skepticism from that of Foucher, who not only did not attack the foundations of
Descartes’s epistemology, but who even tried to defend those foundations against
skeptical arguments like those just outlined.
5 SIMON FOUCHER: LIFE AND WORKS

Simon Foucher (1644–1696) was a historian, apologist, and systematizer of the
philosophy of the ancient Academics, as well as a critic of Cartesian philosophy. His
stated goal was to do for the ancient Academy what Sextus had done for Pyrrhonism
and what Pierre Gassendi had done for Epicureanism (Foucher 1693: 2). Foucher’s
interest in Academic philosophy was not merely exegetical; he modernized and
applied that philosophy to critique the main ideas of leading philosophers of his
time, especially Descartes, Malebranche, and Leibniz.
Foucher was born in Dijon in 1644, received Holy Orders early in life, became
the honorary canon of the Holy Chapel in Dijon shortly thereafter, then soon
left any work this position entailed (but not the title) to study in the faculty of
theology at the Sorbonne, after which he never left Paris. Besides these few points
and the general fact that Foucher was early on a well-connected intellectual with
wide-ranging literary, philosophical, and scientific interests and abilities, most of the
details of Foucher’s early life have been lost. The latter half of Foucher’s biography
is indistinguishable from his bibliography. He is said to have died from exhaustion
from studying and writing.5
Adrien Baillet, in his biography of Descartes, alleges that, when Descartes’s
remains were returned to Paris in 1667, the notable Cartesian, Jacques Rohault,
asked Foucher to deliver the funeral oration. The truth of this claim has been
questioned (see Rabbe 1867: 4–5; Watson 1987: 33–34). Whether or not Foucher
was given the honor early in life of eulogizing Descartes, we know beyond any
doubt that Foucher’s later life was spent burying Descartes’s and his followers’
metaphysics deep in the ground. Foucher is remembered most of all for his
polemic with Malebranche, which began with a coincidence. In 1673, Foucher
began circulating among his closest friends several copies of a short treatise of his
entitled Dissertation on the Search after Truth, or on the Logic of the Academics. The
coincidence is that in the following year Malebranche published the first volume of
his best-known work with a very similar title, On the Search after Truth. Foucher
believed that Malebranche published his work as a response to Foucher’s call for
fellow laborers in the search after truth (Watson 1987: 35), and so he felt obliged
to correct Malebranche’s errors. A decades-long debate surrounding foundational
elements of Cartesianism followed the publication of Foucher’s 1675 Critique of the
Search after Truth (Watson and Greene 1995: 13–57).
The debate with Malebranche established Foucher as one of the foremost critics
of Cartesianism and brought him into dialogue and debate with a number of leading
philosophers. First to respond to Foucher was the Benedictine monk and defender
of Cartesianism, Dom Robert Desgabets, who published his Critique of the Critique
of the Search after Truth, also in 1675. This work forced Foucher to clarify his
understanding of and relation to ancient Academic skepticism, which Foucher did in
his 1687 Apology on Behalf of the Academics and in his 1693 History and Principles
of the Academics, which are the best sources for Foucher’s own philosophical ideas.
Other orthodox Cartesians, including Louis de la Forge, Pierre-Sylvain Régis, and
Antoine Arnauld, were moved by Foucher’s objections to Malebranche to revise
core doctrines of Cartesianism, especially relating to the nature of ideas and their
capacity to represent objects external to the mind (Watson 1987: 79–100). Foucher
was also a close friend and frequent correspondent of Leibniz and was the first to
publish criticisms of Leibniz’s monadology and theory of pre-established harmony
(Watson 1987: 131–147).
Richard A. Watson’s The Breakdown of Cartesian Metaphysics (Watson 1987, in
which Watson 1966 is reprinted) is the authoritative study of Foucher’s critiques of
Descartes, Malebranche, and Leibniz, and will remain so for many years. There is
still much work to be done on Foucher, however, especially in relation to skepticism.
In particular, there are three areas of research that deserve further attention and that
will be treated below: Foucher’s interpretation of the ancient Academics; the nature
and extent of Foucher’s metaphysical skepticism; and Foucher’s anti-skeptical
defense of the criteria of truth attacked by Huet.
330 SKEPTICISM
6 FOUCHER’S CARTESIAN INTERPRETATION

OF THE ANCIENT ACADEMY
Foucher’s philosophical career is unified by his systematization and application of
ancient Academic philosophy. A brief look at the titles of Foucher’s works (especially
1675, 1687, and 1693) demonstrates that he was interested in being recognized
as the modern champion of the Academy, and this desire of his was not lost on
contemporaries like Baillet and Leibniz, who gave Foucher the title of “restorer of
the Academy” (Rabbe 1867: 22, Appendix VIII).
Some of Foucher’s contemporaries, like Desgabets, saw in Foucher’s Critique of
1675 only the spirit of the most skeptical Academics. Desgabets accused Foucher
of filling his mind and his book with the “old arguments” of the Academics who
desired “to suspend their judgment on all things” (Desgabets 1675: 6), who claimed
“not to know either body or soul, first principles or their consequences, or even the
foundations of mathematics” (Desgabets 1675: 15–16), and who “made a trophy
of their ignorance by banishing from the world all true knowledge” (Desgabets
1675: 27). However, other contemporaries closer to Foucher, including Huet, saw
little trace of Academic skepticism in Foucher’s writings. According to Huet, who
met with Foucher on a regular basis for many years, “[Foucher] hardly knew the
names of Arcesilaus and Carneades, let alone anything about Pyrrhonism” (Letter to
Nicaise April 19, 1697; quoted from Rabbe 1867: Appendix X). In Huet’s judgment,
Foucher intended to rehabilitate Platonic philosophy, not Academic skepticism.
Foucher’s last writings strike a balance between Desgabets’s and Huet’s
assessments. Huet was certainly right that Plato’s Academy was the one Foucher was
most interested in reviving; Desgabets was also right, however, that in Foucher’s
view the Academy had an important skeptical dimension from the outset. In the
opening pages of his last work, Foucher refers to himself as “Academico-Platonicus,”
and vows to follow Plato, not blindly and by the letter, but in his method (1693: 2).
Foucher distinguishes two kinds of followers of Plato: “The first followers make
Plato affirm a number of things, while the others view Plato as continuing to doubt
and to search for the truth” (1693: 2). The first followers are Platonists, while the
latter followers are the Academics to which Foucher claims to belong.
We must be careful when referring to Foucher as an “Academic skeptic” as
many recent writers have done, not only because Foucher never used that term to
describe himself, but more importantly because he considered the term confused.
Foucher clarified the similarities and differences between Academic philosophy and
skepticism. There are three kinds of philosophers: Skeptics, who search for the truth
but who claim not to have found any truths; Academics, who search for the truth
and who claim to have found many important truths; and Dogmatists, who claim to
have found all the truth they were seeking (Foucher 1687: 4, 28–30; 1693: 239).
Foucher’s Academics differ from Skeptics, therefore, in that they confidently assert
that their search for truth has been successful; they differ from Dogmatists in that
they claim that their search for truth has been only partly successful.
Foucher neither claimed to be, nor wanted to be, the restorer of Academic
skepticism, pace Popkin (2003: 275). Foucher was aware of earlier attempts to revive
Academic skepticism and showed disdain for them. For example, Foucher names
Giovanni Pico della Mirandola (1463–1494) as a proponent of the Academics, but
he condemns Pico’s portrayal of the Academics as skeptics (1693: 71). But as we will
see in the next section, there is still good reason to call Foucher an Academic skeptic,
at least in the realm of metaphysics, since he offered one of the most powerful and
influential skeptical objections to the existence of the external physical world and
did so from a self-styled Academic standpoint. In the rest of this section, I will
show how Foucher’s interpretation of Academic philosophy is like a Trojan horse,
carefully constructed to lure in the Cartesians so that it would be easier to destroy
their philosophy on Foucher’s ground.
Foucher’s Apology (1687) and History (1693) are not presented as rediscoveries of
the Academics presented to the uninitiated; they are instead presented as original and
carefully constructed interpretations of the Academics presented to those who already
know much about that school. Foucher’s goal is not like that of the Renaissance
commentators, to uncover the original spirit of the ancient Academy through
painstaking exegeses of Cicero or Diogenes Laertius; his goal is to build upon key ideas
first discovered by Plato and his successors in order to make use of them in critiquing
the works of his contemporaries. The key to understanding Foucher’s interpretation
of the Academics therefore does not lie in any author of the ancient Academy itself,
but rather in the modern period, in particular in the writings of Descartes. Foucher’s
Academic philosophy is not predominantly Platonic or Carneadean or Ciceronian; it
is anachronistically Cartesian (see Maia Neto 1997, 2003; Charles 2013a).
Foucher’s tendency to interpret the ancient Academics as proto-Cartesians has
already been noted in the literature. Maia Neto (2003: 78–79) has shown that the
role of doubt in Foucher’s philosophy is like the role given to doubt by Descartes
in the first Meditation. Doubt is not the end of Academic philosophizing, but it is
essential “[b]ecause it is necessary to remain as if at one’s post, which provides cover
from preconceptions and errors, until evident truth forces one to leave” (Foucher
1693: 136). A second Cartesian element of Foucher’s reading of the Academics
is his reduction of the Academic way of philosophizing to five laws,6 which were
unmistakably inspired by Descartes’s rules in the Discourse on Method (CSM I
120; AT VI 18–19). Especially Descartes’s first law, “never to accept anything as
true if I did not have evident knowledge of its truth: that is, carefully to avoid
precipitate conclusions and preconceptions,” which is essentially Foucher’s first law,
can be found everywhere in Foucher’s writings, including in the passage quoted
immediately above (Foucher 1693: 136).
But the most striking example of Foucher’s Cartesian reading of the Academics
has yet to be noted in the literature. It occurs in the first chapter of the 1693 History
in which Foucher claims to be writing as a disinterested historian of the ancient
Academy. Yet if one reads Descartes’s Discourse on Method and Meditations, and
then turns to Foucher’s “History of the Academics” (Foucher 1693: 1–72), it is
striking how often and carefully Foucher maps the progress of Descartes’s thought
onto the history of the Academy, from Socrates to Antiochus. Though Foucher
never explicitly says so, each Academic is for him the personification of a stage of
Descartes’s progress toward truth.
332 SKEPTICISM
According to Foucher, Socrates was the first to realize that, in order to give some
order to philosophy, it was necessary to undermine the foundations of the reigning
disorder, which were presumption and preconceptions. Destroying preconceived
opinions was the only thing that Socrates could do in the absence of a criterion
of truth (Foucher 1693: 12). The parallels with Discourse One and Two, and with
the project of Meditations One, are evident here. Socrates takes us into the second
Meditation as well, since it turns out that he discovered the cogito two millennia
earlier than Descartes: “When [Socrates] said that he knew one thing, namely that
he knew nothing, he was acknowledging that he was thinking, and that he knew
he was thinking because he was doubting all other things” (Foucher 1687: 111).
Plato’s method was simply that of Socrates, which Foucher suggestively describes as
“meditations on first principles” (Foucher 1693: 15). These meditations led Plato
to discover some basic starting points, foremost of which were that the “senses are
not the proper judges of the truth of things that are outside of us,” that what we
perceive by the senses are merely “modifications of our soul,” and that our souls,
though initially unknown to us, are “nevertheless better known to us than our own
bodies and any other bodies external to us” (1693: 16–17). In other words, Plato
discovered the highlights of the second Meditation wax example.
The immediate successors of Plato—Speusippus, Xenocrates, Polemon, Crates,
and Crantor—strayed from the method of Socrates and Plato and rendered the
Academy dogmatic. They remind us of the primary intellectual vice identified in the
first Meditation: “My habitual opinions keep coming back” (CSM II 15; AT VII 22).
Arcesilaus was therefore faced with the task of renewing the Academy, which he
did by imitating Socrates by undermining precipitous judgments through doubt
(Foucher 1693: 31). Carneades then realized that for the Academics “a provisional
morality was necessary to guide their lives until they arrived at the knowledge they
were seeking” (Foucher 1693: 43, emphasis mine). Like Descartes in the third
Discourse (CSM II 122; AT VI 22–23), Carneades realized that, in the absence of
certain moral knowledge and faced with the necessity of acting, we must follow the
available probable guides, such as laws, customs, and religion.
Philo of Larissa set the Academy again on the search after scientific truth. He
gave the Academics hope that despite the doubts of Arcesilaus, some truths about
ourselves and the external world might be discovered by reason and understanding
(Foucher 1693: 54). Philo inspired his successor Antiochus to work on establishing
certain truths. Antiochus’s main contribution was to establish a rational criterion of
truth, which Foucher, like Descartes, calls évidence (Foucher 1687: 111). Antiochus
takes the Academy into the early third Meditation. This is where the ancient
Academy comes to a close in Foucher’s view. Not surprisingly, Foucher thinks that
Descartes begins to go astray in the rest of the third Meditation.7
Foucher’s history of the Academics provides a historical basis for the argumentative
strategy that he consistently used against Descartes’s philosophy:
Eager to defend antiquity against the disdain of innovators, [Foucher] believed

that it would suffice to convict the reigning philosophy of plagiarism, and to find
in that same antiquity antecedents of the new philosophy. . . . Whatever Descartes
taught that was rigorous and solid was borrowed from others; if Descartes ever
showed any originality, it was only to distance himself from the path of truth.
(Rabbe 1867: 21)
Descartes took the essentials of his first through early third Meditations from
the Academics; afterwards, he went astray so that now “we must bring Descartes
back to the Academy, rather than leave the Academy to join Descartes” (Foucher
1693: 113).
Foucher’s history also demonstrates that Academic skepticism is just one humble
part of the Academic philosophy that Foucher wished to rehabilitate. The doubts of
Socrates and Arcesilaus, and the probabilism of Carneades, play a role in Foucher’s
method, but doubt is not the end of his philosophy, nor does Foucher resist
affirming numerous truths with the greatest certainty: “It is true that the Academics
must doubt a great number of things, but this is because these things are doubtful.
Nevertheless, the Academics know the principal truths, such that their doubts
concern only scientific matters and dogmatic propositions about subjects of pure
human speculation” (Foucher 1687: ix).
7 FOUCHER’S CRITIQUE OF DESCARTES

Popkin blurs important differences between Huet and Foucher when he writes that
“Huet and Foucher . . . destroyed Cartesianism at its epistemological heart” (Popkin
2003: 282). In this section and the next, I will argue that Foucher, unlike Huet,
was not interested in destroying the heart of Cartesian epistemology, but in fact
assumed that foundation in his skeptical writings. Foucher’s Academic skepticism
is limited to the realm of metaphysics, while Huet’s Academic skepticism is all-
encompassing. In this section, I give an overview of Foucher’s skeptical critique of
Cartesian metaphysics, while in the next I argue that Foucher’s 1693 History can be
read in part as a Cartesian defense of reason against the skepticism of Huet.
Foucher acknowledges that his Academic philosophy shares much in common
with Descartes’s philosophy, especially relating to method. In particular, there are
three important similarities: (1) “they begin to philosophize by means of doubting
things in general, and by the universal examination of our judgments”; (2) “they
agree that we must consider the senses incapable of judging by themselves the
truth of things outside of us”; and (3) “they agree in following approximately the
same method of philosophizing” (Foucher 1693: 187–188). Contained in the third
similarity, that of basic method, is the shared criterion of rational evidence, which is
the basis of the first law of both Descartes and of Foucher’s Academics.
On the basis of this commitment to an initial position of doubt, mistrust of the
senses, and a basic method of philosophizing that primarily involves assenting only to
what is most evident, Foucher launches a skeptical attack on Descartes’s philosophy
(though Malebranche was the occasional cause of Foucher’s objections, Foucher
is clear that Descartes is his target—see Foucher 1675: 3). The skeptical argument
in general is that, if Descartes had remained committed to the three points above
(doubt, mistrust of the senses, method), then he would have realized that he never
334 SKEPTICISM
attained any certain knowledge of an external, physical world. Moreover, as long as

Descartes remained committed to his ontological dualism, he could never hope to
attain knowledge of the external world. Foucher’s main skeptical accomplishment
is to have shown more rigorously than anyone before him that ontological dualism
entails external world skepticism.
Foucher’s skeptical strategy in particular is to break down the primary-secondary
quality distinction of which Descartes made extensive use.8 The aim of Foucher’s
strategy was to force Cartesians to regard the primary quality of extension, which
they considered the essence of the material world, in the same way they regarded
sensible secondary qualities like color, smell, and taste, namely, as mere modes of
the mind that represent nothing real outside the mind. Foucher’s goal was to restrict
knowledge to the mind’s ideas.
According to Descartes’s ontological dualism, there are two created substances:
mind, which is thinking, non-extended substance, and body, which is non-thinking,
extended substance. The mind is immediately aware only of its own modes—
ideas, sensations, imaginings, memories, judgments. Consequently, if the mind is
to become aware of bodies, then it can do so only by the indirect way of its ideas
and sensations of bodies. Cartesian philosophy of mind distinguished ideas from
sensations by claiming that the former modes of the mind represent real external
objects, while the latter modes are merely effects of those objects on the sense organs
and ultimately on the mind (see Watson 1987: 47–53 for a concise summary of
orthodox Cartesianism in Foucher’s time). It is with this distinction between ideas
and sensations that Foucher begins his main attack against Descartes:
These two sorts of ideas belong equally to us and are, properly speaking, nothing
other than our soul disposed in such-and-such a manner. But it is always our soul,
and since the soul [according to Descartes’s dualism] has nothing in itself which
resembles matter and extended beings, it is difficult to conceive how it could
represent anything other than its own ideas. (Foucher 1675: 45)
Two important implicit premises do much of the work in Foucher’s objection: (1)
that in order to represent an object, an idea must resemble the object, and (2) if ideas
are essentially distinct from bodies (as Cartesian dualism posits), then ideas cannot
resemble, and therefore cannot represent, bodies. Foucher defends (1) by arguing
that “we understand nothing else by representing than rendering a thing present to
us” (Foucher 1675: 52). Foucher is clear that no modification of our soul can make
a body present to us: “It is necessary, in order to represent [bodies] as they are in
themselves, for our ideas to dispose us exactly as if these things were actually in us
and present to us immediately” (Foucher 1675: 53). But the Cartesian way of ideas,
in conjunction with ontological dualism, renders such immediate access to physical
objects impossible, since only extensionless thoughts can be immediately present to
the mind. Foucher continues:
All that we know by the senses [according to Descartes] are merely modes of
our soul that belong entirely to us and that do not resemble anything in material
objects. But we know extension also by means of sensation, from which we must
conclude that extension is nothing but a mode of our soul, and that there is
nothing similar to extension in material objects. (1675: 65)
Foucher reports that he first put this objection to Rohault in 1667 and that Rohault’s
response was to deny that we know extension by means of sensation. Foucher claims
that he was baffled by this response, which he refuted by the following argument:
“When I see a red square, for example, I simultaneously perceive its shape and color,
on the one hand, and its extension on the other, since I judge its size; consequently,
we know the square’s extension doubly by means of sensation: for we know it by
sight and by touch, whereas we know colors only by the eyes” (Foucher 1675: 66).
With extension relegated to the realm of sensible qualities along with color, the real
mind-independent existence of the extended world becomes as doubtful as that of
colors.
Rohault’s attempt to evade Foucher’s argument was just one of many that
contemporary Cartesians advanced. Malebranche denied that ideas are modes of
the human mind and argued instead that we perceive ideas in the mind of God,
which Foucher considered to be a very pious, but hardly a very evident position to
espouse (Foucher 1675: 116). More orthodox Cartesians tried to save Descartes
by developing sophisticated accounts of how ideas can represent material objects
without resembling them in essence (Watson 1987: 89–99).
8 FOUCHER’S DEFENSE OF DESCARTES’S

CRITERIA OF TRUTH
In his Apology, Foucher identifies Huet as another philosopher who believed that
the Academic philosophy was the most solid and most compatible with Christianity
(Foucher 1687: 36). In this same passage, Foucher expresses his hope that Huet
will soon publish a book on Academic doubt. There is also an extant letter written
in 1685 by Foucher, urging Huet to publish on the Academics (see Lennon 2008:
34–37). The work that would ultimately satisfy Foucher’s requests, Huet’s Treatise,
would never be published in Foucher’s or Huet’s lifetime.
Foucher would not have been happy with the Treatise had he read it. We can
be sure of this because Foucher must have been disturbed by Huet’s earlier anti-
Cartesian work, the Censura, which first appeared in 1689, between the publication
dates of the last part of Foucher’s Apology and the first part of Foucher’s History.
Foucher read the Censura (see Foucher 1693: 92), and though he never criticizes
Huet by name, Foucher was clearly opposed to the radical nature of Huet’s
skepticism. In Foucher’s History, there is an extended refutation of anti-Cartesian
skeptical arguments that had been elaborated by Huet in the Censura and that would
be offered again in the Treatise. These skeptical arguments target the foundation of
both Descartes’s and Foucher’s epistemology, the natural light of reason, and the
general criterion of evidence.
As we have seen, Huet reduced evidence to a mere psychological phenomenon, an
appearance that can be as deceptive as any sensory appearance (see Huet 1694/2003:
130; 1723: 76). There is no infallible evidence for Huet; there is only the varying
336 SKEPTICISM
subject-relative appearances of evidence. When Foucher begins to combat skeptics

who are opposed to the criterion of evidence, he identifies their principal strategy
as the relegation of evidence to mere persuasion (Foucher 1693: 141). Foucher
describes these skeptics as “animals that trouble the world,” “rebels against the
light,” “beasts,” “condemned by God and man,” “incurably sick,” “unwilling to
open their eyes” (1693: 120–121).
Foucher begins his response to this skepticism by noting that nobody would deny
that the proposition “two plus two equals four” appears to be necessarily true. But
if the skeptic can recognize the appearance of necessary truth, then the skeptic must
possess a general idea of truth: “In order to know whether some proposition seems
to be true, it is necessary for us to know at least in one instance what it is to be
true” (Foucher 1693: 134). Ironically, the Academic Foucher takes this classic anti-
skeptical argument from the anti-Academic, Augustine. Foucher believes that, if the
skeptic meditates on this initial idea of truth, he will see that it contains a general
criterion of truth, namely, evidence (1693: 83–88, 133, 208).
The skeptic will respond to Foucher by claiming that there is a difference between
“true evidence” and “apparent evidence” and that we need a further criterion in
order to distinguish these things (1693: 140–141). In the absence of such a criterion
for the criterion, we are left merely with degrees of persuasion, from weak persuasion
up to invincible persuasion, but without any guarantee that the highest degree of
persuasion corresponds with the possession of truth. Foucher does concede that we
lack a precise definition of the criterion of truth and an advanced art of teaching
others to recognize the criterion (1693: 133–134); the first goal of Academic logic
is to establish that art. But to search for a separate criterion beyond the perception
of evidence would be misguided in Foucher’s view: “It would be like searching to
clarify light itself” (1693: 133). Foucher apparently wants his reader to find absurd
the idea of clarifying light by means of another light; however, Huet (1723: 83)
demanded just that. It was probably with the obstinate skepticism of Huet in mind
that Foucher attempted the task of shedding light on the criterion of evidence.
Foucher does so by giving some further criteria by which to guarantee evident
truths, and these criteria are not further qualities of the mind’s ideas but rather are
rational capacities and incapacities. We are in possession of the “highest degree of
evidence”—a state that Huet argued we cannot achieve—when we cannot conceive
of a thing in any other way than we presently do (Foucher 1693: 142); when we do
not hesitate in the conception that we have of the thing (1693: 142); when the
conception we have is “incontestable,” that is, cannot be doubted (1693: 81, 213);
when we are determined, whether against our will or in conformity with it, to
believe in our conception by the evidence that we have for it (1693: 137); when
we cannot even conceive that the contrary of our belief is true (1693: 141). Most
of Foucher’s examples of evident truths are mathematical, such as “two plus two
equals four.”
The skeptic might object, however, that “two plus two equals four” can be
doubted by means of the omnipotent God objection from Descartes’s first
Meditation and that consequently the proposition does not count as an evident
truth on Foucher’s account. But Foucher responds by arguing that Descartes’s
appeal to the omnipotence of God to cast into doubt common notions of

mathematics is an example of excessive, not reasonable, doubt that ultimately
undermines Descartes’s system:
We should not think that piety requires us to speak in this way of divine power.
For it serves only to overturn all the certainty that we can have concerning God,
and to destroy not only theology, but also religion. That is what the Cartesians
need to consider, they who take their master at his word rather than interpreting
him charitably. (1693: 201)
Foucher’s skepticism clearly has both epistemological and theological limits.
Foucher’s defense of rational evidence ultimately rests on the distinction and
connections between three species of reason: there is reason in itself, or divine
reason, which is immutable, eternal, infallible, and necessary; there is right reason,
which is reason in itself as shared by all human beings, and which is, like divine
reason, infallible; and there is our reason or particular reason, which is a mode of the
human mind, and is subject to error because of the interference of preconceptions
and bad habits (1693: 195–199; 1687: 16). Foucher argues that right reason must
be infallible using what he calls an a posteriori argument. Right reason must be
infallible or else we could not explain the constancy, the accuracy, and the stability of
the technical arts and sciences that are based in right reason. Buildings which outlast
their architects, the invention of scientific instruments, and algebraic formulae all
point to the infallibility of right reason (1693: 197–198). The unity and infallibility
of right reason are the guarantors of mathematical truths: “It is true that the same
Reason enlightens all men and extends to every mind in which it produces ideas that
are perfectly similar. . . . It is certain that this unity of species [of reason] suffices to
support the most solid demonstrations of the mathematicians” (1693: 207).
Foucher’s argument against Huet’s skepticism is that human reason is based in
divine reason. Moreover, evidence is more than invincible persuasion; it is the root
of the fruitfulness of the sciences, which could be witnessed every day in Foucher’s
time. This is also Foucher’s answer to dogmatists who complain that his insistence
on contemplating basic evident truths will not get us very far:
We easily perceive that if two triangles have two sides and one angle equal to
each other, then they are congruent. Many people scorn this truth as if it had
no importance. Yet it is by means of this truth that we discovered the art of
measuring inaccessible places, of judging the sizes of the stars and their distances
from the earth. (1693: 85)
The experience of evident truths is, as Huet alleges, primarily psychological:
evidence is perceived by individuals as an inner “torch in the fog” (Foucher 1693:
205). The goal of philosophy is to clear away the fog of preconceived notions and
precipitous judgments in order to render the light of evidence even brighter. But in
addition to the subjective experience of evidence, which Huet is right to criticize
as potentially deceptive, there are also the incontestable triumphs of science and
technology, which are based on these evident truths, and which ought to convince
us that there is more to evidence than the modes of our minds. Foucher’s a posteriori
338 SKEPTICISM
argument for the infallibility of reason anticipates our own contemporary trust in
reason that stems from the successes of science and technology.
9 CONCLUSION
Both Huet and Foucher are notable among later seventeenth-century philosophers
for their extensive interest in the ancient Academy. They each wrote a history of
the Academy, carefully noting its differences from and superiority to other ancient
sects. They each paid close attention, in particular, to the nature and extent of the
skepticism of the Academy, and they employed skeptical arguments against their
common foe, the Cartesians, whom they ultimately defeated by means of their
skepticism.9 There is good reason, therefore, to consider Huet and Foucher Academic
skeptics, and the label “modern Academic skepticism” is useful for referring to the
skeptical revival of the late seventeenth century that undermined the system of one
of Western philosophy’s greatest thinkers, Descartes.
However, modern Academic skepticism was not a unified movement, which may
be partly responsible for its success. Huet’s Academic skepticism aimed primarily at
Cartesian epistemology, while Foucher’s aimed primarily at Cartesian metaphysics.
Because of their fundamental differences, Huet and Foucher never did, and they
never could, harmonize their versions of Academic philosophy or coauthor a single
anti-Cartesian treatise. But this did not prevent readers familiar with the works of
both authors from deriving a single conclusion from them: Descartes’s philosophy
is fundamentally and fatally flawed.
If Popkin is right that the seventeenth century began with a crise pyrrhonienne,
should we say that the century ended with a crise académicienne? If we mean by
this a rise of skepticism that took its inspiration from a careful study of the ancient
Academics, then yes, we can say that the century ended with an Academic crisis. On
the other hand, to Huet and Foucher, as well as to many of their contemporaries,
Academic skepticism was not the problem, let alone a crisis: it was a solution. The
real crisis to which Academic skepticism was the remedy was the crise cartésienne,
which Huet and Foucher could agree was the twofold problem of the improper use
of reason in supporting the Christian faith and of the infiltration of the Christian
faith into philosophy. Malebranche’s Search after Truth epitomized this crisis in
the minds of both Huet and Foucher, and herein lies the greatest source of unity
in modern Academic skepticism. It is not a common ancient Academic source,
or common Academic skeptical trope, but a common target that makes “modern
Academic skepticism” a meaningful label worth preserving in our histories
of skepticism.
NOTES
1. Huet was aware that he was a representative of the Republic of Letters; he justified
the prima facie vain project of his autobiography by arguing that he was not writing
about himself alone, but primarily about the history of the time in which he lived (see
Shelford 2007: 17).
2. For a complete bibliography of Huet’s works, along with English summaries, see
Huet (1810: vol. 2, 465–490).
3. For the reasons that led Huet not to publish the Treatise in his lifetime, see Shelford
(2007: 140–141).
4. It is common, though not uncontroversial, in the English secondary literature on
Huet and Foucher to translate évidence as “evidence,” which is sensible considering
that it was the English word used in the seventeenth century to translate the French
term évidence. A good alternative would be “evidentness,” but I have not seen the
term used in works earlier than the nineteenth century, so it is not ideal for translating
seventeenth-century works. “Self-evidence” has advantages, but whereas Huet and
Foucher (and others, like Bayle) speak of degrees of évidence, it does not seem natural
to speak of degrees of self-evidence. A final plausible alternative to “evidence” would
simply be “clarity,” but this would make it impossible to distinguish, as Huet does,
“évidence” and “clarité.”
5. For a complete bibliography of Foucher’s works, see Watson (1969: xxxviii–xlii).
6. (1) Proceed only by means of demonstration in philosophy; (2) do not consider
questions that clearly cannot be decided; (3) admit that you do not know that of
which you are ignorant; (4) distinguish the things you know from the things that you
do not know; (5) always search for new truths (Foucher 1687: 5–9).
7. Descartes’s first error, however, occurs in the second Meditation when he declares
that he is a thinking thing. In Foucher’s view, it is correct to acknowledge that the
mind thinks, but it is precipitous to declare that the essence of the mind is thinking
(Foucher 1675: 14–15).
8. In doing so, Foucher anticipates, if not provides, the foundation of George Berkeley’s
idealism. Berkeley probably would have learned of Foucher’s attack on the primary-
secondary quality distinction from Pierre Bayle’s Dictionaire historique et critique, in
the article “Pyrrhon,” remark C. Some see Foucher’s arguments as tending, not toward
metaphysical skepticism, but toward idealism (see Armour 2003).
9. See Watson (1987) for the case that Foucher is primarily responsible for the
downfall of Cartesianism; see Lennon (2003a, 2008) for the case that Huet is mainly
responsible.
REFERENCES
Armour, Leslie. 2003. “Simon Foucher, Knowledge and Idealism: Philo of Larissa and
the Enigmas of a French ‘Skeptic.’” In T. M. Lennon (ed.), Cartesian Views: Papers
Presented to Richard A. Watson, 96–116. Leiden: Brill.
Charles, Sébastien. 2013a. “Entre réhabilitation du scepticisme et critique du cartésianisme:
Foucher lecteur du scepticisme académique.” In S. Giocanti (ed.), La réception des
Academiques à l’Âge moderne. Special issue of Astérion 11 (online): http://asterion.
revues.org/2382 (last consulted July 4, 2014).
Charles, Sébastien. 2013b. “On the Uses of Scepticism against a Certain Philosophical
Arrogance: Huet as a Critic of Cartesian Logic and Metaphysics,” Science et esprit 65:
299–309.
Charron, Pierre. 1601/1986. De la sagesse, edited by B. de Negroni. Paris: Fayard.
340 SKEPTICISM
Cicero, Marcus Tulius. 1979. Academica and De Natura Deorum, translated by H. Rackham.
Cambridge, MA: Harvard University Press.
Desgabets, Dom Robert. 1675. Critique de la Critique de la Recherche de la Vérité, où l’on
découvre le chemin qui conduit aux connaissances solides, pour servir de réponse à la
lettre d’un Academicien. Paris: Jean du Puis.
Foucher, Simon. 1673. Dissertations sur la recherche de la verité, ou sur la logique des
academiciens. Dijon: Publisher unknown.
Foucher, Simon. 1675. Critique de la Recherche de la verité où l’on examine en même-
tems une partie des Principes de Mr Descartes. Lettre par un Academicien. Paris: Martin
Coustelier.
Foucher, Simon. 1687. Dissertation sur la recherche de la verité, contenant l’apologie des
academiciens, où l’on fait voir que leur manière de philosopher est la plus utile pour
la religion, et la plus conforme au bon sens, pour servir de Réponse à la Critique de
la Critique, etc., avec plusieurs remarques sur les erreurs des sens et sur l’origine de la
philosophie de Monsieur Descartes. Paris: Estienne Michallet.
Foucher, Simon. 1693. Dissertations sur la recherche de la verité, contenant l’histoire et les
principes de la philosophie des academiciens. Avec plusieurs réflexions sur les sentimens
de M. Descartes. Paris: Jean Anisson.
Guellouz, Suzanne (ed.). 1994. Pierre-Daniel Huet (1630–1721): Actes du Colloque de
Caen. Biblio 17. Paris, Seattle and Tübingen: Papers on French Seventeenth-Century
Literature.
Huet, Pierre-Daniel. 1679. Demonstratio evangelica. Paris.
Huet, Pierre-Daniel. Censura Philosophiae Cartesianae. Paris: Daniel Horthemels.
Huet, Pierre-Daniel. 1690. Alnetanae quaestiones de concordia rationis et fidei. Caen
and Paris.
Huet, Pierre-Daniel. 1694/2003. Against Cartesian Philosophy, edited, translated,
annotated, and introduced by T. M. Lennon. Amherst, NY : Humanity Books. (The
first English translation of Censura Philosophiae Cartesianae. The translation is of the
revised 1694 edition.)
Huet, Pierre-Daniel. 1723. Traité philosophique de la foiblesse de l’esprit humain.
Amsterdam: Henri Sauzet.
Huet, Pierre-Daniel. 1810. Memoirs of the Life of Peter Daniel Huet, Bishop of Avranches.
2 volumes, translated by John Aikin. M. D. London. (An English translation of Huet’s
Commentarius de rebus ad eum pertinentibus. The Hague: 1718.)
Lennon, Thomas M. 2003a. “Foucher, Huet, and the Downfall of Cartesianism.” In
T. M. Lennon (ed.), Cartesian Views: Papers Presented to Richard A. Watson, 117–128.
Leiden: Brill.
Lennon, Thomas M. (ed.) 2003b. Cartesian Views: Papers Presented to Richard A. Watson.
Leiden: Brill.
Lennon, Thomas M. 2008. The Plain Truth: Descartes, Huet, and Skepticism. Brill Studies
in Intellectual History 170. Leiden and Boston: Brill.
Maia Neto, José R. 2003. “Foucher’s Academic Cartesianism.” In T. M. Lennon (ed.),
Cartesian Views: Papers Presented to Richard A. Watson, 71–95. Leiden: Brill.
Maia Neto, José R. 2008a. “Charron and Huet: Two Unexplored Legacies of Popkin’s
Scholarship on Early Modern Skepticism.” In J. D. Popkin (ed.), The Legacies of Richard
Popkin, 155–172. International Archives of the History of Ideas 198. Dordrecht:
Springer.
Maia Neto, José R. 2008b. “Huet sceptique cartésien,” Philosophiques 35: 223–239.
Maia Neto, José R. 2008c. “Huet n’est pas un sceptique chrétien,” Les Études Philosophiques
2 (85): 209–222.
Maia Neto, José R. 2014. Academic Skepticism in Seventeenth-Century French Philosophy:
The Charronian Legacy, 1601–1662. International Archives of the History of Ideas 215.
Dordrecht: Springer.
Malebranche, Nicolas. 1674/1997. The Search after Truth, translated and edited by
T. Lennon and P. Olscamp. Cambridge: Cambridge University Press.
Malbreil, Germain. 1994. “Le Traité Philosophique de la Foiblesse de l’Esprit Humain, de
Feu Monsieur Huet, ancien Evêque d’Avranches.” In S. Guellouz (ed.), Pierre-Daniel
Huet (1630–1721): Actes du Colloque de Caen. Biblio 17. Paris, Seattle and Tübingen:
Papers on French Seventeenth-Century Literature.
Popkin, Richard H. 2003. The History of Scepticism: From Savonarola to Bayle. Oxford:
Quantin, Jean-Louis. 1994. “La Raison, la Certitude, la Foi: quelques remarques sur les
préliminaires de l’acte de foi selon Huet.” In S. Guellouz (ed.), Pierre-Daniel Huet
(1630–1721): Actes du Colloque de Caen. Biblio 17, 83–98. Paris, Seattle and Tübingen:
Papers on French Seventeenth-Century Literature.
Rabbe, Felix. 1867. Étude Philosophique. L’Abbé Simon Foucher Chanoine de la Sainte
Chapelle de Dijon. Paris: Didier et Cie.
Rapetti, Elena. 2003. Percorsi anticartesiani nelle lettere a Pierre-Daniel Huet. Florence:
Leo S. Olschki.
Schmitt, Charles B. 1972. Cicero Scepticus: A Study of the Influence of the Academica in the
Renaissance. The Hague: Martinus Nijhoff.
Sextus Empiricus. 2000. Outlines of Scepticism, edited by Julia Annas and Jonathan Barnes.
Shelford, April G. 2007. Transforming the Republic of Letters: Pierre-Daniel Huet and
European Intellectual Life, 1650-1720. Rochester, NY: University of Rochester Press.
Watson, Richard A. 1966. The Downfall of Cartesianism, 1673–1712: A Study of
Epistemological Issues in Late 17th Century Cartesianism. The Hague: Martinus Nijhoff.
Watson, Richard A. 1969. “Introduction.” In Simon Foucher, Critique de la Recherche de la
verité, v–xlix. New York and London: Johnson Reprint Corporation.
Watson, Richard A. 1987. The Breakdown of Cartesian Metaphysics. Indianapolis and
Cambridge: Hackett.
Watson, Richard A., and Marjorie Grene. 1995. Malebranche’s First and Last Critics:
Simon Foucher and Dortous de Mairan. Carbondale and Edwardsville: Southern Illinois
University Press.
CHAPTER TWENTY-FOUR
Spinoza on Skepticism
ALISON PETERMAN
It remains now to investigate the doubtful idea—i.e., to ask what are the things
that can lead us into doubt, and at the same time, how doubt is removed. I am
speaking of true doubt in the mind, and not of what we commonly see happen,
when someone says in words that he doubts, although his mind does not doubt.
For it is not the business of the Method to emend that. That belongs rather to
the investigation of stubbornness, and its emendation.
(Spinoza, Treatise on the Emendation of the Intellect §77)
1 INTRODUCTION
The skeptic is one of Spinoza’s favorite targets of abuse. Whatever truth Spinoza
might manage to discover, he complains, “Some skeptic would still doubt it.”
“There is no speaking of the sciences” with skeptics; instead, “they must be regarded
as automata, completely lacking a mind” (TIE 41).1 But Spinoza contrasts such
mindless obscurantists, “who ha[ve] no other end than doubting,” with Descartes,
who wielded skepticism instead “to free his mind from all prejudices, so that in the
end he might discover firm and unshakable foundations of the sciences” (TIE 41).
And Spinoza begins his geometrical reconstruction of Descartes’s Principles of
Philosophy with an approving exposition of that “great man’s” methodological
skepticism and its response, arguing that we can attain certainty only once we have
a clear and distinct idea of the essence of God.
But this agreement turns out to be superficial, because Spinoza and Descartes have
deeply different stories about how the idea of God grounds certainty. I outline this
story in Section 2, focusing on Spinoza’s early work, where I argue that he responds
to the Cartesian skeptic by denying the very distinction between true ideas and
ideas that are clearly and distinctly perceived. However, Spinoza’s characterization
of truth in the Ethics as agreement of an idea with its object complicates matters,
leaving it unclear whether Spinoza has responded in a way that would satisfy a
Cartesian skeptic. In Section 3, I discuss Spinoza’s three kinds of knowledge, or
cognition. I argue that Spinoza’s answer to skepticism vindicates only one special
kind of knowledge and that he ultimately denies that empirical, mathematical,
moral, and revelatory knowledge of nature are possible. In Section 4, I conclude
SPINOZA ON SKEPTICISM 343
by showing that, according to Spinoza, Descartes salvages knowledge from the

Cartesian skeptic only by reducing all natural knowledge to revelation.
2 CERTAINTY
In the Prolegomenon to his Principles of Cartesian Philosophy, Spinoza provides an
account of Descartes’s deployment of and response to methodological skepticism.
Superficially, it is a faithful account: we can have certain beliefs if, and only if, we
cultivate a clear and distinct idea of God. But the demon is in the details.
Spinoza starts by asking: while meditating with Descartes, why can we be certain
of our own existence despite the possibility that God is a deceiver, but we cannot be
certain, for example, that the angles of a triangle add to 180 degrees? In the former
case, he answers:
Wherever we turned our attention—whether we were considering our own nature,
or feigning some cunning deceiver as the author of our nature, or summoning up,
outside us, any other reason for doubting whatever—we came upon no reason for
doubting that did not by itself convince us of our existence. (C 237)
When entertaining the proposition that I exist (call this proposition “cogito”), no
other idea furnishes me with a reason to doubt it, including the idea that God is a
deceiver. Not so when I entertain the proposition that the three angles of a triangle
sum to two right angles (call this “triangle”). Instead, while I am “compelled to
infer” triangle when I attend to the nature of a triangle,
We cannot infer the same thing from [the supposition] that perhaps we are
deceived by the author of our nature. . . . So here we are not compelled, wherever
we direct our attention, to infer that the three angles of a Triangle are equal
to two right angles. On the contrary, we discover a ground for doubting, viz.
because we have no idea of God which so affects us that it is impossible for us to
think that God is a deceiver. (C 237)
I can render mathematical truths like triangle certain by cultivating a clear and
distinct idea of God:
When we have formed such an idea, that reason for doubting Mathematical
truths will be removed. Wherever we then direct our attention in order to doubt
some one of them, we shall come upon nothing from which we must not instead
infer that it is most certain—as happened concerning our existence. (C 237)
These passages suggest that, according to Spinoza, S’s belief that p is certain if and
only if all three of the following are true:
(a) p is a “mathematical truth.” A “mathematical truth” in the PPC seems to be a
truth that we are “compelled to infer” from a clear and distinct idea, and not
necessarily one that concerns mathematics (C 238).
(b) S cannot discover a “ground for doubting” p. (According to Spinoza,
a “ground for doubting” p is an idea from which a doubt that p can be
344 SKEPTICISM
inferred—so ultimately we will want to know: what is a doubt that p,

for Spinoza?)
(c) S can infer p (or, S can “infer that p is most certain” (C 237)) from every
other idea that S can attend to.
Spinoza claims that these criteria are satisfied for mathematical truths like triangle
if and only if we have a clear and distinct (“C&D”) idea of God’s essence, and in
particular of God’s veracity (C 237). What is his justification for this? Let’s focus
on (b) and first consider how Descartes guarantees it through God’s veracity. Once
I have the C&D idea of God’s essence, I can conclude that God exists and that
God lacks the will to deceive. That means that God has actually created me, and
has created me with faculties that would not ineluctably cause me to believe what I
shouldn’t.
This argument is in no way open to Spinoza, and he does not make it. Spinoza
denies that we have faculties at all (E IIp42), and he denies that God has a will
whose quality can be evaluated (E Ip37). So he cannot rely on God’s veracity being
actualized in God’s well-intentioned creation of our faculties to get us from the
C&D idea of God’s essence to the claim that we cannot generate any doubts about
triangle. Indeed, in the PPC, Spinoza takes pains to stress that I can be certain of
triangle before I know that my author is not a deceiver—indeed, before I know that
God exists. All that is required for my ideas of mathematical truths to be certain is
that I have a C&D idea of God’s essence.
Descartes’s claim that once I have this idea, I can become sure that I will never
generate a doubt of any C&D idea, p, required first showing that I should believe
that p, since a non-deceiving God created me. That is what guarantees not just that I
cannot currently generate any reasons to doubt that p from my unclear and indistinct
ideas but that it is impossible for me ever to generate a reason to doubt that p, if p is
C&D. It is the epistemic status that, for Descartes, guarantees that certain ideas are
doubt-proof—their being doubt-proof does not confer epistemic status.
Given that Spinoza does not rely on God’s guarantee and the epistemic status it
provides to achieve (b), there are two questions to answer:
A. How does Spinoza think that we can proceed from the C&D idea of
God’s essence to Spinozistic certainty? After developing the C&D idea
of God’s essence, it seems like I can still come to doubt triangle on some
other grounds—say, some fuzzy recollection that there are non-Euclidean
geometries. According to the Prolegomenon, having any ideas that are
not clear and distinct should place me in perpetual peril of doubt, since
there is no telling what doubts I can infer from them—not to mention that
I have ideas—even, it seems, clear and distinct ones—from which I cannot
infer triangle.
B. Even if Spinoza can establish Spinozistic certainty of triangle, has he
established that he should believe it, or that it is true?
Spinoza appears to provide an answer to (A) in the Treatise on the Emendation of
the Intellect (TIE), where he explains that the C&D idea of God’s essence ensures
that our ideas are indubitable if they are deduced in the proper order from God’s
essence, which is “the origin of all things”:
If someone proceeds rightly, by investigating [first] those things which ought to
be investigated first, with no interruption in the connection of things . . . he will
never have anything but the most certain ideas. (TIE 80; see also TIE 44)
Since “true knowledge proceeds from cause to effect” (TIE 85), if I begin with a
C&D idea of God’s essence, carefully deduce what follows from it in the proper
order, and populate my mind only with the resultant ideas, I will never generate a
ground for doubting that I know any of those ideas. This is clearly impossible for a
finite mind, which will always have unclear and indistinct ideas (see, e.g., E IIIp1d,
E IVp4, TIE 44). But Spinoza seems to think that we can approximate it and that we
can in some sense “insulate” this God-derived chain or network of clear and distinct
ideas from the rest of our minds and, to that extent, make those ideas immune to
doubt. In this picture, God’s essence guarantees that I have certain ideas in virtue of
God’s being the cause and origin of all things, and not in virtue of God’s being the
cause of my cognitive and epistemic faculties.
On the TIE account, then, it looks like knowledge of God’s essence guarantees
the Spinozistic certainty of our ideas of mathematical truths only if those ideas are
deduced in the right way from the clear and distinct idea of God’s essence. This
explains the third, seemingly unattainable, criterion that an idea is certain only if
it is deducible from every idea in the mind. Spinoza seems to assume the further
point that, if every one of a person’s ideas were certain, each certain idea would be
deductively connected with every other. I have not found a place where Spinoza
asserts anything that strictly entails this, so I can only offer: it sounds exactly like
something Spinoza would say.
In fact, I think that this third criterion is sufficient for certainty, according to
Spinoza, if we add that we also have the C&D idea of God (which Spinoza, at E
IIp47, claims that we all do). There are a few reasons to think this. It makes sense of
the claim in the TIE that “if there should be only one idea in the soul . . . whether
it is true or false, there will be neither doubt nor certainty, but only a sensation of
a certain sort” (TIE 77). Certainty is not simply absence of doubt but essentially
concerns the relationship of an idea with other ideas in the mind. Furthermore,
Spinoza himself defines doubt in terms of certainty. At TIE 78 Spinoza writes that
“doubt will arise though another idea which is not so clear and distinct that we can
infer from it something certain about the thing concerning which there is doubt.”
Finally, Spinoza writes that “by certainty we understand something positive, not
the privation of doubt” (E IIp49). This gives us an alternative characterization of
doubt: I doubt that p if I have any idea from which I fail to “infer from it something
certain” about p. Although Spinoza tries to retain Cartesian language in the PPC
by suggesting that the possibility of God’s deception furnishes a reason to doubt
triangle, it is clear that he is really concerned there with whether or not we can
“infer that [triangle] is most certain” (C 237).
This is an absurdly high standard, but certainty, for Spinoza, is a matter of degree.
The more C&D ideas we have deduced from God’s essence, from which we can
346 SKEPTICISM
deduce the idea in question, and the fewer unclear and indistinct ideas we have
that we cannot connect up with it, the more certain the idea. The perfectly certain
idea, which can be deduced from every idea in the mind, is to be found only in
God’s intellect.
Now the Cartesian skeptic can still ask whether our C&D ideas are true, or
whether we are justified in believing them, starting with the C&D idea of God. Or as
Spinoza gripes, “Perhaps . . . some Skeptic would still doubt both the first truth itself
and everything we shall deduce according to the standard of the first truth” (TIE 47).
Spinoza responds that this person “speaks contrary to his own consciousness” and
his mind is “completely blinded” (TIE 48). Doubt must always have another idea as
a reason or a cause (TIE 78). To have an idea is just to (pro tanto) believe or affirm
that idea, because ideas essentially involve affirmation (E IIp49), although other ideas
in the mind may undermine that affirmation. If there are no ideas in the mind other
than the C&D idea of God and those deduced from it, the skeptic has nothing that
could generate a doubt, and is just “saying in words” that he doubts.
But this is still only a psychological diagnosis of the skeptic. Very well: being
certain in the Spinozistic sense entails that someone is unable to doubt her ideas. But
can it ensure that she is justified in believing them, or that they are true?
Recall that, on the account of certainty that we have provided for Spinoza, that
an idea is certain entails, among other things, that it is C&D. Spinoza stresses in the
PPC that we can form a C&D idea of God’s essence before we know whether our
author is a deceiver—a point that Descartes also clearly accepts. What does it mean
to perceive something clearly and distinctly? In the case of Cartesian immutable
natures, it entails that the properties that I C&Dly perceive to belong to that essence
really do belong to that essence. As Descartes puts it: “The mere fact that I can
produce from my thought the idea of something entails that everything which I
clearly and distinctly perceive to belong to that thing really does belong to it” (CSM
II 91).2 In the same place, Descartes claims that it is this sort of understanding that
he has of the truths of mathematics.
But then what, exactly, are we doubting when we doubt whether triangle is
true? It cannot be that triangle in fact follows from the essence of a triangle—the
truth of this is just exhausted by the fact that the truth of triangle can be clearly
and distinctly perceived to belong to the essence of a triangle. It is not clear that
Descartes can give any sense to the claim that he doubts the truth of triangle, once
he grants that he C&Dly perceives the essence of a triangle and C&Dly perceives
that triangle belongs to it. Descartes tries to drive a wedge between the assertion
that triangle belongs to the nature of a triangle and the assertion that triangle can be
truly affirmed of a triangle in the Second Replies (CSM 106) where he distinguishes
between the following claims:
1. “That which we clearly understand to belong to the nature of something
can be truly asserted to belong to its nature,” which claim Descartes deems
“tautological,” and
2. “That which we clearly understand to belong to the nature of something can
truly be affirmed of that thing,” which he seems to think is more substantive.
But Descartes’s own Definition IX in the Second Replies (CSM II 74) comes close
to an admission that (1) and (2) are equivalent: “When we say that something is
contained in the nature of the concept of a thing, this is the same as saying that it is
true of that thing, or that it can be asserted of that thing [my emphasis].”3 Descartes
can try to rely on a distinction between asserting something of a thing and affirming
something of a thing, but I do not see what sort of distinction he has in mind.
Spinoza clearly believes that Descartes cannot maintain this distinction. All that
the truth of triangle consists in is that it follows from the essence of a triangle.
Spinoza applies this concept of truth even to essences that do not seem to be simple
or immutable natures in the Cartesian sense:
If some architect conceives a building in an orderly fashion, then although such a
building never existed, and even never will exist, still the thought of it is true, and
the thought is the same, whether the building exists or not. (TIE 69)
This is Spinoza’s point when he insists in the PPC that I can form a C&D idea of
God’s essence and deduce ideas from it, even before I know that I am not authored
by a deceiver: truth does not depend on a well-intentioned creator matching up my
ideas with the world. In these passages, an idea is true if it does in fact follow from
a given possible essence. Sometimes it sounds like this may be a finite essence: in
the case of the architect’s blueprint, Spinoza seems to suggest that “the architect’s
building has a spiral staircase” is true if and only if it is in fact a feature of that
blueprint. Other times, as in Spinoza’s discussion of certainty above, it sounds like
this could be true only if the architect’s blueprint itself was deduced in the right way
from God’s essence—the cause and origin of nature.
It might look like, while this gets us an (ideally) perfectly coherent network of
ideas, we have not established that anything outside of these ideas, including God,
actually exists.4 But Spinoza thinks that existence claims, like claims about essences,
also follow from God’s essence (E Ip25) and are independent of truths about
essences (e.g., E IIp8s). It is not clear how truths about which things exist and in
what order follow from God’s essence, by Spinoza’s lights, and Spinoza apparently
thinks that we shouldn’t even bother trying to deduce the series of finite existences
from God’s essence, since we are in no position to know it (e.g., Ep. 32). But at least
in principle it can be so deduced.5
However, Spinoza also articulates a different conception of truth than this one,
on which p is true, not if it can be clearly and distinctly perceived to follow from a
given essence, but if it corresponds to its object. In the Ethics, it is axiomatic that
“a true idea must agree with its object” (E Ia6), which axiom serves as the very
definition of truth at Short Treatise II 15. How could this agreement be guaranteed
by certainty, on the account of certainty we attributed to Spinoza?
In the Ethics, Spinoza replaces talk of clarity and distinctness with talk of
adequacy, using them interchangeably in some places (E IIp36, 38). Like clarity and
distinctness, it is an “intrinsic denomination”; an adequate idea
insofar as it is considered in itself, without relation to an object [objectum], has
all the properties, or intrinsic denominations of a true idea . . . I say intrinsic to
exclude what is extrinsic, viz. the agreement of the idea with its object. (E IId4)
348 SKEPTICISM
It seemed like Spinoza had managed to rid himself of his dreaded skeptic by
denying him the required distinction between truth and what is clearly and distinctly
perceived of an essence. But now Spinoza has asked for it. How do you know that
what has the intrinsic denominations of adequacy has the extrinsic denominations
of agreement with an object?
Spinoza asks this question at E IIp43: “How can a man know that he has an idea
that agrees with its object?” “He who has a true idea,” Spinoza answers, “at the same
time knows that he has a true idea, and cannot doubt the truth of the thing.” The
proof claims that if the human mind has an adequate idea of X, then there also is an
adequate idea of the idea of X in that mind, concluding:
And so he who has an adequate idea, or (by P34) who knows a thing truly, must at
the same time have an adequate idea, or true knowledge, of his own knowledge.
I.e. (as is manifest through itself), he must at the same time be certain, q.e.d.
Spinoza’s proof that true ideas carry certainty moves from a mind’s having an
adequate idea of X to its “knowing a thing truly,” which is justified by E IIp34:
“Every idea that in us is absolute, or adequate and perfect, is true.”6 However, the
proof’s reliance on IIp32, which in turn appeals to IIp7c, suggests that the sense of
“agreement” that is being established between an idea and its object is not, prima
facie, the kind of agreement with which the skeptic is concerned. It is instead the
agreement of a mode of thought with its parallel mode, and while adequate ideas do
agree with their objects in this sense, so do all ideas. But this is not what we want to
know: to use Spinoza’s example, when we ask whether Peter’s idea of Paul is true,
we want to know whether it agrees with what it represents—Paul—and not whether
it agrees with its parallel object in God, which is some state of Peter’s body.
This correspondence account of truth seems to be in tension with the account
of truth in at least parts of the TIE; in addition to the passages above, in the TIE
69–71, Spinoza claims that “the form of the true thought must be placed in the
same thought itself without relation to other things.” Spinoza means something
different by “truth” here than he does in the Ethics, where he separates the intrinsic
and extrinsic aspects of an idea from one another. Spinoza goes on to articulate
his epistemology almost exclusively in terms of adequacy and not truth. There are
a number of excellent works that treat Spinoza’s account of truth and adequacy
in greater depth.7 But as long as he maintains a distinction between truth and
adequacy, Spinoza’s response to the skeptic in IIp34 will strike her as a dodge and
not a response, since it does not relate the intrinsic properties of the idea, like
adequacy, to agreement with its object.
3 KNOWLEDGE
Skepticism is usually formulated as a question about the possibility of knowledge,
but we have not really talked about knowledge so far. Actually, Spinoza uses the
word ‘cognitio’, which is the word that is usually translated as “knowledge,” in
contexts that involve very different types of cognitive contents and in contexts that
do not involve truth or even adequacy. There seem to be no special constraints on
what counts as cognitio besides having an idea (in IIp20, e.g., Spinoza refers several
times to “an idea, or knowledge [sive cognitio]”). Spinoza does not have a concept
that plays the role of Descartes’s knowledge or of our knowledge—something that
requires, say, belief, truth, evidence, or certainty. Not only does ‘cognitio’ signify a
variety of kinds of perceptions or conceptions with varying degrees of adequacy, but
no more is required for cognitio than having an idea. The only reason anyone would
doubt this, Spinoza continues in IIp43s, is if they conceived of an idea “as something
mute, like a picture on a tablet,” so that it would be possible to be in possession of a
true idea without believing it. As we have seen, Spinoza denies that this is possible.
Rather than a single distinction between what counts as knowledge and what fails
to qualify, Spinoza proposes a hierarchy of kinds of cognitio. Our highest epistemic
goal is to know the essences of particular things in nature, or to “reproduce the
formal character of nature, both as to the whole and as to the parts” (TIE 91). There
is only one kind of cognition that serves this goal: the kind that deduces the essence
of a thing from the essence of its cause, and ultimately from the “source and origin
of nature,” or God. It is about this kind of knowledge that it is fair to call Spinoza
anti-skeptical. But Spinoza admits that he knows very little in this way (TIE 20).
While every person does have adequate knowledge of God’s essence (E IIp47), it is
difficult to see how we can get from that to knowledge either of the laws of nature
or of the essences of particular things.
Most cognitio is through experientia vaga, or “experience that is not determined
by the intellect” (TIE 20), which includes all the ideas acquired through the senses,
or what Spinoza calls imagination. But this cognition cannot give us adequate
knowledge of nature “as it is in itself” (TIE 9) and never involves certainty (TIE 26).
There is a kind of cognition in between these two that, according to the Ethics,
gives us adequate knowledge of the properties of things in general but not of the
essences of particular things. However, the most natural reconstruction of Spinoza’s
account of their adequacy seems to entail that they are a kind of accidentally
adequate imaginative knowledge. We think that we are perceiving the properties of
external bodies, while we are actually perceiving a combination of properties of our
own bodies and the ones that are affecting us. Usually this would make those ideas
inadequate, but since those properties happen to be the same in certain cases, we
perceive them adequately. This would seem to be precisely the kind of case where
skeptical worries would arise—sometimes those ideas are adequate, and sometimes
they are not, but how can we be sure which one is which? In fact, in the earlier TIE,
Spinoza writes that only intuition “comprehends the adequate essence of the thing
and is without danger of error” (TIE 29). So, at least early Spinoza would have
denied that we know these common properties with certainty.
Given Spinoza’s accounts of truth, certainty, and belief, it is difficult to evaluate
Spinoza as a skeptic or an anti-skeptic in those terms. But if we redefine skepticism
in Spinozistic terms, we find that Spinoza counts as skeptical about a variety of kinds
of possible knowledge: those are the areas of inquiry in which we are not able, for
Spinoza, to form adequate ideas. The reason is not that we cannot have certainty
in these realms, or that our ideas of their subject matters are fallible. Rather, in all
of these cases, Spinoza argues that there are systematic reasons why the ideas that
350 SKEPTICISM
are formed are inadequate. In this way, Spinoza denies that we can possibly have
knowledge of nature of the following kinds: revealed, moral, applied mathematical,
and empirical.
If we understand scientific knowledge as empirical or mathematical knowledge
of nature, then Spinoza seems to be a skeptic about scientific knowledge, belying
the widely accepted story, as expressed by Popkin, that “if Spinoza was an irreligious
sceptic, he was most un- or anti-sceptical in the areas of scientific and philosophical
knowledge” (Popkin 2003: 246). Any cognition from sense experience counts
as the first—inadequate—kind of knowledge, even the most carefully controlled
experiments. In his exchange with Oldenburg about Robert Boyle’s experiments
with niter, Spinoza argues that whatever properties of matter an experiment might
make manifest to the senses, we can never know the deeper causes at play (Ep. 13).
“The way in which things are really ordered and interconnected,” Spinoza admits,
“is quite unknown to us” (TTP II 58). While experiments can usefully catalog the
sensible properties of physical objects, such a catalog is of nature “as it is related to
the human senses” and not “as it is in itself” (Ep. 6).
What about mathematical physics? Spinoza does not doubt the adequacy of
mathematical claims, and he identifies mathematics as a model on which all inquiry
should be based (E I app). But he straightforwardly denies that applied mathematics
provides us with knowledge of nature. Treating bodies in terms of their geometrical
properties, Spinoza claims, involves abstraction, and abstractions do not describe
“real and physical beings” (Ep. 12). Abstraction by its nature leads to confusion,
since it elides particular differences, and all that exists are particulars (TIE 75–76,
93, 99). While Descartes makes a similar critique of abstraction, unlike Spinoza, he
claims that our perception of bodies in terms of their geometrical properties involves
a “clear and distinct intellectual operation” (First Set of Replies, CSM II 84). Spinoza
makes his critique of applied mathematics explicit in his “letter on the infinite”
(Ep. 12) and in E Ip15s, where he argues that “Measure, Time, and Number” are
“only aids of the Imagination.” What is organized by measure, time, and number,
however, are dimensionality, duration, and classes, so these latter, the passages
suggest, are themselves generated by the imagination and thus do not represent
nature as it is in itself.
So much for the limits of natural knowledge in its sphere. Spinoza also denies
that we can have natural knowledge of moral and religious truths. In the Ethics,
Spinoza writes that good and evil, praise and blame, and sin and merit, understood
as absolute, are “nothing but modes of imagining” that “do not indicate the nature
of anything, only the constitution of the imagination” (E I app). It is possible to make
sense, for Spinoza, of something’s being beneficial for another thing if it contributes
to an increase in its power, but what contributes to the power of one finite thing
detracts from the power of another. In fact, Spinoza identifies the reification of
these beings of imagination as itself a cause of skepticism in the Appendix to the
Ethics Part I, since what “seems good to one” turns out to be bad for another.
Spinoza proposes defining revelation as “the sure knowledge of some matter
revealed by God to man” (TTP 1). However, “all that we clearly and distinctly
understand is dictated to us by the idea and nature of God,” according to
Spinoza, so if by “revelation” we mean certain knowledge that comes from God,

then all knowledge, including all natural knowledge, counts as revelation. “[F]or
the knowledge that we acquire by the natural light of reason depends solely on
knowledge of God and of his eternal decrees” (TTP 1).
Prophecy and scripture, however, are divine in a different way. The prophet
interprets revelation for those who cannot achieve certain natural knowledge of
nature and must rely on faith instead to guide their lives toward beatitude. The
goal of prophecy, then, is obedience and not genuine knowledge. This is achieved
through the prophets, who are not individuals with more powerful rational insight,
but with “more lively imaginative faculties.” An examination of scripture, Spinoza
says, reveals that “everything that God revealed to the prophets was revealed either
by words, or by appearances, or by a combination of both” (TTP 1). We saw above
that imagination does not yield adequate knowledge, but Spinoza confirms in the
TTP that “imagination by itself, unlike every clear and distinct idea, does not of
its own nature carry certainty with it” (TTP 2). For this reason, those who rely on
prophecy must demand a sign in order to be certain of what is revealed in prophecy.
Miracles, also, provide us with no understanding of nature or of God: the power
of God and the power of nature are identical, so an event that cannot be explained
through natural causes cannot be explained at all, and so cannot provide us with any
kind of knowledge (TTP 6).
Insofar as there is any true knowledge in scripture, according to Spinoza, it must
also be natural knowledge. Those teachings of scripture that are not have “nothing
to do with philosophy” (Preface, TTP), and they do not represent “the sort of
knowledge that derives from the natural light of reason.” The goal of scripture is
to convey, with moral certainty, the “simple conception of the divine mind” that
teaches obedience to God, justice, and charity. Such cognition is “completely distinct
from natural knowledge in its purpose, its basis, and its method,” such that “these
two have nothing in common.”
4 CONCLUSION
As we saw in the last section, prophecy does not carry certainty with it, but, unlike
natural knowledge, it requires a sign. Prophecy and natural knowledge share in
common that they, like any knowledge, are “dictated to us, as it were, by God’s
nature” (TTP 1). But all of our clear and distinct ideas, for Spinoza, are “dictated to
us by the idea and nature of God—not indeed in words, but in a far superior way
and one that agrees excellently with the nature of mind, as everyone who has tasted
intellectual certainty has doubtless experienced in his own case” (TTP 1). That
revelation requires a sign rather than carrying certainty with it makes it inferior to
natural knowledge, according to Spinoza (TTP 3). Similarly, in the TIE, Spinoza
stresses that the “true method does not consist in seeking the signs of truth after the
acquisition of the idea. . . . The truth needs no sign” (TIE 61).
Let us return to Spinoza’s account of Descartes’s response to the skeptical
scenario. Spinoza, like Descartes, sees some kind of knowledge of God as part of
the guarantee of knowledge that the demand for certainty represents. But according
352 SKEPTICISM
to Spinoza, this is knowledge of God’s essence, not of God’s existence and veracity.
We saw that Spinoza appreciated Descartes’s foundational project when it was
understood as an attempt to examine all of our ideas in order to clarify them, so
that they either lost their power or furnished evidence for rather than against the
doubted belief. But Spinoza rejected the idea that we must know our origin in order
to know that our beliefs are not caused in a way that leads us into error.
Now, this is not to say that Spinoza is not concerned with the causal origin of our
ideas. Like for Descartes, in order for ideas to be true, they must be caused in the
right way. God’s essence grounds our knowledge because knowledge of an effect,
for Spinoza, is grounded in knowledge of its cause (E Iax4). Since God’s essence is
the cause of everything in nature, “all our conceptions involve God’s nature and are
conceived through God’s nature, thus we can accept finally, that everything that we
adequately conceive is true” (TIE 63). Moreover, according to Spinoza, our true
ideas involve or express that cause: “Nothing can be conceived without God” and
“everything in Nature involves and expresses the conception of God in proportion
to its essence and perfection; and therefore we acquire a greater and more perfect
knowledge of God as we gain more knowledge of natural phenomena” (TTP 6).
This suggests that just by attending to a true idea, the knowledge of its cause, God,
is evident, and this is a mark of its truth.
For Descartes, to treat the origin of our faculties as dependent on the will of
God independently of their place in nature is, according to Spinoza, to treat them as
miracles and to treat the idea of God as a sign. God’s goodness acts as a guarantee
of the agreement of the idea with its object, but without making any clearer the
nature of truth. This is illustrated by the fact that for Descartes to prove that our
clear and distinct ideas are true because God guarantees their veracity does not
at all require that we characterize or understand the nature of truth. All that is
important for that proof is that God would not allow that our clear and distinct
ideas were false, whatever truth or falsity might be. According to Spinoza, this turns
all natural knowledge into revelation—an inferior kind of knowledge. Spinoza
takes his response to make the nature of truth plain, and to relate it necessarily to
God’s essence.
Spinoza adopts the Cartesian skeptical method, if by that we understand that he
appreciates that we must have knowledge of God to have knowledge of anything.
But we also require knowledge of God to understand the nature of truth. Being
able to be derived from God’s essence just is their truth, for Spinoza. He is an anti-
skeptic insofar as we can do that. But since we know very little in this way, he is a
skeptic about many types of knowledge claims.
NOTES
1. I use the following abbreviations for Spinoza’s works: E = Ethics, where p =
proposition, d = demonstration, def = definitions, c = corollary, a = axiom, app =
appendix, l = lemma; KV = Short Treatise on God, Man, and his Well-Being; TIE =
Treatise on the Emendation of the Intellect; TTP = Theologico-Political Treatise;
CM = Appendix Containing Metaphysical Thoughts; Ep. = Epistle. For quotations
from the Principles of Cartesian Philosophy (PPC), I provide the page number from
Curley (1985): for example, “C 237.” All quotations are from Curley (1985). I am
very grateful to Baron Reed for his insightful comments on this chapter and for
enlightening conversation.
2. He mentions the fact that he has shown God’s veracity, but seems to go on to suggest
that their certainty predates this—equivocating, I believe, between psychological and
normative certainty.
3. And see also the Seventh Replies (CSM II 310): “No matter who the perceiver is,
nothing can be clearly and distinctly perceived without its being just as we perceive it
to be, without being true.”
4. Della Rocca (2007) argues that Spinoza resists skepticism by identifying the truth
of an idea, its intelligibility, and the existence of what it represents. I think that the
account of truth and certainty I have described so far dovetails nicely with his and is
inspired by many of the same considerations. However, as I go on to suggest, I do not
think this can be the full story, since Spinoza seems elsewhere to suggest that there is
a distinction between the intelligibility of a thing’s essence and its existence.
5. Note that all this can be done without appealing to any experience of the cogito or
anything else that actually exists. But Spinoza does allow that we have an immediate
knowledge of at least one finite existence—each of us has immediate knowledge of
the existence of her own mind and body (E IIp2 and 4). As Perler (2018) argues, this
precludes skepticism about the existence of the physical world, and also the external
physical world, because we “feel [our] body to be affected in many ways” (E IIp4).
Perler convincingly argues that we can develop these ideas that we get in this way
about the things that actually affect us, connecting them up to some extent with
the C&D ideas that we deduce from the essence of God. Given that these ideas are
infinitely confused, while they give us knowledge that things exist, it is hard to know
how to connect this knowledge of particular existences with the knowledge from
essences that we get by deducing ideas from the clear and distinct idea of God’s
essence. I see E IIp4 as Spinoza’s explanation of how we have anything approaching
knowledge of finite existences in our local environment, given that there is no chance
that we could deduce them from God’s essence, rather than as a response to the
skeptic.
6. Not to mention that Spinoza seems to be offering a new account of certainty here—
one that requires having an idea of an idea.
7. See, for example, Curley (1975) and Della Rocca (1994).
REFERENCES
Bolton, Martha. (1985). “Spinoza on Cartesian Doubt,” Noûs 19: 379–395.
Curley, Edwin. (1975). “Descartes, Spinoza and the Ethics of Belief.” In E. Freeman and
M. Mandelbaum (eds.), Spinoza: Essays in Interpretation, 159–189. La Salle, IL: Open
Court.
Curley, Edwin. (1985). The Collected Works of Spinoza, Volume 1. Princeton: Princeton
University Press.
Della Rocca, Michael. (1994). “Mental Content and Skepticism in Descartes and Spinoza,”
Studia Spinozana 10: 19–42.
354 SKEPTICISM
Della Rocca, Michael. (2007). “Spinoza and the Metaphysics of Scepticism,” Mind 116:
851–874.
Doney, Willis. (1971). “Spinoza on Philosophical Skepticism,” The Monist 55: 617–635.
Perler, Dominik. (2018). “Spinoza on Skepticism.” In M. Della Rocca (ed.), The Oxford
Handbook of Spinoza, 220–239. Oxford: Oxford University Press.
Popkin, Richard. (2003). The History of Scepticism from Savanarola to Bayle. Oxford:
CHAPTER TWENTY-FIVE
Pierre Bayle
JOHN CHRISTIAN LAURSEN
1 INTRODUCTION
The scholarship on Pierre Bayle (1647–1706) seems designed to reproduce the
ancient skeptical tradition: equipollent arguments on both sides of any issue. On
the one side, he has been described as a skeptic, and even a “superskeptic” (Popkin
2003: 283, 300–301). On the other side, he has been described as a thoroughgoing
rationalist and dogmatic philosopher (Mori 1998; Israel 2006). He wrote so much,
and returned to every issue multiple times and from different angles, that there is
some justification, somewhere in his writings, for almost every possible interpretation
of him. This very prolific many-sidedness may deserve its own place in the history
of skepticism: the refusal to leave any issue alone, the constant return to arguments
and opinions from every point of view, the prolixity which was perhaps matched by
no other author in the tradition. Perhaps we should call it “zetetic” skepticism, after
the ancients who kept on inquiring, as Sextus Empiricus put it (PH I 7). No issue
was finished, closed, or immune to reopening as long as Bayle was alive. He was not
a philosopher of closure.
There are ways in which Bayle’s writings are similar to the skepticism of the
Pyrrhonian and Academic traditions from Pyrrho, Arcesilaus, and Carneades
through Cicero, Sextus Empiricus, and many others to Montaigne. These traditions
need not be interpreted as incoherent, inconsistent, or immoral, as has often been
the case, and hence Bayle’s skepticism also need not be dismissed on those grounds
(Laursen 2004). And there are also times in which Bayle’s writings express dogmatic,
and especially moralistic, claims. This chapter will compare and contrast these
opposing positions. We shall begin with his understanding of famous philosophers
such as Pyrrho, Descartes, Spinoza, and Malebranche in the entries of his Historical
and Critical Dictionary (1697). A substantial part of what he wrote was history
of philosophy, and, as one scholar has put it, there was “in the case of Bayle, an
indissoluble link between philosophical skepticism and history as an empirical
science” (Smith 2013: 30). Then we will turn to his practical philosophy. He wrote
many books that were philosophical interventions into the raging issues of the day
such as religious persecution and toleration, and in some of them we can detect the
influence of his skepticism.
356 SKEPTICISM
2 BAYLE ON THE ANCIENT SKEPTICS

There is no doubt that Bayle was aware of the skeptical traditions. In a letter of
1673 to his friend Vincent Minutoli he paraphrased Sextus Empiricus to the effect
that “the most general division between all of the sects of philosophers is between
those who think they have found the truth, those who think it cannot be found,
and those who think they have not yet found it and keep searching,” and reviewed
the teachings of the Academics and Pyrrhonians (Labrousse et al. 1999: vol. 1,
185). One of the most famous entries in his Historical and Critical Dictionary was
“Pyrrho.”1 However, as we shall see, it is probably fair to say that Bayle never tried
to follow closely any school of skepticism, and better to say that throughout his life
he improvised his own variations on the theme of skepticism.
The article on Pyrrho surveys some of the chief elements of Pyrrhonism. They
can be usefully divided into the critiques of reason and the claim that custom and
education determine most of what we think. Pyrrho supplied a series of underminings
of reason: “the minds of men are unstable”; “the inconstancy of human opinions
and passions is so great that it might be said that man is a small republic that often
changes its magistrates” (Bayle 1991: 209 note F); the Pyrrhonians “were always
seeking, and always unsteady, and ready at all times to reason after a new manner,
according to the various occurrences they met with” (Bayle 1734–1738: vol. 4,
657 note F). On the role of custom and education, “Pyrrho maintained that there
was not really any one thing that was this or that, and that the nature of things
depended upon laws and customs; that is to say, that men by their laws and customs
made some things good, laudable, ill, blameable &c” (1734–1738: 658 note I). So
the alternative to reason was largely custom and habit. In “Jonas” he observed that
“ideas of virtue depend on education and custom” (Bayle 1991: 106).
Pyrrhonism is mentioned or discussed in many other articles, usually to make one
of the above points about the weakness of reason and the strength of custom and
education (Bayle 1734–1738: vol. 5, 103 Index). The same points are often made
without reference to Pyrrhonism. In “Bunel” he quotes Cardinal Pole for saying
that one should “admit that he knows nothing; that this is the non plus ultra of
philosophy” and thus that one should turn to religious faith for truth (Bayle 1991:
42). In his own voice Bayle says that
I find that Pole’s judgment is the most sensible one that can be made about
philosophy; and I am extremely pleased that such an author gives me an
opportunity of confirming what I have set down in various places, namely, that
our reason is suitable for making everything perplexing and for raising doubts
about everything. No sooner has it built something than it provides the means for
destroying it. Reason is a veritable Penelope, unraveling during the night what
she has been weaving during the day. (1991: 42)
In the entry, “Manicheans,” he observes that human reason “is a principle of

destruction . . . [and] is only fit to make man aware of his own blindness and
weakness” (1991: 151). His critique of the Socinians in “Socinus, Faustus” is that
their reliance on reason is “hardly suitable for converting the people . . . [and]
PIERRE BAYLE 357
is more appropriate for leading persons . . . into skepticism (pyrrhonisme)” than

anything else (Bayle 2000a: 266). Such observations are repeated in a wide range of
variations on the theme in Bayle’s vast texts.
Pyrrhonism was often accused of fostering immorality and atheism in Bayle’s
day, and he later added a “Clarification concerning Pyrrhonism” to the second
edition of his Historical and Critical Dictionary (1702) in order to defend it against
these charges. It was not subversive, he wrote, because he was saying nothing in
it about religion, only about philosophy, and that Jesus himself had endeavored
to “confound all philosophy and show its vanity” just like the Pyrrhonians (Bayle
1991: 423). Bayle often made a major distinction between faith and reason, which
insulated religion from the inroads of reason and also insulated philosophy from
the inroads of religion. His argument was that one could not know anything in
philosophy—a version of skepticism—but that religious faith did not require anyone
to know anything.
Bayle’s article on Pyrrho sends us elsewhere for more on his understanding of
skepticism. Of Pyrrho, he says: “His opinions did not differ much from those of
Arcesilaus,” one-time head of the Academy and one of the founders of Academic
skepticism (Bayle 1991: 194). The only difference was that the Academics affirmed
the incomprehensibility of things where the Pyrrhonians suspended judgment even
about that: “In every thing else they were perfectly alike” (Bayle 1734–1738: vol.
4, 653 note A). In the article “Arcesilaus,” Bayle wrote that this Academic skeptic
approved of the verse of Hesiod that “the Gods keep human Understanding behind
the Veil” (Hickson 2010: 561). The article on Arcesilaus’s successor, Carneades,
reports that the latter “allowed of nothing but Probabilities for the use of Life; and
beyond that he believed there was no Certainty or Evidence” (1734–1738: vol. 2,
326 note B). When he says of the Academics that “their Speculation hung betwixt
two Contraries; but their Practice stuck fast to one of them” and observes that
“It is the way of the World: Men do not live up to their Principles,” he is further
undermining rationalism, which would require one to resolve which of the two
contraries is right before acting on one of them, and that men should live up to their
principles (1734–1738: vol. 2, 330 note G). This was one of the crucial features of
Bayle’s skepticism: even if we do know something, there is a disconnect between
reason and life, and we do not live by our knowledge. Based on these and other
remarks, some scholars have concluded that Bayle was an Academic skeptic (Maia
Neto 1999; Lennon 2002).
It is possible to interpret Bayle in somewhat narrowly Pyrrhonian or Academic
terms. It is also possible, to follow a remark by Thomas Lennon, that “Bayle’s
approach is thus less skeptical than ‘critical’” (1999: 10). Given the prominence of
this term in the title of Bayle’s Dictionary and in the titles of the three most substantial
volumes of Immanuel Kant’s later philosophy, it is surprising how little historical
and analytical attention that word has received. In the “Projet” or proposal for his
Dictionary, Bayle claimed that “critical research” would help us avoid mistakes and
give us “a true notion of the weakness of the mind, and of the nothingness and
vanity of the sciences” (Bayle 2000a: 13). Bayle’s skepticism certainly included a fair
measure of critical thinking. It is a remarkable paradox that, as one scholar has put
358 SKEPTICISM
it, Bayle seems to have had an obsession with the truth, a “critical obsession” which
led him to test and take apart every idea in the history of philosophy, such that no
truth survived his analysis (Simonutti 2011).
3 BAYLE ON THE MODERN PHILOSOPHERS

The above can serve as a background for Bayle’s readings of famous philosophers
such as Descartes, Spinoza, and Malebranche. They are skeptical and critical
deconstructions of these thinkers. It is well known that Descartes based his answer
to the hypothesis of the malin génie on the claim that God would not lie to us. In his
entry “Rimini,” Bayle pointed to medieval and contemporary Socinian arguments
that God did indeed have the power to lie (Paganini 2008a: 368–370). This led to
all sorts of paradoxes such as, if God cannot lie, then is he not all-powerful? And if
he can lie, how can we possibly tell when he is telling the truth and when he is lying?
Bayle provided no answers, and the questions lingered into the eighteenth century
(Laursen 1994). After Bayle, the Cartesian principle of divine veracity did not have
much purchase.
This was not all that Bayle had to say about Descartes. He may have begun as a
Cartesian philosopher, but he moved far away. In the Dictionary he reported that
Descartes was not original: he borrowed his theory of the vortices from the Greek
philosopher Leucippus (Bayle 1991: 125). Bayle gave balancing counterarguments
to what Descartes had said about the vacuum, noting that this favored the
Pyrrhonians (1991: 136–137). He lined up the counterarguments to Descartes’s
view of animals as machines (1991: 214–245). He delved into the pros and cons
of Cartesian claims about the extension of spirit (1991: 280–284) and the nature
of God (1991: 369–374, 385), and he found contradictions in all of the works of
the Cartesians of his day. The foregoing is only a small fraction of what Bayle wrote
to undermine Descartes. Much of his critique of Descartes and the Cartesians was
not direct refutation, but in the classic skeptical form of providing examples of
counterarguments from the history of philosophy.
Spinoza fared no better with Bayle. He opposed Spinoza largely because the
latter was a rationalist, and rationalism was one of Bayle’s bêtes noires. Bayle wrote
that Spinoza’s doctrine of monism, that there is one substance and that it is identical
with God, meant “giving the words a completely new signification” and has absurd
consequences (Bayle 1991: 313, 330). He argued that religious faith was perfectly
justifiable because “there is no contradiction between these two things: (1) The
light of reason teaches me that this is false; (2) I believe it nonetheless because I am
convinced that this light is not infallible” (1991: 298). In this and in later polemics
against the rationalist theologians of his era, Bayle insisted that rationality had no
place in religion (Brogi 1998; Paganini 2008b).
Bayle associated Malebranche with Chrysippus, who raised more objections
against his own doctrine than he could answer. In Bayle’s reductio ad absurdum,
Malebranche is accused of subjecting God to fate and implied that God selfishly
pursues glory at the expense of human suffering (Paganini 2008a: 371–379; 2008b:
115–118). The upshot of Bayle’s analyses of the modern philosophies is that none
PIERRE BAYLE 359
of them can stand. Assertions are met by counterassertions, arguments are shown to
imply absurdities. No philosophy emerges unscathed. This surely must qualify as a
combination of several sorts of skepticism.
4 BAYLE ON ATHEISM
Skepticism is often associated with atheism, although they are not the same thing:
atheism usually means a dogmatic theory that there is no God, which the more
consistent skeptics could not admit. They are closer to the agnostics when they
suspend judgment about the Gods. Bayle’s first important work, Diverse Thoughts
on the Comet of 1682, was not really about comets but about the nature of belief,
the effects of atheism on society, and more. He undermines those who think we can
come up with reliable and true principles of behavior by pointing out, as he would
often repeat, that even if we did have such principles, they would not affect our
behavior. Most of the time one does not live by one’s principles, but according to
one’s “temperament, the natural inclination toward pleasure, the taste one contracts
for certain objects, the desire to please someone, a habit gained in the commerce of
one’s friends” (Bayle 2000b: 169). People also live for their passions, their pride,
and in accordance with the “mechanical constitution of their nature” (2000b: 210–
211), which track fairly closely to the first two of Sextus Empiricus’s four rules by
which the skeptic lives after suspending belief (PH I 23). Showing that we live by the
alternatives to knowledge and truth is one way of indicating that whether or not we
have access to truth may be irrelevant.
Perhaps the most notorious implication of this argument, also followed up in
some of Bayle’s later works, was the case for the possibility of a virtuous society of
atheists. It was widely accepted in Bayle’s time that, without religion, people would
violate all of the rules of morality and kill each other. Bayle argued on the contrary
that people’s actions are determined by their temperament and taste, not by their
opinions about religion (Bayle 2000b: 178–180). Rewards and punishments, pride,
reputation, vanity, are more than enough to keep social peace (2000b: 212–213,
223). We have records of atheists who have been good people (Bayle 2000b: 214).
Bayle also noted that the record of religious people was just as bad as what was
conjectured of atheists: one example was the court of Catherine de Medici, where
“although they believed in God, they were capable of every sort of wickedness”
(2000b: 193).
Many years later, when the Dictionary was also accused of atheism, Bayle
answered with a “Clarification concerning atheists” in the second edition of the
work. There he pointed out that, in the absence of moral truths derived from truths
about God, people might nevertheless behave appropriately on the basis of their
“temperament, education, liveliness of ideas of virtue, love of glory, or dread of
dishonor” (Bayle 1991: 407). He does not retract what critics perceived as a defense
of atheism, but repeated his view that it could be harmless to society. And a few
years after that, in some of his last works, Bayle reaffirmed the value of the example
of the ancient Chinese as combining atheism with good morals and politics (Bayle
1966a: 227–229, 397; Bayle 1966b: 958, 966, 983, 988).
360 SKEPTICISM
Does the foregoing mean that Bayle was an atheist? Scholars have been divided.
Ever since his own time, many have thought that he was an atheist, and many still do
so today (Mori 1999, 2003; Israel 2006; McKenna 2010, 2012). Others, however,
have found him to be a “fideist,” or someone whose religion is based on faith,
not knowledge (Popkin 2003); a “complicated Protestant” (Bost 2006, 2008); and
religious enough to write theodicy (Hickson 2010, 2011, and 2013). Mori makes
the convincing argument that Bayle’s natural philosophy from the time of his early
courses at Sedan leaves no room for a deity (Mori 1999). And it seems likely that,
if Bayle had been an atheist, he would have concealed it for the sake of safety
(McKenna 2010). But Bayle often reiterated the observation that one’s philosophy
may have no connection with how one lives, what religion one follows, and other
matters. Arguing that philosophy and knowledge have no effect on life and that we
live by other means is a form of skepticism. And in this case, without trying to figure
out what the “real” Bayle thought, we can observe that like other skeptics of the
tradition, he left many clues and arguments on each side of the issue.
In his time and in the eighteenth century, most readers believed Bayle was
an atheist, and his influence was wide. His Dictionary was the most widely held
book in French private libraries in the eighteenth century (Lennon 1999). G. W.
Leibniz thought it important to refute him, David Hume took notes from Bayle
while writing his Treatise of Human Nature, and virtually every other significant
eighteenth-century thinker engaged with his thought. His contributions to shaping
our views of the history of philosophy and of skepticism cannot be overestimated.
We owe to him our understanding of much of this history as a debate between
dogmatic rationalism and skepticism.
5 RATIONALISM AND SKEPTICISM IN BAYLE’S

MORE PRACTICAL WRITINGS
Bayle’s Philosophical Commentary of 1686, widely accepted as one of the greatest
defenses of religious toleration in the history of philosophy, also contains a tension
between skepticism and rationalism (Bayle 2005). It may be internally contradictory,
and perhaps even incoherent. But the history of philosophy includes far more than
the perfectly consistent and coherent, and even works that do not meet the highest
standards of consistency have had their influence. A skeptic of Bayle’s stripe may not
be too worried about this, because he is occupied with juxtaposing and analyzing
all sorts of arguments, and in this case his determination to support toleration
and condemn persecution means that he is willing to use any arguments, even
contradictory ones.
Bayle’s book is in the first instance a commentary on a parable of Jesus (Luke,
ch. 14) that was used throughout the seventeenth century by both Catholics and
Protestants to justify religious persecution (Marshall 2006: 201–212, 404, 465,
643–646). In that parable, a man prepares a celebration and invites his friends.
They are unable to come, so he orders his servant to “go out into the highways and
hedges,” find people, “and compel them to come in.” This is a story about inviting
PIERRE BAYLE 361
people to a feast, and yet it was used as early as St. Augustine to justify violent
persecution of heretics. The feast was interpreted to mean the church, and the word
“compel” was taken to justify the use of force to bring people into the church. Jesus
is often quoted for the things he says about peace and gentleness, but he also chased
the moneychangers out of the temple (Jn 2:15) and said: “I come not to send peace
but a sword” (Mt. 10:34), so persecutors did not have to rely on Lk. 14:23 alone.
Bayle, a French Protestant by birth who by that time was living in exile in
Rotterdam, responded to Augustine and the persecutors with one of the most
thoroughgoing arguments for religious toleration in the early modern Western
tradition. At the beginning of the Philosophical Commentary, he asserts the priority
of what he variously calls the “Light of Nature,” the “first Principles of Reason”
(Bayle 2005: 65), “inward Light,” “natural Light,” and “interior Truth” (2005: 70).
All theologians rely on reason whenever they make arguments. That means “That
Reason, speaking to us by the Axioms of natural Light, or metaphysical Truths, is
the supreme Tribunal, and final Judg without Appeal of whatever’s propos’d to the
human Mind” (2005: 67). Bayle asserts that reason naturally comes up with the
idea of moral equality and fairness. Thus “all moral Laws, without exception, ought
to be regulated by that natural Idea of Equity” (2005: 69). There is a “Conscience
of Right and Wrong imprinted on the Souls of all Men,” an “interior Light which
communicates it self immediately to all Spirits” (2005: 71). This is a dogmatic,
rationalist argument, with no trace of skepticism.
We may begin to suspect the purity of Bayle’s rationalism when we realize that
he is actually promoting Christian theology. Relying on the literal reading of the
word “compel” to justify violence, Bayle asserts, would be “repugnant to the purest
and most distinct Ideas of natural Reason” (Bayle 2005: 75). But it is a very specific
theology that emerges as the product of reason: “By the purest and most distinct Ideas
of Reason, we find there is a Being sovereignly perfect, who rules over all things,
who ought to be ador’d by Mankind, who approves certain Actions, and rewards
‘em” (2005: 76). This is not pure light and reason but a Christian interpretation of
religion. Bayle asserts that several of the principles of Christianity are products of
reason: “Natural Light distinctly inform[s] every Soul which attentively consults it,
that God is just, that he loves virtue,” etc., and everyone knows “That ‘tis honest
and praise-worthy to forgive Enemys” (2005: 81). He also cites the complaints
of the first Christians against violent persecution (2005: 124), but these cannot
be considered universal or rationalist arguments. They are arguments from one
particular religion.
In Bayle’s hands, Protestant theology undermines the argument that violent
conversion is justified by law: “All laws which oblige against Conscience, are made
by a Person, having no Authority to enact it, and who manifestly exceeds his Power,”
and are thus unjust (Bayle 2005: 113). Conscience “is the Voice and Law of God
in” each “particular Man” (2005: 113). Princes have “no Power, either from God
or Man, of enjoining their Subjects to act against Conscience” (2005: 114). Note
that this Protestant definition of conscience is not universal, and not shared by
most Catholics. Bayle also uses non-universal Christian arguments such as that
allowing Christians to make forcible conversions deprives them of one of their chief
362 SKEPTICISM
objections to the Muslims. Christians always asserted that false religions relied on
force where the true religion could rely on truth, and if they rely on force, they are
conceding that theirs might not be true.
The argument proceeds with references to “the Dictates of right Reason” (Bayle
2005: 195) and the “clearest principles of Metaphysics” (2005: 220) as if these
were unproblematic. But Bayle began to undermine them by the claim that “every
one’s determin’d by his own private Lights” (2005: 74). If the light of reason is not
universal, but rather individual, then all of his arguments from natural light and
reason are refuted. But now we get to the most interesting part: in chapters 8 to
10 of Part Two, Bayle undermines his own assertions of the rationality of religious
toleration. He pursues a skeptical strategy instead of a dogmatic strategy in order to
argue against the idea that Jesus intended us to persecute in order to compel people
to join his church. Bayle may have no better reason for his stance than his own gut
feelings, but that is an accepted Pyrrhonian justification. In a few earlier places he
reminded people that they had been wrong before and might still be wrong now. But
now he appeals openly to human ignorance and ineptitude in his famous analysis of
the erroneous conscience (see Laursen 2001).
Up until this point, Bayle has generally been relying on a strong notion of truth.
Now he argues that “whatever a Conscience well directed allows us to do for the
advancement of Truth, an erroneous Conscience will warrant for advancing a
suppos’d Truth” (Bayle 2005: 220). But if correct and erroneous consciences cannot
agree, how can we know which is the real truth? He insists that “the Law of never
violating the Lights of our Conscience is such that God himself can never dispense
with” it because it is an “eternal and immutable Law” (2005: 227). But clearly many
people’s consciences have not recognized that law.
Bayle realizes that his arguments justify conscientious persecutors. His answer is
only that “it does not follow, that they act without Sin,” and that “this ought to not
hinder our crying out loudly against their false Maxims, and endeavoring to enlighten
their Understandings” (Bayle 2005: 242). But by his theory they are required to act
on their wrong understandings. He asserts that no one would ever preach “Sodomy,
Murder, and Rapine” because they are patently against “natural Rectitude” and
scripture (2005: 244), but we know that many people have considered those actions
justifiable, perhaps under different labels.
Bayle’s epistemological arguments undermine much of what he has said before.
“God has not printed any Characters or Signs on the Truths which he has reveal’d,
at least not on the greatest part of ‘em, by which we might certainly and infallibly
discern ‘em” (Bayle 2005: 259). If “Conscience is given us as a Touch-stone of
Truth” and our consciences err, then we will not agree on truths (2005: 261). He
concludes that “God in the present Condition of Man exacts no more from him
than a sincere and diligent Search after truth,” not actually finding it (2005: 264). In
one of his most devastating arguments, he even concludes that consciences are not
very different from matters of taste: “No matter whether this Conscience presents
to one Man such an Object as true, to another as false. . . . Does not one man’s Tast
tell him that such Food is good, and the Tast of another tell him it’s bad?” (2005:
271). But truths and light were the foundation of his claims for toleration up to this
PIERRE BAYLE 363
point, and if they are now just like matters of taste, they cannot make a universal
case for toleration.
The fact is, Bayle insists, we believe what we believe based on our education.
“Education is undoubtedly capable of making the Evidence of Truths of Right
utterly disappear” (Bayle 2005: 275). But if education can make truths disappear,
what has become of the universal truths of light and reason? The Supplement to
Bayle’s book, published two years later, reinforces this side of Bayle’s argument. He
demonstrates repeatedly that it is “the imperious Force of Education” (2005: 470),
“Custom and Education” (2005: 471), that determines beliefs (see also 2005: 473–
474), not reason, a universal conscience, and natural light. As for reason, it emerges
that there are reasons on both sides of every issue. Yet another reason to doubt the
universality of reason is that we do not know why one set of reasons appeals to one
person and another to another person. Bayle is reduced to saying that “naturally his
Turn of Understanding is such, that he’s struck much more by one sort of Reasons
than by another” (2005: 490). Anyone may be attracted to one set of reasons “by an
odd Turn of Understanding” or by a “relish” (2005: 490). As Bayle puts it, “I can’t
for my part see, that there must necessarily be a criminal Passion at heart, to make
a Man prefer the Reasons which are alleg’d against some Points of Orthodoxy, to
the Reasons which support ‘em” (2005: 490). But if “reasons” can go either way,
what does that imply for the appeals to reason and natural light in the first parts of
the book?
If anyone thinks that Bayle might distinguish univocal philosophical truth
from debatable theological truth, it is worth observing that for Bayle not only
theological but also philosophical reason can be wrong. Another chapter of the
Supplement, titled “Whether the Arguments for the Truth are always more solid
than those for Falsehood” (Bayle 2005: 531–538), undermines any claims for
philosophical reason. There, he says, “necessary truths . . . [are] so evident . . .
[that they] carry their own Reason along with ’em which no one contests” (2005:
531). But if anyone does contest them, then they are not such necessary truths.
Needless to say, the persecutors not only contest the ideas of the people they
persecute, but they just as obviously contest the idea that persecution is wrong. So
these are cases “when a necessary truth is not evident” (2005: 531). And people
decide what these are according to the way “they are determin’d by Temper
and Complexion to some Notions rather than others” (2005: 533). In all cases,
“each have had their Patrons and Partys in different Countrys, and different Ages
of the World” (2005: 533). If a person “inclines to one, ‘tis more from some
Prevalence of Temper than that of the Arguments” (2005: 534). This means that,
on contested issues like the value of persecution, it is not abstract and universal
reason but rather temperament that determines our judgment. Bayle has been
reduced to claiming that, if one wants to (and Bayle does), one may be tolerant,
but there does not appear to be any way to argue conclusively and convincingly
against a temperament that prefers persecution.
On the one hand, Bayle has written of the clearest truths of reason and natural
light. On the other, he has also raised epistemological barriers to knowing what
these truths are, or if there really are any, and argued that we are captive to
364 SKEPTICISM
our consciences, whether right or erroneous, and are products of our education
(Laursen 2011).
Much of what Bayle says would seem to vindicate Richard Popkin’s judgment
that “Bayle’s all-out attack on everything that is said and everything that is done
carried scepticism to its ultimate extreme” (2003: 300). Popkin is aware of the
recent reinterpretations that see Bayle as a “positive moralist” (2003: 294) relying
on reason, and recognizes that he “adopted positive moral views” (2003: 297). But
set in the context of recognition of the rights of the erroneous conscience, the fact
that some consciences are right and some are wrong does not get much traction.
Bayle may believe his reason has taught him some truths, but he also recognizes
that other people’s reason teaches them different truths. This may be interpreted
as a sort of suspension of judgment about the universal truth of moral ideas and
amount to following the traditional skeptical reliance on customs and education in
determining how to live.
The foregoing makes at least a prima facie case that Bayle carried out the
Pyrrhonian skeptical project in his Philosophical Commentary, even though he did
not explicitly say that that was what he was doing. Later, in explicit discussions
of Pyrrhonism he described, often with the same words, what he had done in the
Philosophical Commentary. The upshot is that a book which starts out taking for
granted universal truths and conscientious morals ends up arguing that one reason
we cannot be meant to persecute the people who are wrong is the good Pyrrhonian
reason that we can rarely be sure who is right and who is wrong. Pyrrhonian reasons
justify this conclusion: reason is weak and works itself into paradoxes, and we are
products of our education. So who is the real Bayle, the dogmatic rationalist or the
Pyrrhonian who recognizes the weakness of reason and the power of education and
conscience?
Jonathan Israel brings out the rationalist side of Bayle in Enlightenment
Contested (Israel 2006). He asserts that “for Bayle . . . philosophical reason more
geometrico is the only criterion of what is true” (2006: 72). “Bayle actually bases his
thought . . . on a radical Cartesian mathematico-historical rationalist foundation”
(2006: 77). And “On the fundamental rationality of human morality . . . Bayle
refuses to compromise and is never skeptical” (2006: 558). Israel says that Bayle’s
toleration theory is “based solely on philosophical reason and notions of ‘equity,’”
and he accepts Bayle’s assertion that reason must lead to “equity, equality, individual
liberty, and a comprehensive toleration” (2006: 147, 680). But this means ignoring
the places where Bayle says that reason may tell us other things, too. An optimistic
Enlightener like Israel thinks that, if people will only reason rightly, they will agree
with him on the democratic virtues. But Bayle’s Part Two, chapters 8 to 10 and the
Supplement undermine that hope. If reason can be deconstructed, we are blinded
by our education, and things are much more complicated: the same corrosive reason
that undercuts the claims of Christianity and monarchy may also undercut equality
and democracy. Israel’s interpretation requires ignoring Bayle’s critique of the
rationalist theologians Jaquelot, Le Clerc, and Bernard, who argued that reason
leads us to Christian principles. Bayle argued that it does not, and that from the
point of view of reason there are serious paradoxes in Christianity (Brogi 1998).
PIERRE BAYLE 365
He must have realized that there are paradoxes in equality and democracy as well.
Too many of Bayle’s texts undermine reason and may be fatal to Israel’s project of
placing Bayle and reason at the heart of the Radical Enlightenment.
Israel knows Bayle’s discussion of the erring conscience in the Supplement,
but he does not see any serious implications (2006: 149–154). For him, it is all
part of his argument for Bayle’s rationality: “Nothing could be less Pyrrhonist
than Bayle’s system of moral, social, and political thought” (2006: 670). But
notice that he has to interpret Bayle by adding something. When Israel recognizes
Bayle’s view of the power of education, he adds that Bayle, “presumably, means
the reader to infer, without his spelling it out, that education itself must therefore
be fundamentally reformed” (2006: 569). And Bayle’s problems would be solved
if all erring consciences could be educated into correct, tolerant opinions. But
Bayle does not show any confidence that this can be done, and that may be why he
does not supply the recommendation for educational reform that Israel suggests
he should.
Any attempt to harness Bayle’s Philosophical Commentary to a straightforward,
one-dimensional “Enlightenment project” of reason and natural light must take
into account the arguments that undermine that rationalism that we have reviewed.
However, this does not mean that Bayle’s argument for toleration must fail. Bayle
is unquestionably against persecution—even against persecution of persecutors;
although that does not mean he endorses them—and the wide range of his
arguments, even if some of them undermine others, could nevertheless have the
effect of encouraging toleration. Perhaps the Pyrrhonian undermining of all sides
could even prevent toleration from becoming a rigid dogma.
Bayle’s analysis of toleration in the Philosophical Commentary was not his
only contribution to the case of religious toleration. The Historical and Critical
Dictionary also contains an arsenal of arguments for toleration. Some of them were
grounded in skeptical arguments, and some of them were dogmatic. Several of his
political pamphlets, such as France Wholly Catholic (1686) and Important Warning
to the Refugees (1692) were strident and moralistic cries for justice without any trace
of skepticism (Stanley et al. 2014). This has been called his “moral rigorism,” and
it is undeniable that he has strong moral commitments (Labrousse 1964). Whether
they have unequivocal rational justification is another question. They may be more
the product of his education and the customs and traditions around him than of any
philosophical proof, as he often put it.
It is sometimes asserted that skepticism must lead to political conservatism
because one cannot fight for reform or revolution without truths, or that it must
lead to revolutionary politics because the established realm is not backed by truth.
But Bayle allowed his gut feelings of indignation at persecutors of all stripes to
drive his politics of toleration. And his skepticism helped him find a middle way
between revolutionary politics and arbitrary power: his was a politics of absolute
but not arbitrary power, as one scholar has put it (Simonutti 2007). He wanted the
Protestants to earn an invitation to return to France by developing a reputation as
loyal subjects, and he saw great value in the rule of law. So in his case skepticism
implied neither radicalism nor conservatism, but a middle way.
366 SKEPTICISM
6 CONCLUSION
After reviewing all of the materials above, it is not surprising that the status of
Bayle’s skepticism has been much debated (McKenna 2008). There is no doubt
that he is not a slavish follower of any preceding school of skepticism. There is no
doubt that he was not a complete moral skeptic or relativist but held some moral
positions with all possible dogmatic conviction. But on the whole, given his use of
all of the tropes and practices of the ancient and modern skeptics up to his time, his
industrious inquiries into every historical and philosophical issue of the ages, and
given his refusal to take any philosophical issue as settled, it is at least defensible
to consider him one of the more original as well as influential skeptics of all time.
NOTE
1. English translations of this entry can be found in Bayle (1734–1738) and Bayle (1991).
It will be cited to the easily available 1991 abridged version unless the relevant passage
is only available in the complete 1734–1738 version.
REFERENCES
Bayle, Pierre. 1734–1738. The Dictionary Historical and Critical of Mr. Peter Bayle.
London: Knapton et al. (Originally published as Dictionaire historique et critique.
Rotterdam: Leers, 1697; second edition 1702.)
Bayle, Pierre. 1966a. Continuation des pensées diverses écrites à un Docteur de Sorbonne, à
l’occasion de la Comète. In Œuvres diverses, vol. 3, 187–417. Hildesheim: Olms. (First
published in 1705.)
Bayle, Pierre. 1966b. Rèponse aux questions d’un provincial. In Œuvres diverses, vol. 3,
449–1084. Hildesheim: Olms. (First published in 1706.)
Bayle, Pierre. 1991. Historical and Critical Dictionary, edited by R. Popkin. Indianapolis:
Hackett.
Bayle, Pierre. 2000a. Political Writings, edited by S. Jenkinson. Cambridge: Cambridge
University Press.
Bayle, Pierre. 2000b. Various Thoughts on the Occasion of a Comet, edited by R. Bartlett.
Albany: State University of New York Press. (Originally published as Pensées diverses à
l’occasion d’une comete. Rotterdam: Leers, 1682, second edition 1683.)
Bayle, Pierre. 2005. A Philosophical Commentary on the Words “Compel Them to Come
In”. . ., edited by J. Kilcullen and C. Kukathas. Indianapolis: Liberty Fund. (Originally
published as Commentaire philosophique sur ces paroles de Jésus-Christ, “Contrain-les
d’entrer.” Amsterdam: Abraham Wolfgang, 1686–1688.)
Bost, Hubert. 2006. Pierre Bayle. Paris: Fayard.
Bost, Hubert. 2008. “Pierre Bayle, un ‘protestant compliqué’.”In W. Van Bunge and H.
Bots (eds.), Pierre Bayle (1647–1706): Le philosophe de Rotterdam: Philosophy, Religion
and Reception, 83–102. Leiden: Brill.
Brogi, Stefano. 1998. Teologia senza verità: Bayle contro i ‘rationaux’. Milano: Angeli.
PIERRE BAYLE 367
Hickson, Michael. 2010. “The Message of Bayle’s Last Title,” Journal of the History of
Ideas 71: 547–567.
Hickson, Michael. 2011. “‘Reductio ad malum’: Bayle’s Early Skepticism about Theodicy,”
The Modern Schoolman 88: 201–221.
Hickson, Michael. 2013. “Theodicy and Toleration in Bayle’s Dictionary,” Journal of the
History of Philosophy 51: 49–73.
Israel, Jonathan. 2006. Enlightenment Contested: Philosophy, Modernity, and the
Emancipation of Man 1670–1752. Oxford: Oxford University Press.
Labrousse, Elisabeth. 1964. Pierre Bayle. Vol. II: Hetérodoxie et rigorisme. The Hague:
Nijhoff.
Labrousse, Elisabeth, Antony McKenna, et al. (eds.). 1999. Correspondance de Pierre Bayle.
Oxford: Voltaire Foundation.
Laursen, John Christian. 1994. “The Beneficial Lies Controversy in the Huguenot
Netherlands,” Studies on Voltaire and the Eighteenth Century 319: 67–103.
Laursen, John Christian. 1998. “Baylean Liberalism: Tolerance Requires Nontolerance.”
In J. C. Laursen and C. Nederman (eds.), Beyond the Persecuting Society: Religious
Toleration Before the Enlightenment, 197–215. Philadelphia: University of Pennsylvania
Press.
Laursen, John Christian. 2001. “The Necessity of Conscience and the Conscientious
Persecutor: The Paradox of Liberty and Necessity in Bayle’s Theory of Toleration.” In
L. Simonutti (ed.), Dal necessario al possibile: determinismo e libertà nel pensiero anglo-
holandese del XVII secolo, 211–228. Milan: Angeli.
Laursen, John Christian. 2004. “Yes, Skeptics Can Live Their Skepticism and Cope with
Tyranny as Well as Anyone.” In J. Maia Neto and R. Popkin (eds.), Skepticism in
Renaissance and Post-Renaissance Thought, 201–223. Amherst: Humanity Press.
Laursen, John Christian. 2005. “Skepticism, Unconvincing Anti-skepticism, and Politics.”
In M. A. Bernier and S. Charles (eds.), Scepticisme et modernité, 167–188. Saint-
Étienne: Publications de la Université de Saint-Étienne.
Laursen, John Christian. 2011. “Skepticism against Reason in Pierre Bayle’s Theory of
Toleration.” In D. Machuca (ed.), Pyrrhonism in Ancient, Modern, and Contemporary
Philosophy, 131–144. Dordrecht: Springer.
Lennon, Thomas. 1999. Reading Bayle. Toronto: University of Toronto Press.
Lennon, Thomas. 2002. “What Kind of Skeptic Was Bayle?” Midwest Studies in Philosophy
26: 258–279.
Maia Neto, José R. 1999. “Bayle’s Academic Scepticism.” In J. E. Force and R. Watson
(eds.), Everything Connects, 264–276. Leiden: Brill.
Marshall, John. 2006. John Locke, Toleration, and Early Modern Culture. Cambridge:
McKenna, Antony. 2008. “Pierre Bayle in the Twentieth Century.” In W. Van Bunge and H.
Bots (eds.). 2008. Pierre Bayle (1647–1706): Le philosophe de Rotterdam: Philosophy,
Religion and Reception, 253–267. Leiden: Brill.
McKenna, Antony. 2010. “Pierre Bayle et le bouclier de Charron.” In H. Bost and
A. McKenna (eds.), Les ‘Eclaircissements’ de Pierre Bayle, 299–319. Paris: Honoré
Champion.
368 SKEPTICISM
McKenna, Antony. 2012. “Pierre Bayle: Free Thought and Freedom of Conscience,”
Reformation and Renaissance Review 14: 85–100.
Mori, Gianluca. 1999. Bayle Philosophe. Paris: Honoré Champion.
Mori, Gianluca. 2003. “Pierre Bayle on Scepticism and ‘Common Notions’.” In G. Paganini
(ed.), The Return of Scepticism from Hobbes and Descartes to Bayle, 393–413.
Dordrecht: Kluwer.
Paganini, Gianni. 2008a. Skepsis: Le débat des modernes sur le scepticisme. Paris: Vrin.
Paganini, Gianni. 2008b. “Bayle et les théologies philosophiques de son temps.” In W. Van
Bunge and H. Bots (eds.) 2008. Pierre Bayle (1647–1706): Le philosophe de Rotterdam:
Philosophy, Religion and Reception, 103–120. Leiden: Brill.
Sextus Empiricus. 1933. Outlines of Pyrrhonism, translated by J. B. Bury. Cambridge, MA:
Harvard University Press.
Simonutti, Luisa. 2007. “‘Absolute, not Arbitrary, Power’: Monarchism and Politics in
the Thought of the Huguenots and Pierre Bayle.” In H. Blom, J. C. Laursen, and L.
Simonutti (eds.), Monarchisms in the Age of Enlightenment, 45–59. Toronto: University
of Toronto Press.
Simonutti, Luisa. 2011. “Obsesión con la verdad: paradojas de Pierre Bayle.” In
M. J. Villaverde and J. C. Laursen (eds.), Forjadores de la Tolerancia. Madrid: Tecnos.
Smith, Plínio. 2013. “Bayle and Pyrrhonism: Antinomy, Method, and History.” In
S. Charles and P. Smith (eds.), Scepticism in the Eighteenth Century: Enlightenment,
Lumières, Aufklärung, 19–30. Dordrecht: Springer.
Stanley, Charlotte and John Christian Laursen. 2014. “Pierre Bayle’s The Condition of
Wholly Catholic France,” History of European Ideas 40: 312–324.
Van Bunge, Wiep and Hans Bots (eds.) 2008. Pierre Bayle (1647–1706): Le philosophe de
Rotterdam: Philosophy, Religion and Reception. Leiden: Brill.
CHAPTER TWENTY-SIX
Berkeley and Skepticism

MARGARET ATHERTON
1 INTRODUCTION
Getting a handle on the role of skepticism in the thought of George Berkeley
(1685–1753) is surprisingly difficult. As is well known, Berkeley’s case for idealism
includes old and familiar skeptical arguments. And since the conclusions Berkeley
wants to reach themselves seem so incredible, it is sometimes thought that an
important motive for Berkeley must be to show that his central theses don’t lead
to skepticism. Finally, Berkeley himself presents skepticism neither as a tool nor as
a potential menace to his theory but as a target. He claims we have been led into
skepticism by the views of others. The task he sets himself is to dispel the pernicious
skepticism these others have encouraged. Part of the difficulty in pinning down
skepticism’s role for Berkeley is also that there are too many skeptical beliefs that
might be the focus of his interest in skepticism. Skepticism might be expressed in the
belief that there is neither heat nor cold in the water, or that we are unable to have
perceptual knowledge of ordinary physical objects, or that we are unable to know
the inward essences of things.1 While any or all of these beliefs might be implicated
in Berkeley’s attack on skepticism, it is important to straighten out the nature of
the skepticism Berkeley takes himself to be primarily concerned with. A stress on
different aspects of skepticism—ordinary objects as opposed to inward essences, for
example—results in different accounts of Berkeley’s overall project. My goal here
is to achieve greater clarity on this issue by answering three questions: (1) What are
the skeptical beliefs Berkeley wants to dispel? (2) What are the false principles that
he takes to lead to these skeptical beliefs? And (3) what is Berkeley’s goal in ridding
us of this skepticism? In particular, is he seeking to return us to common sense or a
version of it, or is he trying to instill a better theory based on principles that do not
lead to skepticism?2
A hurdle anyone must face is that Berkeley wrote two books laying out his
immaterialism, books that do not follow exactly the same line of argument. Three
Dialogues between Hylas and Philonous is conspicuously structured around the issue
of skepticism. The two characters agree to debate the question, who is the greater
skeptic, and they return frequently to examine the implications of what they have
been saying concerning the matter of skepticism. In Principles of Human Knowledge,
on the other hand, Berkeley lays out his positive doctrine (PHK: 1–33)3 without any
370 SKEPTICISM
mention of skepticism at all, and when he enlarges on this doctrine in his replies
to objections (PHK: 34–85), he talks about skepticism only once.4 It is natural that
commentators looking to examine Berkeley’s position with respect to skepticism
have focused on the presentation in Three Dialogues.5 There is nevertheless good
reason to take skepticism to be equally important to the argument of the Principles.
Just as he does in Three Dialogues, Berkeley signals in the full title that skepticism
is a central concern.6 More importantly, he lays out his project in the Introduction
to the Principles as addressed to the issue of skepticism, and makes good on this
promise in the final section of the Principles (PHK: 85–156), where he examines the
consequences of what he terms his principles with respect to skepticism. Any attempt
to answer the questions raised about Berkeley’s position on skepticism is going to
need to start with the Principles. What we discover, I am going to maintain, is that
the position against skepticism to be found in the Principles is not only interesting in
itself, but provides a useful framework by which to read Three Dialogues.
2 THE PRINCIPLES
In the Introduction to the Principles, Berkeley sets up the matter of skepticism as
the target of his work through a contrast between ordinary people of common sense
and philosophers. Ordinary people are never troubled by skepticism, whereas when
we “reason, meditate, and reflect on the nature of things,” we soon fall prey to
“uncouth paradoxes, difficulties and inconsistencies” until we “sit down in a forlorn
skepticism” (PHK: Intro 1). In this state, we blame our faculties as not being suited
to “penetrate into the inward essence and constitution of things” (PHK: Intro 2).
But perhaps, Berkeley suggests, this is because we misuse the faculties God gave us.
As he puts it, “We have first raised a dust, and then complain we cannot see” (PHK:
Intro 3). Berkeley’s goal is to identify those principles he suspects to be false that
philosophers have introduced and that have produced this belief, that our mental
faculties are not capable of providing knowledge.
There are some important points that emerge here from Berkeley’s account of
his project. The first is obvious: Berkeley takes his task to be undoing skepticism.
The second is equally obvious. Berkeley sets up a contrast between the “illiterate
bulk of mankind,” who have common sense and are not prone to skepticism, and
another group, philosophers, who are responsible for raising skeptical doubts.
The obvious conclusion is that Berkeley is taking philosophers as his target and
the skepticism he is worried about might be called “philosophical skepticism.” It is
the false principles of philosophers he will expose. It is also worth observing that
there is a “we” moving through these sections that clearly excludes men of common
sense. Berkeley is not just taking philosophers as his target but also includes them
as his audience. He is writing for those who are likely to fall prey to philosophical
skepticism. The skeptical beliefs Berkeley seeks to dispel will be beliefs likely to be
held by that narrow segment of mankind who might have meddled with philosophy.
The final section of the Principles proves to be extremely helpful for identifying
the principle that is “the very root of skepticism” (PHK: 86). It is the supposition
of “a twofold existence of the objects of sense, the one intelligible, or in the mind,
BERKELEY AND SKEPTICISM 371
the other real and without the mind; whereby unthinking things are thought to
have a natural subsistence of their own, distinct from being perceived by spirits”
(PHK: 86). Earlier, in PHK 56, Berkeley has explained how philosophers acquire
this belief in a twofold existence, so this principle leads to philosophical skepticism.
True, we begin with an ordinary belief everyone shares: that they have some
ideas of which they are not the cause. Everyone assumes these “ideas or objects of
perception” exist mind-independently, “without ever dreaming that a contradiction
was involved in those words.” This is a mistake that anyone is prone to, namely,
taking our sensible ideas to exist without the mind in a kind of unreflective naïve
realism.7 But, Berkeley says, philosophers see through the contradictions embedded
in this widespread misapprehension and conclude that “the immediate objects of
perception do not exist without the mind.” Philosophers correct the false belief
of “mankind,” but then fall themselves into the false belief “that there are certain
objects really existing without the mind, or having a subsistence distinct from being
perceived, of which our ideas are only images or resemblances, imprinted by these
objects of the mind” (PHK: 56). The troublesome twofold existence principle is
introduced by philosophers to solve the same problem that led to the naïve realism
of ordinary people, that of identifying a cause for ideas we do not cause.
Berkeley’s account of how the philosopher’s principle of twofold existence leads
to skepticism is two-pronged. The first centers on the philosophical belief that ideas
resemble their causes, so that real knowledge occurs when ideas conform to real
things. This principle drives philosophers to skepticism because it seems to follow
that it is impossible to determine either the nature or the existence of real things. We
cannot determine the nature of any real thing because “things remaining the same,
our ideas vary, and which of them, or even whether any of them at all represent
the true quality really existing in the thing, it is out of our reach to determine”
(PHK: 87). Thus, philosophers conclude that what we see or otherwise sense, being
relative to our senses, cannot inform us of the absolute nature of things. Nor can
we even establish the existence of real things, inasmuch as there is no demonstrable
connection between our sensible qualities and any alleged absolute qualities
belonging to a thing.
Philosophers are therefore left in the uncomfortable position of holding that
our sensible ideas have an external cause, but unable to provide any account of that
cause. Berkeley suggests that it is practitioners of natural philosophy who are driven
to greater depths of skepticism. They propose that “there is in each object an inward
essence, which is the source whence its discernible qualities flow, and whereon they
depend” (PHK: 102). They assume there must be a cause of our sensible ideas the
nature of which necessitates the nature of the properties we discern. Being unable to
produce a viable account of the nature of this inward essence or of the causal process
by means of which qualities are supposed to flow, it is they who proclaim that our
faculties are unable to reveal the inner constitution of anything, being “amused
only with the outside and shew of things” (PHK: 101). This second set of beliefs
can produce what Berkeley likes to present as increasingly desperate attempts to
bolster incoherent accounts of this real but inaccessible cause. And the principle this
deeper form of skepticism depends on is the belief that “everything includes within
372 SKEPTICISM
itself the cause of its properties: or that there is in each object an inward essence”
(PHK: 101). What Berkeley provides in the Principles is a dialectic of increasingly
skeptical beliefs. From the true belief that the immediate objects of perception exist
only in the mind, philosophers adopt the false belief in the twofold existence of
ideas and their causes. They first maintain that ideas conform to their causes, and
then, when faced with the skeptical consequences of this principle, adopt a further
false principle that ideas are an outward show concealing the inward essence of real
things. At this point we have achieved the complete denigration of the senses and of
our intellectual faculties as well.
In the way Berkeley develops his account of his skeptical target in the Principles, the
real villain of his story is not just philosophers in our sense, but natural philosophers,
those physicists-cum-philosophers attempting to provide an account of the order of
nature. In the Principles, again, the remedy Berkeley is urging involves rethinking
the nature of what we take to be a cause of sensible ideas. Berkeley explains that
progress in natural philosophy is actually achieved through “a greater largeness of
comprehension” (PHK: 105) that allows natural philosophers to subsume natural
phenomena under laws of increasing generality. We should understand these general
laws connecting ideas, not as capturing a causal relation or as grounded in an
unknown cause but as “a mark or sign with the thing signified. The fire which I
see is not the cause of the pain I suffer upon my approaching it, but the mark that
forewarns me of it” (PHK: 65). Such explanations of phenomena derive from the
orderly and hence predictive nature of our ideas, and do not require a belief in a
twofold existence of ideas and their material causes. When we recognize the efficacy
of such an explanatory procedure, we will forego a search for non-existent natural
efficient causes. Berkeley has replaced the false principle of twofold existence with
his own true principle, that the only things that exist are minds and ideas, on the
basis of which he shows us how to develop a new non-skeptical science or natural
philosophy. We no longer identify the reality of an idea as one that is caused in the
“right” way, but locate reality within the realm of ideas displaying a lawful order.
3 THREE DIALOGUES
When we turn our attention to Three Dialogues where skepticism is brought to
the fore, a different picture emerges. The motivation launching the discussion is
Hylas’s conviction that Philonous is a greater skeptic than he, and debating this
question allows the implication that Berkeley’s own theory is as likely to lead to
skepticism. Also in Three Dialogues, the problematic beliefs identified typically
concern sensible things like cherries and gloves. This has led commentators to
assume that the skepticism under attack in Three Dialogues concerns the sense
perception of ordinary objects. As George Pappas puts it, a primary skeptical belief
is that “[w]e do not gain knowledge by means of perception of either the nature or
existence of ordinary objects” (Pappas 2000: 287). In Three Dialogues it looks as
though the way to dispel skepticism is to show that we do indeed have perceptual
knowledge of ordinary objects. Finally, the frequent references to common sense
throughout Three Dialogues suggest Berkeley’s goal is to return us to commonsense
beliefs, for example, about the existence and nature of cherries and gloves. Three
Dialogues, with its emphasis on commonsense beliefs about ordinary objects, seems
like a different enterprise than the one uncovered in the Principles. A careful reading
of Three Dialogues , however, reveals that Berkeley is following the same program
and reaching similar conclusions as the Principles.
Undoubtedly, the issue addressed in Three Dialogues is framed differently and in
a more open-ended fashion than in the Principles. In the Principles, Berkeley’s target
is specifically philosophers, but is Hylas, Philonous’s interlocutor, a philosopher?
The question they discuss, who denies the existence or knowledge of sensible things,
sounds like an investigation into ordinary beliefs, such as, that cherries exist and I
know about them. So is Hylas a person of common sense with straightforwardly
realist views about the sensible things? Hylas, unfortunately, is not consistently
either one. There is perhaps an explanation why Berkeley tends to present Hylas as a
moving target. Berkeley is using Hylas to recapitulate the dialectic he outlined in the
Principles. In order to reach his starting point, Berkeley needs Hylas to hold initially
the commonsense view Berkeley identified in the Principles: that “ideas or objects of
perception had an existence independent of, or without the mind” (PHK: 56). Thus,
Berkeley has Hylas start by saying: “whatever degree of heat we perceive by sense,
we may be sure the same exists in the object that occasions it” (3D: I 175). Hylas
is asserting a commonsense conviction that in having ideas, we just are perceiving
qualities in the object. But this is not Hylas’s final position. Berkeley shows Hylas
becoming convinced of what in PHK 56 he had identified as the philosopher’s true
belief that the immediate objects of perception exist only in the mind. Once Hylas
learns to be a philosopher, we see him subsequently developing the positions that
lead him further and further into skepticism.
The first step in Hylas’s conversion to philosophy is accomplished in the first
half of the First Dialogue, ending with Hylas agreeing “that all sensible qualities
are alike to be denied existence without the mind” (3D: I 194). In reaching this
conclusion, Berkeley is often supposed to be deploying skepticism to shake Hylas
loose from his conviction that sensible qualities also exist mind-independently.
Berkeley offers a number of different arguments based on phenomena of perceptual
variability, arguments that are often identified with the Argument from Illusion.
The first concerns a hoary chestnut, the vat of water that feels hot to one hand and
cold to the other. “On your principles,” Philonous points out, “it is really both hot
and cold at the same time” (3D: I 179) But what is this argument intended to show?
The principle Hylas is often taken to be using here, “Heat is a quality of the fire,”
would indeed encourage a skeptical conclusion, namely, we don’t know whether the
water is hot or cold, but not the conclusion Hylas seems prepared to draw: “There
is no heat in the fire.” But notice the principle Hylas initially asserted was actually:
“Whatever degree of heat we perceive by sense, we may be sure that the same exists
in the object that occasions it.” This is the belief Berkeley attributes to ordinary
people. It is not a belief about hotness and water per se, but rather a belief about
perceiving by sense, that perceiving by sense just is perceiving the quality there in
the world. This belief is undermined when we realize we simultaneously perceive
both hot and cold, since it would mean that both qualities exist simultaneously in
374 SKEPTICISM
the water. The purpose of this case (and others Philonous offers) is to move Hylas
from the ordinary person’s position to that of the philosopher, “that the immediate
objects of sensation do not exist without the mind.” Berkeley continues deploying
the fleeting, variable nature of the sensible qualities to show it is not only temperature
and colors that cannot be identified with a “real” quality in the object, but also the
so-called primary qualities. Shape and size are subject to perceptual variation so it is
equally inappropriate to identify the size now experienced with the real size of the
object. The conclusion, then, of the extended discussion of sensible qualities finds
Hylas abandoning his initial position for the more correct, philosophical one: “that
all sensible qualities are alike to be denied existence without the mind” (3D: I 194).
Hylas is no longer a person of commonsense.
Converting Hylas does not instantly make him a skeptic. Nevertheless, by the end
of the First Dialogue, Philonous makes this accusation: “You are therefore by your
principles forced to deny the reality of sensible things, since you made it to consist
in an absolute existence exterior to the mind. That is to say, you are a downright
sceptic” (3D: I 206). Philonous here blames Hylas’s newfound skepticism on his
“principles.” This suggests that somewhere between the passage in which Hylas
gives up his old commonsense principle (3D: I 194) and the passage just quoted
(3D: I 206), Hylas has acquired a new principle. What is that new principle and how
did Hylas come by it?
It is important to notice that, in this intervening passage, Berkeley is using the
fictional Hylas to show how philosophers fall into skepticism once they become
convinced that sensible qualities exist only in the mind, and distinguish between
mind-dependent ideas and mind-independent real things. Hylas, who is ostensibly
raising objections to Philonous, about half-way through enunciates a position of
his own:
To speak the truth, Philonous, I think there are two kinds of objects, the one
perceived immediately, which are likewise called ideas; the other are real things
or external objects perceived by the mediation of ideas, which are their images
and representations. Now I own, ideas do not exist without the mind; but the
latter sort of objects do. (3D: I 203)
Hylas has just articulated what in the Principles Berkeley has called the two-fold
existence principle and that he blamed for skepticism. Hylas is now committed to the
view that immediately perceived mind-dependent ideas enable us to have knowledge
of mind-independent real things. And this view will lead to skepticism if in fact such
ideas cannot convey knowledge of mind-independent things. In the remainder of
the First Dialogue, Philonous demonstrates that these skeptical consequences of this
principle do indeed follow. Briefly, he argues that, as the experience of hearing a
coach coming down the street shows, we can be said to have mediately perceived
some sensible qualities of the coach, such as its familiar shape, when they are
suggested by the immediately perceived sound of the coach. But we cannot be said
to have mediately perceived Julius Caesar by immediately perceiving the light and
colors of his statue, unless we have had the sort of experiences that would allow ideas
of the statue to suggest Caesar himself. The immediate perceptions alone cannot lead
us to the idea of Julius Caesar. Nor can ideas intelligibly be said to be images of the
real things they are supposed to represent. Fleeting and variable ideas do not have
the right characteristics to represent stable, unchanging real things. Finally, sensible
ideas cannot be said to be like insensible real things. Thus, with what is now called
the Likeness Principle, Berkeley concludes that Hylas’s adoption of the philosophical
principle of the twofold existence of ideas and things leads to skepticism.
At the end of the First Dialogue, Hylas has become a philosopher, inasmuch as
he endorses the twofold existence of ideas and things, but he has not yet, despite
Philonous’s charge, become a skeptic. This changes in the Second Dialogue. There,
Hylas and Philonous explore what in the Principles Berkeley had identified as the
motive for adopting the twofold existence principle, the recognition that not all
of my ideas are caused by me. The Second Dialogue suggests that it is the search
for a cause for our ideas that tumbles philosophers into deep natural philosophical
skepticism. At the beginning of the dialogue, Hylas puts forward a version of the
psychophysical account of the causation of sensible ideas that many philosophers of
his time found appealing. He is brought up short when Philonous reminds him that
all of the entities in the account, the bodies that impact on vibrating nerves that leave
traces in the brain, are in fact so many sensible things or ideas. Hylas’s account is
therefore an account of relations among contingently connected ideas. Confronted
with the fact that he has not actually identified a cause of ideas, Hylas admits to
skepticism but takes comfort in the observation that Philonous too is a skeptic.
This theme, stemming from the original question about who is the greater skeptic,
is now pursued to the end of Three Dialogues. Hylas is assuming that, since both
he and Philonous have agreed they only perceive ideas in the mind, they are both
skeptics, not capable of arriving at knowledge of real things. Philonous, however,
identifies a different source of Hylas’s skepticism. “You indeed,” Philonous points
out to Hylas,
said, the reality of sensible things consisted in an absolute existence out of the
minds of spirits, or distinct from their being perceived. And pursuant to this
notion of reality, you are obliged to deny sensible things any real existence: that
is, according to your own definition, you profess yourself a sceptic. But I neither
said nor thought the reality of sensible things was to be defined in that manner.
(3D: II 211–212)
In this speech, Philonous has redefined the principle to which Hylas is committed:
it is now the thoroughly philosophical principle that the reality of sensible things
depends upon an absolute, that is, mind-independent existence. In the last part of
the Second Dialogue, Hylas accepts the principle of independent existence that
Philonous assigns to him but is gradually brought to admit that all his attempts to
make sense of the notion are either contradictory or meaningless. It is at this point,
when Hylas has been deprived of any meaningful version of what Berkeley believes
to be the central concept of natural philosophers, independent existence, that Hylas
ultimately succumbs to skepticism.
At the start of the Third Dialogue, then, we find Hylas in just such a state of
“forlorn scepticism” Berkeley describes in the Introduction to the Principles. Hylas
376 SKEPTICISM
has decided that “our faculties are too narrow and too few” for us to have any
knowledge of sensible things, “their inner constitution, their real and true nature” lie
hidden from us. Hylas now is one of those people he described in the First Dialogue
who distrusted their senses, denied the real existence of sensible things, and pretend
to know nothing of them. But what of Philonous? He denies he is skeptical about
the existence of things, but on what grounds? He associates himself with the vulgar
and says he trusts his senses. Does he think Hylas’s current forlorn skepticism can
be defeated simply by showing we have immediate perceptual knowledge of things,
that is, by defending a version of commonsense, or does he have something else
in mind?
There are many passages, particularly early in the Third Dialogue, that suggest
Berkeley is indeed endeavoring to defend a kind of perceptual common sense, as
when he has Philonous says: “I cannot for my life help thinking that snow is white
and fire hot” (3D: III 229). Philonous might here be saying, so long as we agree
that sensible things are just bundles of sensible qualities and that we immediately
perceive sensible qualities, then we know the existence and nature of sensible things
when we immediately perceive them. Skepticism about the existence or nature of
sensible things is impossible. But when the skepticism issue is raised later in the
dialogue, Philonous gives a more complicated defense. Hylas accuses Philonous of
skepticism because, he says, Philonous is changing things into ideas (3D: III 244).
Philonous’s retort is that he is actually changing ideas into things because he denies
things have hidden essences. Hylas raises objections to this claim, that things have
no hidden essence or inner constitution, relying on the observation that we are not
presented with sets of stable sense qualities. The apple we experience is not a bundle
of one color, one shape, and one size, but a highly variable set of shifting colors,
shapes, and sizes. If this variable set constitutes one apple, it can only be because
the varying sensible qualities derive from a single inner constitution. It is this stable
inner constitution, too, that explains why we think microscopes give us a better view
of something. Philonous endorses the variability of sensations that Hylas points to.
“Strictly speaking, Hylas,” he says, “we do not see the same object that we feel;
neither is the same object perceived by the microscope, which was by the naked eye”
(3D: III 245). In giving this answer, Philonous is not asserting that a sensible thing
is a relatively stable group of qualities, such as could be grasped immediately and at
once, but is instead a highly variable set of qualities which are “observed however to
have some connection in Nature either with respect to co-existence or succession”
(3D: III 245). Similarly, with the microscope, we are not getting a better view of
the thing, but learning more about what ideas go together. As Philonous puts it,
“The more a man knows of the connection of ideas, the more he is said to know
of the nature of things” (3D: III 245). What defeats skepticism for Berkeley is the
recognition that the vast shifting mass of our ideas are not to be identified as real
because of a relationship to some stable thing, but instead because of their relations
with one another when they are regularly connected together.
Berkeley concludes this section by giving Philonous a long speech in which he
walks Hylas through the steps that led to his skepticism, ending with his inability
to conceive an independently existing substance. Philonous concludes: “Lastly,
whether the premises considered, it be not the wisest way to follow Nature,
trust your senses, and laying aside all anxious thought about unknown natures or
substances, admit with the vulgar those for real things, which are perceived by the
senses?” (3D: III 246). It has been important throughout Three Dialogues that real
things are perceived, but this is not all of importance. As in the Principles, it is
equally important that real things be perceived as orderly. The point is spelled out
when Berkeley revisits for the last time the issue of skepticism:
Is not that opposition to all science whatsoever, that phrensy of the ancient and
modern sceptics, built on the same foundation? Or can you produce so much as
one argument against the reality of corporeal things, or in behalf of that avowed
utter ignorance of their natures, which doth not suppose their reality to consist in
an external absolute existence? Upon this supposition indeed, the objections from
the colours of a pigeon’s neck, or the appearances of a broken oar in the water,
must be allowed to have weight. But those and the like objections vanish, if we
do not maintain the being of absolute external originals, but place the reality of
things in ideas, fleeting indeed, and changeable; however not changed at random,
but according to the fixed order of Nature. For herein consists that constancy
and truth of things, which secures all the concerns of life, and distinguishes that
which is real from the irregular visions of the fancy. (3D: III 258)
In this passage, Berkeley is returning to the position of the Principles, that the reality
of things lies not just in their being immediately perceived, but also in their being
perceived in a fixed order, thus allowing us to know what to expect. In the often-
quoted passage at the end of the Third Dialogue, where Berkeley says he is reconciling
the views of the vulgar and the philosophers, the claim that he is attributing to the
vulgar, “that those things they immediately perceive are the real things,” is rendered
acceptable to Berkeley only on his own notion of what it is to be real, that is, for
ideas to occur in a fixed and orderly manner so that they can be connected by us.
The position that Berkeley has in the end argued for in Three Dialogues is in fact
that of the Principles, and his ultimate target is the skepticism that results from a
belief in mind-independent essences. His solution to this skepticism again lies in
the regularities among ideas understood according to the laws of nature. We are
now back at the anti-skeptical position of the Principles: what enables us to have
knowledge of the natural world is that our ideas occur in a stable law-governed
manner, such that present ideas signify to us what to expect. It is this law-governed
nature of our ideas that allows us knowledge beyond whatever commonsense objects
are before us, up to and including “the whole system immense, beautiful, glorious
beyond expression and beyond thought” but which is as well “linked in a mutual
dependence and intercourse with each other” (3D: II 211).
4 CONCLUSION
Let me now return to the questions raised at the beginning. (1) What are the
skeptical beliefs Berkeley wants to dispel? These are the beliefs that he assigns to
Hylas at the beginning of the Third Dialogue after he has succumbed to skepticism
378 SKEPTICISM
and that he raises in the Introduction to the Principles: that our nature and our
faculties are inadequate to allow us knowledge of the nature and existence of things.
Berkeley wishes to get rid of the notion that our Creator has equipped us with a
limited set of cognitive faculties. (2) What are the false principles that he takes to
lead to these skeptical beliefs? The false principles Berkeley identifies as the culprit
in giving rise to skepticism are (a) the twofold existence of ideas and things and (b)
the identification of real things with something that is independently existing and
unperceiving. The principles Berkeley wants to reject like the skepticism it gives rise
to are endemic to philosophy. Finally (3) What is Berkeley’s goal in ridding us of this
skepticism? It is not to vindicate common sense but to build a better philosophical
theory, which reveals that our God-given mental faculties are thoroughly sufficient
to provide knowledge of the existence and nature of things.
NOTES
1. For a thorough discussion of the first kind of skepticism, see Dicker (2011: ch. 5),
and for the second, see Pappas (2000: ch. 9). I will be defending the third possibility
below.
2. The importance of common sense to Berkeley has been recently stressed by Pappas
(2000) and by Roberts (2007). Their accounts have been criticized by Bordner
(2011).
3. References to the Principles will be to PHK and section number, or to PHK Intro and
section number. References to Three Dialogues will be to 3D, Dialogue Number and
page number in the Luce and Jessop edition.
4. Berkeley (PHK: 40) denies that he is endorsing a skepticism of the senses.
5. This is the case, for example, with George Pappas, whose work has been important in
calling attention to the centrality of skepticism in Berkeley’s theorizing (Pappas 2000:
esp. ch. 9).
6. The full title of the Principles is: A Treatise Concerning the Principles of Human
Knowledge wherein the Chief Causes of Error and Difficulty in the Sciences, with
the Grounds of Scepticism, Atheism, and Irreligion, are inquired into. Even the
shorter title of Three Dialogues introduced in the third edition stress the matter of
skepticism: Three Dialogues between Hylas and Philonous In opposition to Sceptics
and Atheists.
7. He is here presumably talking about the same strangely prevailing opinion identified
in Berkeley (PHK: 4): that “houses, mountains, rivers and in a word all sensible
objects have an existence, natural or real, distinct from their being perceived by the
understanding.”
REFERENCES
Berkeley, George. 1949a. The Principles of Human Knowledge. The Works of George
Berkeley Bishop of Cloyne, volume II, edited by T. E. Jessop. London: Thomas Nelson
and Sons Ltd. (First published in 1710.)
Berkeley, George. 1949b. Three Dialogues between Hylas and Philonous. The Works of
George Berkeley Bishop of Cloyne, volume II, edited by T. E. Jessop. London: Thomas
Nelson and Sons Ltd. (First published in 1713.)
Bordner, Seth. 2011. “Berkeley’s ‘Defense’ of Common Sense,” Journal of the History of
Dicker, Georges. 2011. Berkeley’s Idealism. New York: Oxford University Press.
Pappas, George. 2000. Berkeley’s Thought. Ithaca: Cornell University Press.
Roberts, John Russell. 2007. A Metaphysics for the Mob. New York: Oxford University
Press.
CHAPTER TWENTY-SEVEN
A Pyrrhonian
Interpretation of
Hume on Assent
DONALD L. M. BAXTER
1 INTRODUCTION
How is it possible for Scottish philosopher David Hume (1711–1776) to be both
withering skeptic and constructive theorist? I recommend an answer like the
Pyrrhonian answer to the question how it is possible to suspend all judgment yet
engage in active daily life. Sextus Empiricus distinguishes two kinds of assent: one
suspended across the board and one involved with daily living. The first—active
endorsement—is an act of will based on appreciation of reasons; the second—
passive acquiescence—is a causal effect of appearances. I suggest that Hume makes
the same distinction, only he extends the sort of assent involved in the Pyrrhonian’s
daily life to theoretical matters as well.1 He is a skeptic both in finding no reason to
grant the one sort of assent and in being subject to the other.
Understanding my suggested Hume interpretation requires understanding the
Pyrrhonian distinction in Michael Frede’s way, which I will explain. To bring out
the relevant features of Frede’s version, I defend it against the rival version of Myles
Burnyeat. Next, I suggest that Hume’s version of the distinction is that between
belief as an act “of the cogitative part of our natures” and as an act of the “sensitive”
part (Treatise [T] 1.4.1.8).2 Hume’s version can be fleshed out in the way I suggest
by regarding it, as Dugald Stewart did, to be derived from a lengthy discussion of
Ralph Cudworth’s. I then summarize how Treatise 1.4.1, “Scepticism with regard
to reason,” shows both that active endorsement is never merited and that passive
acquiescence cannot be avoided.
Julia Annas hints at an account like the one I give, but does not develop it and
ends up reading Hume as a dogmatist (Annas 2000: 276, 279). Instead, I recommend
reading Hume as finding no sufficiently good reason for active endorsement of
any view while yet yielding passive acquiescence in even some theoretical matters.
In theorizing thus concerning the science of man, Hume remains just as much a
A PYRRHONIAN INTERPRETATION OF HUME ON ASSENT 381
skeptic as Sextus does in daily life, such as when avoiding precipices or engaging in
religious devotions.
2 PYRRHONIAN KINDS OF ASSENT

The characterization of Pyrrhonism I will give depends heavily on the work of Michael
Frede (1997a; 1997b) on kinds of assent. Frede’s interpretation is a minority view;
the majority view follows the interpretation of Myles Burnyeat (1997a). It will help
in understanding Frede’s difficult view to contrast it with Burnyeat’s. In explaining
the key difference between them, as I understand it, I’ll focus on Sextus’s Outlines
of Pyrrhonism. Both Frede and Burnyeat take Sextus to have an answer to the
traditional challenge to the skeptic, one famously reiterated by Hume in his Enquiry
concerning Human Understanding (EHU 12.23).3 Here is the challenge: if you are
utterly passive, then you will quickly perish, but if you act to preserve your life,
then you must believe something, for all action requires some belief. For instance,
you would not eat in response to hunger, if you did not believe that eating would
relieve your hunger. This challenge is generally posed by those who assume that the
Pyrrhonian eschews any kind of assent. That assumption is attacked in different ways
by both Frede and Burnyeat. Their attacks rely on passages from Sextus such as these:
When we say that Sceptics do not hold beliefs, we do not take ‘belief’ in the
sense in which some say, quite generally, that belief is acquiescing in something;
for Sceptics assent to the feelings forced upon them by appearances. . . . Rather,
we say that they do not hold beliefs in the sense in which some say that belief is
assent to some unclear object of investigation in the sciences; for Pyrrhonists do
not assent to anything unclear. (PH I 13)4
But the main point is this: in uttering these phrases they say what is apparent to
themselves and report their own feelings without holding opinions, affirming
nothing about external objects. (PH I 15)
As we said before, we do not overturn anything which leads us, without our
willing it, to assent in accordance with a passive appearance. (PH I 19)
Remember too that we say we neither posit nor reject anything which is said
dogmatically about what is unclear; for we do yield to things which passively
move us and lead us necessarily to assent. (PH I 193)
For his interpretation, Frede distinguishes two kinds of assent. In one way, you
do not assent to the view that eating relieves hunger, but in another way you do.
Burnyeat does not distinguish two kinds of assent, but rather distinguishes two classes
of proposition liable to assent. A rough way to make the distinction is to say that
one class consists of propositions beginning with “It appears to me now that . . .,”5
where this phrase is used in the course of stating that something is appearing to you
and describing how it appears. The other class consists of all the other propositions.
Thus the Pyrrhonian who acts to relieve his hunger assents to the proposition “It
appears to me now that eating relieves hunger” understood in this way, and not to
the proposition “Eating relieves hunger.” In making their respective distinctions,
382 SKEPTICISM
both Frede and Burnyeat hold that there are assentings of a sort that is necessary to
survival, and assentings of a sort that is not necessary to survival. It is this latter sort
of assenting that is at issue in theoretical discussions about unclear matters and that
the Pyrrhonian avoids. Thus one can be a Pyrrhonian without perishing.
Four considerations incline me against Burnyeat’s view. Appreciating them helps
one appreciate the nuances of Frede’s interpretation.
It seems to me, first, that assent to “It appears to me now that eating relieves
hunger,” used to state that one is being appeared to and to describe how, is not
sufficient to motivate eating when one is hungry. It can only motivate eating if it
leads to some kind of assent to “Eating relieves hunger.” One is motivated, generally,
not by realizing how the world appears to one, but by taking some sort of stand on
how the world is. So Burnyeat’s proposal does not solve the problem with survival.
Perhaps Burnyeat need not say that assent to “It appears to me now that eating
relieves hunger” motivates eating. Rather, the appearance that prompts that report
also directly motivates eating.6 However, for the direct motivation to happen, the
content of the appearance must be “Eating relieves hunger.” Being moved by this
appearance will be just like the kind of assent necessary for survival that Frede
defends. So this defense of Burnyeat would be a concession to his opponent.
It seems to me, second, that the Burnyeat interpretation is not forced by the text.
True, Sextus says that the skeptics will only say “what is apparent to themselves, and
report their own feelings, affirming nothing about external objects.” But I think that
often the phrase “It appears that . . .” is not used to state and describe, but rather is
used to qualify one’s assent to the proposition in its scope. If I say, “It appears to me
now that eating relieves hunger,” I am assenting to “Eating relieves hunger” with
reservations. Perhaps I suspect that I have not established its truth conclusively. Or,
more to the point, perhaps I am just reporting the view which is striking me now—
that eating relieves hunger—without making any commitments to there being any
good reason to hold that view. So, when skeptics say what is apparent to themselves,
they need not be unreservedly endorsing that they are appeared to a certain way;
they may rather with maximum reservation be venting the view that things are
that way (PH I 13, 15).7 Thus there is a natural alternative to Burnyeat’s way of
interpreting Sextus’s frequent appeals to how things appear.
This alternative is not exactly what Frede terms the “epistemic sense” of “appears,”
if in that sense “it appears to me” is synonymous with “I think” or “I believe.” These
locutions can sometimes express assent with reservations, but do not always express
that. I am reluctant to use the word “belief” to characterize assent with reservations
as I elucidate it. In his earlier paper, Frede suggests that the skeptic has beliefs,
though it becomes apparent that Frede is using an elastic sense of “belief” that could
include even an impression we don’t object to (Frede 1997a: 10, 21–22). I think this
way of putting things went too far and was corrected in his later paper.
It seems to me, third, that Frede’s view makes better sense than Burnyeat’s of
Sextus’s claims that the skeptical phrases “cancel” themselves (PH I 14–15, 206).
Phrases such as “In no way more” or “No more this than that” mean that anything
and its negation are equally convincing, and so anything is unconvincing—it leaves
the judgment suspended. As Sextus says, “Everything is no more so than not so.”
This utterance applies to itself, and so the skeptics make clear that they suspend
judgment even concerning their own utterances. This is the sense in which the
phrases “cancel” themselves. Each phrase says in part that it itself is unconvincing.
However, if skeptical phrases implicitly have the content that Burnyeat would think
they do, then there would be no self-cancelling. The sentence “It appears to me now
that everything is equally convincing and unconvincing,” if true, does not entail that
everything is unconvincing. So it does not entail that it itself is unconvincing. So it
does not cancel itself.
In a footnote, Burnyeat notices a related problem with self-canceling. He says
that Sextus’s self-canceling propositions have to make a truth claim to be self-
canceling, whereas, according to Burnyeat elsewhere, reports of appearances make
no truth claim (1997a: 30–31).8 Therefore, one infers, Sextus’s self-canceling
phrases cannot be reports of appearances interpreted Burnyeat’s way. However,
Burnyeat sees it as a problem for Sextus, not for his own interpretation (Burnyeat
1997a: 54 n. 52).
Fourth, while both interpretations grant some kind of assent to what is apparent,
Frede’s makes better sense of the skeptic’s practice of continuing to investigate
whether what is apparent is real. If, say, “Honey sweetens” is what is apparent,
then it makes sense to investigate whether “Honey sweetens” is really true (PH
I 19). However, if “It appears to me now that honey sweetens” is what is apparent,
understood the way Burnyeat intends, then there is nothing to investigate beyond
the fact that you are appeared to in that way.
Burnyeat might reply that what one investigates is whether the state or impression
avowed when one says “It appears to me now that honey sweetens” conforms to
reality. However, this response acknowledges my point. For a state to conform to
reality is for the proposition it contains to be true. So the object of investigation is
whether “Honey sweetens” is true, not whether “It appears to me now that honey
sweetens” is true.
Thus, I incline to Frede’s interpretation of Sextus. In any event, Frede’s approach
to Sextus provides an extremely fruitful way of resolving a perennial problem in
Hume interpretation—how to reconcile Hume the skeptic with Hume the scientist
of man—in a way that does justice both to the power of Hume’s skeptical arguments
and to the scientific aspirations of Hume’s psychological theory. So I recommend
assuming that Frede has Sextus right, for the purpose of understanding Hume.
Sextus distinguishes two kinds of assent. Here I will give my own understanding
of the distinction rather than try accurately to represent Frede’s. Call the two kinds
of assent “active endorsement” and “passive acquiescence.” The first is to actively
endorse a view as true for a reason that one appreciates to be sufficiently good. The
second is to passively acquiesce in a view forced upon one by appearances. “View”
can mean the mental state one has in entertaining a proposition that is a candidate
for being assented to, but I mean the proposition itself.9
An example of active endorsement might be the firm conviction of a mathematician
in the result of a proof on which he has labored long and carefully and which has
received the approbation of his peers. Another might be the unshakeable conviction
of a religious zealot who considers his faith’s scriptures to be self-justifying. These
384 SKEPTICISM
examples are not conclusively examples. Active endorsement will, for Hume, turn
out to be a fiction that we and his opponents often believe in, but without good
reason. In any event, the examples are helpful at this stage in seeing what the
contrast with passive acquiescence is supposed to be. Because of the self-conscious
attempt to reason as one ought to, active endorsement has greater pretensions than
everyday belief. Its natural home, if it has any, would be enterprises in which one
investigates unclear matters beyond the mundane concerns of daily life. Crucially,
belief is supposed to be chosen—chosen because one understands that it is what one
ought to believe. In this sense it is called active.
Passive acquiescence, on the other hand, has no more pretension than everyday
belief, and sometimes less. In passively acquiescing one goes along with the view that
strikes one without making any decision one way or the other about its legitimacy.
For example, if a crowd stampedes by shouting, “The dam has broken!” you might
well run along without stopping to consider whether or not this could be true.
Circumstances just sweep you into acting as if the view that the dam has broken
were indeed true (cf. Thurber 1942: 22). Passive acquiescence is acting as if a belief
were true without deciding that it is. In this respect, it is like conjecture, pretense,
supposition, and taking as a working hypothesis. When a two-year old pretends
that a countertop is a sink, going through the motions of washing her hands and
declaring, “This is a real sink,” no one, least of all her, is fooled. She has not decided
that what she says is true, but acts as if it is true for all that. So do people taking the
other attitudes I’ve listed. A member of a committee, who uncharitably suspects a
fellow member of being disingenuous, might take as a working hypothesis that the
other member is acting in good faith. The result might predictably be greater peace
of mind for hypothesizer and a better outcome for the committee. Pursuing these
practical benefits is not the same as deciding in favor of the truth of the hypothesis.
Similarly, passive acquiescence in a view is a matter of acting as if the view is true.
However, there is a major difference from the attitudes I just mentioned. They are
chosen. Passive acquiescence, on the other hand, is not the result of willing. It is a
causal effect of how things strike one. In this sense it is passive.
The best examples of passive acquiescence would be the “beliefs” of animals—for
instance, those of the little pig admired by Pyrrho (DL IX 68, 481). Animal beliefs
are beliefs only in an analogous sense, not being governed in any way by normative
considerations. Instead, they are brought about by some sort of instinctual response
to circumstances.
It is with respect to active endorsement that a Pyrrhonian suspends judgment.
He finds that for any argument for any conclusion, he is able to find an apparently
equally good argument for a conflicting conclusion. This balancing of arguments
leaves him disinclined to either side, resulting in a suspense of judgment, that is, no
active endorsement either way.10 After all, active endorsement is meant to be the
kind of assent that results from reasoning as you ought to in the recognition that you
are reasoning as you ought to. One cannot arrive at this kind of assent when one has
an equally good reason to endorse something conflicting.
This method of balancing conflicting arguments is central to the Pyrrhonian
enterprise, even when Pyrrhonians take pains to lay out all-purpose arguments
against any conclusion. For instance, take the argument that (i) nothing should be
believed without a reason, and (ii) any conclusion is believed either for no reason, or
a circular reason (which is no reason), or a vicious infinite regress of reasons (which
is no reason) (cf. PH I 166, 168, 169). In giving such an “all-purpose” argument, the
Pyrrhonist would be taking for granted that there would be some consideration in
favor of the other side, which needed to be counterbalanced. The result is suspense
of active endorsement.11
Generally, it is investigations into unclear matters, such as those investigated in
philosophy, science, mathematics, and religion, in which one would seek to actively
endorse a view. In such investigations one is self-reflectively trying to find good
reasons for granting assent. In daily life, when one is for the most part just reacting
for practical purposes to the way things appear, one does not strive to meet the high
standards of active endorsement.
As the Pyrrhonian continues to investigate, his ability to balance every argument
with something conflicting has its effect on him. It begins to appear as if no
proposition deserves active assent. As Sextus puts it, “Thus what is said is this: ‘All of
the unclear matters investigated in dogmatic fashion which I have inspected appear to
me inapprehensible’” (PH I 200). He, of course, does not actively endorse this view,
but it governs his actions nonetheless. Now, it seems natural to say that he passively
acquiesces in this view. However, it seems to me that the view concerns a theoretical
matter about something that is not obvious, even if giving an appearance. Note that
not all appearances are sensory appearances (Frede 1973: 809). For example, Sextus
says that the Pyrrhonians accept the appearance “that piety is good and impiety bad”
(PH I 23). Thus, the door seems opened to Hume’s extension of passive acquiescence
into theoretical matters beyond active daily life (Popkin 1966: 95; Norton 1982: 268
n. 48).12 The skeptic’s view seems forced on him by appearances. Hume develops and
extends this notion of views forced by appearances.
3 HUMEAN KINDS OF ASSENT

Hume’s account echoes the Pyrrhonian’s distinction between two kinds of assent
and the appeal to appearances forcing views on one. Hume (T 1.4.1.8) explicitly
makes the distinction between two kinds of assent: belief as an act of the “cogitative
part of our natures” or as a “simple act of the thought” versus “belief as an act of
the sensitive . . . part of our natures.”
It is not immediately apparent that Hume’s distinction maps onto active
endorsement and passive acquiescence in the way I’m suggesting. However, evidence
is provided by historical context. Dugald Stewart (2005: 219) points the way,
noting that the distinction in Hume is familiar from the work of Ralph Cudworth.
Cudworth spends much time developing the distinction:
Wherefore it is evident from what we have declared, that there are two kinds
of Perceptive Cogitations in our Soul: the one Passive, when the Soul perceives
by suffering from its Body, and the Objects without; the other Active, when it
perceives by its exerting its own Native Vigor from within it self. (1731: 4.1.7)13
386 SKEPTICISM
Cudworth explicitly connects the first with appearance, when he observes that “the
Nature of Sense consists in nothing else but meer Seeming or Appearance” (1731:
3.4.2). In the previous section, he connects this sense of “appearance” with Sextus.
He then concludes,
And therefore we can have no Certainty by Sense alone either concerning the
Absolute Natures of Individual Corporeal things without us, nor indeed of their
Existence; but all the Assurance that we have thereof arises from Reason and
Intellect judging of the Phantasms or Appearances of Sense, and determining in
which of them there is an Absolute Reality, and which of them are but meerly
Relative or Phantastical. (1731: 3.4.8)
As opposed to the passive appearances of sense,
Knowledge is an Inward and Active Energy of the Mind it self, and the displaying
of its own Innate Vigour from within, whereby it doth Conquer, Master and
Command its Objects, and so begets a Clear, Serene, Victorious, and Satisfactory
Sense within it self. (1731: 4.1.1)
Cudworth finds the origin of the distinction in Plato’s Theatetus (1731: 3.3.6). He
may well have found some confirmation and development in Descartes’s Meditations:
When I say “Nature taught me to think this,” all I mean is that a spontaneous
impulse leads me to believe it, not that its truth has been revealed to me by some
natural light. There is a big difference here. (AT VII 38)14
Three paragraphs later Descartes distinguishes the basis of belief formed by “blind
impulse” from that formed by “reliable judgment.” He seems to be making a point
captured by Cudworth as follows:
For Knowledge is not a Knock or Thrust from without, but it consisteth in the
Awakening and Exciting of the Inward Active Powers of the Mind. (1731: 3.3.5)15
Hume (T 1.4.1), in contrast, when making this kind of distinction will be taking the
side of the “Knock or Thrust from without.” He contended that belief in persons
is like that in animals: an instinctual effect of observation and experience given
the principles of nature (T 1.3.16). The argument in section 1.4.1 supports his
earlier contention. His purpose in this section is to argue that if we reasoned as
we ought to, and if belief were a cogitative act, we would not assent to anything.
Since we do assent to some things, there must be some other explanation for it than
reasoning as we ought to. Belief is rather an act of sensation—an addition of “force
and vivacity”—naturally caused in us by customary connections observed in the past
(T 1.3.7). Here is Hume’s innovation. The sensitive part of our nature responds
to the impulse from without not only with images, but also with a feeling of their
forcefulness. In his argument, he intends “to make the reader sensible of the truth of
my hypothesis, that all our reasonings concerning causes and effects are deriv’d from
nothing but custom; and that belief is more properly an act of the sensitive, than of
the cogitative part of our natures” (T 1.4.1.8).
Think of it this way: judgment consists of a content plus assent (see Descartes,
Meditations, AT VII 56–58). The question is, is the assent an action one does or
something one suffers? Hume picks the latter.
Hume’s argument begins with the Diminution Argument that all belief depends on
probable reasoning,16 and after every case of probable reasoning “we are oblig’d by
our reason” to weigh in a small measure of uncertainty about whether the reasoning
in that case was properly done. Drawing a conclusion, by weighing considerations
for and against it, always introduces an additional consideration against it—one that
ought to be weighed in. That is because our faculty for weighing considerations has
introduced its own errors in the past, independent of the original considerations
themselves. The additional consideration needs to be weighed in. So every weighing
requires a reweighing. Weighing in an additional consideration against a conclusion
diminishes one’s confidence in the conclusion. So every reweighing diminishes
confidence in the conclusion. But every reweighing is itself a weighing, so requires a
further reweighing. So, confidence in the conclusion will be diminished to extinction
during the successive reweighings. For Hume, like any skeptic, diminishment of
confidence in a conclusion does not increase confidence in its negation. Nor, for
Hume, is there some kind of negative assent, that is, dissent, that can begin to
increase; there is no negative vivacity of ideas. So, the result will be suspense of
judgment. Hume concludes, therefore, that reasoning as we ought will leave us in
suspense of judgment: “all the rules of logic require a continual diminution, and at
last a total extinction of belief and evidence” (T 1.4.1.6).
Don Garrett argues “that Hume offers no epistemic evaluation in concluding
that ‘all the rules of logic’ require a total extinction of belief.” Hume, rather, is
merely explaining how our natural faculty of reason would operate if unhindered by
“the imagination’s difficulty in following the ‘subtle and refin’d reasoning’ of such
repeated reflections” (1997: 228). However, Hume is indeed offering an epistemic
evaluation: one that it strikes him we should and would endorse if we were able to
reason as we ought. The whole section, as well as the Diminution Argument itself,
begins by considering how we ought to reason in our pursuit of truth. To minimize
our chance of error, we ought to take our own fallibility into account when settling
on the degree of assurance we should have in our conclusion. “In every judgment,
which we can form concerning probability, as well as concerning knowledge, we
ought always to correct the first judgment, deriv’d from the nature of the object,
by another judgment, deriv’d from the nature of the understanding” (T 1.4.1.5).
Note the “ought” and the ensuing “correct.” It is an epistemic “ought” because the
goal is truth and avoidance of error (T 1.4.1.1). Again, Hume concludes that, if we
reason as we ought to in pursuit of truth, we will lose all assurance through our
repeated corrections.
Many scholars think Hume’s Diminution Argument fails, but that is beside
the point here.17 It strikes him forcefully that it succeeds. It is an “all-purpose”
argument that can counterbalance the considerations in favor of any other view.
Consequently, like the Pyrrhonians, Hume is unable to find anything that one ought
to choose actively to endorse.
388 SKEPTICISM
Note that the conclusion itself is one of the propositions it applies to. The
conclusion itself is subject to diminution. So even in his own conclusion Hume has
found nothing one ought to assent to.
Suppose that belief were active endorsement. We would believe just as we
ought. So without anything we ought to assent to, there would be no belief. “If
belief, therefore, were a simple act of the thought, without any peculiar manner of
conception, or the addition of a force and vivacity, it must infallibly destroy itself,
and in every case terminate in a total suspence of judgment” (T 1.4.1.8).
However, we cannot and do not suspend all belief as we ought to. We have lots
of beliefs. Among other things, Hume assents to the conclusion of the Diminution
Argument, at least during T 1.4.1 and at 1.4.7.7. Still, he says he is not able to believe
it “sincerely and constantly,” which may mean that sometimes he does not believe it,
but more likely means that he is never able to act on it (T 1.4.1.7).18 In any event, we
have beliefs, therefore belief is not an act of the cogitative part of our soul but is rather
an act of the sensitive part. Belief is a matter of being caused to have lively ideas. It is
a brute fact about us that, however much we might see the justice of the Diminution
Argument, we cannot consistently follow it out for any given proposition. We quickly
become overtaxed and almost as quickly return to belief in the convictions we were
questioning. That is because “Nature, by an absolute and uncontroulable necessity has
determin’d us to judge as well as to breathe and feel” (T 1.4.1.7). We cannot suspend
all assent whatsoever any more than a sighted person can prevent himself from seeing
his environs with eyes directed to them and open in the sunshine. In making the origin
of all belief causal, Hume is echoing the Pyrrhonian view of passive assent as assent to
views forced upon us by appearances. The “forcing” is a matter of finding ourselves to
have the view in a way independent of the exercising of our will.19
One might wonder why Hume would propose the rules for causal reasoning
in 1.3.15 if he thinks we cannot reason as we ought. Why would he propose
standards for reasoning if we cannot follow such standards? The answer is that
his proposed standards for judging of causes and effects are ultimately practical
standards, extensions of Sextus’s standards of action (PH I 21–24). They are how
we should reason given Hume’s Title Principle, as Garrett has termed it: “Where
reason is lively, and mixes itself with some propensity, it ought to be assented to.”
Whence the “ought”? In order not to “be a loser in point of pleasure; and this is
the origin of my philosophy” (T 1.4.7.11–12). The rules for causal reasoning are
meant to guide our natural propensity to use induction. But they do not address the
skeptical problem that we can find no sufficiently good reason to regard the results
of induction as true (T 1.3.6.11). We are able to follow standards gleaned from
natural ways of reasoning in service of practical goals. These practical standards
for regarding some things as true or false are the “measures of truth and falsehood”
that Hume implies that we have (T 1.4.1.7). What we are unable to follow with any
stable result are the epistemic standards we aspire to. Instead of losing all assurance,
as we epistemically should, we retain some. As he puts it later, “After the most
accurate and exact of my reasonings, I can give no reason why I shou’d assent to it;
and feel nothing but a strong propensity to consider objects strongly in that view,
under which they appear to me” (T 1.4.7.3).
Thus in 1.4.1 Hume finds himself in the position of unavoidably believing

things that it strikes him he wouldn’t believe if he reasoned as he ought, and
so epistemically ought not believe. Readers of Hume who resist the Pyrrhonian
interpretation crucially assume that no one can consistently be in such a position.20
It seems to such readers that assenting to a view is implicitly committing oneself
to its epistemic merit.21 However, the distinction between the two kinds of assent
is precisely intended to explain how one consistently can believe things that one
realizes he epistemically ought not assent to, whether actively or passively. I’ll
discuss the two possibilities in turn.
First, one can passively acquiesce in things for which one finds no reason
sufficiently good for active endorsement. Passively acquiescing in views forced upon
one by appearances just is a way to assent without regard to whether one epistemically
ought to assent. So one violates no implicit commitment to the epistemic merit of
the views in recognizing that one can find none.
Now, if one is to be such a “fool,” Hume observes, as are “all those who reason
or believe,” he should at least be so in a way that is “natural and agreeable” as
captured by the Title Principle (T 1.4.7.10–11).22 By acquiescing in the Title
Principle, Hume decides for practical reasons not to try to conform to onerous
epistemic requirements. He is better off ignoring such requirements for the sake
of “service to mankind” and his “own private interest” (T 1.4.7.10). So, secondly,
he is able to passively acquiesce in views even while also passively acquiescing
in the view that he epistemically ought to lose all assurance whatsoever. The
Diminution Argument with its general conclusion strikes him as successful, and yet
he finds he cannot use that result to extinguish any given actual belief. He finds no
practical reason why he should even make the effort. And not doing so violates no
commitment implicit in his assent, as it would if he were actively endorsing. Again,
active endorsement is committing oneself to following the strictures of what one
epistemically ought to hold. It would be inconsistent to actively endorse a view
when one thought one epistemically shouldn’t. But in passively acquiescing in views
forced upon one, one does not make such commitments. So in passively acquiescing
in views despite passively acquiescing in the view that he shouldn’t, Hume violates
no implicit commitment.
That beliefs are caused by appearances, not chosen for their merits, remains true
even if we make theoretical investigations. We can voluntarily make observations,
do experiments, weigh opposing considerations, etc.23 But Hume’s point is that
the process will cause us to come to a conclusion. It will not be the occasion for
our freely choosing which conclusion to come to. And we cannot just will away
the resulting belief. Granted, our will could set in motion a process of further
weighing of other considerations that could be expected to reduce our assent, or
even eliminate it. But the assent itself happens to us, rather than is willed by us. In
that way it is passive.
To appreciate this point, the content of the proposition assented to must be
distinguished from the kind of assent. Theoretical content does not entail active
endorsement. Nor does normative content. Hume can even assent to propositions
of the form, “It is evident that . . .,” or “It is true that . . .,” or “One ought to believe
390 SKEPTICISM
that . . .” without actively endorsing them (T 1.4.7.15). His assent is wrung from
him by appearances, by how matters strike him, and so is passive acquiescence.
This passive view of assent is nicely summarized, in the Treatise:
We may, therefore, conclude, that belief consists merely in a certain feeling or
sentiment; in something, that depends not on the will, but must arise from certain
determinate causes and principles, of which we are not masters. (T Appendix 2)
And also in the first Enquiry:
It follows, therefore, that the difference between fiction and belief lies in some
sentiment or feeling, which is annexed to the latter, not to the former, and which
depends not on the will, nor can be commanded at pleasure. It must be excited
by nature, like all other sentiments; and must arise from the particular situation,
in which the mind is placed at any particular juncture. (EHU 5.11)
Note that passive acquiescence ought not be thought of as weak assent. The assent
may be normatively weak, but may well be causally “strong and steady” (T 1.3.7.5
n. 20). Hume allows, as Sextus apparently did not, that some views forced upon
one by appearances may be causally hard to dislodge (PH I 230). Force and vivacity
come in degrees and sometimes in a very high degree (T 1.3.13.19).
I’ve said that Hume is like the Pyrrhonians in finding insufficiently good reason
for active endorsement and in yielding to passive acquiescence. But where exactly
is the skeptic’s suspense of judgment? It might seem trivial to claim that Hume
suspends active endorsement, when he thinks there is no such thing. Why does
withholding an assent impossible to grant make him a skeptic? The answer is that,
when he reasons as carefully as he can with truth as the goal, he finds no sufficiently
good reason why he ought to believe anything. This process leads to temporary
suspension of passive acquiescence, until his natural propensities to believe reassert
themselves. But it also leads to stable passive acquiescence in the view that those
skeptical arguments are unanswerable, even if they produce no lasting suspension
of judgment as passive acquiescence. So were active endorsement possible, Hume
would find it suspended. And for those who believe in active endorsement, they
either ought to suspend it or else ought to admit with Hume that their assent is
really nothing more than passive acquiescence, after all.
In summary, reasoning as we ought to would result in total suspense of judgment.
Were there such a thing as active endorsement, we would totally suspend it, since
we can find nothing worthy of it. Nonetheless, passive acquiescence is compelled
in us by appearances, even appearances that result from theoretical investigation.
In this way Hume can combine his radical skepticism with his constructive science,
mathematics, and philosophy. He does so in a way very like the way the Pyrrhonian
combined radical skepticism with the assent required to live a normal life.
It is true that Hume resists calling his skepticism Pyrrhonian, and calls it mitigated.
But Hume erroneously took the Pyrrhonians to recommend no assent whatsoever,
which is why he thought that a true Pyrrhonian must “remain in total lethargy”
and soon “perish.” As we have seen, the Pyrrhonians had a ready answer to this
charge in the distinction between two kinds of assent. They found no reason to
grant active endorsement, but yielded to passive acquiescence in order to survive.

Likewise Hume. There is no mitigation of his failure to find a sufficiently good
reason to actively endorse anything; there is much natural mitigation of any great
loss of passive acquiescence that might result from his skeptical arguments (EHU
12.23–25). Hume’s own one-sided view of the Pyrrhonians should not prevent our
seeing him as their natural successor concerning assent.24
NOTES
1. My reading follows Popkin (1966) (on Popkin on the crucial influence of Pyrrhonism
on modern philosophy in general, see Maia Neto’s chapter in this book). See also
Baxter (2006; 2008: ch. 1; 2009a; 2009b). Note that Hume does not follow the
Pyrrhonians in their centrally important concern with tranquility. For a detailed
discussion of Hume’s familiarity with Sextus’s works, see Fosl (1998).
2. I cite Hume’s Treatise by book, part, section, and paragraph. I use the text in
Hume (2007).
3. I cite Hume’s Enquiry by section and paragraph. I use the text in Hume (2000).
4. I follow the translation by Annas and Barnes in Sextus Empiricus (2000).
5. Or, perhaps better, propositions perspicuously paraphrased by propositions
beginning with “It appears to me now that . . . .”
6. I’m grateful to Lionel Shapiro for this suggestion.
7. Cf. Sellars (1956: sec. 16). Wittgenstein (1958: 92) says: “Don’t regard a hesitant
assertion as an assertion of hesitancy.”
8. Barnes (1997: 65 n. 21) disagrees. Burnyeat uses the term “self-refuting” instead
of “self-cancelling.”
9. “Belief” encompasses a similar ambiguity between mental state and its
propositional content.
10. I disagree with Burnyeat’s claim that this process could happen only if the skeptic
comes to the belief that “contrary claims are equal” (Burnyeat 1997a: 56) The
skeptic need at most acquiesce in this appearance, and perhaps, as Barnes urges, the
suspense of judgment “simply happens to us” (Barnes 1997: 58–59).
11. For a similar approach to “Agrippa’s Trilemma,” see Williams (1988: 573–583),
though Williams sees the results of the modes of Agrippa as denials of specifically
epistemological theses that would be held by a dogmatist, whereas I see them as in
conflict with any conclusion assented to for a reason.
12. “In vain would the skeptic make a distinction between science and common life, or
between one science and another. The arguments employed in all, if just, are of a
similar nature and contain the same force and evidence” (Hume 1980: 9–10).
13. I cite Cudworth (1731) by book, chapter, and section.
14. I follow the translation in Descartes (1984).
15. “[T]hat all our Cogitations, are Obtruded, and Imposed upon us from without;
and that there is no Transition in our Thoughts at any time, but such as had been
before in Sense; (which the Democritick Atheist averrs) this is a Thing, which we
absolutely deny. For, had we no Mastery at all over our Thoughts, but they were all
392 SKEPTICISM
like Tennis Balls, Bandied, and Struck upon us, as it were by Rackets from without;
then could we not steadily and constantly carry on any Designs and Purposes of
Life” (Cudworth 1964: 845).
16. Even beliefs not initially arrived at by reason are subject to its scrutiny.
17. For some crucial discussions of the 1.4.1 argument in addition to Garrett’s, see
Morris (1989: 39–60), Owen (forthcoming), and Fogelin (1985: ch. 2).
18. Cf. the belief that one ought not believe in body, where “carelessness and in-
attention” prevent the opinion from being held constantly (T 1.4.2.56–57).
19. Thus I agree with Morris that Hume attacks “the intellectualist model of the rationally
reflective epistemic agent” in order to replace it with a naturalistic model (Morris
1989: 56). See also Baier (1991: 59–61) and Owen (1999: 120–121). However, all
three focus on a conception of reason as a discoverer of necessary connections. I
take Hume to rather be arguing that an active endorser—someone who arrives at
assent by choosing to reason as he ought—will not arrive at assent. Only assent that
is a feeling caused by appearances can survive reasoning. Baier (1991: 96) contends
in contrast that reason on Hume’s own account has normative authority because
it can bear its own survey. Thus Hume discovers “which mental causes are good
reasons.” I agree instead with Owen (1994: 198) that 1.4.1 precisely shows that
even Humean reason cannot bear its own survey.
20. For instance, Garrett (2007: 7, 8, 11) assumes (i) that in providing psychological
explanations, Hume has no reason to dispute them, (ii) that coming to believe
according to the Title Principle could satisfy one’s curiosity only if it epistemically
ought to, and (iii) that concluding that one’s own beliefs were not probable would
undermine one’s ability to make judgments about usefulness. All these make the
deep assumption that one can’t believe things without believing one epistemically
ought to believe them.
21. My thinking here was sparked partly by Patrick Greenough’s lecture entitled
“Pragmatics in Thought: The Case of Moore’s Paradox,” University of Connecticut,
October 2010.
22. His belief in body is another example of Hume’s believing something that he realizes
he epistemically ought not believe. “Carelessness and in-attention” are our only way
to retain any sort of assent in the vulgar view involving “gross illusion” and the
philosophical view “loaded” with an “absurdity” (T 1.4.2.56–57). Even were the
Title Principle an epistemic one, it could not justify the conclusion that forgetting
or ignoring a problem is a solution to it.
23. In this respect Hume is allowing for a Pyrrhonian way to accommodate the
procedures of the New Academy (PH I 227–229).
24. I’m grateful to Richard Bett for penetrating comments in a symposium at the
American Philosophical Association Central Division Meetings, 2012, as well as to
Diego Machuca and Baron Reed for their helpful suggestions.
REFERENCES
Annas, Julia. 2000. “Hume and Ancient Scepticism,” Acta Philosophica Fennica 66: 271–
285.
Baier, Annette. 1991. A Progress of Sentiments. Cambridge, MA: Harvard University Press.
Barnes, Jonathan. 1997. “The Beliefs of a Pyrrhonist.” In M. Burnyeat and M. Frede (eds.),
The Original Sceptics: A Controversy, 58–91.
Baxter, Donald L. M. 2006. “Identity, Continued Existence, and the External World.” In
S. Traiger (ed.), The Blackwell Guide to Hume’s Treatise, 114–132. Oxford: Blackwell
Publishing.
Baxter, Donald L. M. 2008. Hume’s Difficulty: Time and Identity in the Treatise. New York:
Routledge.
Baxter, Donald L. M. 2009a. “Hume’s Theory of Space and Time in its Skeptical Context.”
In D. F. Norton and J. Taylor (eds.), The Cambridge Companion to Hume, 2nd ed.,
Baxter, Donald L. M. 2009b. “Précis of Hume’s Difficulty,” Philosophical Studies 146:
409–411.
Burnyeat, Myles. 1997a. “Can the Sceptic Live his Scepticism?” In M. Burnyeat and
M. Frede (eds.), The Original Sceptics: A Controversy, 25–57. Indianapolis: Hackett.
Burnyeat, Myles. 1997b. “The Sceptic in His Place and Time.” In M. Burnyeat and M. Frede
(eds.), The Original Sceptics: A Controversy, 92–126. Indianapolis: Hackett.
Cudworth, Ralph. 1731. A Treatise Concerning Eternal and Immutable Morality. London:
James and John Knapton.
Cudworth, Ralph. 1964. The True Intellectual System of the Universe. Stuttgart-Bad
Cannstatt: Friedrich Frommann Verlag.
Descartes, René. 1984. “Meditations on First Philosophy.” In J. Cottingham, R. Stoothoff,
and D. Murdoch (eds.), The Philosophical Writings of Descartes, Vol. 2, 1–90.
Fogelin, Robert. 1985. Hume’s Skepticism in the Treatise of Human Nature. London:
Routledge & Kegan Paul.
Fosl, Peter. 1998. “The Bibliographic Bases of Hume’s Understanding of Sextus Empiricus
and Pyrrhonism,” Journal of the History of Philosophy 36: 261–278.
Frede, Michael. 1973. Review of C. L. Stough, Greek Skepticism: A Study in Epistemology,
The Journal of Philosophy 70: 805–810.
Frede, Michael. 1997a. “The Sceptic’s Beliefs.” In M. Burnyeat and M. Frede (eds.),
The Original Sceptics: A Controversy, 1–24. Indianapolis: Hackett.
Frede, Michael. 1997b. “The Sceptic’s Two Kinds of Assent and the Question of the
Possibility of Knowledge.” In M. Burnyeat and M. Frede (eds.), The Original Sceptics: A
Controversy , 127–151. Indianapolis: Hackett.
Garrett, Don. 1997. Cognition and Commitment in Hume’s Philosophy. New York: Oxford
University Press.
Garrett, Don. 2007. “Reasons to Act and Believe: Naturalism and Rational Justification in
Hume’s Philosophical Project,” Philosophical Studies 132: 1–16.
Hume, David. 1980. Dialogues Concerning Natural Religion, edited by R. H. Popkin.
Indianapolis: Hackett Publishing.
Hume, David. 2000. An Enquiry Concerning Human Understanding, edited by
T. L. Beauchamp Oxford: Clarendon Press.
Hume, David. 2007. A Treatise of Human Nature, edited by D. F. Norton and M. J. Norton.
Morris, William Edward. 1989. “Hume’s Scepticism about Reason,” Hume Studies 15:
39–60.
394 SKEPTICISM
Norton, David Fate. 1982. David Hume: Common-Sense Moralist, Sceptical Metaphysician.
Princeton: Princeton University Press.
Owen, David. 1994. “Reason, Reflection, and Reductios,” Hume Studies 20: 195–210.
Owen, David. 1999. Hume’s Reason. Oxford: Oxford University Press.
Owen, David. 2015. “Scepticism with Regard to Reason.” In D. Ainslie (ed.), The Cambridge
Companion to Hume’s Treatise, 101–134. Cambridge: Cambridge University Press.
Popkin, Richard. 1966. “David Hume: His Pyrrhonism and His Critique of Pyrrhonism.”
In V. C. Chappell (ed.), Hume, 53–98. Garden City, NY: Anchor Books.
Sellars, Wilfrid. 1956. “Empiricism and the Philosophy of Mind.” In H. Feigl and M. Scriven
(eds.), Minnesota Studies in the Philosophy of Science, vol. 1, 253–329. Minneapolis:
University of Minnesota Press.
Sextus Empiricus. 2000. Outlines of Scepticism, translated by J. Annas and J. Barnes.
Stewart, Dugald. 2005. “Dissertation on the Progress of Philosophy.” In J. Fieser (ed.),
Early Responses to Hume’s Metaphysical and Epistemological Writings, II, 2nd edn.
Bristol: Thoemmes Press.
Thurber, James. 1942. “The Day the Dam Broke.” In his My Life and Hard Times.
New York: Harper Brothers.
Williams, Michael. 1988. “Scepticism without Theory,” The Review of Metaphysics 41:
547–588.
Wittgenstein, Ludwig. 1958. Philosophical Investigations, 3d ed., translated by
G. E. M. Anscombe. New York: MacMillan.
CHAPTER TWENTY-EIGHT
Thomas Reid’s
Engagement with Skeptics
and Skepticism
RENÉ VAN WOUDENBERG
1 INTRODUCTION
Engagement with skeptical problems, as a means of resisting skepticism, may take
various forms. First, one may try to convince skeptics that their skeptical conclusion
is false. Second, one may try to identify certain assumptions underlying the skeptical
argument that may initially seem unobjectionable but that cannot withstand close
scrutiny.1 Third, one may try to take up the skeptical challenge and offer an
argument for the denial of the skeptical conclusion, or for the weaker thesis that
it is not irresponsible or irrational to accept the denial of the skeptical conclusion.
I call these engagements convincing the skeptic, exposing the skeptical argument, and
countervailing the skeptical conclusion, respectively.
In this chapter, I will review Thomas Reid’s engagement with skeptics and
skepticism under these three headings. Before starting out, I first provide some basic
information about Reid, and next reflect on the interrelations between the three
forms of engagement.
After having served as a minister in the Church of Scotland, Thomas Reid (1710–
1796) became a professor of “moral philosophy” at the universities of Aberdeen
and Glasgow. During his lifetime he published three books, An Inquiry into the
Human Mind on the Principles of Common Sense (1764), Essays on the Intellectual
Powers of Man (1785), and Essays on the Active Powers of Man (1788). Reid became
the father of what has come to be called “the Scottish school of common sense
philosophy.” His work was widely influential across the continent, but especially in
the United States, where his work inspired Thomas Jefferson and other founding
fathers of the New World (Redekop 2004). Over the last forty years Reid has again
being recognized as a major philosopher. His work has had a lasting influence on
philosophers such as Roderick Chisholm, Keith Lehrer, Alvin Plantinga, William
Alston, and Nicholas Wolterstorff, along with many others.
396 SKEPTICISM
As to the interrelations between the three forms of engagement with skeptics and
skepticism, one could think that one way to convince the skeptic is to expose the
skeptic’s argument by laying bare at least one problematic assumption underlying
the argument. But this need not be sufficient. After all, the skeptic may simply retort
that though his argument is problematic, this does not entail that the skeptical
conclusion is false. And here the skeptic is, of course, right. In order to convince the
skeptic, it would be required to expose all arguments for the skeptical conclusion.
And even this may not be sufficient for conviction. For, unfortunately, it isn’t a
characteristic of good arguments that they convince everybody. Here we enter the
mires of human psychology. As to the relation between convincing the skeptic and
countervailing the skeptical conclusion, we should note that a skeptic may remain
totally unconvinced by countervailing arguments. This is again a point about human
psychology. So, there is no direct relation between convincing the skeptic, on the one
hand, and exposing skeptical arguments and countervailing the skeptical conclusion
on the other. Nor is there a direct relation between exposing skeptical arguments and
countervailing the skeptical conclusion. For countervailing the skeptical argument,
that is, advancing an argument for the denial of the skeptical conclusion, needn’t
bring to light objectionable assumptions that underlie the skeptical argument. So,
the three sorts of engagement are not related in any simple way. But both exposing
the skeptical argument and countervailing the skeptical conclusion at least stay
clear from the muddle of human psychology. These latter engagements would be
meaningful even if there were no skeptics. The first engagement, by contrast, would
not, for in order to convince skeptics, there must be skeptics.
2 CONVINCING THE SKEPTIC

Reid relates a number of stories about supposedly real-life skeptics who profess not
to believe their senses, but who nevertheless usually don’t bang their heads against
posts, don’t step into ditches and fires, and generally stay out of harm’s way. About
Pyrrho, the great skeptic, Reid writes, for example, that he “is said once to have
been in such a passion with his cook, who probably had not roasted his dinner to his
mind, that with the spit in his hand, and the meat upon it, he pursued him even into
the market place” (1764: 21). Stories like these made Reid suspicious about what
these skeptics tell about themselves. His comment is that such skeptics either “act the
hypocrite” or else “impose upon themselves” (1764: 170). Some find this comment
“shallow” and “superficial” (De Bary 2002: 7, 12). But is it? For surely someone may
profess that he doesn’t believe his eyes, while in fact he does believe them. A bystander
who observes that someone who professes not to believe his senses carefully stays out
of harm’s way, has an excellent reason for thinking that the skeptic doesn’t believe
what he professes he believes. The skeptic’s behavior betrays him; he does believe his
senses. One plausible principle about belief ascriptions to others involves observed
behavior. Someone who says he likes to have a drink and walks to the refrigerator
should, in ordinary situations, surely be ascribed the belief that the refrigerator
contains drinkable items. Reid’s comment, then, needn’t be deemed shallow at all; it
can be taken to point to a rather sensible principle of belief ascription.2
THOMAS REID’S ENGAGEMENT WITH SKEPTICS AND SKEPTICISM 397
If we suppose that Reid’s comment is true, and so suppose that the skeptics
who figure in the stories are misleading themselves as to what they believe, doesn’t
this imply that they are not really skeptics, as they do, in fact, believe their senses?
So, doesn’t his comment, if true, mean that there are no skeptics? No, but it does
mean that in order to define being a skeptic, we need to move one level upwards.
A skeptic should not be defined as someone who doesn’t believe his senses—for,
if Reid’s remark is right, he does believe them. Rather, a skeptic should be defined
as someone who professes not to believe his senses. Accordingly, the program of
convincing the skeptic really is the program of convincing someone who professes
not to believe his senses of the fact that he does believe them after all.
Reid clearly recognized that there is a class of skeptics whom he cannot possibly
convince by arguments of the wrongness of their conclusion, namely, the class of
those who believe that our powers of reasoning and judgment are fallacious:
If a skeptic should build his skepticism upon this foundation that all our reasoning
and judging powers are fallacious in nature, or should resolve at least to withhold
assent until it be proved that they are not; it would be impossible by argument to
beat him out of this strong hold, and he must be left to enjoy his own skepticism.
(1785: 480)
Does this indicate that according to Reid total skepticism is irrefutable, as De Bary
(2002: 16) claims? I think it is better to say that what this quote indicates is that in
Reid’s view the total skeptic cannot, by arguments, be convinced of the falsehood of
his skeptical thesis. Reid’s point here is about persons, skeptics, not about arguments
or theses, skepticism.
But what about skeptics who do argue for skeptical conclusions? Could they,
perhaps, be convinced by an argument that Reid might bring against them? Some
reasoning skeptics, as we might call them, urge us not to accept any proposition
unless it is “founded upon reason.” This skeptic finds fault with those ordinary
people who “believe undoubtedly that there is a sun, moon, and stars; an earth
which we inhabit; country, friends, and relations, which we enjoy; lands, houses,
and moveables, which we possess,” even though they have no decent argument or
proof for any of this. The skeptic, however, “pitying the credulity of the vulgar,”
resolves “to have no faith but what is founded upon reason” (1764: 18). He thinks it
is wrong to believe what is not founded upon reason. In effect, the reasoning skeptic
Reid has in mind affirms the following Principle of Belief:
PoB
It is wrong to believe proposition p unless p is founded upon reason.
And, Reid explains, a proposition is “founded upon reason” if it can be proved by
reasoning, that is, if it can be made the conclusion of a sound argument.
One point that Reid makes about PoB is that there are propositions the skeptic
believes in a way that violates PoB. Hume, for example, says Reid,
appears to me to be but a half-skeptic. He hath not followed his principles so
far as they lead him; but, after having, with unparalleled intrepidity and success,
398 SKEPTICISM
combated vulgar prejudices, when he had but one blow to strike, his courage
fails him, he fairly lays down his arms, and yields himself a captive to the most
common of all vulgar prejudices—I mean the belief of the existence of his own
impressions and ideas. (1764: 71)
Reid’s point, then, is that Hume believed that there exist such things as impressions
and ideas but that this belief is not founded upon reason.
Reid makes a similar point about Descartes, who also endorsed PoB. Descartes
tried to prove his own existence in the Cogito argument. That argument concludes
to the existence of a mind, or subject of thought, from the existence of thought.
But, Reid asks, “why did he [Descartes] not prove the existence of thought?
Consciousness, it may be said, vouches that. But who is voucher for consciousness?
Can any man prove that his consciousness may not deceive him? No man can”
(1764: 17). In effect, Descartes believes in the existence of thought even though it
is not founded upon reason.
So, Reid argues, the skeptic accepts at least some propositions (propositions about
the existence of impressions and ideas, or thought) in a way that violates PoB. Still,
he continues, the skeptic’s adherence to PoB really is an expression of his allegiance
to the faculty of reason, and his fundamental distrust of other faculties, perception
included. But why, Reid asks, should reason and perception be treated differently?
Reason, says the skeptic, is the only judge of truth, and you ought to throw off
every opinion and every belief that is not grounded on reason. Why, sir, should
I believe the faculty of reason more than that of perception?—they came both
out of the same shop, and were made by the same artist; and if he puts one piece
of false ware into my hands, what should hinder him from putting another?
(1764: 169)
Reid, then, accuses the skeptic of unwarranted partiality.
Will a skeptic be convinced by this line of thought? Not necessarily. For as noted,
it isn’t a hallmark of sound arguments that they convince. Still, a reasoning skeptic,
bothered by inconsistency, could be convinced by Reid’s charge that he actually
violates a principle that he officially adopts. But even if the skeptic isn’t convinced
by the charge, it does make explicit that the skeptic’s argument assumes a certain
principle, PoB, that can be contested. In that sense, what we have heard Reid say so
far can be viewed as an effort to expose the skeptical argument—which is the topic
of the next section.
3 EXPOSING THE SKEPTICAL ARGUMENT

Important strands of Reid’s thought rather naturally fit the heading exposing the
skeptical argument. In the Letter of Dedication of the Inquiry, Reid says:
For my own satisfaction, I entered into a serious examination of the principles
upon which this skeptical system [Hume’s] is built; and was not a little surprised
to find, that it leans with its whole weight upon a hypothesis . . . of which I could
find no solid proof. The hypothesis I mean is, That nothing is perceived but what
is in the mind which perceives it: That we do not really perceive things that
are external, but only certain images and pictures of them imprinted upon the
mind, which are called impressions and ideas. If this be true; supposing certain
impressions and ideas to exist in my mind, I cannot, from their existence, infer
the existence of anything else. (1764: 4)
This passage indicates that Reid’s diagnosis of what is problematic in the skeptical
argument for the conclusion that we can’t have knowledge of the external world
is that it assumes the correctness of what he elsewhere calls the “theory of ideas.”
And it furthermore suggests that Reid held that in order to dispose of the skeptical
argument it is sufficient to explode that theory.
Reid identifies a further assumption that underlies the skeptical argument—one
that we have touched on at the end of the previous section. The assumption is
that our thoughts, our sensations, and everything of which we are conscious,
hath a real existence, is admitted in this system as a first principle; but everything
else must be made evident by the light of reason. Reason must rear the whole
fabric of knowledge upon this single principle of consciousness. (1764: 210)
This assumption has been called classical, or Cartesian, foundationalism.3 Its basic
idea is that the only beliefs with enough certainty to qualify as foundational are
incorrigible or indubitable “ideas” about our own current mental states. All other
beliefs (about the external world, for example, or the past) can be justified only if
they are demonstrable from these foundational beliefs. This assumption obviously
has a skeptical implication, for, as the history of philosophy between Descartes and
Hume has made clear, it is impossible to prove the existence of mind-external items
solely on the basis of premises about one’s current mental states. And hence, on
the assumption of classical foundationalism, none of our external world beliefs are
justified or warranted. Here, too, Reid is suggesting that in order to expose the
skeptical argument it is sufficient to explode Cartesian foundationalism.
Let us take a look at what Reid thinks is wrong with both of these assumptions
and next turn to the question whether rejection of them is sufficient to dispose of all
skeptical arguments. Let us start with the foundationalist assumption.
As intimated in the previous section, Reid’s main objection to Cartesian
foundationalism is that it is arbitrary. Why should consciousness, the faculty that
generates beliefs about one’s own current mental states, and reason be trusted, while
all other faculties (perception and memory, for example) should be trusted only after
their trustworthiness has been vouched for by consciousness and reason? There is
no relevant difference among our faculties that would warrant the preference of
consciousness and reason over the others. Says Reid:
The faculties of consciousness, of memory, of external sense, and of reason, are
all equally gifts of nature. No good reason can be assigned for receiving the
testimony of one of them, which is not of equal force with regard to the others.
(1785: 463)4
400 SKEPTICISM
Reid criticizes the assumption of the theory of ideas—that ideas are the immediate
objects of perception—in a number of ways. I single out one line of criticism: that
there are no good arguments for the existence of ideas. Hume’s famous argument for
the existence of ideas from the relativity of perception, Reid argues, is a failure. The
argument is this: “The table, which we see, seems to diminish, as we remove farther
from it; but the real table, which exists independently of us, suffers no alteration;
it was, therefore, nothing but its image, which was present to the mind” (Hume
1902: 152). Reid criticizes this argument by drawing on the distinction between
real and apparent magnitude. These two have very different natures even though
both are named “magnitude.” The real magnitude of a table has three dimensions, it
is measured on a linear scale, and it is an original object of touch, not of sight. The
table’s apparent magnitude, by contrast, has only two dimensions, it is measured by
the angle which it subtends at the eye, and is an original object of sight, not of touch.
Given this distinction, Reid argues that Hume’s argument equivocates between
these two magnitudes. Hume starts out by saying “that the table seems to diminish
as we remove farther from it.” This, says Reid, is a claim about the table’s apparent
magnitude. But when Hume continues that “the real table suffers no alteration,” he
is making a claim about the table’s real magnitude. But then Hume’s conclusion that
it is not the real table that we see doesn’t follow from the premises because “the
syllogism has what the logicians call two middle terms” (1785: 182).
In the quotation taken from the Letter of Dedication to the Inquiry given above,
Reid seems to say that, since the theory of ideas is both necessary and sufficient to
generate skeptical conclusions, any successful reply to skepticism must reject the
theory of ideas. But, as John Greco has argued, although the theory of ideas may
be sufficient to generate skeptical conclusions, it is not necessary in order to arrive
at such conclusions (2004: 142–143). And this means that it is not sufficient for a
successful reply to the skeptical argument that one rejects the theory of ideas, for
there are skeptical arguments that do not invoke the theory of ideas at all.5
4 COUNTERVAILING THE
SKEPTICAL CONCLUSION 6
To show that some problematic assumptions underlie the (or: a) skeptical argument
may help to expose the argument. But it does not show that the skeptical conclusion
is false. It only shows that certain arguments for that conclusion can be dispelled.
Reid’s most important contribution to the project of countervailing the skeptical
conclusion is his theory of First Principles. This section clarifies what Reid thinks
is the nature of these principles and his reason for supposing such principles exist.
This will allow us to see how Reid countervailed the skeptical conclusion.
Reid thinks about first principles as true propositions that are “basic,” that is,
as propositions that cannot be based on other, more fundamental, propositions
that constitute grounds or evidence for them. First Principles are propositions we
don’t (have to) reason to but may reason from. They don’t need support because
“they support themselves” (1785: 463). They are, so to speak, “properly basic,”
they “need no proof and . . . do not admit of direct proof” (1785: 39). The reason
for this is that “their evidence is not demonstrative, but intuitive. They require not
proof, but to be placed in a proper point of view” (1785: 41). First principles are
propositions such that anyone whose faculties are functioning properly will accept
and believe them.
Reid’s argument for the existence of first principles is the following
regress argument:
When we examine, in the way of analysis, the evidence of any proposition, either
we find it self-evident, or it rests upon one or more propositions that support
it. The same may be said of the propositions that support it, and of those that
support them, as far back as we can go. But we cannot go back in this track to
infinity. Where then must this analysis stop? It is evident that it must stop only
when we come to propositions which support all that are built upon them, but are
themselves supported by none—that is to self-evident propositions. (1785: 455)
As Alston has pointed out (Alston 1987: 436), Reid’s category of self-evident
propositions must be understood as including propositions that are directly evident.
That there is a tree in front of one (if there is one) is never self-evident; still, it can
be directly evident.
The examples Reid offers of first principles are of different kinds. Some have
moral content (e.g., “What is done from unavoidable necessity may be agreeable
or disagreeable, useful or hurtful, but cannot be the object either of blame or of
moral approbation”), others are broadly metaphysical (e.g., “that there are some
things that cannot exist by themselves, but must be in something else to which they
belong,” 2002: 43), and still others are epistemological. I will be concentrating on
examples of the last kind, since these are pertinent to the problem of knowledge of
the external world. Following is an incomplete list of epistemological first principles
that Reid adumbrates:
Principle 1
It is a First Principle “that, in most operations of the mind, there must be an
object distinct from the operation itself. I cannot see, without seeing something.”
(1785: 44)
Principle 2
I shall also take for granted [i.e., as a First Principle] such facts as are attested to
the conviction of all sober and reasonable men, either by our senses, by memory,
or by human testimony. Although some writers on this subject have disputed
the authority of the senses, of memory, and of every human faculty, yet we find
that such persons, in the conduct of life, in pursuing their ends, or in avoiding
dangers, pay the same regard to the authority of their senses and other faculties,
as the rest of mankind. (1785: 46)
Principle 3
Those things do really exist which we distinctly perceive by our senses, and are
what we perceive them to be. (1785: 476)
402 SKEPTICISM
Principle 4
Another first principle is, that the natural faculties, by which we distinguish truth
from error, are not fallacious. If any man should demand a proof of this, it is
impossible to satisfy him. For, suppose it should be mathematically demonstrated,
this would signify nothing in this case; because, to judge of a demonstration, a man
must trust his faculties, and take for granted the very thing in question. (1785: 480)
In order to focus subsequent discussion, I start out by noting that Reid’s quest for
First Principles has not always been looked upon favorably. Kant, for instance, in
the Preface to the Prolegomena, says:
They [namely, Reid and some others philosophers of common sense] found a
more convenient method of being defiant [towards Hume’s doctrines] without
any insight, namely the appeal to common sense. It is indeed a great gift of God
to possess right (or as they now call it) plain common sense. But this common
sense must be shown in action by well-considered and reasonable thought and
words, not by appealing to it as an oracle when no rational justification for one’s
position can be advanced. To appeal to common sense when insight and science
fail, and no sooner—this is one of the subtle discoveries of modern times, by
means of which the most superficial ranter can safely enter the lists with the most
thorough thinker and hold his own. . . . Seen clearly, it is but an appeal to the
opinion of the multitude, of whose applause the philosopher is ashamed, while
the popular charlatan glories and boasts in it. (1783: 7)
Here Kant explicitly states that appeal to first principles (he refers to them as
common sense) has no justificatory power, because such an appeal is nothing “but
an appeal to the opinion of the multitude,” and such an appeal is something a
philosopher should be ashamed of—probably because there is no reason to suppose
that the opinion of the multitude is true.
In one respect Kant was right here. For Reid indeed held that there is merit in a
proposition’s receiving (nearly) universal consent. He says, for example, that
(A) we ought to take for granted, as first principles, things wherein we find
universal agreement, among the learned and unlearned, in different nations and
ages of the earth. (2002: 44)
As Reid thinks of them, however, first principles have other features besides
being (nearly) universally consented to. Kant does not mention these, but they are
nonetheless part of Reid’s actual view. These other features are of great importance
to Reid. For he does not forget to observe that “there seems to be great difference of
opinion among philosophers about first principles. What one takes to be self-evident,
another labours to prove by arguments, and a third denies altogether” (2002: 453).
One example he offers of this is the thesis that there is an external world. Before
Descartes, he notes, the existence of an external world was taken as a first principle,
Descartes sought to prove it, and Berkeley and Hume, noticing the weakness of the
“proofs,” denied its existence altogether. This leaves us wondering how to identify,
or recognize, a genuine first principle. Reid’s answer is that they are to be identified
by certain features. One of those we have already noticed: (A) universal agreement.
Another feature is that it is almost impossible not to believe that these Principles are
true; they force themselves upon us, we cannot resist believing them. Here are some
statements of this point:
(B) The belief which is implied in [perception] is the effect of instinct. (1764: 172)
I cannot demonstrate that two quantities which are equal to the same quantity,
are equal to each other; neither can I demonstrate that the tree which I perceive
exists. But by the constitution of my nature, my belief is irresistibly carried along
by my perception of the tree. (1764: 172)
Closely related to (B) is the feature that first principles are indispensable for
conducting practical life. Maybe the skeptic thinks she can dispense with giving
assent to these Principles, but the practice of her life shows otherwise. First principles
are such
(C) that every man of common understanding readily assents to them, and finds it
absolutely necessary to conduct his actions and opinions by them, in the ordinary
affairs of life. (1785: 40–41)
It is one thing to profess [that the testimony of sense is fallacious], another seriously
to believe it and to be governed by it in the conduct of life. It is evident that a man
who did not believe his senses could not keep out of harm’s way one hour of his
life; yet, in all the history of philosophy, we never read of any skeptic that ever
stepped into fire or water because he did not believe his senses. (1785: 99)
Another feature of a first principle is
(D) that no man ever pretended to prove it, and yet no man in his wits calls it in
question. (2002: 474)
That sense perception is reliable, says Reid, is not something that can be proved,
nor something that can seriously be called into question. Feature (D) says that it
is our predicament that first principles, such as Principle 3, which says that sense
perception is reliable, can never be proved. Reid provides the reason why this is
so (in the previous section we already touched upon this): in order to prove that
perception is reliable, one has to take for granted that the faculty that is operative in
the proof is reliable. Discussing the reliability of reason, Reid says:
If a man’s honesty were called in question, it would be ridiculous to refer to
the man’s own word, whether he be honest or not. The same absurdity there
is in attempting to prove, by any kind of reasoning . . . that our reason is not
fallacious, since the very point in question is, whether reasoning may be trusted.
(1785: 480)
The point can be generalized over all of our faculties:
Every kind of reasoning for the veracity of our faculties, amounts to no more
than taking their own testimony for their veracity; and this we must do implicitly,
until God give us new faculties to sit in judgment upon the old. (1785: 481)
404 SKEPTICISM
We cannot prove that our faculties are reliable, without pragmatically presupposing
what we try to prove. And the flipside of this is that we cannot disprove their
reliability either, without pragmatically presupposing their reliability.
Still another feature of a First Principle is that
(E) opinions that contradict them “are not only false but absurd.” (2002: 462)
Furthermore, belief in First Principles is not determined by experience or education
but “by our nature.” This is a prevalent theme throughout Reid’s writings:
(F) I consider this instinctive belief [that my faculties are trustworthy] as one of
the best gifts of Nature. I thank the Author of my being, who bestowed it upon
me before the eyes of my reason were opened, and still bestows it upon me, to be
my guide where reason leaves me in the dark. And now I yield to the direction
of my senses, not from instinct only, but from confidence and trust in a faithful
and beneficent Monitor, grounded upon the experience of his paternal care and
goodness. (1764: 170)
original and natural judgments (of perception) are . . . a part of that furniture
which nature hath given to the human understanding. They are the inspiration of
the Almighty. . . . They are a part of our constitution. (1764: 215)
First principles, then, have certain features that enable us to identify them. Kant, in
the quotation given earlier, attributed to Reid the view that a proposition’s having
those features (Kant, as noted, only mentions feature (A)) is evidence for the truth
of the purported first principle proposition. According to Kant, Reid holds that
universal agreement can be used in an argument that seeks to prove the truth of
the principle. But, as we can now see, this was not what Reid actually held. First
principles resist direct proof. For Reid, these features don’t prove the truth of
supposed first principles but are indicators of a proposition’s being a first principle.
Or, as Keith Lehrer puts it, they are “criteria . . . for determining whether something
is a first principle” (Lehrer 1989: 87). And Reid, although he nowhere explicitly
says so, held that the proposition that sense perception is reliable, displays all, or
nearly all, the features mentioned and therefore qualifies as a first principle.
Reid’s position on first principles can, then, be summarized as follows: (1) as the
regress argument shows, there must exist first principles; (2) one such principle is
that we can gain perceptual knowledge of the external world; (3) first principles
have various features, (A)–(F), that enable us to identify or recognize them; these
features cannot be used in an argument to show or prove the truth of the principles;
(4) although the proposition that sense perception is reliable cannot be proven,
it is nonetheless reasonable to accept it as a first principle; (5) we are justified in
believing first principles until their falsehood has been established; (6) the falsehood
of the proposition that sense perception is reliable (and of the belief that there is an
external world) has not been established.
NOTES
1. Examples of this kind of engagement are Williams (1996) and Greco (2000).
2. Pyrrhonist skeptics would, of course, deny what the bystander says about them. They
would say that they are following appearances in the absence of having beliefs. If there
appears to be some danger, they will stay out of harm’s way, but this doesn’t mean,
they say, that they believe there is some danger. However, even if the bystander’s
interpretation is wrong by the lights of the Pyrrhonists, this doesn’t invalidate the
bystander’s point; she may still have a very good reason for thinking that Pyrrhonists
do believe what they say they don’t.
3. Wolterstorff (1987) argues that when Reid spoke of the “theory of ideas” he had
in mind a commitment not only to mental representationalism but also to classical
foundationalism. I will discuss these assumptions separately.
4. No one, to my mind, has appreciated this point better than Alston (1993).
5. Likewise, Van Cleve (2015: 328–332) has argued that Reid’s direct realism, his
alternative to the theory of ideas, by itself does not give sufficient resources for
answering so-called closure-based skeptical arguments. I myself have argued for a
somewhat different evaluation of the anti-skeptical power of Reid’s direct realism in
Van Woudenberg (2000).
6. This section has lifted a couple of paragraphs from Van Woudenberg (2004).
REFERENCES
Alston, William P. 1993. The Reliability of Sense Perception. Ithaca: Cornell University
Press.
Alston, William P. 1986. “Thomas Reid on Epistemic Principles,” History of Philosophy
Quarterly 2: 435–452.
De Bary, Philip. 2002. Thomas Reid and Skepticism: His Reliabilist Response. London and
Greco, John. 2004. “Reid’s Reply to the Skeptic.” In T. Cuneo and R. van Woudenberg
(eds.), The Cambridge Companion to Thomas Reid. Cambridge: Cambridge University
Press: 134–155.
Greco, John. 2000. Putting Skeptics in Their Place. Cambridge: Cambridge University Press.
Hume, David. 1902. Enquiries Concerning the Human Understanding and Concerning the
Principles of Morals, edited by L. Selby-Bigge. Oxford: Clarendon Press.
Kant, Immanuel. 1783 [1950]. Prolegomena to Any Future Metaphysics, edited by L. White
Beck. Indianapolis: Bobbs-Merrill.
Lehrer, Keith. 1989. Thomas Reid. London and New York: Routledge.
Redekop, Benjamin. 2004. “Reid’s Influence in Britain, Germany, France and America.” In
T. Cuneo and R. van Woudenberg (eds.), The Cambridge Companion to Thomas Reid,
Reid, Thomas. 1764 [1997]. An Inquiry into the Human Mind on the Principles of Common
Sense, edited by D. R. Brooks. Edinburgh: Edinburgh University Press.
Reid, Thomas. 1785 [2002]. Essays on the Intellectual Powers of Man, edited by
D. R. Brooks. Edinburgh: Edinburgh University Press.
Van Cleve, James. 2015. Problems from Reid. New York: Oxford University Press.
Van Woudenberg, René. 2000. “Perceptual Relativism, Scepticism, and Thomas Reid,”
Reid Studies 3: 65–90.
406 SKEPTICISM
Van Woudenberg, René. 2004. “Reid and Kant against the Skeptic.” In J. Houston (ed.),
Thomas Reid: Context, Influence, Significance. Edinburgh: Dunedin Academic Press:
155–180.
Van Woudenberg, René. 2013. “Thomas Reid between Externalism and Internalism,”
Journal of the History of Philosophy 51: 75–92.
Williams, Michael. 1996. Unnatural Doubts. Princeton: Princeton University Press.
Wolterstorff, Nicholas. 2001. Thomas Reid and the Story of Epistemology. Cambridge:
Wolterstorff, Nicholas. 1987. “Hume and Reid,” The Monist 70: 398–417.
CHAPTER TWENTY-NINE
Johann Georg Hamann

GWEN GRIFFITH-DICKSON
1 INTRODUCTION
Johann Georg Hamann (1730–1788) was a Prussian scholar, writer, and editor,
considered to be one of the most well-read scholars of his time (greatly aided by his
fluency in many languages). He was born in Königsberg, the son of a midwife and a
barber-surgeon, and began study in philosophy and theology at the age of sixteen.
He left university without completing his studies and became the governor to a
wealthy family on a Baltic estate. During this time he continued his extraordinarily
broad reading and private research. He took up a job in the family firm of a friend
from his Königsberg days, Christoph Berens, and was sent on an obscure mission
to London, in which he evidently failed. He then led a high life until he ran out of
friends, money, and support. In a garret, depressed and impoverished, he read the
Bible cover to cover and experienced a religious conversion. He returned to the
House of Berens in Riga, where they evidently forgave him his failure. He fell in love
with Christoph Berens’s sister, Katharina, but was refused permission to marry her
by his friend, on the grounds of his religious conversion; Berens was an enthusiastic
follower of the Enlightenment and was nauseated by the more pious manifestations
of Hamann’s new-found religiosity. Smarting from this blow and its motivations,
Hamann returned to his father’s house in Königsberg, where he lived for the rest
of his life until his final months. During this period in Königsberg, he lived with a
woman whom he never married (despite his committed Christianity) but to whom
he remained devoted and faithful, having four children on whom he doted and who
occasionally feature in his writings (principally as unruly distractions to the author’s
scholarship). At the end of his life, he accepted an invitation to Münster from one
of his admirers, Princess Gallitzin; he died there in 1788.
Hamann had a profound influence on the German “Storm and Stress” movement,
and on other contemporaries such as Herder and Jacobi; he impressed Hegel and
Goethe (who called him the brightest head of his time) and was a major influence on
Kierkegaard. He also maintained a friendly, mutually respectful relationship with
Kant despite their sharp philosophical disagreement and the several critiques of
Kant’s work penned by Hamann—indeed, it was through the intercession of Kant
408 SKEPTICISM
that Hamann was able to attain employment as a civil servant in the tax office of
Frederick the Great.
Hamann never held an official academic post, nor an ecclesiastical one; this may
in part have been due to his pronounced speech impediment, which inhibited him
from either lecturing or preaching. Yet he was highly respected in his time for his
scholarship and breadth of learning. His writings were notorious even in his own time
for the challenges they threw down to the reader. These challenges to interpretation
and understanding are only heightened today. Nevertheless, an increasing number
of scholars from philosophy, theology, aesthetics, and German studies are finding his
ideas and insights of value to contemporary concerns. His central preoccupations
are still pertinent: language, knowledge, the nature of the human person, sexuality
and gender, and the relationship of humanity to God. Meanwhile, his views, which
in many respects anticipate later challenges to the Enlightenment project and to
modernity, are still relevant and even provocative.
Hamann is sometimes portrayed, particularly in the English-speaking world, as an
opponent of the Enlightenment, an enemy of reason, and an irrationalist. Hamann
studies have moved on from this interpretation. He had a complex, admittedly
combative relationship to his time, but given the fundamental disjunctions of
Enlightenment thought—rationalism, materialism, etc.—he could hardly be a single-
minded opponent. Both these tendencies of his time displeased him. Positively put,
his underlying insight and conviction was the wholeness, the integration, and above
all the relations of the phenomenon in question.
Hamann held that the various striking features of our experience—ourselves,
others, language, the world, ultimately God—are inextricably related. It was
his conviction that a true and deep understanding of any of these can only arise
following consideration of its relationship to the others. So, for instance, an analysis
of the philosophy of language would have to include a discussion of God, which in
turn must make reference to sexuality, and vice versa; such assertions have led to
some considering Hamann to be a mystic.
This emphasis on relationships between the most basic and important facets of
our experience must be borne in mind when exploring Hamann’s viewpoint on
skepticism, for it meant that he eschewed any conventional relationship to positions
on knowledge and skepticism that were held at the time. Hamann was skeptical,
yet a decided anti-skeptic. He concerned himself with the limits and possibilities
of knowledge, yet not in the manner of Locke or Kant: for Hamman this was not
a technical or a technique question—his interest was in the human implications of
knowing and not knowing. Hamann also had a very different story to tell: one in
which history was best understood as “‘mythology”; one in which the cause-and-
effect relationship of our ingestion and elimination had to be taken on faith, but
attaining reliable knowledge of God was considered to be as easy as child’s play.
Hamann does appear as a skeptic, or a so-called irrationalist, in several respects.
First, he had an ironist’s delight in how certain philosophers deployed skepticism
to undermine faith in reason. To an extent, this was a position he agreed with—
overweening faith and confidence in our epistemological powers was for Hamann
unwarranted and almost ethically dangerous. Second, he was a skeptic vis-à-vis
JOHANN GEORG HAMANN 409
“reason” and the qualities attributed to it by the typical figures of the time. Third, he
was a skeptic when it came to some epistemological claims to knowledge, namely,
its certainty and reliability, and its unity and systematicity.
Yet Hamann was nonetheless a “True Believer” in a certain conception of
knowledge, “rightly understood” as: piecemeal and iterative; relational rather than
rational, where we know best by intimacy rather than by inference; inseparable from
experience, tradition, and language; and as ultimately consisting in self-knowledge,
which itself rests on being known. Hamann’s skepticism, then, did not come close
to playing the nihilistic, ultimately destructive role it played for other philosophers
of the Enlightenment.
2 “I DON’T KNOW”: THE TEST CASE OF SOCRATES

Hamann’s epistemology called for a realism concerning our ability to acquire and
hold information in isolation, twinned with immense positivity regarding the ability
to understand our experience in a holistic fashion. This is well demonstrated by his
use of Socrates to illustrate his conception of human knowledge.1 Plato portrayed
Socrates, whether sincerely or disingenuously, as continually protesting his lack of
knowledge—a phenomenon well known in the literature as “Socratic Ignorance.”
In an act of wanton provocation, Hamann took the self-proclaimed ignorance of
Socrates, the poster boy of the Enlightenment, as his strategy for asserting the limits
to human knowledge. The fact that the most intelligent of individuals persistently
claimed to know nothing challenges any claim to knowledge that his hearers—or
eighteenth-century admirers—might have made and calls the powers of human
knowledge itself into question.
In Hamann’s hands, Socratic ignorance is not the foundation stone of skepticism;
it was not a feigned agnosticism as a sophisticated argumentative position affected in
an elenctic dispute, nor a tactic. Socrates’s ignorance was “Empfindung”: a feeling,
perception, intuitive sense for what is the case;2 it is a recognition of the reality of
human knowing.
Hamann deploys this theme in multiple ways. In one maneuver, he portrays
Socrates’s avowal of ignorance as an ethical stance, almost a renunciation, a
refusal to participate in a distasteful activity. If a skilled card player responds to an
invitation from a pack of cheats by saying, “I don’t play,” we recognize this, not as
a lack of skill, but as a rejection of their immoral proclivities. So too with Socrates:
on his lips “I don’t know” means “I don’t play the game of competitive knowing
the way you Sophists do.” Hamann depicts the Athenians’ curiosity, their lust for
the novelty of knowledge, as “satyriasis,” a sex addiction if you will. Socrates’s
insistence on his ignorance was a rejection of their epistemological promiscuity, a
refusal to join in their philosophical vices: both the lust for uncommitted knowing
and the pride in conquest (76:28f.). For Hamann, knowledge properly understood
is a relationship; and thus there is a way to know deeply, reverently, and ultimately
deliciously. But there is also a style of snatching and quitting in a shallow exchange;
and knowing something for the sake of power or even domination of the object
is, in Hamann’s metaphor, akin to passing acquaintance for the purposes of bodily
410 SKEPTICISM
pleasure. Both signify greed at the expense of the other—a lack of perception,
sensitivity, commitment.
If there is such a thing as epistemological promiscuity for Hamann, there is also
epistemological rape. Hamann describes Socrates as “indifferent” to “that which
is called truth” (76:5). In personal letters at the time, Hamann portrays himself as
similarly indifferent to “no other interest than the interest in knowing the truth (do
not be frightened by my blunt confession)—of this hyperbolic interest I have neither
a concept nor a feeling” (Letters 3, 7:12–14).3 Contemporary letters to Kant make
clear that more damage can be done with truth than with falsehoods, depending on
the use made with them (Letters 1, 377:10–15); and in a striking image:
Truth would not let herself be approached too closely by highwaymen, she wore
one garment on top of another so that one doubted being able to find her body.
How frightening if they had their way and saw that frightful ghost, the truth,
before them. (Letters 1, 381:8–11)
Hamann’s provocative rejection of lusting for the truth should be seen of a piece
with his rejection of Athenian promiscuous knowing; as knowledge is found in
faithful commitment, so too truth is like a woman protecting herself from the
violation and domination of a predator. In contrast, Hamann’s ideal “heroic spirit
of the philosopher” possesses “thirsting ambition for truth and virtue, and a fury
to conquer all lies and vice.”4 The task of knowledge confronts us, then, with the
ethical choice of being a knight questing for knowledge as one in service of the
truth, the fair lady he serves, while rejecting lies and vice—or being the highwayman
or seducer who snatches and robs.
Thus, the activity of knowing is not the acquisition and holding of information,
but being in a relationship to reality and to truth. Not only can we not dominate the
object of knowledge, we cannot even know our own knowing: “As our skeleton is
concealed, because we are made in secret, because we are built below in the earth,
how much more so are our concepts made in secret, and can be seen as the limbs of
our understanding” (66:10–13).
More importantly, knowledge, ignorance, and skepticism are all ultimately ethical
matters to Hamann. The person closest to the truth is the one who displays fidelity,
commitment, and humility—and renounces the power and pride of knowledge in
favor of something else: the humbling of self-knowledge and being known.
The Delphic Oracle commanded those who entered the temple to “Know
Thyself.” It also, when asked who was the wisest of all, commended Sophocles and
Euripides but proclaimed Socrates as the rightful bearer of this title. Hamann runs
these themes together with Socrates’s frequent claim to ignorance:
Sophocles and Euripides would not have become such great models for the stage
without the art of dissecting the human heart. Socrates, however, surpassed them
both in wisdom, because he had progressed further in self-knowledge than they,
and knew that he knew nothing. (71:7–11)
With this move, Hamann repositions the Enlightenment debate: what is, or should be,
the real content of knowledge, the mark of wisdom? If his Enlightenment contemporaries
prize the Greeks as much as they claim to, they ought to agree: the sign of wisdom is
neither omniscience nor sophistry, but is above all found in self-knowledge.
In “Socratic Memorabilia,” Hamann goes on to make what might seem an extreme
declaration of the impossibility of knowledge; that is, our inability to “know” the
most self-evident, experiential facts: “Our own being and the existence of all things
outside us must be believed and can be made out in no other way” (73:21–22). All
knowledge, whether of the unknowable Transcendent or of our own mortality (as
Socrates demonstrated in a syllogism), thus rests on the same ground; not reason, but
the relationship with the known that is Hume’s “belief”—or faith (the same word in
German: Glaube).5 It is widely recognized that Hamann makes strategic use of Hume’s
skeptical critique as an attack on reason. At the same time, Hamann has also been
portrayed as a fideist. To conceive of Hamann as either a Humean skeptic or a pious
fideist, however, is a simplistic misunderstanding of a complex and subtle thinker.
Fundamentally, he does not oppose believing and knowing but sees them as inextricably
related in our epistemological affairs, as is clear from this letter written late in his life:
It is pure idealism to separate believing and feeling from thinking. Companionship
is the true principle of reason and language, by which our sensations and
representations are modified. This and that philosophy always separates things
which by no means can be divided. Things without relationships, relations
without things. There are no absolute creatures, as little as there is absolute
certainty. . . If we believe our sensations, our representations, then perhaps all
divisions cease. We cannot dispense with these witnesses for ourselves, but can
refute no one with their agreement. (Letters 7, 174:12–18, 22–24)
Recognizing the epistemological role played by belief subverts the demand for
absolute certainty that triggers the Enlightenment slide into skepticism. Our senses,
our existence, and the existence of all around us rest on a relationship of faith
and trust. Paradoxically, this portrait of knowledge as consisting in a relationship
to reality and grounded ultimately in faith or belief enables his cynicism about
Enlightenment claims to knowledge, but also paves the way for his ultimate rejection
of skepticism. Either way, a methodological doubt of the senses or a sophisticated
philosophical position collapsing into solipsism is undermined.
Hamann’s discussion of Socratic ignorance reaches its climax with a Pauline
quote from I Corinthians 8, which he proclaims as the “venerable seal” and the
“key to Socrates’s testimony to his own ignorance”: “If anyone thinks he knows
something, he knows nothing as he should know it. But if someone loves God, he
will be known by him—as Socrates was known by Apollo [the Delphic Oracle] to
be a wise man” (74:20–27). This unites the two themes of Socratic ignorance and
the relational basis for knowledge: to know oneself to be ignorant but a lover of
God is to be known by God. Our natural wisdom will decay; but from this “death,”
this “nothingness” of humble, self-knowing ignorance, “life and being sprout forth
newly-created in a higher knowledge” (ibid.).
Hamannian skepsis, in fact, gives way to an extraordinary optimism about
knowledge. What can be known by the humble knight of philosophy is far greater
than what can be known by reason: deep self-knowledge, and in so doing, knowing
412 SKEPTICISM
oneself to be in love with God, and thus to know God himself in relationship: “a
truth which (lies) concealed,” “a secret wisdom.” All other forms of “knowledge”
Hamann derides as “labyrinths of learned sophists” (77:6–8); and of these Socrates
can disavow all acquaintance, while being the wisest of all:
Thus, as all our powers of knowledge have self-knowledge as their object, so too
all our inclinations and desires have self-love. The first is our wisdom, the latter
our virtue. As long as it is not possible to know oneself, it remains impossible to
love oneself. Therefore only the truth can set us free. . . . (from “Fragments,”
Writings 1, 300:17–22)
3 REASON’S MISUNDERSTANDING
Hamann then is not to be considered a skeptic regarding the possibility of
knowledge—only a skeptic regarding an incorrect conception of it. One could still
consider him critical, in the context of the Enlightenment, of some conceptions of
reason. Yet despite some interpreters’ reading of him, he clearly was not a mere
“irrationalist”; for one thing, he saw that position as ultimately self-refuting: “All
forms of irrationalism presuppose the existence of reason and its misuse” (Writings
3, 191:21–27). Certainly there are passages where he sought to articulate a positive
function for reasoning in relation to other phenomena:
God, nature and reason have as intimate a relation to one another as light, the
eye, and all that the former reveals to the latter, or like the centre, radius and
periphery of any given circle, or like author, book and reader. (Letters 5, Nr. 784,
272:14ff.)
Hamann subjected the Enlightenment concept to a more careful deconstruction; no
mere “for or against reason,” but a more searching inquiry into the way the concept
was characterized and what claims were made on its behalf. “For what is celebrated
Reason with its universality, infallibility, unsurpassability, certainty and evidence? An
Ens rationis, an idol, to whom the screeching superstition of irrationality ascribes
divine attributes” (from “Konxompax,” Writings 3, 225:3–6). As a human activity,
inseparable as he maintained from passion, language, experience, indeed from
revelation, it was subject to human fallibility; his rejection was of the infallibility
ascribed to it by the Enlightenment, as well as injecting a more modern, indeed
postmodern insight: refuting the “universal healthy reason” that failed to take into
account the role that history, culture, language, and lived experience all play in
shaping and conditioning human cognition. “People talk about Reason as if it were a
real being, and the good Lord as if the same were nothing but a concept” (Letters 7,
26:34–35). Reason for Hamann is not a thing, at the end of the day; it is an activity,
subject to the same conditions that affect all our other undertakings. As he says in a
letter to Jacobi: “Our reason must wait and hope—and desire to serve nature, not
legislate it” (Letters 5, Nr. 782, 265:30). He adds:
With me it is not so much the question: what is reason? but far more: what is
language? and this I suppose to be the ground of all paralogisms and antinomies of
which the former is accused. Hence, it happens that one takes words for concepts,
and concepts for the things themselves. In words and concepts no evidence is
possible, which is ascribed purely to things and states of affairs. (Letters 5, Nr.
782, 264:34–265:1)
Hamann’s skepticism of overblown Reason, the knowledge acquired by it, and the
philosophy based on them are best seen in his unpublished deconstruction of Kant’s
first Critique: “Metacritique of the Purism of Reason.” Here Hamann accuses
Kant (in his Critique of Pure Reason) of conducting a “threefold purification of
philosophy”: purifying philosophy from tradition, experience, and finally from
language itself. All three purifications are highly undesirable in Hamann’s view; and
he alleges that, in so doing, reason becomes not merely the object, source, or mode
of knowledge (as Kant discusses), but in fact “is the foundation of all objects, sources
and kinds of knowledge; is itself none of the three.” This, however, subverts the role
of experience and the senses being the ground of all knowledge in an integrated and
holistic way.
Once you have purged tradition, that is, human communities over history, from
philosophy, and secondly have maintained that knowledge without experience
is in some sense possible, you have implicitly declared that reason can function
without language, “whose only credentials are tradition and use” (“Metacritique
of the Purism of Reason,” Writings 3, 283:22). The first two purifications entail
the third.6 Of these three purifications, the most intolerable is the “purification”
of philosophy from language; for “Not only the entire ability to think rests on
language . . . but language is also the crux of the misunderstanding of reason with
itself” (“Metacritique of the Purism of Reason,” Writings 3, 287:7–9).
Language and reason are two moments in the same activity of human knowing.
This insistence on the integration of reason and language took root in his early
writings: “Without language, we would have no reason; without reason no religion,
and without these three essential components of our nature, neither spirit nor
the bond of society” (“Two Mites on the Latest German Literature,” Writings 3,
231:10–12). This insight persisted, and deepened, through to his last personal
letters. As he explained to his friend, the philosopher Jacobi, Hamann thought of
reason much as St. Paul thought of the Law and its justification: as something that
can give knowledge of error, but not the way to truth and life. He had eschewed
investigation into reason in favor of the inquiry into language: “Without word, no
reason—no world. Here is the source of creation and providence” (Letters 5, Nr.
721, 95:13–22).
Similarly, then, to his view on epistemology, Hamann should not be regarded as
a wholesale skeptic rejecting reason. Once again, he is merely in search of a more
accurate portrayal of the role it plays in our human experience:
If only I was as eloquent as Demosthenes, I would have to do no more than
repeat a single word three times. Reason is language—Logos; I gnaw on this
marrowbone and will gnaw myself to death over it. It is still always dark over
these depths for me: I am still always awaiting an apocalyptic angel with a key to
this abyss. (Letters 5, Nr. 753, 177:16–21)
414 SKEPTICISM
4 REVELATIONS AND TRADITIONS

In another letter, he points out: “Our knowledge is piecemeal—no dogmatist is in a
position to feel this great truth, if he is to play his role and play it well; and through
a vicious circle of pure reason skepsis itself becomes dogma” (Letters 5, Nr. 833).
For Hamann, belief and knowledge, indeed knowledge and skepticism, are not
divided by hard boundaries. Belief and knowledge both rest on a foundation of trust;
and with the rejection of absolute certainty for knowledge under any circumstances,
the threat that skepticism about our epistemological powers poses is reduced:
Belief needs reason just as much as the latter needs the former. Philosophy is
of idealism and realism: as our nature is composed of body and soul. Qui bene
distinguit, optime definire potest—. . . Only scholastic reason divides itself into
idealism and realism. Correct and authentic reason knows nothing of this imagined
difference, which is not grounded in the nature of things, and contradicts their
unity, which lies at the foundation of all our concepts, or at least, should do so.
Every philosophy consists of certain and uncertain knowledge, of idealism and
realism, of sensuousness and deductions. Why should only the uncertain be called
belief? What then are—rational grounds? Knowledge without rational grounds
is as little possible as sensus sine intellectu. . . . Complex beings are not capable
of simple emotion, still less simple knowledge. Feeling can as little be separated
from reason in human nature as can reason from sensuousness. . . . Emotion
must be delimited by rational grounds. Knowledge from belief is fundamentally
identical to: Nil in intellectu. (Letters 7, Nr. 1060, 165:7–166:3, 10–11)7
Both idealism and realism share an undesirable picture of the human being standing
over against the world, in Hamann’s view; whereas for him the knower and the
known share an underlying unity. Qui bene distinguit: only one who wisely makes
such distinctions can rightly define body and soul, knowledge and world.
Hamann’s positive account of knowledge ultimately rests on his anthropology: a
holistic epistemology is the activity of diverse yet integrated human beings, a union
of the senses and understanding, reasoning and passions, Baroque gentility and
Homerian lust for vengeance.8 His lament was that philosophers have always given
truth a bill of divorce, separating what nature intended to be joined together, such
as thinking and feeling (cf. Writings 3, 40:3f.):
The understanding and experience are fundamentally the same thing: as the
understanding and application are. From where do we get this distinction into
opposites? Does the entire mystery of our reason, its antitheses and analogies,
rest on nothing but a poetic license to separate what Nature has joined and to
couple what she wanted to distinguish, to mutilate and then mend again? (Letters
6, Nr. 1011, 534:14–20)
Instead, a little epistemological metaphor makes clear his preferred, holistic, and
multivalent picture of human knowing:
The human being therefore is not only a living field but also the son of the field,
and not only field and seed (according to the system of the materialists and
idealists) but also the king of the field, to cultivate good seed and enemy weed on
his field; for what is a field without seed and a ruler without land and revenue?
These three in us therefore are one. (“Philological Ideas and Doubts,” Writings 3,
40:14–20)
In knowing, we passively receive, like a field receiving seed (a “materialist”

epistemology); we actively generate and germinate ideas and knowledge (as
the “idealists” imagine); we are the product and the result (son) of our own
epistemological activity; but knowledge is also a question of wise governance
(rule) with ethical implications (good seed and enemy weed).9 With this we are
charged.
Meanwhile, the “stamen and menstruation” of our cognition (what we germinate
and what in turn flows from us) “are therefore in the truest sense revelations and
traditions which we take for our own property, transform with our juices and powers
and thereby fulfill our commission . . . as a political animal, partly to reveal and partly
to transmit” (“Philological Ideas and Doubts,” Writings 3, 39:7–18). Hamann’s
epistemology is also social, even political. Even if we come into the world as a tabula
rasa—or as he prefers, “an empty wineskin”—this lack makes us “all the more able
to enjoy nature through experiences and to enjoy community with [our] race through
traditions. Our reason arises, at the very least, from this twofold lesson of sensuous
revelations and human testimonies” (“Philological Ideas and Doubts,” Writings 3,
39:23–28).
From his earliest journals, he insisted that knowledge rests on self-knowledge;
and self-knowledge is ironically dependent on another:
In order to make easier the knowledge of our self, my own self is visible as in
a mirror in every neighbor. As the image of my face is reflected in water, so
my self is reflected in every neighbor. That this self may be as dear to me as
my own, Providence has sought to unite as many advantages and amenities in
human society as possible. God and my neighbor are therefore a part of my self-
knowledge, my self-love. (Writings 1, 302:16–23)
Human knowledge begins and ends with another. “Self-knowledge begins with
the neighbor, the mirror, and just the same with true self-love; that goes from the
mirror to the matter” (Letters 6, 281:16–17). To know oneself, in other words, is to
know oneself as existing only within a relationship.
In sum, as neither a skeptic nor a rationalist, the possibilities of knowledge
are both more restricted and more generous for Hamann than for many of his
contemporaries. By “pure” reason, by inference and ratiocination, our knowledge
is far less certain and far less ample than for other Enlightenment thinkers. But
our ignorance, or lack, is the gateway for something far richer and deeper than
epistemological certitude. It is the sign, and the opportunity, of our openness and
need for each other. Indeed, for Hamann, knowledge ultimately opens out to the
transcendent with perfect confidence; the cheerful confidence of a child certain
of being known and loved. Ourselves, human history, the whole of creation is
“mythology,” a “book” we read by an author who addresses us. Human knowing is
416 SKEPTICISM
thus located in the context of a fundamental relationship with the divine, revealing
to us the highest knowledge, which is being known:
Not Cogito; ergo sum, but vice versa, and more Hebraic: est; ergo cogito, and
with the inversion of such a simple principle perhaps the whole system might
receive another language and direction. (Letters 5, Nr. 840, 448:26–28)
NOTES
1. All translations of Hamann are mine. References are to Josef Nadler’s edition of
his Schriften (cited hereafter as Writings) (Hamann 1949–1957) and in English
translation by Gwen Griffith-Dickson (1995). The latter have been laid out such that
the same numbered reference refers both to the English translation in Dickson and
to the original in Nadler’s edition. Unless otherwise stated, quotations in this first
section are from Hamann’s essay “Socratic Memorabilia,” Writings 2, and appear
with page and line number only.
2. I argue that Hamann intends to translate Hume’s “feeling”, that which is foundational
to reason and philosophy (Dickson 1995: 47).
3. The standard edition of his letters is Henkel (1955–1975). They are cited here as
Letters, volume, page: line.
4. “A little vision and superstition here would not merely merit indulgence, but rather
something of this leaven is necessary to start off the soul’s fermentation towards
a philosophical heroism. A thirsting ambition for truth and virtue, and a fury to
conquer all lies and vice, which precisely are not recognized as such, nor wish to be;
in this consists the heroic spirit of a philosopher” (63:30–35).
5. Hamann was arguably the first to identify the tactical potential of Hume’s account of
belief for theology. Hamann later writes of himself in a letter to Jacobi: “I was full of
Hume when I wrote SM, and p. 49 of my little book refers to this,” and “I studied
him before I wrote SM, and am indebted for my doctrine of faith to the same source”
(Letters 7, 167:9).
6. For thorough discussions of these points, see Dickson (1995) and Piske (1989),
who are in broad agreement. In the former, I argue: “The experience excepted
by a workable definition of ‘a priori’ is precisely the experience in question, and
precisely the experience which subverts the authority, autonomy, and universality
of the Enlightenment theory of reason” (1995: 306). And also: “for reason to be
as universal and as historically unconditioned as this strand of the Enlightenment
maintains, it must be free of any language which is the product and embodiment of
its time and place. . . If reason does always involve language while being purified of
experience and tradition, and therefore able to exist as universal and autonomous,
language must not be a matter of experience and tradition and sociality, but must lie
within the structures of the mind, indeed, share its structure, its syntax or grammar,
with the structures of judgement. . . . The possibility of discovering the true structure
or form of a phenomenon from self-reflection, apperception, the analysis of our own
thought, implies—or possibly requires, if one holds that thought is impossible without
language—the possibility of doing the same for or with language. A priori thought,
arising from the structuration of the mind, requires either an a priori language arising
from the structure of the mind, or else no language at all” (1995: 304–305).
7. “Sensuousness” translates Sinnlichkeit (Kant’s “sensibility”).
8. In an ironic rebuttal of Rococo aesthetics of art as the imitation of beautiful nature,
Hamann quipped: “A lust for vengeance (Rachsucht) is the beautiful nature Homer
imitated” (Letters 2, 157:12).
9. “Good seed and enemy weed” is a reference to one of the parables of the sower (Mt.
13:24-30).
REFERENCES
Alexander, W. M. 1966. Johann Georg Hamann. Philosophy and Faith. The Hague:
Martinus Nijhoff.
Bayer, Oswald. 2012. A Contemporary in Dissent. Grand Rapids: Wm. B. Eerdmans
Publishing Company.
Bayer, Oswald and Bernd Weißenborn. 1993. Johann Georg Hamann. Londoner Schriften.
Munich: C. H. Beck.
Beech, Timothy. 2010. Hamann’s Prophetic Mission. A Genetic Study of Three Late Works
against the Enlightenment. London: Maney Publ.
Beiser, Frederick C. 1987. The Fate of Reason: German Philosophy from Kant to Fichte.
Cambridge, MA: Harvard University Press.
Berlin, Isaiah. 1993. The Magus of the North: J. G. Hamann and the Origins of Modern
Irrationalism, edited by Henry Hardy. London: John Murray.
Betz, John R. 2009. After Enlightenment: The Post-Secular Vision of J. G. Hamann. Oxford:
Wiley-Blackwell.
Brose, Thomas. 2004. Johann Georg Hamann und David Hume. Frankfurt am Main: Peter
Lang.
Dahlstrom, Daniel. 2000. “The Aesthetic Holism of Hamann, Herder and Schiller.” In
K. Ameriks (ed.), The Cambridge Companion to German Idealism, 76–94. Cambridge:
Dickson, Gwen Griffith. 1995. Johann Georg Hamann’s Relational Metacriticism. Berlin:
de Gruyter.
Dunning, Stephen. 1979. The Tongues of Men: Hegel and Hamann on Religious Language
and History. Missoula: Scholars Press.
Hamann, Johann Georg. 1949–1957. Sämtliche Werken, 6 volumes, edited by Josef Nadler.
Vienna: Verlag Herder.
Hamann, Johann Georg. 1955–1975. Briefwechsel, 8 volumes, edited by W. Ziesemer and
A. Henkel (from volume 4 on, edited by Henkel alone). Wiesbaden and Frankfurt: Insel
Verlag.
Haynes, Kenneth. 2007. Hamann: Writings on Philosophy and Language. Cambridge:
Hegel, G.W. F. 2008. Hegel on Hamann, edited and translated by Lisa Marie Anderson.
Evanston: Northwestern University Press.
418 SKEPTICISM
Jørgensen, Sven-Aage. 1962. Johann Georg Hamann: Fünf Hirtenbriefe das Schuldrama
betreffend. Einführung und Kommentar. Copenhagen: Ejnar Munksgaard.
Jørgensen, Sven-Aage. 1968. Sokratische Denkwürdigkeiten. Aesthetica in Nuce. Mit einem
Kommentar. Stuttgart: Reclam.
Leibrecht, Walter. 1966. God and Man in the Thought of Hamann, translated by James H.
Stam and Martin H. Bertram. Philadelphia: Fortress Press.
Lowrie, Walter. 1950. Johann Georg Hamann, an Existentialist. Princeton: Princeton
Theological Seminary.
Nisbet, H. B. (ed.) 1985. German Aesthetic and Literary Criticism: Winckelmann, Lessing,
Hamann, Herder, Schiller, Goethe. Cambridge: Cambridge University Press.
O’Flaherty, James C. 1952. Unity and Language: A Study in the Philosophy of Johann Georg
Hamann. Chapel Hill: University of North Carolina Press.
O’Flaherty, James C. 1967. Hamann’s Socratic Memorabilia: A Translation and a
Commentary. Baltimore: The Johns Hopkins Press.
O’Flaherty, James C. 1979. Johann Georg Hamann. Boston: Twayne Publishers.
O’Flaherty, James C. 1988. The Quarrel of Reason with Itself: Essays on Hamann, Michaelis,
Lessing, Nietzsche. Columbia: Camden House.
Piske, Irmgard. 1989. Offenbarung, Sprache, Vernunft. Zur Auseinandersetzung Hamanns
mit Kant. Frankfurt am Main: Regensburg.
Schmidt, James (ed.). 1996. What is Enlightenment? Eighteenth-Century Answers and
Twentieth-Century Questions. Berkeley and Los Angeles: University of California Press.
Smith, Ronald Gregor. 1960. J.G. Hamann, 1730–1788. A Study in Christian Existence.
With Selections from his Writings. London: Collins.
Sparling, Robert. 2008. Johann Georg Hamann and the Enlightenment Project. Toronto:
University of Toronto Press.
CHAPTER THIRTY
Kant and External

World Skepticism
GEORGES DICKER
1 INTRODUCTION
There are many varieties of philosophical skepticism—skepticism about induction,
about other minds, about morality, and more. But the brand of skepticism that has
most preoccupied philosophers, at least since the time of Descartes, is skepticism
about the external world. Ever since Descartes suggested that, for all we can tell
from our perceptual experiences, all the things we seem to perceive by our senses
may be parts of an all-encompassing hallucination foisted on us by God or by some
powerful evil deceiver, philosophers have wrestled with the question of how we
can know that we live in a world populated by objects that exist in space and time
independently of our minds.
Among these philosophers, Immanuel Kant (1724–1804) occupies an important
but ambiguous position. The ambiguity stems mainly from his notorious doctrine
of transcendental idealism, which, if taken literally, implies that space and time are
nothing but forms imposed by our minds on what we perceive, and thus that at
least some very basic properties of so-called physical objects have no reality apart
from our minds. But even if Kant’s doctrine is interpreted less literally (for example
as holding, à la Peter Strawson, that we cannot make intelligible to ourselves the
notion of an experience that is not in time and space)—as I would urge it should be
if we seek to find some truth in it—problems of interpretation remain. Perhaps chief
among these is how to interpret the crucial section of Kant’s Critique of Pure Reason
(hereafter CPR) commonly referred to as “the Transcendental Deduction.”1 Not only
did Kant completely rewrite this section for the Second Edition of CPR, but neither
version makes it clear whether Kant’s argument is “regressive” or “progressive.”
According to the regressive interpretation, the argument begins by assuming that
we have knowledge of the external world and seeks to explain how this knowledge
is possible. According to the progressive interpretation, the argument attempts to
prove that we have knowledge of the external world, by deriving that knowledge
from some uncontroversial premise about purely subjective experience. Aside from
the vexed matter of textually establishing either of these interpretations, both face
420 SKEPTICISM
difficulties. The regressive interpretation seems to rob the argument of any power
to defeat an external world skeptic; for if the argument starts by assuming that
we have knowledge of the external world, then obviously it cannot prove that we
have such knowledge. The progressive interpretation faces the very difficult task
of constructing a convincing argument, from some slender premise about purely
subjective experience, to the weighty conclusion that we know the existence of the
physical world.2
There is one section of CPR, however, where at least the question of whether Kant
is arguing progressively or regressively is easy to answer, and where his argument is
detachable from (some would even say incompatible with) transcendental idealism.
This is the famous section entitled “Refutation of Idealism,” which Kant added to
the Second Edition. In announcing this new section, Kant famously declared that
it . . . remains a scandal of philosophy and universal human reason that the
existence of things outside us . . . should have to be assumed merely on faith, and
that if it occurs to anyone to doubt it, we should be unable to answer him with a
satisfactory proof. (B xxxix note)3
Kant then proceeded, clearly taking on the progressive task, to argue that “the mere,
but empirically determined, consciousness of my own existence proves the existence
of objects in space outside me” (B 275–276). This chapter expounds and defends
Kant’s argument. It draws on my previous work, which is strongly influenced by the
writings of Paul Guyer and Jonathan Bennett.4
2 KANT’S ANSWER TO SKEPTICISM IN THE

CRITIQUE OF PURE REASON
The key passage of the Refutation in CPR is this:
I am conscious of my existence as determined in time. All time-determination
presupposes something persistent in perception. But this persisting element
cannot be an intuition in me. For all the determining grounds of my existence
that can be encountered in me are representations; and as such they themselves
need something persisting distinct from them, in relation to which their change,
and thus my existence in the time in which they change, can be determined. Thus
the perception of this persistent thing is possible only through a thing outside me
and not through the mere representation of a thing outside me. Consequently, the
determination of my existence in time is possible only by means of the existence
of actual things that I perceive outside myself. (B 275)5
The opening premise that “I am conscious of my existence as determined in time”
requires explanation. For Kant agrees with David Hume that it is impossible to find
a persisting self in introspection; as Kant says: “inner sense . . . gives no intuition of
the soul itself, as an object” (A 22/B 37); “the consciousness of oneself in accordance
with the determinations of our state in internal perception is merely empirical,
forever variable; it can provide no standing or abiding self in this stream of inner
appearances” (A 107). I shall take Kant’s premise to mean, as suggested by the third
KANT AND EXTERNAL WORLD SKEPTICISM 421
sentence of his proof, that I am aware that I have representations or experiences that
occur in a specific temporal order. For these experiences are the sole “determining
grounds of my existence”; it is precisely and only by being aware of them, as they
succeed each other in time, that I can be said to be aware of my own existence and
its continuation in time.
In line with this interpretation of Kant’s premise, we can formulate his proof
this way:
(1) I am aware that I have experiences that occur in a specific temporal order.
(2) I can be aware of having experiences that occur in a specific temporal
order only if there is some persisting element by reference to which I can
determine their temporal order.
(3) No conscious state of my own can serve as this persisting frame of reference.
(4) Time itself cannot serve as this persisting frame of reference.
(5) If (2) and (3) and (4), then I can be aware of having experiences that occur in a
specific temporal order only if I perceive persisting objects in space outside me
by reference to which I can determine the temporal order of my experiences.
Therefore,
(6) I perceive persisting objects in space outside me by reference to which I can
determine the temporal order of my experiences.
Kant does not state premise (4), but the argument requires it in order to rule out the
possibility that this premise rejects, and it is a premise that he insists upon, on the
grounds that “time itself cannot be perceived” (A 166/B 207, A 177/B 219, A 182/B
225, A 183/B 226, A 211/B 257).
This argument, however, is open to a serious objection. The temporal order of
experiences mentioned in (2) is simply the order in which I have those experiences
or “representations.” So why should I need any “persisting element” or enduring
thing in order to know what this order is? As Guyer puts this objection:
It remains unclear why anything more than mere acquaintance with
representations which in fact succeed one another in otherwise uninterpreted
experience, or anything other than the mere occurrence of such representations,
should be necessary for one to judge that there has been such a succession. . . .
Why is the successiveness of consciousness insufficient for its own recognition?
(Guyer 1987: 285–286, 298)
Nevertheless, Guyer thinks that we can extract a successful version of the argument
from the following “Reflexion,” written in 1790, after the appearance of the Second
Edition of CPR (1787), and contained in Kant’s notes, the Handschriftlicher Nachlass:
Since the imagination (and its product) is itself only an object of inner sense,
the empirical consciousness (apprehensio) of this condition can contain only
succession. But this itself cannot be represented except by means of something
which endures, with which that which is successive is simultaneous. This enduring
thing, with which that which is successive is simultaneous . . . cannot in turn be a
422 SKEPTICISM
representation of the mere imagination but must be a representation of sense, for

otherwise that which lasts would not be in the sensibility at all. (Kant 1902: vol.
18, Reflection 6313, p. 614, as translated by and quoted in Guyer 1987: 305)
Guyer explains the basic idea of this passage as follows:
The starting point of this argument is clearly that the mere occurrence of
a succession of representations is not sufficient for the representation or
recognition of this succession. But Kant’s further claim that such recognition
can be grounded “only on something which endures, with which that which
is successive is simultaneous” can only mean that successive representations in
one’s own experience can be judged to be successive only if they are judged to
be severally simultaneous with the severally successive states of some enduring
object. (Guyer 1987: 306)
In what follows, I reconstruct and defend this version of Kant’s argument.
3 KANT’S ANSWER TO SKEPTICISM AFTER THE

CRITIQUE OF PURE REASON
We each have what might be called an “experiential history”—a series of subjective
experiences or conscious states that stretches back in time. Now ask yourself: what
enables you, right now, to set many of your past experiences in their proper time
order, thereby giving you knowledge of your own past? After all, the only thing you
have to go on right now is your present memory of the experiences, which have
come and gone. Furthermore, recollections of your earlier experiences, considered
purely as subjective conscious states or “seemings,” need not be fainter or come with
a greater feeling of “pastness” than recollections of your more recent ones; a fortiori
the recollected members of a series of increasingly temporally remote experiences
need not be increasingly fainter or possess a greater sense of pastness. What then
enables you to know the order in which they occurred? Kant’s answer, according to
Guyer, is that you correlate the remembered experiences with successive states of an
enduring reality that exists independently of the experiences—you take them to be
“severally simultaneous with the severally successive states of some enduring object.”
Here is a schematic way of putting the basic argument. Suppose that you have
had two successive experiences, E1 that occurred at time t1, and E2 that occurred
at the later time t2. Kant takes it as a datum that at a still later time, t3, you can
know that you had E1 before E2. But he sees that there is a question about how you
know this. For at t3, both E1 and E2 have come and gone, so that you cannot then
tell by perception whether they occurred in the order E1, E2 or in the order E2, E1.
In other words, your present state at t3—call it E3—does not in itself contain E1
or E2, but only the memories of E1 and of E2. But these memories could represent
these states as having occurred in either order. What then prevents you from now
thinking that they occurred in the order E2, E1 rather than E1, E2? Only that
you correlate E1 with an objective state of affairs S1 that occurred before another
objective state of affairs S2.
There is a possible misunderstanding that can obscure the force of this

argument. It may seem that the argument assumes that knowledge of one’s own
past experiences rests on memory, but that one has some kind of direct access,
unmediated by memory, to past objective states of affairs. But since any knowledge
of the past must rest on memory, the argument may seem to be based on a false
assumption. But this is not so. Rather, the argument’s core idea is that one set of
memories, namely, memories of the order of one’s subjective experiences, rests on
another set of memories, namely, memories of the order of successive states of an
enduring reality. This idea should not be confused with saying that any knowledge
of time orders based on memory depends on some other knowledge of time orders
also based on memory. Such a view would lead to an infinite regress, and it would
entail that no knowledge of time orders based on memory is possible.6 Rather, the
idea is that knowledge by memory of the time order of one’s subjective experiences
depends on knowledge by memory of the time order of states of an enduring reality.
The reasons for this dependence are complex; they include that the physical world
changes in regular, “clocklike” ways, and that objects in it can be re-encountered,
regardless of these changes, after intervals of time during which they were not
perceived.7 By contrast, subjective experiences are fleeting and transitory; once
gone they cannot be re-encountered, and even if they had certain clocklike features,
the mere having of these features need not be recorded in memory (more about
this later).
It may be objected, however, that sometimes one can remember the order of
one’s experiences “straight off,” with no appeal to an objective world. Indeed,
Jonathan Bennett has identified three types of cases in which this seems to be true:
(a) If Y occurred so soon after X that one can recall a specious present containing
both, then one can simply recall that X preceded Y. If this were not so, one
could not simply recall hearing someone say “damn” rather than “mad.” (b)
From this it follows that one can simply recall that X preceded Y if one can
recall a continuous sequence of happenings starting with X and ending with
Y . . . . (c) One may simply recall that X preceded Y by recalling a time when one
experienced Y while recalling X. (Bennett 1966: 228–229)8
As Bennett immediately goes on to note, however,

These counter-examples to Kant’s thesis cover only a small fragment of all the
temporal orderings we wish to establish. If I am asked “Did you clean your shoes
before or after you went for a walk?”, I may be able—in the manner of (a)—
to recall the moment when I straightened my back from shoe-cleaning, put the
brush down and strode off into the street; or (b) to relive the detail of my day
from the shoe-cleaning episode through to the walk two hours later; or (c) to
recall thinking, while on my walk, that it had been a mistake to clean my shoes
before going out in the mud. But it is far more likely that my answer will have
to be based on my recollection that I cleaned my shoes while the sun rose and
walked as it was setting, or that I heard the one o’clock news while cleaning my
shoes and arranged to go for a walk at three o’clock, or something else equally
424 SKEPTICISM
dependent for its relevance on the truth of . . . statements about the objective
realm. It would be—to put it mildly—a queer personal history which could be
ordered solely in the manner of (a), (b) and (c). (Bennett 1966: 228–29)
The argument, nevertheless, should be formulated so as to allow for such relatively
rare cases, which I shall call “B-related experiences” (short for “Bennett-related
experiences”).9 I offer the following reconstruction:
(1′) I can correctly determine the order in time of experiences of mine that are
not B-related, that is, that did not (a) occur within a specious present, and
that do not (b) belong to a remembered sequence of continuous experiences,
or (c) include a recollection of the sequence of all the earlier experiences
within each of the later ones.
(2′) When I remember two or more past experiences that are not B-related, my
recollection of those experiences does not itself reveal the order in which
they occurred.
(3′) If (2′), then I cannot correctly determine the order of two or more past
experiences of mine that are not B-related just by recollecting them.
(4′) I cannot correctly determine the order of two or more past experiences of
mine that are not B-related just by recollecting them. [from (2′) and (3′)]
(5′) If I cannot correctly determine the order of two or more past experiences
of mine that are not B-related just by recollecting them, then I can correctly
determine their order only if I know that at least some of my experiences are
caused by successive objective states of affairs that I perceive.
(6′) I can correctly determine the order of two or more past experiences of mine
that are not B-related only if I know that at least some of my experiences
are caused by successive objective states of affairs that I perceive. [from (4′)
and (5′)]
(7′) I know that at least some of my experiences are caused by successive
objective states of affairs that I perceive. [from (1′) and (6′)]
The term “objective” introduced in (5′) begs no questions, since (5′) spells out a
necessary condition for correctly determining the order of my past experiences, not
just for assigning some order or other to them. Note, however, that (1′) does not
mean that for any two experiences of mine that are not B-related, I can correctly
determine their order; the premise allows that there may be cases in which I am
unable to make this determination. Even so, a memory skeptic would of course
object to (1′); I address this point immediately below.
4 OBJECTIONS AND REPLIES

Critics have raised several interesting and challenging objections to the above
reconstruction, and I have offered replies.10 There is not enough space to review
this debate here, so I shall consider only two possible objections.11
The first is that a philosopher who adopts a radically skeptical stance, as

Descartes provisionally does in his First Meditation, would not grant the argument’s
first premise, for she would doubt the reliability of memory. Now it is true that
Kant intended his Refutation of Idealism to refute what he called “the problematic
idealism of Descartes” (B 274). It is also true that in one sentence of his Second
Meditation, Descartes supposes that perhaps his memory always deceives him.
Nevertheless, Descartes does not seriously doubt memory. Indeed, the reliability
of memory must be assumed in order for his most severe doubt even to arise, since
that doubt concerns whether conclusions that are correctly remembered to have
been properly deduced from self-evident premises are true (see Frankfurt 1970:
160–62). In any case, Kant’s argument is not intended to refute a memory skeptic,
any more than Descartes’s attempted proof of the physical world in his Sixth
Meditation. The doubt that Kant is out to conquer in the Refutation of Idealism is
not about the reliability of memory; it is a doubt arising from the contrast between
the inner and the outer, appearance and reality, the subjective and the objective.
This is the only kind of doubt that Kant calls “problematic idealism,” and the
sole purpose of his argument is to refute it, by showing that “even our inner
experience, undoubted by Descartes, is possible only under the presupposition of
outer experience” (B 275).
The second possible objection, which targets premise (5) of the argument, is
that I could determine the order of my non-B-related experiences by reference to
certain “clocklike” features of my purely inner experience. Katherine Gasdaglis
puts it well:
Why . . . couldn’t there be some persisting “clocklike” feature of our conscious
experience that could be used to keep track of the temporal order of our states?
Consider a variation on Strawson’s famous example of the “master sound.”
Imagine a person who, from birth until death, hears a persistent tone that
gradually increases in pitch throughout her life. She could simply refer to the
sound’s pitch, a purely subjective feature of her experience, as a temporal marker
of her representational states. She could recall, for example, that she saw the
movie at a B and parked her car at an F. The possibility of clocklike features
constitutes a counter-example to Dicker’s premise (5). His only response is that,
luckily, there are in fact no such features. (Gasdaglis 2013: 25–26)
In a similar vein, Andrew Chignell writes:
Suppose there is a digital clock in the corner of your visual field (it remains when
you close your eyes), or a “voice in your head” audibly counting off seconds
throughout your conscious life . . . . The fact that, by hypothesis, these “clock-
like” features of your experience count off the seconds in proper order is all
that is required for them to serve as the measure for your other remembered
states . . . . No external objects are necessary. (Chignell 2011: 185)
Let us consider these two scenarios in turn. Even from a common sense, “empirical”
point of view, the rising tone scenario is too fantastic to pose a real difficulty. Human
426 SKEPTICISM
lives normally last for years, and a tone that perceptibly rises in pitch from birth to
death of a human being, without ever “reversing” course, is quite unimaginable.
Further, in order for such a rising tone to serve as a temporal marker of all of
one’s past experiences, it would have to be divisible into numberless distinguishable
sounds, each of which would have to be remembered as a distinct tone, and in the
right order. This is not possible for any human being.
There are other difficulties as well. Suppose that at time t, I remember hearing
both a high note and a low note: can I tell, solely on the basis of my memory of
those tones, which came first? In some cases, the answer is yes. For example, if
the high note and the low note are part of a melody that I have been continuously
following up to t, then these are B-related experiences, since I remember an entire
sequence of sounds rising from the low note to the high one. But suppose that I
do not hold in memory the entire sequence of tones between the low note and
the high note. Here it is important to notice that the fact that the sequence did in
fact rise that way, and that I heard each segment of the sequence as it occurred,
does not show that I now remember the entire rising sequence. The hypothesis of
a gradually rising sound may encourage one to think that the memory was thus
unbroken, but in the case that I am envisaging, which is both logically possible and
realistic, the memory is not continuous. When I remember hearing the high note, I
still also remember hearing the low note, but I can’t decide whether the tones that
separated them, which I heard but perhaps paid scant attention to (maybe I was
struggling to suppress a cough at a concert) were gradually rising, or rising and
falling irregularly. How then might I know that the high note came after the low
note, rather than vice versa? Well, I might know it by remembering that the low
note sounded when the cellos began playing and the high note sounded when the
violins began playing, and that the cellos made their entrance before the violins—
objective facts about a sequence of events that I observed. The key general point
here is that the mere fact that my inner experience had clocklike features does not
show that these features were registered, recorded, or retained in my memory, and
so cannot be appealed to in order to determine the order of the experiences. Rather,
this order must be determined by the remembered order of objective events, even if
the inner experiences had clocklike features.
Consider next Chignell’s scenario. Suppose that there were a digital clock
permanently present in the corner of my visual field. Could I order my past
experiences by reference to this clock? To be able to do that, two conditions would
have to be satisfied. First, the numbers on the digital display must track the passage
of time; it must not be the case, for example, that the number 8:00/1/1/2015 (8:00
A.M. on January 1, 2015) appears before the number 8:00/1/1/2014. Second, I must
know that the numbers on the display track the passage of time. If I have no idea
whether the 8:00/1/1/2014 display annexed to a remembered experience X occurred
earlier or later in time than the 8:00/1/1/2015 display annexed to a remembered
experience Y, then I cannot use those displays to determine the order in which X
and Y occurred. The trouble lies with this second condition: absent perceptible
external objects whose states I can correlate with the numbers in the display, and
given the point that memories of more remote past experiences are not necessarily
fainter than memories of more recent ones, I have no way of knowing that the
order in which the numbers occurred does correspond to the passage of time; my
ability to know this is subject to exactly the same difficulty as my knowledge of the
order in which any of my other experiences occurred. For example, the number
8:00/1/1/2014, for all I can tell purely on the basis of my memory of the display,
may have appeared later than 8:00/1/1/2015. Of course, I know that this reversal
would not happen on a real digital clock in good working condition, but that only
illustrates the need for an objective reality by reference to which I can order my past
experiences. Analogous considerations apply to the voice counting off seconds in
my head.
It might be replied that I could know that my mental clock tracks time without
appeal to external reality, by induction.12 This would work as follows. In cases
of B-related experiences, as we have seen, one can determine the order of one’s
past experiences “straight off.” Suppose then that during stretches of B-related
experiences, my mental clock manifestly tracks time correctly: it always displays a
later time after an earlier one. Then, it might be said, I can infer inductively that it
also tracks time correctly when my experiences are not B-related.
My reply is that such an inductive inference could at best show that the clock
always tracks time correctly when our experiences are B-related. Since we necessarily
can never observe it tracking time when two experiences are separated by not
being B-related, we do not have a sample of similar cases from which we can infer
inductively that the clock tracks time during cases of the same kind. Perhaps there
is a psychological mechanism such that the clock “engages,” so to speak, only for
experiences that are B-related; for other remembered experiences it just “runs
freely.” Perhaps there is a constant conjunction between (a) only the clock’s readings
and the awareness of temporal succession but not between (b) those readings and
succession in time per se. I do not see by what good inductive argument (b) could be
shown to be less probable than (a).
The reason why the rising tone and the digital clock scenarios may seem
plausible is that the “internality” of the auditory experiences and of the mental
clock, combined with the hypotheses of a gradually rising sound and progressively
increasing numerical sequence, encourages us of think of the experiences as being
B-related. If we had a wholly B-related sequence of internal experiences from birth
to death, then presumably we could use it to order our other experiences. But
human experience is manifestly not like that. Now although some Kant scholars
would disagree, I believe that Kant is concerned to state conditions under which
we can order our experiences, given the kind of experience humans actually and
unquestionably have. Like other so-called transcendental arguments in Kant’s
philosophy, the Refutation of Idealism starts from certain very general though
contingent premises about the kind of experience that humans unquestionably have,
and argues that such experience is possible only under certain conditions. If all the
premises of Kant’s argument had to be necessary truths, then its conclusion that
I know that some of my experiences are caused by successive objective states of
affairs that I perceive would have to be a necessary truth as well. I see no textual or
philosophical reason to saddle Kant with such an implausible view.
428 SKEPTICISM
NOTES
1. In addition to this problem, Kant’s multifaceted approach has given rise to
widely divergent interpretations of both the nature of his response to external world
skepticism and its importance within his philosophy as a whole. For example, Caranti
(2007) argues that while Kant was deeply concerned from the beginning of his career
to refute the sort of skepticism evinced in Descartes’s evil-demon scenario, his most
sustained and strongest answer to it is not to be found, as will be assumed here,
in his “Refutation of Idealism,” but rather in the section of CPR entitled “Fourth
Paralogism of Ideality” and, contrary to the view just implied, that it presupposes
a robust version of Transcendental Idealism, found in Kant’s account of time and
(especially) space in the section entitled “Transcendental Aesthetic.” By contrast,
Forster (2008) argues that Kant was only marginally concerned with external world
skepticism, and much more deeply interested in developing a chastened metaphysics
in the wake of Pyrrhonian skepticism, Humean Skepticism, and his own Antinomies.
For some critical discussion of Caranti, see Guyer (2007); for some critical discussion
of Forster, see Dicker (2010).
2. I defend a progressive reading of both the first edition and the second edition versions
of the Deduction in Dicker (2004: 84–144; 2013: 410–425).
3. In all references to CPR, I use “A” to refer the first edition and “B” to refer to the
second edition; thus A 50 would refer to page 50 of A, B 50 to page 50 of B, and
A 50/B74 to page 50 of A and page 74 of B.
4. See Dicker (2004: 24–25, 194–211; 2008: 80–108; 2011: 175–183; 2012:
191–195; 2013: 410–431, esp. 425–429). Parts of this chapter include material from
these works, especially from Dicker (2008; 2012; 2013).
5. The third and fourth sentences were substituted by Kant, in a footnote at B xxxix, for
a single sentence in the original B 275 text of the “Refutation of Idealism” that reads:
“This persistent thing, however, cannot be something in me, since my own existence
in time can first be determined only through this persistent thing.”
6. James Van Cleve alerted me to this point in correspondence.
7. The need for re-encounterability stems from this: Given that remembered experiences
can be ordered only by reference to something with “clocklike” features, this item
must be something that has preserved its identity over time. For it must be something,
call it X, that has undergone some kind of change or alteration, and by reference
to whose successive states we can date or determine the positions in time of the
remembered experiences. But if X ceases to exist before it can be re-encountered, so
that anything we encounter after encountering X must be an item with a different
beginning of existence than X, then it cannot be numerically identical with X, since
nothing can have two beginnings of existence. So, the item we then encounter cannot
be something that has merely undergone some kind of change or alteration while
remaining identical with X, and therefore it cannot serve as something by reference
to whose successive states we can determine the positions in time of our remembered
experiences (or of anything else).
8. Bennett numbers these possibilities as (1), (2), (3); I have substituted (a), (b), (c) to
avoid confusion with the numbering of the steps in the argument.
9. This label was suggested by Mark Brown in discussion.

10. For example, for a reply to the possible objection that “correlated with,” rather than
“caused by,” should be used in (5) and (6), see Dicker (2008: 94–95).
11. For a much fuller discussion of possible objections and replies, see Chignell (2009;
2011) and Dicker (2008; 2011; 2012).
12. Chignell suggested this in conversation.
REFERENCES
Bennett, Jonathan. 1966. Kant’s Analytic. Cambridge: Cambridge University Press.
Caranti, Luigi. 2007. Kant and the Scandal of Philosophy: The Kantian Critique of Cartesian
Scepticism. Toronto: University of Toronto Press.
Chignell, Andrew. 2009. “Causal Refutations of Idealism,” Philosophical Quarterly 60:
487–507.
Chignell, Andrew. 2011. “Causal Refutations of Idealism Revisited,” Philosophical
Dicker, Georges. 2004. Kant’s Theory of Knowledge: An Analytical Introduction. New York:
Dicker, Georges. 2008. “Kant’s Refutation of Idealism,” Noûs 47: 80–108.
Dicker, Georges. 2010. “Review of Kant and Skepticism, by Michael Forster,” European
Journal of Philosophy 18: 609–615.
Dicker, Georges. 2011. “Kant’s Refutation of Idealism: A Reply to Chignell,” Philosophical
Dicker, Georges. 2012. “Kant’s Refutation of Idealism: Once More Unto the Breach,”
Kantian Review 17: 191–195.
Dicker, Georges. 2013. “Transcendental Arguments and Temporal Experience.” In
A. Bardon and H. Dyke (eds.), A Companion to the Philosophy of Time, 410–431.
Oxford: Wiley-Blackwell.
Frankfurt, Harry. 1970. Demons, Dreamers, and Madmen: The Defense of Reason in
Descartes’s Meditations. New York: Bobbs-Merrill.
Gasdaglis, Katherine. 2013. “The Refutation of Idealism and the Priority of Perception.”
Unpublished manuscript.
Guyer, Paul. 1987. Kant and the Claims of Knowledge. Cambridge: Cambridge University
Press.
Guyer, Paul. 2008. “Kant and the Scandal of Philosophy: The Kantian Critique of Cartesian
Scepticism,” Philosophy and Social Criticism 34: 825–836.
Kant, Immanuel. 1997. Critique of Pure Reason, translated and edited by P. Guyer and
A. W. Wood. Cambridge: Cambridge University Press.
Kant, Immanuel. 1902. Kant’s Gesammelte Schriften, 29 vols. Berlin: Walter de Gruyter.
CHAPTER THIRTY-ONE
Hegel: Philosophy as a
Kind of Skepticism
DIETMAR H. HEIDEMANN
1 INTRODUCTION
Hegel is not a skeptic, although he believes philosophy to be a kind of skepticism.
This view looks paradoxical, but it isn’t. Hegel, as many philosophers, holds that
skepticism can take different forms. According to his account of philosophical
doubt, there are untrue, misconceived forms of skepticism, that is, ordinary versions
of skepticism according to which we cannot know whether our beliefs are true,
and there is genuine skepticism that coincides with true philosophy itself. True
philosophy realizes that skeptical doubt is to be conceived as an integral part of
rationality and that, although skepticism as such cannot be overcome, it is not flying
in the face of reason either, since genuine skepticism has the positive function of
productively contributing to true philosophical cognition.
Already in his infamous essay On the Relation between Skepticism and Philosophy
(1801), Hegel distinguishes “skepticism that is one with philosophy” from
“skepticism that is sundered from it.” The latter “can be divided into two forms,
according to whether it is or is not directed against Reason” (Hegel 2000: 330,
translation modified). For Hegel, “genuine skepticism” is “skepticism that is one
with philosophy” because this form of skepticism represents, in his terminology, the
“negative side” that each true philosophy has in that it is “directed against everything
limited” or “the whole soil of finitude” (2000: 322–323). Genuine skepticism is all-
comprising and hence not limited to doubts about the existence of the external
world, sense perception, or human reasoning in general. It also has speculative
reason or metaphysics as its object but, unlike skepticism that is sundered from
philosophy, not as its target. Unlike ordinary kinds of skepticism, genuine skepticism
does not just call into doubt arbitrary epistemic claims but takes skeptical doubt
to systematically demonstrate that certain types of cognition are misconceived in
principle. Further, according to Hegel, the systematic skeptical proof that certain
types of cognition are necessarily untrue belongs to the idea of what true philosophy
consists in because in the end skeptical doubt contributes to the constitution of the
true as a whole. In what follows, I present the arguments Hegel puts forward in
HEGEL: PHILOSOPHY AS A KIND OF SKEPTICISM 431
favor of this rather challenging account of skepticism. In Section 2, I discuss the

celebrated conception of “self-fulfilling skepticism” of the Phenomenology of Spirit
(1807) that is supposed to overcome untrue types of cognition in order to promote
“absolute knowing.” In Section 3, I debate Hegel’s more advanced view according
to which genuine skepticism must be construed as dialectic.1
2 SKEPTICISM AND HISTORY

OF SELF-CONSCIOUSNESS
The Phenomenology of Spirit, which Hegel conceives as a systematic introduction
to metaphysics as true philosophical science, deals with the problem of justifying
true knowledge. The project of justifying true knowledge is directly connected with
skeptical doubts because epistemic justification presupposes a criterion by means
of which it can be decided whether knowledge is justified; however, for the skeptic
the criterion itself is questionable. Although theories of epistemic justification like
coherentism or contextualism claim to be able to do without a criterion and thereby
to dissolve skepticism, Hegel, in the Phenomenology at least, explicitly affirms that
a criterion of justification is indispensable. The problem of the non-justifiability
of the criterion originates in ancient (Pyrrhonian) skepticism, which for Hegel is
substantially more sophisticated than modern skepticism (see Hegel 1969–1971:
XIX 358–403; Forster 1989: 9–35; Heidemann 2007: 117–198). The Pyrrhonist
Agrippa models it as a skeptical trilemma. The general argument runs as follows:
the logical “criterion of truth” serves as “the standard regulating belief in reality or
unreality”; however, it is unprovable. For to settle the philosophical dispute over
the existence of a criterion of truth, a criterion is required to determine whether
there is one or not. Now, this criterion can
(i) be proven in a circular way because the proof of a criterion already requires
an accepted criterion by means of which it can be proven; or
(ii) be a dogmatic “assumption” that involves simply presupposing its
existence; or
(iii) be proven by another criterion, which in turn will be proven by another one,
and so on ad infinitum. (cf. Sextus Empiricus 1933: I 21–24, II 18–21)
In any of the three cases, the justification of the criterion necessarily fails. Hegel,
who was familiar with the Pyrrhonists’ writings, claims in the Phenomenology to
have found a solution to it. In particular, he thinks that he has developed a procedure
that justifies the criterion without making unjustified external presuppositions.
According to Hegel, the aporia of the criterion, or “standard” (Maßstab) as he most
frequently calls it, arises because the “investigation and examination of the reality
of cognition” cannot manage “without some presupposition that can serve as its
underlying criterion” (1977: sec. 81). From this there results a “contradiction”
(1977: sec. 82) or an aporia. Hegel argues as follows:
An examination consists in applying an accepted standard, and in determining
whether something is right or wrong on the basis of the resulting agreement
432 SKEPTICISM
or disagreement of the thing examined; thus the standard as such (and Science
likewise if it were the criterion) is accepted as the essence or as the in-itself (an
sich). But here, where Science has just begun to come on the scene, neither
Science nor anything else has yet justified itself as the essence or the in-itself;
and without something of that sort it seems that no examination can take place.
(1977: sec. 81)
This version of the aporia essentially resembles Sextus Empiricus’s approach, even
though Hegel does not explicitly mention the Pyrrhonian background here (cf.
Claesges 1981: 68–74 and 77–96). Hegel’s solution to the problem is based on
the analysis of epistemic consciousness. It can be outlined as follows: in order to
prevent the unjustified presupposition of the criterion, in the introduction to the
Phenomenology of Spirit, Hegel introduces the concept of Science as it “comes on the
scene” (1977: sec. 76). Science is characterized at this stage first as an “appearance,”
that is, a form of cognition emerging alongside other epistemic claims. Secondly, as
an “emerging” science, it is “not yet Science in its developed and unfolded truth.”
As this nascent Science competes with other concepts of cognition, due to skeptical
equipollence it cannot simply declare its superiority since this would be nothing but a
dogmatic “assurance” (1977: sec. 76). Therefore, Hegel’s strategy is to show that the
emerging Science develops into true Science by turning against the mere appearance
of knowledge and thereby abandoning its provisional status in order to overcome
untrue modes of cognition. However, to the extent that the “exposition of how
knowledge makes its appearance” (1977: sec. 76) is nothing but the “examination of
the reality of cognition” (1977: sec. 81), the Phenomenology obviously cannot avoid
relapsing into the aporia, since that examination presupposes a “standard” of its
own. In other words, Science must provide the standard of truth although it is just
about to emerge and is not yet true and real, that is, justified. Hegel now argues that
a standard of justification is indispensable but cannot be dogmatically presupposed
as an external criterion. It therefore must fall into the sphere of consciousness itself.
The standard is an internal factor of consciousness, for “truth” and “knowledge”
prove to be “abstract determinations” that “occur in consciousness” itself (1977:
sec. 82). Examination of the relation between these determinations yields the insight
that “[c]onsciousness provides its own criterion from within itself” by performing
nothing more than a “comparison” with itself (1977: sec. 84). Hegel conceives of
this as the self-creation of the standard for the distinction between knowledge and
truth, between that which is for consciousness and that which is in itself outside the
relation, for that very distinction is one that is made within the epistemic structures of
consciousness itself. Whenever consciousness “knows” an object, this object is both
for it and at the same time in itself or true, since consciousness regards the object as
given outside of the cognitive relation. The in itself, however, is a determination that
consciousness “affirms” within its knowledge. For Hegel these epistemic structures
generate the standard: “Thus in what consciousness affirms from within itself as
being-in-itself or the True we have the standard that consciousness itself sets up by
which to measure what it knows” (1977: sec. 84).
Thus, according to Hegel’s solution to the skeptical trilemma, the examination of
knowledge consists in a comparison of consciousness with what it takes to be true.
If consciousness corresponds to what it takes to be true, that is, if it is in accordance

with itself, then its knowledge meets the standard of true knowledge, produced by
itself. This stage is achieved by consciousness at the level of true self-consciousness
or subjectivity, that is, “absolute knowing.” In the course of its development toward
fully fledged subjectivity, consciousness is already following the path of science
and hence does not have to presuppose an external standard. It is the “history of
self-consciousness” in terms of self-fulfilling skepticism by means of which Hegel
explains how consciousness proceeds through that development.
In the Phenomenology, Hegel conceives of the “history of self-consciousness” as
the “history of the education of consciousness itself to the standpoint of Science”
(1977: sec. 78).2 The purpose of the history of self-consciousness is not to describe
the empirical coming into being of self-consciousness on the basis of the natural,
cultural, or social development of mankind. Rather, the conception of the history
of self-consciousness, as Fichte and Schelling have developed it, is about the ideal
or systematic genesis of cognitive capacities in the human mind. The idea is to
show how the human mind develops such capacities starting with primitive forms
of epistemic consciousness or knowledge, which in the sequel is enriched by more
complex forms until they finally culminate in the fully developed self-conscious
cognitive subject.
The Phenomenology is arranged as an interconnected “series of forms” of
consciousness (1977: sec. 79). Hegel presents it as a development of epistemic
claims. A form of consciousness is an idealized epistemic shape or structure of
consciousness within a specific field like sense certainty, perception, reason, or
spirit. The exposition of forms of consciousness is not, however, about the way
consciousness has sense-impressions, perceptual representations, etc. Rather it deals
with the specific epistemic claims involved in sense certainty, perception, etc., and the
question whether consciousness meets these claims so as to acquire true cognition.
According to this method, each form of consciousness must have its own standard
that makes it possible to decide whether an epistemic claim is satisfied. For example,
the standard of sense certainty as a form of consciousness is “immediacy” (1977: sec.
90–93), that is, non-inferentiality is the standard that natural consciousness sets up as
the criterion for the examination of the truth of its knowledge claims. So at the level
of sense certainty, consciousness examines whether its epistemic claim, that is, beliefs
based on sense-impressions, meet the standard of immediacy. The examination of
this form of consciousness then proves that sense certainty does not meet its standard
since it turns out to be inferential knowledge after all and hence, according to its own
standard, one of “the forms of the unreal consciousness” (1977: sec. 79). Thus sense
certainty as the form or claim of immediate knowledge has to be given up, because
all knowledge based on sense-impressions is conceptually mediated. The standard
as such is not just an external presupposition since it is intrinsic to each form of
consciousness and set up by consciousness itself.3
This seems to be a legitimate conception to the extent that it is possible to typify
different classes of epistemic claims according to different criteria of epistemic
justification. However, the systematic interconnection of these claims in the
Phenomenology does generate a problem. According to Hegel, with the exception of
434 SKEPTICISM
absolute knowing, no shape of consciousness meets its standard as the fully developed
form of self-consciousness or self-knowledge. However, the series of the “formative
stages” (1977: sec. 28) of consciousness is not arbitrary but is said to be complete as
well as necessary (1977: sec. 29, 77–80). Hegel tries to guarantee this on the basis
of the following argument: the skeptically motivated non-satisfaction of a specific
standard not only leads to a modification of the epistemic claim of consciousness but
also “the criterion for testing is altered,” so that the examination of consciousness
is also an examination of “the criterion of what knowledge is” (1977: sec. 85).
From this alteration, Hegel claims, “the new true object issues” by a “dialectical
movement that consciousness exercises” (1977: sec. 86). This new object is the new
epistemic claim of consciousness, including its new standard. That is to say, with the
exception of sense certainty as the first form of consciousness, each of the ensuing
forms of consciousness necessarily (logically) follows from the preceding one and
hence always contains what consciousness has learned from its previous shape. By
this “historical” or, to be more precise, developmental and skeptical process, natural
consciousness continuously develops into true “absolute knowing” or the completed
subject, encompassing the entire experience consciousness has made before.
Although subjectivity constitutes the unity of the entire development and is
thus the thread of this process, from beginning to end, the question of how the
developing consciousness itself can have knowledge of the logical interconnection
and epistemic features of the different stages remains an open one. Indeed, the
developing consciousness does not have such knowledge. This is due to the fact that
the conception of a history of self-consciousness presupposes the methodological
differentiation between the developing consciousness at the first level and the
phenomenological philosopher who establishes the theoretical links between the
forms of consciousness at the second level. The developing consciousness ignores
the logical relations between its different epistemic claims since the theoretical
assessment takes place from the already fully developed philosophical standpoint.
This is the reason why Hegel maintains that the Phenomenology as “the way to
science is itself already Science” (1977: sec. 88) and therefore does not lay claim to
unjustified external standards of epistemic examination.4
Consequently, the history of self-consciousness shows, first, how the standard
is generated by the skeptically self-examining consciousness, secondly, what the
standard is in each particular case, and thirdly, that Hegel’s solution does not come
down to mere subjectivism since the philosophical evaluation of the “pathway”
is “something contributed by us” (1977: sec. 87), the knowing philosopher, and
therefore made from an objective standpoint. Nevertheless, this argument as a
whole can be reproached with circularity because the conclusion, the standpoint of
philosophical truth, is a constitutive element needed to make sense of the premises,
that is, the history of self-consciousness. For, on the one hand, the history of self-
consciousness constitutes the skeptical justification of true philosophical knowledge,
but, on the other hand, the theory already makes use of it prior to the completion
of the whole developmental process before it is itself justified. The history of
self-consciousness is forced to make the external presupposition of philosophical
knowledge that explains what is going on within the skeptical process of justifying
the sequence of epistemic claims that consciousness raises. This difficulty might be
the reason why Hegel in his later work further developed his account of skepticism
by accommodating it with skepticism.
3 SKEPTICISM AND DIALECTIC

The place of true philosophy in Hegel’s system is logic conceived as metaphysics.
Unlike the Phenomenology, logic, as depicted in the Science of Logic (1812–1816)
and the Encyclopaedia of the Philosophical Sciences (1817–1830), does not operate
on the basis of self-fulfilling skepticism. In the logic, skepticism rather appears
in connection with dialectic. Hegel makes two crucial claims with respect to the
relation between skepticism and dialectic: (a) “philosophy . . . contains the skeptical
moment as a moment within itself—specifically as the dialectical moment” (1991:
§81) and (b) “skepticism” as “the dialectical moment itself is an essential one in the
affirmative Science” (1991: §78). Both of these claims have to be understood in
light of the problem of “total presuppositionlessness” (1991: §78). This problem is
subsequent to the problem of a systematic introduction to metaphysics or logic. For
Hegel acknowledges the skeptic’s claim that the demonstration of “Science” (1991:
§78), that is, philosophical science, must not make any unjustified presuppositions.
However, employing skeptical doubt as the method of a systematic introduction
to science would be question-begging since “Science” itself is supposed to be self-
reliant, that is, it cannot depend on a preceding justification through skepticism.
Hence, it cannot be legitimate to construe the demonstration of “Science” as a
procedure of skeptical justification. This, of course, means that Hegel can no longer
hold onto the conception of the Phenomenology as a systematic introduction to
metaphysics or “Science.”5
The fact that skepticism, particularly Phenomenology’s “self-fulfilling skepticism,”
cannot perform the function of a systematic introduction to “Science” raises two
questions: First, is there an alternative way of obtaining access to “Science,” a
systematic way that does not violate the claim of “presuppositionlessness”? Second,
if skepticism cannot accomplish the task of a systematic introduction, wouldn’t
skeptical doubt be independent of, and therefore a fundamental threat to, “Science”?
Hegel’s answer to the first question is that we can obtain access to “Science” without
making presuppositions “by the freedom that abstracts from everything, and grasps
its own pure abstraction, the simplicity of thinking—in the resolve to think purely”
(1991: §78, translation modified). It is by no means clear what the “resolve to think
purely” consists in, not least because Hegel does not further explain this concept.
But he seems to believe that this concept represents the appropriate methodological
procedure that enables an individual to get immediate access to the proper content
of “Science” by pure thought. Let us assume, for the sake of argument, that “Science”
can indeed be grasped by this procedure. How then does Hegel respond to the
second question? He clearly denies that independent skepticism is threatening to
“Science” since according to his theory skepticism is contained within philosophy as
“the dialectical moment” (1991: §81). In order to see Hegel’s point, one has to recall
his conception of “Science” or logic as metaphysics. In Hegel’s speculative idealism,
436 SKEPTICISM
logic represents “the pure truth itself” (1991: §19). As such, it contains “objective
thoughts” that “express the essentialities of things,” though not as a collection of
abstract terms but as dialectically evolving “thought-determinations.” Since Hegel’s
logic is therefore to be construed in the traditional terms of ontology, it “coincides
with metaphysics” (1991: §24).
In order to demonstrate that truth can be cognized through pure thought, it first
has to be shown that the one-sided, finite “forms of cognition,” as represented,
for example, through Kant’s transcendental account of cognition, are deficient,
for otherwise logic or metaphysics wouldn’t represent the only possible “Science”
of “the pure truth itself” (1991: §19). This is compatible with Hegel’s idea of
the nullification of the finite “forms of cognition” by means of skeptical doubt.
However, this does not conflict with the view that skepticism cannot function as
an introduction to metaphysics. For here Hegel explicitly refers to the destructive
role of skepticism with regard to the finite “forms of cognition” beyond any kind of
introductory function as in the Phenomenology: “The high skepticism of antiquity
accomplishes this by showing that every one of those forms contained a contradiction
within itself” (1991: §24, Addition 3). As we shall shortly see, “showing that every
one of those forms contained a contradiction within itself” indicates the point of
intersection between dialectic and skepticism.
In On the Relation between Skepticism and Philosophy, Hegel has already argued
that skepticism is innocuous “[s]ince every genuine philosophy has this negative side”
(Hegel 2000: 325). The “negative side” of “every genuine philosophy” is a positive,
systematic component “implicit in every genuine philosophical system.” In Hegel’s
eyes, Plato’s Parmenides along with Spinoza’s Ethics are ideal representatives of this
skeptical, “free side of philosophy,” since these works exhibit antinomies according
to “the principle of skepticism: panti logôi logos isos antikeitai” [to every argument
an equal argument is opposed] (Hegel 2000: 324f).6 The presentation of skepticism
as the “negative side” of true philosophy indicates Hegel’s mature view of positively
integrating skepticism into philosophy by assimilating it with dialectic. By means of
this integrative strategy Hegel intends to immunize metaphysics or, more broadly,
philosophy against skeptical attacks. This connection between skepticism and
dialectic is not at all arbitrary, but rather systematically motivated. For according
to Hegel,
the dialectical, taken separately on its own by the understanding, constitutes
skepticism, especially when it is exhibited in scientific concepts. Skepticism
contains the mere negation that results from the dialectic. (1991: §81)
The connection of skepticism and dialectic is justified by means of “negation,” that is,
in Hegelian terms, through contradiction. However, skepticism is not identical with
dialectic. Dialectic shows the “one-sidedness and restrictedness of the determinations
of the understanding” (1991: §81), and this means their “negation.” This is also
accomplished by skepticism in that the skeptic raises doubts and therefore negates
epistemic claims by opposing them according to the aforementioned principle,
“panti logôi logos isos antikeitai.” Unlike dialectic, skepticism does not go any further
than demonstrating the “negation” of the “determinations of the understanding.”
However, skepticism is capable of recognizing the positive content of its “negation.”

For this reason skepticism and dialectic coincide only inasmuch as both reveal finite
understanding’s intrinsic negativity. Nonetheless, finite understanding cannot be
aware of dialectic itself but exclusively of the skeptical negation of epistemic claims.
Since finite understanding is incapable of reaching dialectical insights, it must accept
skepticism as a fundamental threat to its epistemic claims, for without the help
of dialectical means finite understanding cannot escape from its natural skeptical
doubts and contradictions. By contrast, “genuine philosophy” itself is not affected
by skeptical doubts:
Skepticism should not be regarded merely as a doctrine of doubt; rather, it is
completely certain about its central point, i.e., the nullity of everything finite. . . .
Philosophy, on the other hand, contains the skeptical as a moment within itself—
specifically as the dialectical moment. But then philosophy does not stop at the
merely negative result of the dialectic, as is the case with skepticism. (1991: §81,
Addition 2)
Hence, skepticism and dialectic accord with respect to “negation” while being vitally
distinct at the same time. Although both likewise rely on “negation,” skepticism
performs “abstract negation” by simply doubting, that is, nullifying epistemic
claims, whereas from the point of view of dialectic the “result” of the “negation”
preserves the positive content which it “resulted from.” Here Hegel draws on the
Phenomenology’s distinction between two kinds of negation. Skepticism’s, that is,
finite understanding’s (“abstract”) negation merely “wait[s] to see whether something
new comes along and what it is, in order to throw it too into the same empty abyss”
(1977: sec. 79). By contrast, true, “determinate negation” realizes that,
nothingness is, in a determinate fashion, specifically the nothingness of that from
which it results. For it is only when it is taken as the result from which it emerges,
that it is, in fact, the true result; in that case it is itself a determinate nothingness,
one which has a content. (1977: sec. 79, translation modified)
Consequently, it is the method of counter-position or negation that makes skepticism
dialectical and thereby a positive element of rationality. From the immanent point of
view of human understanding, skeptical counter-position or negation of epistemic
claims has devastating consequences for rationality since finite understanding’s
cognitive performances lead naturally and inescapably to skeptical doubt. For finite
understanding it therefore looks as if the skeptic, by means of counter-position or
negation, is able to prevent the human mind from successfully achieving positive
metaphysical knowledge. Hegel does not deny the negative consequences that
skepticism has for finite understanding. However, he criticizes this destructive view
of skepticism as one-sided. For him the true positive meaning of skepticism is the
“dialectical moment,” that is, skepticism’s ability to nullify finite reflection by means
of skeptical doubt in order to promote all-comprising metaphysical cognition of
the true.
It is important to see that Hegel is not operating with two distinct types of
skepticism, negative skepticism on the one hand and positive philosophical
438 SKEPTICISM
skepticism on the other. On the contrary, negative and positive philosophical

skepticism are identical. It is one and the same skeptical doubt that—on the level
of finite understanding—nullifies reflection while—on the level of dialectic—it
promotes metaphysical knowledge. This distinction itself is, of course, made from
a philosophical point of view, that is, it is an insight that cannot be achieved from
within finite reflection. However, from a dialectical angle, skeptical doubt has to be
taken as the documentation of true philosophical knowledge.
3 PHILOSOPHY, SKEPTICISM, AND EVERYDAY LIFE

Hegel’s account of skepticism is a philosophical challenge in many ways. First
of all, self-fulfilling skepticism claims to be a methodological construction of a
series of skeptical doubts that promotes true philosophical knowledge. Although
one can grant Hegel that he is aware of one of the most fundamental problems
of epistemology, that is, the problem of the criterion, it is hard to see how the
construction of a series of skeptical doubts is able to face this problem. For the
skeptic might still question the objectivity of the systematic connections between
the forms of skeptical doubts that supposedly lead consciousness necessarily to the
standpoint of true cognition. On the other hand, within the systematically evolved
forms of skeptical doubts, Hegel discusses many of the skeptical arguments that
are still debated in the present, like skeptical doubts concerning external world
beliefs, cognition a priori, identity, first-person beliefs, morality, or realism, and
offers astute critical analyses of these arguments.7
Secondly, with respect to the account of skepticism in Hegel’s later work, one
might question whether “dialectical” skepticism is at all a kind of skepticism. For
Hegel’s assimilation of skeptical doubt and dialectic seems to violate our ordinary
understanding of skeptical doubt. Whereas skeptical arguments are supposed to
support the claim that we cannot know whether our beliefs are true, dialectic is
designed to bring about metaphysical knowledge. Hence, the skeptic would insist
that doubt and dialectic are intrinsically incompatible. Hegel would respond that
the skeptic’s insistence is symptomatic of finite understanding’s incapacity to
judge metaphysical truth, that is, that true philosophy and skepticism coincide.
Accordingly, “Science” shows that this kind of criticism is limited to a one-sided
finite perspective that is to be overcome in the direction of all-comprising true
philosophy. From the Hegelian angle, to repudiate skeptical doubts would even be
inefficacious since skepticism is a natural, inescapable problem for human cognition,
a problem that can never be eliminated from our philosophical examination of the
question of knowledge. This is a point of intersection between Hegel’s and the more
recent discussion of skepticism. According to Michael Williams’s interpretation,
contemporary “New Skeptics” like Stroud and Nagel follow Hume in that they
“are rightly impressed by the simplicity, brevity and seeming transparency of typical
skeptical arguments” (Williams 1996: xv), that is, they take skepticism to be a natural
problem. However, the crucial question to be asked is: “Is skepticism really a natural
or intuitive problem?” For Williams it isn’t since, if it is, “then Hume is right, and we
will never reconcile everyday attitudes with the results of philosophical reflection”
(Williams 1996: xv). In this case, from the point of view of everyday life, philosophy
would be irrelevant since what we commonsensically believe to be true in everyday
life does not carry the heavy burden of fundamental skeptical doubt. The best way
to deal with this situation is, Williams argues, to offer a “theoretical diagnosis” of
skepticism that identifies “theoretical ideas” the skeptic holds “that we are by no
means bound to accept” (Williams 1996: xvii). If we do so, we have a good chance
to overcome skepticism by an alternative account of cognition. Hegel’s strategy
is similar to Williams’s since, according to his view, skeptical doubts arise from a
misconceived conception of cognition, that is, one-sided finite cognition that is to be
overcome by true philosophical “Science.” Philosophical “Science” then prevents us
from accepting those “theoretical ideas” that naturally force us into skepticism. That
is, according to Hegel, skeptical doubts are natural as long as we reflect within the
theoretical framework of, for example, Kantian subjectivism. For Kant’s restriction
of cognition to appearances as ordinary objects of everyday life leads naturally into
skepticism about things in themselves, that is, about things as they really are. In
order to escape this kind of skepticism and to therefore eliminate unnatural doubts,
Hegel asks us to give the theoretical framework that like Kantian subjectivism makes
skepticism unavoidable. With respect to the claim that skeptical doubts are unnatural
as long as we subscribe to certain kinds of theoretical frameworks, Hegel therefore
concurs with Williams’s theoretical diagnosis. However, unlike Williams, Hegel
argues that skepticism itself makes indeed a positive contribution to this insight
in that skeptical doubts should be understood as the flipside of true philosophical
“Science.” Skeptical arguments like Agrippa’s trilemma or Kant’s antinomies of
pure reason point to the finiteness of ordinary human cognition that Hegel aims
to overcome by absolute idealism, according to which true philosophical cognition
is possible. For absolute idealism skeptical doubts are unnatural, on the one hand,
but they remain everlasting historical facts on the other. Although as historical
appearances they are unnatural as such, they are at the same time manifestations
of the shortcomings of the human mind and hence move rationality toward true
philosophy. In this sense true philosophy is a kind of skepticism.
NOTES
1. This article will not be concerned with Hegel’s early views of the relation between
logic and metaphysics that pave the way for his later account of skepticism. For
more details, see Heidemann (2011: 81–84), from which I draw in my current
interpretation of Hegelian skepticism. For the broader context, see Düsing (1995:
chs. 2 and 3).
2. On the “history of self-consciousness,” see also Breazeale (2001), Düsing (1993), and
Heidemann (2009: 16–19).
3. For a more detailed analysis of the skeptical argument in “Sense certainty” and
“Perception,” see Heidemann (2011: 88–92).
4. On the methodological differentiation between the standpoint of consciousness in
process and the standpoint of the philosopher, see Heidemann (2007: 245–271).
440 SKEPTICISM
5. Hegel, after 1807, in fact gives up the view that the Phenomenology serves as such
an introduction, although he does not give up the Phenomenology itself. On Hegel’s
later reappraisal of the relation between the Phenomenology of Spirit and the Science
of Logic, see Heidemann (2007: 327–339).
6. By “principle of skepticism” Hegel alludes to Sextus Empiricus (1933: I 12). The
use of “antinomy” not only refers to the antinomies of the Parmenides but also to
the antinomies of pure reason in Kant’s first Critique and possibly to the antithetical
structure of the principles in Fichte’s Doctrine of Science.
7. I cannot elaborate here on Hegel’s worthwhile discussion of these arguments, but
I have done so in Heidemann (2007).
REFERENCES
Breazeale, Daniel. 2001. “Fichte’s Conception of Philosophy as a Pragmatic History of
the Human Mind and the Contribution of Kant, Platner, and Maimon,” Journal of the
History of Ideas 62: 685–703.
Claesges, Ulrich. 1981. Darstellung des erscheinenden Wissens. Systematische Einleitung in
Hegels Phänomenologie des Geistes. Bonn: Bouvier.
Düsing, Klaus. 1993. “Hegels Phänomenologie und die idealistische Geschichte des
Selbstbewußtseins,” Hegel-Studien 28: 103–126.
Düsing, Klaus. 1995. Das Problem der Subjektivität in Hegels Logik. Systematische und
entwicklungsgeschichtliche Untersuchungen zum Prinzip des Idealismus und zur
Dialektik. 3rd ed. Bonn: Bouvier.
Forster, Michael N. 1989. Hegel and Skepticism. Cambridge, MA: Harvard University
Press.
Hegel, Georg Wilhelm Friedrich. 1969–1971. Theorie-Werk-Ausgabe, edited by
E. Moldenhauer and K. M. Michel, 20 vols. Frankfurt: Suhrkamp.
Hegel, Georg Wilhelm Friedrich. 1977. Phenomenology of Spirit, translated by A. V. Miller.
Hegel, Georg Wilhelm Friedrich. 1991. The Encyclopaedia Logic. Part I of the Encyclopaedia
of Philosophical Sciences with Zusätze, translated by T. F. Gerates, W. A. Suchting, and
H. S. Harris. Indianapolis: Hackett.
Hegel, Georg Wilhelm Friedrich. 2000. “On the Relation between Skepticism and
Philosophy. Exposition of His Different Modifications and Comparison of the Latest
Form with the Ancient One,” translated by H. S. Harris. In G. di Giovanni and
H. S. Harris (eds.), Between Kant and Hegel. Texts in the Development Of Post-Kantian
Idealism, 311–362. Indianapolis: Hackett.
Heidemann, Dietmar H. 2007. Der Begriff des Skeptisismus. Seine systematischen Formen,
die pyrrhonische Skepsis und Hegels Herausforderung. Berlin and New York: De
Gruyter.
Heidemann, Dietmar H. 2008. “Substance, Subject, System: The Justification of Science
and the History of Self-Consciousness in Hegel’s Phenomenology of Spirit.” In
D. Moyar and M. Quante (eds.), Hegel’s Phenomenology of Spirit. A Critical Guide,
Heidemann, Dietmar H. 2011. “Hegel on the Nature of Scepticism,” Bulletin of the Hegel
Society of Great Britain 63: 80–99.
Sextus Empiricus. 1933. Outlines of Pyrrhonism, translated by R. G. Bury. Cambridge,
MA: Harvard University Press.
Williams, Michael. 1996: Unnatural Doubts: Epistemological Realism and the Basis of
Scepticism. Princeton: Princeton University Press.
CHAPTER THIRTY-TWO
Friedrich Nietzsche
ANDREAS URS SOMMER
1 INTRODUCTION
Friedrich Nietzsche (1844–1900) is generally labeled as a “skeptic” because he
attacked traditional (Christian) religion vigorously and declared God to be dead,
but also as a thinker frequently transgressing traditional intellectual and political
boundaries (Shaw 2007). Beyond this widely accepted point of view, however, he
seems to defend in his experimental philosophy strong skeptical assumptions, not
only a “mitigated skepticism” (as Magnus 1980 claims). Recent research has shown
that Nietzsche does not hold hard philosophical propositions or doctrines (Sommer
2012; 2013). Instead, he uses such propositions or doctrines—like the Übermensch
(“overman”), the Ewige Wiederkunft des Gleichen (“eternal recurrence of the same”)
or the Wille zur Macht (“will to power”)—as instruments to subvert traditional (not
only philosophical) knowledge and beliefs. Nietzsche might be a skeptic as someone
who never has enough reason to believe that something is true. His doubts about the
knowability of truth should not be ignored: skepticism, which often has a positive
connotation in Nietzsche’s writings, can easily result in a suspension of belief in
reality altogether. However, the skeptical refutation of knowledge does not imply
an apodictic denial of the moral order of reality, although Nietzsche’s concept
of extreme nihilism in his later writings appears to indicate such a form of moral
skepticism. This type of nihilism is postulated by the motto of the Assassins: “Nothing
is true, everything is permitted” (Nietzsche 1999: IV 340, V 399), but it is neither
free of negative dogmatism nor reflects Nietzsche’s very own position. Even when
he promulgates his so-called main doctrines (“overman,” “eternal recurrence,” or
“will to power”), a Pyrrhonian tendency is evident, namely, taking these pretended
truths only as situational assertions, expressing ephemeral perspectives (see Kain
1983: 283–287; Sommer 2000: 524–541; Wotling 2006).
It has also been remarked that Nietzsche’s aphoristic style, often avoiding a final
decision between contradictory conceptions, resembles the skeptical mode of speaking
developed by Sextus Empiricus and Montaigne: a mode full of reservations, aware
of the fact that silence would perhaps be more adequate (Bett 2000: 83). Nietzsche
is, similar to Montaigne, an ardent traveler and an ardent philosophical traveler,
detesting any definite positions or goals. Skepticism is one of Nietzsche’s means
of avoiding final systematization, and it helps to neutralize existential ambiguities.
FRIEDRICH NIETZSCHE 443
However, Nietzsche’s singular type of skepticism significantly deviates from ancient

and modern approaches: in the late 1880s, he argues in favor of an experimental
form of skepticism that is heroic, active, and after all judgment-friendly.
Skepticism is prevalent in the whole of Nietzsche’s writings, from his first years
in Basel in the early 1870s to the last months of his conscious life in Torino in
1888. The understanding of truth in its metaphoricity established in Nietzsche’s
unpublished early work Ueber Wahrheit und Lüge im aussermoralischen Sinne
(On Truth and Lie in the Non-moral Sense) may already be interpreted as a neo-
Pyrrhonian attempt to subvert one’s own speech in speech itself (cf. Berry 2006).
Thus, truth appears as a mere product of speech, “the laws of speech provide
the first laws of truth” (Nietzsche 1999: I 877).1 In this text, the “aporia of the
missing epistemological place of a critique that has turned self-referential” (Kaiser
1994: 71) becomes obvious. This aporia can be seen as a metamorphosis of the
Pyrrhonian aporia subverting its own point of view. Yet in the following, this article
will focus on Nietzsche’s later, more decided, and explicit statements about skeptics
and skepticism. First, I shall review Nietzsche’s knowledge of historical types of
skepticism (Section 2), then investigate the character and forms of skepticism in
Nietzsche’s writings (Section 3), and finally evaluate the consequences of skepticism
for religion, morality, and the “revaluation of all values” (Umwerthung aller Werthe)
according to Nietzsche (Section 4).
2 NIETZSCHE’S VIEW OF HISTORICAL SKEPTICISM

Nietzsche was well aware of the ancient skeptical traditions.2 A major part of his early
philological work was devoted to the prominent doxographer of Greek philosophy,
Diogenes Laertius. He certainly knew very well the chapters on Pyrrho and Timon
in the Vitae philosophorum (DL IX 11 and 12). Nevertheless, his knowledge of
ancient philosophy and skepticism was also based on comprehensive overviews of
Greek literature and philosophy. Nietzsche read, for example, Rudolf Nicolai’s
Geschichte der gesammten griechischen Literatur, Friedrich Ueberweg’s Grundriß
der Geschichte der Philosophie von Thales bis auf die Gegenwart, and Eduard
Zeller’s Die Philosophie der Griechen in ihrer geschichtlichen Entwicklung.3 Beyond
this second-hand reading, there is no evidence that Nietzsche ever studied the main
source of Greek skepticism, namely, the oeuvre of Sextus Empiricus.4
Nietzsche did also read contemporary philosophy, although it was not his
predominant field of occupation. In the books of Friedrich Albert Lange, Afrikan
Spir,5 and Gustav Teichmüller,6 he found different skeptical reflections that
presented him with deep challenges. Apart from Arthur Schopenhauer, Lange’s
Geschichte des Materialismus, first published in 1866, was the most influential
philosophical book for young Nietzsche and also important for his knowledge of
classical skeptical positions (Tongeren 1999). It proposes a radicalized criticism
with some skeptical aspects, but it also extensively discusses recent philosophers
of the skeptical tradition: Michel de Montaigne, Pierre Bayle, and David Hume.
Nietzsche also came across other modern skeptics while reading several volumes
of Kuno Fischer’s Geschichte der neuern Philosophie.7 He recurred to the original
444 SKEPTICISM
texts of Montaigne, Blaise Pascal,8 François de La Rochefoucauld, and other more

or less skeptical French moralists, yet continued to frequently consult secondary
surveys of the French history of ideas. Nietzsche even urged Ida Overbeck, the wife
of his closest friend, Franz Overbeck, to translate—for him and for the general
public—some selected Causeries by Sainte-Beuve on major exponents of the French
Enlightenment, particularly on the skeptic Diderot.9
As a skeptic of language, Georg Christoph Lichtenberg and his collected
works were well present in Nietzsche’s intellectual horizon (Lichtenberg 1867;
see Stingelin 1996). The same applies to Gustav Gerber’s Die Sprache als Kunst
in Nietzsche’s early writings (Gerber 1871–1874; see Meijers and Stingelin 1988).
Nietzsche could find skepticism about religious claims to truth in Hume’s Dialogues
Concerning Natural Religion, of which he possessed a German translation (Hume
1781); in William Edward Hartpole Lecky’s History of the Origin and Influence of
the Enlightenment in Europe (Lecky 1873)—a book he studied intensively; in Jean-
Marie Guyau’s L’irréligion de l’avenir (Guyau 1887); and in Eugène de Roberty’s
L’ancienne et la nouvelle philosophie (Roberty 1887).
In the same year Guyau’s and Roberty’s books were published, Nietzsche also
bought a copy of Victor Brochard’s Les sceptiques grecs (Brochard 1887; see Bett
2000, and Busellato 2012). Nietzsche expressly praises this study and its author
(Nietzsche 1999: VI 284). Some posthumous fragments from 1888 are linked
to Brochard’s work: they deal with the founding father of ancient philosophical
skepticism, Pyrrho of Elis (cf. Brobjer 1997). Before Nietzsche reads Brochard,
the skeptics of Antiquity barely appear in Nietzsche’s writings (Brobjer 2001: 14).
Certainly Heraclitus, the exemplary philosopher of becoming, is a central figure of
reference already for young Nietzsche, but he is not primarily seen by him as a proto-
skeptic. It is an ongoing dispute in Nietzsche studies whether the Sophists, Gorgias,
Hippias, Callicles, and Protagoras had any philosophical importance for Nietzsche
(Mann 2003) or not (Brobjer 2001, 2005). According to a draft periodization of
Greek philosophy from 1872/73, Socrates is expressly classified as skeptic (1999:
VII 540) and anti-dogmatic (1999: VII 544): “Socratic skepticism is a weapon against
the traditional culture and knowledge” (1999: VII 555). According to Nietzsche,
there is also both skepticism in early Plato, who saw everything in “flux” (1999:
VII 550; see Ghedini 1999: 272–273), and a skepticism reloaded in the refutation
of Plato’s doctrine of ideas in the dialogue Parmenides. Several years later, in 1885,
Nietzsche distinguishes an exoteric and an esoteric point of view in Plato. No other
philosopher, he claims, cultivated such an “absolute skepticism towards all given
concepts” as Plato, although he “taught the opposite” (Nietzsche 1999: XI 147).
Incapable of art, the Academic skeptic Arcesilaus stands in this tradition (1999: VII
97; see Zeller 1880: 496–497 n. 4).
Pyrrhonian skepticism is very rarely mentioned in Nietzsche’s early and middle
oeuvre. In the second part of Menschliches, Allzumenschliches, a figure named
Pyrrho appears in a dialogue with an “old man.” He is presented as someone who
wants to promulgate mistrust of “everything and everyone,” because this seems to
be the “only path to truth” (1999: II 645). In the dialogue, Pyrrho has to admit
that he is forced to silence if he wants to be mistrustful of all words, but that he
also has to tell people that he must keep silent: “Saying nothing and laughing—is
that your whole philosophy? Pyrrho: it would not be the worst” (1999: II 646).
According to Diogenes Laertius (DL IX 61 and IX 76), aphasia—the wordlessness,
the silence in light of the impossibility of making a decision—is typical of Pyrrhonian
skepticism. A better informed approach to the historical Pyrrho can be traced
after Nietzsche’s reading of Brochard in 1887/88. Under this influence, Pyrrho is
now presented as a “Greek Buddhist” (1999: XIII 264), even as “Greek Buddha”
(1999: XIII 347), performing a fundamental criticism of a philosophy unsuitable
for life (1999: XIII 311). In another note, Pyrrho’s own philosophy looks like
“decadence,” defending itself against “science and mind,” denying the “importance
of all things” and leading to a “final self-overcoming, final indifference” (1999: XIII
276–278). Nietzsche compares Pyrrho to another décadent, Epicurus, yet finds the
former “more travelled, more worn out, more nihilistic,” perhaps more radical. In
Nietzsche’s view, he challenged the very bases of philosophy: “His life was a protest
against the great doctrine of identity.” Pyrrho even gave up the pursuit of happiness
(1999: XIII 276–278).
Not primarily interested in happiness was also René Descartes. His methodological
doubt in the philosophy of knowledge did not seem radical enough to Nietzsche:
Descartes still seeks and intends to find evidence for an absolute and evident truth.
For Nietzsche, this is just “nonsense”: “cogito, ergo sum pretends that one knows
what ‘thinking’ is and secondly, what ‘being’ is” (Nietzsche 1999: XI 640). However,
we don’t know what these two things are. While Nietzsche does not show any
particular interest in Hume,10 he holds Kant liable for translating the “Englishmen’s
epistemological skepticism” to the Germans by meeting their moral and religious
needs, “just as Pascal even made use of moral skepticism to excite the need for faith”
(1999: XII 340). In fact, Nietzsche considers Kant a dogmatist to the core, who, “by
means of a clever and mischievous skepticism,” has even tried to make Christianity
irrefutable (1999: VI 176).
3 CHARACTER AND FORMS OF SKEPTICISM

In his first philosophical reflections, Nietzsche is well aware that a fundamental
criticism of knowledge has skeptical consequences. Typical is his early response
to such a challenge: he underlines the “necessity of art and illusion” in order to
reject “absolute skepticism” (Nietzsche 1999: VII 458, also XI 649). Philology in
particular, Nietzsche’s original profession—he was a professor of Greek philology,
never of philosophy—appears in his notes as a skeptical discipline, destroying all
kinds of cultural and intellectual securities. This view, already present in 1875
(1999: VIII 37, 55), prevailed until Nietzsche opted for a philosophical “ephexis in
interpretation” (1999: VI 233) in 1888: this is meant as a giving-up of all projections
and premature judgments while reading texts or also the book of nature (see also
Benne 2005).
On the other hand, Nietzsche fights against a cheap skepticism of easy comfort
connected with fatigue and feeling weak (Nietzsche 2003: V 153). While, according
to Nietzsche, traditional skeptics claimed to own the truth that nothing is true, he
446 SKEPTICISM
claims for his own standpoint “that we do not have truth” (Nietzsche 1999: IX
52). Nietzsche proclaims a new form of skepticism against the skeptical attitude of
the contemporary décadence and probably also against the academic skepticism of
the nineteenth-century German university: “Skepticism! Yes, but a skepticism
of experiments! Not the lethargy of despair” (1999: IX 287). This motif of an
experimental skepticism that takes its legitimation out of non-knowledge remains
dominant in several posthumous notes, although the real shape of this specific form
of skepticism is not quite clear. It can also be seen as “a temptation” (1999: X 526).
However, in the same context Nietzsche takes life itself as “an experiment”: its
“happiness” consists “in guessing or experimenting (skepticism)” (1999: X 527).
He seems to be attracted to the therapeutic effect of the assumption that things are
unknowable: “Despair abolished via skepticism,” giving way to “the right to create”
(1999: X 232). Such reflections in his notebooks lead to statements in his published
books like: “I salute any form of skepticism to which I am allowed to reply: ‘let’s
try!’” (1999: III 415).
Nevertheless, Nietzsche attempts to guard himself against “skeptical randomness”:
the recently gained liberty “from the tyranny of ‘eternal’ concepts” should teach us
“to view concepts as experiments, with the help of which certain types of humans can
be bred” (Nietzsche 1999: XI 526). This is the way Nietzsche’s own new concepts,
will to power and eternal recurrence of the same, are to be understood: as intellectual,
but also as existential, experiments. Existence has a “perspectivist character,” and
the world “contains within itself endless interpretations” (1999: III 626–627). This
endorses the inevitable perspectivism and situationism of knowledge. Taking this
experimental skepticism seriously, Nietzsche’s readers have to take the seemingly
dogmatic late work The Antichrist as ruled by a skeptical strategy of proclaiming
certain interpretations against others.11 Undoubtedly, Nietzsche’s experimental
skepticism is far away from the Pyrrhonian attitude of withholding judgment (cf.
Bett 2000: 81). Instead, his skepticism has a neutralizing effect on the permanent
stating of judgments.
In 1873, Nietzsche noted that there were only “friends of truth,” “enemies of
truth and skeptics” (1999: VII 574). Indifference toward the truth thus characterizes
the skeptic. In Nietzsche’s later diagnosis, his contemporary cultural and intellectual
situation is determined by the abolishment of a metaphysical idea of truth. Skepticism
is therefore revealed as “the afterbirth of the pride” that had been cultivated for a
long time by humanity believing in its own metaphysical worth (1999: IX 200).
Given that such skepticism will first diminish human pride, Nietzsche assumes
that later a period of exhaustion will follow. However, Nietzschean skepticism
is confrontational: “We have become disgusted by the views of the authorities—
rather we would starve!” (1999: IX 226). And this “skepticism is a passion,” too.
Nietzsche uses the word “ephexis” for the Pyrrhonian epochē and postulates it as
a main scientific principle (1999: XI 521).12 The ephectic keeps problems open
and remains reserved toward precipitate explanations. Ephexis turns out to be no
intellectual laziness, but a difficult task of disillusionment.
In 1883/84, Nietzsche establishes a distinction between a “skepticism of weakness”
and a skepticism of “courage,” to be connected with “heroic feelings” (1999: X 662).
Thus, he seems to abandon his former idea of an incompatibility of the tragic and
the skeptical worldviews. In a historical perspective, “the skeptical periods, suffering
from indecision” are precisely those periods tempted to adopt a “rigid faith” (1999:
XII 191). Christianity became powerful in a skeptical world (1999: V 66). The
“suffering from uncertainty” is seen as the very origin of systematic thought.
Nietzsche’s experimental and heroic skepticism tries to withstand the temptation
of systematic thought. He complains that authoritative skeptics have emerged only
rarely up to the present (1999: XI 696), and he seeks for the “overcoming of the
skeptic of weakness” (1999: XI 211). Particularly the sections 208 to 210 of Beyond
Good and Evil articulate this search, where a strong skepticism is hailed against a
weak and comfortable skepticism, a “gentle, pure, soothing opium” (1999: V 137).
The new destabilization has to be “a skepticism of bold masculinity . . . that is
closely related to the genius for war and plunder” (1999: V 141). While Nietzsche
promulgates this skepticism as the relevant way of philosophizing in his last books
(e.g. 1999: VI 178, 236), his critical views on skepticism in the ancient world
influenced by Brochard remain hidden in his posthumous notes (Bett 2000: 79).
In these notes, ancient as well as modern skepticism rather appears as “an effect
of decadence” (Nietzsche 1999: XIII 264). Certainly, the active, experimental
skepticism preferred by Nietzsche is totally different from the passive Pyrrhonism
with its refusal to make a judgment and may come closer to Academic skepticism
(that Nietzsche knew well from Cicero, see note 2). Brochard revitalized Nietzsche’s
interest in the complexity of skepticism in general. Yet, the skepticism propagated
by Nietzsche is the exact opposite of the Pyrrhonian skepticism in its approach
to life.
4 SKEPTICISM ABOUT RELIGION AND MORALITY,

AND THE “REVALUATION OF ALL VALUES”
Challenging the possibility of knowing the truth, skepticism obviously stands in
opposition to religion when understood as an institution that socially administers
the truth (see, e.g., Nicolai 1867: 299). Nietzsche was well aware of this opposition
when, in 1873, he defined skepticism as an “ascetic viewpoint of the thinker”
(Nietzsche 1999: VII 624) who rejects the very idea of believing. Twelve years
later, he focused on “the inwardly audacious skepticism in Germany” in reaction
to the overly close familiarity of philosophers with Protestantism (1999: XI 473).
Nietzsche could observe the way from Protestantism to skepticism personally in
the example of his most loyal friend, the agnostic Franz Overbeck, professor of
theology at the University of Basel (cf. Sommer 1997). Nietzsche stresses, on the
one hand, that also Catholics like Pascal also used skepticism against reason—that
humans cannot access truth through rational operations—for proselytism: according
to them, faith alone opens the path to knowledge (1999: II 383–384). On the other
hand, it was a skeptical insight that eliminated God, especially the Christian one, for
Nietzsche. Thus, in The Antichrist and in the late Nachlass, Jesus is not the founder
of a new religion, but a relative of Epicurus (1999: XIII 276–278) and also of
448 SKEPTICISM
Pyrrho: Jesus and Pyrrho are both akin to Buddha (see 1999: VI 202, XI 247), since
both give up “all animosities, boundaries and distances in feeling” (1999: VI 200),
all judgment and denial. For both, Nietzsche traces this peculiar attitude to their
particular pathological, physiological conditions: Jesus and Pyrrho had to follow
this way of life and thought because they suffered from decadence.
In Nietzsche’s oeuvre, skepticism is not a simple epistemological exercise at all
(see also Welshon 2009). His project of a genealogy of moral sentiments and of
morality itself is fundamentally marked by skeptical reflections (Nietzsche 1999: IX
109 and V 252), while moralism usually rejects the skeptics (1999: II 407). Quite
early, Nietzsche saw himself as a proponent of a “moralistic skepticism” through
which pessimism had deepened and criticized itself (1999: II 370). In Daybreak
(1881), “moral skepticism” mistrustful of man does not have to be accompanied
by bad humor or weakness. The aphorism is a dialogue of two persons, A and B.
According to A, “general moral skepticism” seems to be an elixir of life, making him
healthier. B responds that A has ended up being a skeptic as he says “no” again. A
answers that he has thus learned to say “yes” again (1999: III 284). This short text
illustrates in nuce the performative pattern of Nietzsche’s experimental skepticism,
which is, despite its sympathy for “ephexis in interpretation,” a judgment-friendly
skepticism. It challenges and provokes “value judgments on people and things”: “one
would have to overturn all values by means of a radical skepticism, to have a clear
path” (1999: IX 62). Thus Socrates himself can turn out as the eminent forerunner
of moral skepticism (1999: IX 363, 365). In return, Nietzsche recommends moral
skeptics to subject their “mistrust of morality” to a mistrustful skepticism (1999:
X 67–68), because skeptics attacking morality might simply forget how they are
themselves infected with morality (1999: XI 486). Far-reaching doubt about
morality can be problematic, since the reason for doubt fades away altogether as
soon as the skeptic stops to believe in truth—unless “the will to know” has “a quite
different root yet . . . from that of truthfulness” (1999: XI 510–511). Consequently,
Nietzsche assumes that “no skepticism has arisen without ulterior motives” and that
these ulterior motives have so far been moral ones (1999: XII 143–144). Even to
ancient skeptics, Nietzsche argues, morality remained the highest value (1999: XIII
319–321), when they subordinated their epochē to the “herd instinct,” to “living
like the ‘common man’” (1999: XIII 286). Nietzsche’s own experimental skepticism
fosters an abhorrence of such subordination. His skepticism intends to create
differences, not to make them disappear. A “victory of skepticism” over “morality
as prejudice” may entail “a dawn-like happiness” (1999: XII 195). Of course, only
hard work helps to attain it: “One should reduce and bound the realm of morality
step by step” (1999: XII 476).
In late Nietzsche, moral skepticism is a requirement for the emergence of nihilism.
It also has a crucial part in Nietzsche’s final task of a “revaluation of all values”
(Umwerthung aller Werthe). He pretends at the end of 1888 to have fulfilled this task
in The Antichrist. If the revaluation of values is to avoid a new heteronomy, it cannot
establish a canon of new values, to be obeyed as the former. Otherwise, the liberty
to create values that Nietzsche wants to guarantee (at least to strong individuals)
would be undercut again. “Revaluation of all values” aims for such liberty to shape
one’s self and one’s world. One instrument to attain this goal is the neutralizing of
all differences through diaphoniai. Nietzsche’s conception of revaluation demands of
every individual to make unavoidable personal, situational, and perspectivist decisions
in order to shape his or her life and the world. This concept is radically different
from the Pyrrhonian inclination to accept the given circumstances without making
any decisions or alterations. Ephexis in The Antichrist is a principle of interpretation
against the pretense of one absolute perspective, but it does not imply the Pyrrhonian
suspension of one’s own decision. Nietzsche’s experimental skepticism forbids epochē
as a basic attitude to life (see Obstoj 1985). Philosophers and skeptics should both
become “givers of orders and laws” (Nietzsche 1999: V 145).
Extreme nihilism and experimental skepticism are alternative names for the late
Nietzsche’s strategy of revaluating all values. His alleged “main doctrines” of “will to
power” and “eternal recurrence of the same” are only the means, not the ends, of this
revaluation. Through extreme nihilism and experimental skepticism, humans gain
radical distance from all ideas and doctrines hitherto taken for granted. Skepticism
then enables intellectual and practical freedom. So is freedom for Nietzsche a value
beyond all traditional values, not affected itself in a revaluation of values?13
NOTES
1. See also Hödl (1997) and Reuter (2009).
2. See Bett (2000), Sommer (2000), Brobjer (2001), Berry (2011), and Busellato (2012).
Berry (2011) claims a fundamental difference between ancient and modern, (post-)
Cartesian skepticism and sees Nietzsche in the tradition of classical Pyrrhonism
as well as of Montaigne with his interest in the practical relevance particularly of
epochē (cf. Santini 2013). On the other hand, Leiter (2014) rejects a Pyrrhonian
ground for Nietzsche’s “value Skepticism.” More deeply engaged in questions of
“Quellenforschung” is Busellato (2012). He documents, for example, the interest
Nietzsche had in Cicero’s Academica—a text the young professor of philology used in
one of his courses at the University of Basel in 1870/71. The relevance of the model
of the wise in Hellenistic philosophy for Nietzsche’s Übermensch is made clear by
Bertino (2007) (see already Parush 1975/76).
3. See Nicolai (1867: esp. 298–300, 489–90), Ueberweg (1867); Zeller (1880: esp.
478–527; 1881: esp. 1–69).
4. On parallels between Sextus and Nietzsche, see Conway and Ward (1992).
5. See Spir (1873; 1877). On Nietzsche’s reception D’Iorio (1993) and Small (2001:
1–20).
6. See Teichmüller (1882) and Nietzsche (1999: XIII 579).
7. Nietzsche (1999: XII 259–270) is a mainly unaltered excerpt taken from Fischer. See
Sommer (2012b) on Fischer, Nietzsche, and Spinoza.
8. See Vivarelli (1998) on Montaigne and Pascal.
9. This translation has been published under Nietzsche’s protection by his publisher:
Sainte-Beuve (1880), critical new edition Sainte-Beuve (2014).
10. See Beam (2001) and Kail (2009) for a different perspective on Nietzsche
and Hume.
450 SKEPTICISM
11. Nietzsche (1999: VI 236) reflects this strategy of utilizing “convictions” frankly.
12. See also Nietzsche (1999: VI 232) and Sommer (2013: 244–46).
13. This text is based on Sommer (2006). I would like to thank Daniel Unger, the
editors of this book, and particularly Julia Maas for many very useful corrections,
comments, and suggestions.
REFERENCES
Beam, Craig. 2001. “Ethical Affinities: Nietzsche in the Tradition of Hume,” International
Studies in Philosophy 33: 87–98.
Benne, Christian. 2005. Nietzsche und die historisch-kritische Philologie. Berlin and
New York: Walter de Gruyter.
Berry, Jessica. 2006. “Skepticism in Nietzsche’s Earliest Work: Another look at ‘On Truth
and Lie in an Extra-Moral Sense,’” International Studies in Philosophy 38: 33–48.
Berry, Jessica. 2011. Nietzsche and the Ancient Skeptical Tradition. Oxford: Oxford
University Press.
Bertino, Andrea. 2007. “Nietzsche und die hellenistische Philosophie. Der Übermensch
und der Weise,” Nietzsche-Studien 36: 95–130.
Bett, Richard. 2000. “Nietzsche on the Skeptics and Nietzsche as Skeptic,” Archiv für
Geschichte der Philosophie 82: 62–86.
Brobjer, Thomas. 1997. “Beiträge zur Quellenforschung,” Nietzsche-Studien 26:
574–579.
Brobjer, Thomas. 2001. “Nietzsche’s Disinterest and Ambivalence Toward the Greek
Sophists,” International Studies in Philosophy 33: 5–23.
Brobjer, Thomas. 2005. “Nietzsche’s Relation to the Greek Sophists,” Nietzsche-Studien
34: 256–277.
Brochard, Victor. 1887. Les sceptiques grecs. Paris: Imprimerie Nationale.
Busellato, Stefano. 2012. Nietzsche e lo scetticismo. Macerata: Edizioni Università di
Macerata.
Conway, David and Janet Ward. 1992. “Physicians of the Soul: Peritrope in Sextus and
Nietzsche.” In D. W. Conway and R. Rehn (ed.), Nietzsche und die antike Philosophie,
193–224. Trier: WVT.
D’Iorio, Paolo. 1993. “La superstition des philosophes critiques. Nietzsche et Afrikan
Spir,” Nietzsche-Studien 22: 257–294.
Fischer, Kuno. 1865. Geschichte der neuern Philosophie. Bd. 1: Descartes und seine Schule.
1. Theil: Allgemeine Einleitung. René Descartes. 2. Theil: Descartes’ Schule. Geulinx.
Malebranche. Baruch Spinoza. Zweite völlig umgearbeitete Auflage. Mannheim:
Friedrich Bassermann.
Gerber, Gustav. 1871–1874. Die Sprache als Kunst. 2 vols. Bromberg: Mittler’sche
Buchhandlung.
Ghedini, Francesco. 1999. Il Platone di Nietzsche. Genesi e motivi di un simbolo controverso
(1864–1879). Napoli: Edizioni Scientifiche Italiane.
Guyau, Jean-Marie. 1887. L’irréligion de l’avenir. Étude sociologique, 2nd edn. Paris:
F. Alcan.
Hödl, Hans Gerald. 1997. Nietzsches frühe Sprachkritik. Wien: Universitätsverlag.

Hume, David. 1781. Gespräche über natürliche Religion. Nach der zweyten englischen
Ausgabe. Nebst einem Gespräch über den Atheismus von Ernst Platner. Leipzig: Weygand.
Kail, Peter. 2009. “Nietzsche and Hume: Naturalism and Explanation,” Journal of
Nietzsche Studies 37: 5–22.
Kain, Philip. 1983. “Nietzsche, Skepticism, and Eternal Recurrence,” Canadian Journal of
Kaiser, Stefan. 1994. “Über Wahrheit und Klarheit. Aspekte des Rhetorischen in ‘Ueber
Wahrheit und Lüge im aussermoralischen Sinne,’” Nietzsche-Studien 23: 65–78.
Lange, Friedrich Albert. 1866. Geschichte des Materialismus und Kritik seiner Bedeutung in
der Gegenwart. Iserlohn: J. Baedeker.
Lecky, William Edward Hartpole. 1873. Geschichte des Ursprungs und Einflusses der
Aufklärung in Europa. Mit Bewilligung des Verfassers übersetzt von H. Jolowicz. 2 vols.
Leipzig and Heidelberg: C. F. Winter’sche Verlagshandlung.
Leiter, Brian. 2014. “Moral Skepticism and Moral Disagreement in Nietzsche,” Oxford
Studies in Metaethics 9: 126–151.
Lichtenberg, Georg Christoph. 1867. Vermischte Schriften. Neue Original-Ausgabe. 8 vols.
Göttingen: Dieterich.
Magnus, Bernd. 1980. “Nietzsche’s Mitigated Skepticism,” Nietzsche-Studien 9: 260–267.
Mann, Joel. 2003. “Nietzsche’s Interest and Enthusiasm for the Greek Sophists,” Nietzsche-
Studien 32: 406–428.
Meijers, Anthonie, and Martin Stingelin. 1988. “Konkordanz zu den wörtlichen Abschriften
und Übernahmen von Beispielen und Zitaten aus Gustav Gerber: Die Sprache als Kunst
(Bromberg 1871) in Nietzsches Rhetorik-Vorlesung und in ‚Ueber Wahrheit und Lüge
im aussermoralischen Sinne,’” Nietzsche-Studien 17: 350–368.
Nicolai, Rudolf. 1867. Geschichte der gesammten griechischen Literatur. Ein Versuch.
Magdeburg: Heinrichshofen’sche Buchhandlung.
Nietzsche, Friedrich. 1999. Sämtliche Werke. Kritische Studienausgabe in 15 Einzelbänden,
edited by G. Colli and M. Montinari, 3rd edn. München, Berlin and New York:
Deutscher Taschenbuch Verlag and Walter de Gruyter.
Nietzsche, Friedrich. 2003. Sämtliche Briefe. Kritische Studienausgabe in 8 Bänden, edited
by G. Colli and M. Montinari, 2nd edn. München, Berlin, and New York: Deutscher
Taschenbuch Verlag and Walter de Gruyter.
Obstoj, Dietmar. 1985. Skepsis bei Nietzsche. PhD Thesis, University of Hannover.
Parush, Adi. 1975/1976. “Nietzsche on the Skeptic’s Life,” The Review of Metaphysics 29:
523–542.
Reuter, Sören. 2009. An der “Begräbnisstätte der Anschauung.” Nietzsches Bild- und
Wahrnehmungstheorie in “Ueber Wahrheit und Lüge im aussermoralischen Sinne.” Basel:
Schwabe.
Roberty, Eugène de. 1887. L’ancienne et la nouvelle philosophie. Essais sur les lois générales
du développement de la philosophie. Paris: F. Alcan.
Sainte-Beuve, Charles-Augustin. 1880. Menschen des XVIII. Jahrhunderts nach den Causeries
du Lundi, translated by I. Overbeck. Chemnitz: Ernst Schmeitzner.
452 SKEPTICISM
Sainte-Beuve, Charles-Augustin. 2014. Menschen des XVIII. Jahrhunderts. Übersetzt von

Ida Overbeck, initiiert von Friedrich Nietzsche. Mit frisch entdeckten Aufzeichnungen
von Ida Overbeck neu ediert von Andreas Urs Sommer. Berlin: Die Andere Bibliothek.
Santini, Carlotta. 2013. “Nietzsche und der antike Skeptizismus,” Nietzsche-Studien 42:
368–374.
Shaw, Tamsin. 2007. Nietzsche’s Political Skepticism. Princeton and Oxford: Princeton
University Press.
Small, Robin. 2001. Nietzsche in Context. Aldershot and Burlington: Ashgate.
Sommer, Andreas Urs. 1997. Der Geist der Historie und das Ende des Christentums. Zur
“Waffengenossenschaft” von Friedrich Nietzsche und Franz Overbeck. Berlin: Akademie
Verlag.
Sommer, Andreas Urs. 2000. Friedrich Nietzsches “Der Antichrist”. Ein philosophisch-
historischer Kommentar. Basel: Schwabe.
Sommer, Andreas Urs. 2006. “Nihilism and Skepticism in Nietzsche.” In K. Ansell Pearson
(ed.), A Companion to Nietzsche, 250–269. Malden: Blackwell.
Sommer, Andreas Urs. 2007. “Skeptisches Europa? Einige Bemerkungen zum Sechsten
Hauptstück: wir Gelehrten (Friedrich Nietzsche, Jenseits von Gut und Böse, Aphorismen
204–213),” Nietzscheforschung. Jahrbuch der Nietzsche-Gesellschaft 14: 67–78.
Sommer, Andreas Urs. 2012. Kommentar zu Nietzsches “Der Fall Wagner”. “Götzen-
Dämmerung” (= Historischer und kritischer Kommentar zu Friedrich Nietzsches Werken.
Ed. by Heidelberger Akademie der Wissenschaften. Vol. 6/1.) Berlin and Boston: Walter
de Gruyter.
Sommer, Andreas Urs. 2013. Kommentar zu Nietzsches “Der Antichrist”. “Ecce homo”.
“Dionysos-Dithyramben”. “Nietzsche contra Wagner” (= Historischer und kritischer
Kommentar zu Friedrich Nietzsches Werken. Ed. by Heidelberger Akademie der
Wissenschaften. Vol. 6/2.) Berlin and Boston: Walter de Gruyter.
Spir, Afrikan. 1873. Denken und Wirklichkeit. Versuch einer Erneuerung der kritischen
Philosophie. 2 vol. Leipzig: J. G. Findel.
Spir, Afrikan. 1877. Denken und Wirklichkeit. Versuch einer Erneuerung der kritischen
Philosophie. Zweite, umgearbeitete Auflage. 2 vol. Leipzig: J. G. Findel.
Stingelin, Martin. 1996. “Unsere ganze Philosophie ist Berichtigung des Sprachgebrauchs”.
Friedrich Nietzsches Lichtenberg-Rezeption im Spannungsfeld zwischen Sprachkritik
(Rhetorik) und historischer Kritik (Genealogie). München: Wilhelm Fink.
Teichmüller, Gustav. 1882. Die wirkliche und die scheinbare Welt. Neue Grundlegung der
Metaphysik. Breslau: Koebner.
Tongeren, Paul J. M. van. 1999. “Nietzsche’s Symptomatology of Skepticism.” In B. Babich
(ed.), Nietzsche, Epistemology, and Philosophy of Science. Nietzsche and the sciences II,
61–71. Dordrecht and Boston: Springer.
Ueberweg, Friedrich. 1867. Grundriß der Geschichte der Philosophie von Thales bis auf die
Gegenwart. Erster Theil: Das Alterthum. Dritte, berichtigte und ergänzte Auflage. Berlin:
E. S. Mittler and Sohn.
Vivarelli, Vivetta. 1998. Nietzsche und die Masken des freien Geistes. Montaigne, Pascal
und Sterne. Würzburg: Königshausen and Neumann.
Welshon, Rex. 2009. “Saying Yes to Reality: Skepticism, Antirealism, and Perspectivism in
Nietzsche’s Epistemology,” Journal of Nietzsche Studies 37: 23–43.
Wotling, Patrick. 2006. “‘L’ultime scepticisme.’ La vérité comme régime d’interprétation,”

Revue philosophique de la France et de l’étranger 131: 479–496.
Zeller, Eduard. 1880. Die Philosophie der Griechen in ihrer geschichtlichen Entwicklung.
Dritter Theil, erste Abtheilung: Die nacharistotelische Philosophie. Dritte Auflage.
Leipzig: Fues’s Verlag (R. Reisland).
Zeller, Eduard. 1881. Die Philosophie der Griechen in ihrer geschichtlichen Entwicklung.
Dritter Theil, zweite Abtheilung: Die nacharistotelische Philosophie. Dritte Auflage.
Leipzig: Fues’s Verlag (R. Reisland).
CHAPTER THIRTY-THREE
Russell’s Logical
Construction of the
External World
PETER J. GRAHAM
1 INTRODUCTION
Bertrand Russell (1872–1970) was no skeptic. Like G.E. Moore (1873–1958),
Russell agreed that we know about external, material objects like tables and trees.
But Russell gave a very different explanation than Moore. In his highly influential
1914 Lowell Lectures, Our Knowledge of the External World, Russell “logically
constructed” commonsense tables and trees from sense-data. In this chapter, I
expound Russell’s logical construction of the external world.1
Most contemporary philosophers have only a passing knowledge of Russell’s
work in metaphysics and epistemology, derived from reading his (still popular) 1912
Problems of Philosophy. That book presents a view of our knowledge of external,
material objects as the result of an inference to the best explanation from awareness
of the qualities and characteristics of our sensations. That is not the view of our
knowledge of the external world that Russell presents in his Lowell Lectures and
related papers. Instead he presents a view of sense-data that sounds odd to our
contemporary ears. And instead of inferring the existence of tables and chairs,
Russell will, where possible, “logically construct” external objects out of sense-data.
Did Russell succeed, and thereby explain our knowledge of the external world?
It will not surprise anyone to learn that the answer, as we will see, is no. Russell is
then faced with a choice: abandon his view of knowledge of the external world as a
logical construction out of sense-data or embrace radical skepticism.
Despite its apparent failure, understanding Russell’s particular attempt is worth
the effort. It has roots in the British empiricist tradition (especially Berkeley and
Hume) as well as in Leibniz’s monadology. Russell’s originality then in turn had a
substantial influence on Rudolf Carnap’s (1891–1970) own “logical construction
of the world” (Carnap 1928) as well as a series of other “phenomenalisms.” These
largely logical positivist views then set the terms of debate for W.V.O. Quine’s (1908–
RUSSELL’S LOGICAL CONSTRUCTION OF THE EXTERNAL WORLD 455
2000) explorations in “naturalized” epistemology. The spirit of Russell’s project is

even still with us today, taken up by David Chalmers (2012), among others.
I first present Russell’s view of sense-data. I then say what he means by a “logical
construction.” I then turn to Russell’s logical construction of “things” and then of
space. With his ontological machinery of sense-data and logical constructions in
place, I then turn to his view of our knowledge of the external world as a logical
construction of the “soft data” of tables and chairs out of the “hard data” of sense-
data. I’ll then explain why Russell’s attempt falls short, for it needs to assume an
unearned knowledge of other minds as well as the possibly unknowable existence of
sensibilia. Though Russell’s project falls short, I conclude with a few achievements
that, in the context of various debates with his contemporaries, led Russell to see
value in his efforts nonetheless.
2 SENSE-DATA
I begin with Russell’s view of sense-data. Suppose you and I walk up to a table in the
park. We believe it existed before we saw it and will continue to exist after we walk
away. We see the same thing, and any passerby would see it, too.
Russell did not believe we immediately see the table. Think about your experiences
as you approached and interacted with the table. From a distance it looked rather
gray; you experienced a particular shade of color with a particular shape. As you
got closer, you experienced a different shade of color with a larger shape. Russell
believed these differing visible patches of shape and color—these differing “sensible
objects”—are what we immediately sense in sense-perception:
When I speak of a “sensible object,” it must be understood that I do not mean
such a thing as a table, which is both visible and tangible, can be seen by many
people at once, and is more or less permanent. What I mean is just that patch
of color which is momentarily seen when [you] look at the table, or just that
particular hardness which is felt when [you] press on it, or just that particular
sound which is heard when [you] rap on it. (1914a: 83)
Russell calls these “sensible objects” sense-data. Sense-data are the particular
patches of color, shapes, sounds, and textures that we immediately sense, and may
attend to, in sense-perception. Russell also calls them appearances or aspects of
material objects.2
Russell relied upon the argument from perceptual variation to distinguish sense-
data from material objects. As we move toward and around the table, our sense
appearances change; from one location we are directly aware of something small
and gray and from another something large and brown. But the picnic table does not
change as our appearances change. Thus what we are immediately aware of in sense-
perception is not the same as the table. Since the material object stays the same but
what we immediately sense changes, we do not immediately sense material objects
(1912: ch. 2; 1914a: 76; 1959: 103).3
Russell attributed various properties to sense-data. First, sense-data have the
properties they are sensed to have.4 Russell even believed “there are no such things
456 SKEPTICISM
as ‘illusions of sense’” (1914a: 85). Sensory acquaintance is a kind of direct, infallible

relationship with objects of sense as they really are. To take a familiar example, if
you press one finger against your eyeball you will “see double.” For Russell, this
means you will sense two distinct visual sense-data where before you sensed one,
and those two distinct visual sense-data have exactly the properties they appear
to have; they are just as real as the data you sense when you are not pressing your
eyeball. Your experience is not illusory; there really are two sense-data with just
those properties.5
Secondly, sense-data are in constant flux. Outer experience is like watching a
movie: we are viewing a series of distinct frames, each existing only for a brief
moment before our eyes (1915: 96–97). Relatedly, Russell also rejected the view
that the immediate objects of sense, though “external” to our minds, continue to
exist when we shut our eyes or even turn our eyes in another direction (1915:
96). This belief is “a piece of audacious metaphysical theorizing; objects are not
continually present to sensation, and it may be doubted whether they are there
when they are not seen or felt” (1914a: 107). Echoes of Hume.
Thirdly, Russell believed (and this may come as a surprise to contemporary
readers) that sense-data are physical. “I believe . . . the immediate objects of
[perception] are extra-mental, purely physical” (1914a: 71; 1914b: 111–113;
1915: 96–98). Sense-data are the objects of sensations, not sensations themselves.6
In sense-perception the mind is directly acquainted with external, non-mental,
physical particulars.7 Many take the argument from perceptual variation to show
that sense-data are mental; Russell did not. Sense-perception is then the perception,
by sensation, of sense-data.
Fourthly, Russell held that sense-data are private. (This may also come as a
surprise, for if they are physical, how can they be private?) He argued as follows.
First, a perceiver will perceive “an immensely complex three-dimensional world” of
sensible objects at a time. (This is the subject’s complete, “perceived perspective.”)8
Second, there will always be differences, however slight, in different subjects’ points
of view, and so there will always be differences, however slight, in the subjects’
immediate objects of sense-perception. Third, the places of objects “can only be
constituted by the things in or around them” (for position in space is purely relative).
If we see different objects, we see different spaces. Thus the three-dimensional space
of one subject’s sensible objects and the three-dimensional space of another’s has
no place in common. As a result, there “is absolutely nothing which is seen by two
minds simultaneously” (1914a: 94–95; 1914b: 117; 1915: 103). We all perceive
different “private worlds.”
3 LOGICAL CONSTRUCTIONS
If we immediately perceive private, evanescent sense-data, then how do we know
of public, continually existing material objects? In his 1912 Problems of Philosophy
Russell argued that the commonsense hypothesis “that there really are objects
independent of us, whose actions cause our sensations” best explains features of
our sense-data: “every principle of simplicity urges us to adopt the natural view,
that there really are objects other than ourselves and our sense-data which have an
existence not dependent upon our perceiving them” (1912: 22–24).
In 1914 Russell found this inference dubious and changed course: instead of
inferring commonsense objects, Russell constructed them. Russell announced
“the supreme maxim in scientific philosophizing. . . Wherever possible, logical
constructions are to be substituted for inferred entities” (1914b: 115). In the case
at hand, Russell was worried that if we construe material objects independently
of sense-data, they will turn out to be unknowable “things-in-themselves.” So
to prevent that, we must “construct” the material “things” of commonsense out
of sense-data.
Russell’s and Whitehead’s analysis of numbers in Principia Mathematica
illustrates the idea (1910–1913). According to Russell and Whitehead, a number is
a class of classes whose members bear a one-to-one relation to each other (Russell
1959: 70–71). Zero is the class of all empty classes; one is the class of all singletons;
two is the class of all couples, etc. Numbers are then just “logical constructions” of
equinumerous classes.9
When Russell turned to the external world, Russell’s and Whitehead’s success was
his paradigm (1914b: 115–116). Just as Russell and Whitehead eliminated numbers
as metaphysically distinct objects by Occam’s razor, Russell will “construct” the
material “thing” of commonsense out of classes of sensible objects and thereby
eliminate “things-in-themselves” (1914a: 72). Russell often put this “analysis” of
material objects in terms of sense-data as a “reduction” or an “interpretation” of
statements about material objects to classes of sense-data, where material objects
are “functions” of sense-data (1914a: 85, 86; 1915: 97). The goal is to “reduce”
commonsense “facts” to a “form in which nothing is assumed beyond sensible
facts . . . to what is actually given in sense” (1914a: 85–86).10
4 THE LOGICAL CONSTRUCTION OF “THINGS”

Begin with your own sense-data.11 This is your perceived perspective, your private
world. Add to it any sense-data reported by testimony. Is that enough sense-
data to construct material objects? Surely not, for what we know about material
objects goes well beyond what you and I sense. Russell needs more sense-data. So
he hypothesizes sensible objects that go beyond sense-data: sensibilia (1914a: 89).
Imagine sitting next to another person in a room looking at the tables and chairs.
Next imagine the perspective of a third person if a third person sat between the
two of you (1914a: 95). Now imagine that no third person in fact sits between
the two of you, but hold in place the “unperceived” perspective you just imagined.
Russell hypothesizes the reality of this perspective and says this third perspective
consists of sensibilia, not sense-data. They are not sense-data, for no one perceives
them; sense-data are the relata of sense-perception. Sensibilia, on the other hand,
are “objects which have the same metaphysical and physical status as sense-data
without necessarily being data to any mind . . . a sensibile becomes a sense-datum by
entering into the relation of acquaintance” (1914b: 110). Sensibilia are unperceived
sense-data, the constituents of unperceived perspectives. Russell then hypothesizes
458 SKEPTICISM
an infinity of such unperceived perspectives to accompany perceived perspectives.

Russell no longer faces a shortage of sensible objects.
Russell then correlates sensible objects in one perspective with sensible objects
in another. If you are looking at a table (so to speak), and I am too, then though
you and I are sensing different sense-data in different three-dimensional worlds,
our sense-data can be pretty darn similar for all that. And what’s true for tables is
true for everything else we see. “Thus it is possible . . . to establish a correlation by
similarity between a great many of the things of one perspective, and a great many
of the things of another” (1914a: 95–96, emphasis added).12 (Whether this is in fact
possible is an excellent question for another occasion.)
Russell then constructs commonsense things out of classes of similar
sensible objects:13
Given a [sensible] object in one perspective, form the [class] of all the [sensible]
objects correlated with it in all the perspectives; that [class] can be identified with
the momentary commonsense “thing.” Thus an aspect of a “thing” is a member
of the [class] of aspects which is the “thing” at that moment. (1914a: 96)
“A physical thing consists . . . of the whole set of its aspects . . . in all the different
worlds; thus . . . a thing is a . . . set of aspects” (1914a: 117). What’s “really real”
are all the aspects; the “thing” is merely a logical construction.14 A table is then not
something over and above, or standing behind, all of its appearances; the table just
is all of the appearances across perspectives. “The [table] . . . is a class of particulars,
and what I see when I now look at the [table] is a member of this class” (1915: 102).
Russell thereby razors away “things-in-themselves.”
Russell can now explain how two people can “see” the same table (1914a:
83). When two people see two distinct but correlated aspects in their respective
perspectives, they both see appearances of one thing (1914b: 118). So when two
people immediately see two correlated sense-data, they “see” the same “thing” by
immediately seeing distinct aspects of the same “thing.”
5 THE LOGICAL CONSTRUCTION OF SPACE

If that’s how Russell constructs “things,” then how does he construct the public
space, and so the public locations, of tables and trees? Russell’s idea is to construct
a three-dimensional, “all-embracing” public space out of the system of perspectives
(both perceived and unperceived), each with their own “internal” three-dimensional
structure (1914a: 95, 97). In the all-embracing space, the perspectives are the points.
The points then have a three-dimensional location within the system. Every sense-
datum then has a three-dimensional location within the private perspective of which
it is a part, and the three-dimensional location of the perspective within the system
of perspectives. Russell here acknowledges the influence of Leibniz on his logical
construction of space out of private perspectives, where each mind “looks out upon
the world . . . from a point of view peculiar to itself” (1914a: 94).
How does Russell spatially structure the system of perspectives? He places the
perspectives near or far from one another based on their degrees of similarity and
difference that turns on the similarity or difference of their constituent sensible

objects. The more similar, the closer they are; the more different, the farther away:
Suppose . . . we start from one [perspective] which contains the [circular, not
elliptic] appearance of a . . . penny. We can then form a series of perspectives
containing a graduated series of circular aspects of various sizes: for this purpose
we only have to move (as we say) towards the penny or away from it. The
perspectives in which the penny looks circular will be said to lie on a straight line
in perspective space, and their order on this line will be that of the sizes of the
circular aspects. (1914a: 97–98)
And for any two similar perspectives, we can insert an intermediate perspective,
thus making the all-embracing space continuous and three dimensional.
Russell can specify the location of “things” within the space of physics. Take the
line Russell just formed of circular aspects of the penny. Now take another straight
line of aspects of the penny, “in which the penny is seen ends-on and looks like a
straight line of a certain thickness . . . [T]hese two lines will meet in a certain place in
perspective space, i.e. in a certain perspective.” Russell singles out that perspective
as the location of the penny in perspective space (1914a: 98). The “thing” is located
in the perspective that is at the intersection of two (or more) straight lines of
perspectives that contain series of aspects of the “thing” (1914a: 98).15
A perspective in which a “thing” looks large (where the aspect of the thing
perceived is large) is then closer to the “thing” than a perspective in which the
“thing” looks small, for the perspective in which the “thing” looks larger is in fact
“nearer to the perspective which is the place where the thing is.” “[T]he perspectives
in which the penny looks big will be said to be nearer to the penny than those in
which it looks small” (1914a: 98–99).
Which means where you are—where “here” is, spoken in your voice—is the
location of your current perspective. “Here,” spoken in my voice, is the place in
perspective space occupied by my current perspective (1914a: 99–100). If the
penny is located in the perspective in perspective space right next to your private
perspective, then, on Russell’s view, you are in fact right next to the penny.
Each aspect of a thing now has three locations (1914a: 100; 1914b: 117–119;
1959: 107–108). First, there is the “internal” three-dimensional location of the
aspect in the private perspective of the perceiver. Second, there is the public location
of the thing of which the aspect is a part; that place is where the aspect is at. Third,
there is the public location from which the aspect is perceived. That place is the
location of the perceiver’s perspective.
Russell can now capture another commonsense point: we typically see material
objects from a distance. The place from which I see a table (the place where I am)
is distinct from the place where the table is at. If we see things by seeing aspects of
things, and aspects of things have two locations in the system of perspectives, then
the place from which I see a thing (the place of my perspective) is usually distinct
from the place at which the thing I see resides (the perspective that is the location
of the thing).16
460 SKEPTICISM
To summarize, Russell has constructed inter-subjective, public objects (things)

in an inter-subjective, public space out of a plurality of subjective, private objects
in subjective, private spaces, where the entire construction of material objects and
locations in space turns on relations of similarity and difference among the plenum
of sensible objects: “[We thereby] construct the one all-embracing space of physics”
(1914a: 118–119).
6 EPISTEMOLOGY AS LOGICAL CONSTRUCTION

If you have gotten this far, you are past the hard part. Now it is time to turn
to epistemology.
Russell adopted a largely non-skeptical methodology: he took it for granted that
we know most of what we think we know; there is no point in trying to stand outside
of our knowledge and evaluate all of it at once. Instead Russell claimed epistemology
explains how we know what we know: our existing system of knowledge comprises
the “data” that are the business of the philosopher to analyze.
To “analyze” our system of knowledge, Russell distinguishes “hard data” from
“soft data” by degrees of certainty. “Hard” data include instances of knowledge
that resist “the solvent influence of critical reflection.” “Soft” data include instances
of knowledge that, “under the operation [of critical reflection], become to our
minds more or less doubtful” (1914a: 77–78). Hard data are rationally indubitable;
soft data are not. Hard data need no justification; they are completely self-evident
(1914a: 75, 77). Soft data, however, “though entitled to a certain level of respect,”
are not “on a level with” hard data.
Sorting our knowledge in two is the first step in analysis. The second step is
to see to what extent the soft data may be derived from the hard data; this is the
reconstructive step. Russell’s philosophical hope was that the hard data “may prove”
the soft data “to be at least probable.”
Sense-data fall within our hard data. “The immediate facts perceived by sight or
touch or hearing do not need to be proved by argument, but are completely self-
evident” (1914a: 75, 77). For Russell the “hardest of the hard data [include] . . .
our own sense-data” (1914a: 78–79). “The more we reflect on these, the more
we realize exactly what they are, and exactly what a doubt concerning them really
means, the more luminously certain do they become” (1914a: 77–78).17 Russell
also included “some facts of recent memory . . . introspective facts . . . spatial and
temporal relations . . . and some facts of comparison . . . among our hard data”
(1914a: 19).
Among our soft data, Russell includes our knowledge of material objects.
Knowledge of other minds also belongs in our soft data. And since belief in what
is “reported by the testimony of others, including what we learn from books, is
involved in the doubt as to whether other people have minds at all,” it too falls
within our soft data (1914a: 79).18
Russell thus adopts a traditional, foundationalist ideal for our knowledge. Ideally,
we start with our knowledge of sense-data and then reconstruct our knowledge of
material objects, other minds, and the contents and veracity of testimony.
7 AN EPISTEMOLOGICAL FAILURE?
Pulling this off won’t be easy. “[T]he world from which our reconstruction is
to begin is very fragmentary. The best we can say for it is that it is slightly more
extensive than the world at which Descartes arrived by a similar process” (1914a:
79–80). Can Russell’s hypothetical construction show how we know about external
objects from knowledge of sense-data? To put the issue in Russell’s words, now that
we have “constructed a largely hypothetical picture of the world, which contains
places and experienced facts . . . [do] we have any good reason to suppose that it is
real?” (1914a: 100). Many epistemologists would love to see Russell pull the rabbit
out of the hat. Russell clearly would love to do it, too.19
To do it, Russell would need to infer “the sense-data of other people, in favor
of which there is the evidence of testimony, resting ultimately upon the analogical
argument . . . and the ‘sensibilia’ which would appear from places where there
happen to be no minds” (1914b: 116). Unfortunately, Russell does not think much
of the analogical argument: “[T]he hypothesis that other people have minds . . .
[is not] . . . susceptible of any very strong support from the analogical argument”
(1914a: 101–103).
But then Russell immediately adds:
In actual fact, whatever we may try to think as philosophers, we cannot help
believing in the minds of other people, so that the question whether our belief
is justified has a merely speculative interest. And if it is justified, then there
is no further difficulty of principle in the vast extension of our knowledge,
beyond our own private data, which we find in science and common sense.
(1914a: 103–104, emphasis added)
If it is justified?! But we wanted to know whether it is justified.
Even worse, the very idea of knowing about material bodies on the basis of
testimony, and knowing testimony on the basis of other minds, puts the wrong
foot forward: for we know about other minds on the basis of their behavior, and so
of their bodies. If we didn’t already have some knowledge of material bodies that
already goes beyond our knowledge of our own sense-data, then we’d be very hard
pressed indeed to know about their behavior and so to know about their minds.
(Cf. Soames (2014).)
Even so, Russell does not abandon belief in other minds. Instead he retains it as
a working hypothesis. For the hypothesis of other minds
systematizes a vast body of facts and never leads to any consequences which there
is reason to think false. There is therefore nothing to be said against its truth, and
good reason to use it as a working hypothesis. When once it is admitted, it enables
us to extend our knowledge of the sensible world by testimony, and thus leads
to the system of private worlds we assumed in our hypothetical construction.
(1914a: 103)
If we hypothesize other minds, we can use their testimony to construct material
objects and the space of physics. But then reconstruction as epistemology grinds to
462 SKEPTICISM
a halt, for we cannot just hypothesize other minds when reconstructing “soft data”
from “hard data.”
What about sensibilia? In OKEW he does not return to this question. Perhaps
we know about these from experience, by knowing about what things would
look like were I in a different position. But in RSDP he says the inference “raises
much more serious questions” than the inference to other minds. Instead he says
“I should regard these supposed appearances only in the light of a hypothetical
scaffolding, to be used while the edifice of physics is being raised. . . . These
‘sensibilia’. . . are therefore to be taken rather as an illustrative hypothesis. . .”
(1914b: 117). But then Russell’s epistemological optimism must grind to a
complete halt indeed.
Russell leaves us with two “working hypotheses” on which the entire logical
construction of the external world rests. Dashed is the hope of justifying our
knowledge of the external world from our sense-data, of deriving the “soft” data of
“things” from the “hard data” of sense-perception. The foundationalist project in
epistemology is just as disappointing as it is tempting.
8 CONCLUDING AFTERWORD:
RUSSELL’S ACHIEVEMENTS?
Russell’s logical reconstruction of the external world has often been read solely
as an attempt to reconstruct our knowledge from more certain foundations.
Russell certainly hoped that such a reconstruction might succeed, as we’ve seen.
Part of the influence of OKEW was to inspire others to do just that. But Russell
was clearly aware of the limitations of his own efforts. But if it so evidently fails
as an epistemological reconstruction of our knowledge, a failure that invites the
specter of radical skepticism, should we condemn Russell’s logical construction as
a complete failure?
Reconstructing our knowledge was not his sole motivation. He had at least
four others. One was to explain how sense-data depend upon physiology: if the
objects of sensation depend upon the state of our brains, then sense-data depend,
in some sense, on physiology. Another was to reconcile physics and psychology:
verification depends upon sense-data, but the world of physics radically departs
from the appearances of the senses. The third was to reveal the structure of our
knowledge and relations of dependence of one kind of knowledge on another, a
point he emphasizes in the last lecture of OKEW. A fourth was to solve puzzles that
arise from his direct realist view of sense-data as physical.
Let me say a few words about this last motivation. Russell’s Cambridge teacher,
G.F. Stout (1860–1944), also accepted the argument from perceptual variation. But
he did not think the immediate object of perception could be physical. For one
person may “see” a round sense-datum in one location, and another an elliptical
sense-datum in that very same location, but it would be an impossible contradiction
to suppose that an elliptical thing and a circular thing might exist in the same place
at the same time.20
Russell partly developed his hypothetical construction to avoid this problem for
his direct realism21:
What we call the different appearances of the same thing to different observers are
each in a space private to the observer concerned. No place in the private world
of one observer is identical with a place in the private world of another observer.
There is therefore no question of combining the different appearances in the one
place; and the fact that they cannot all exist in one place affords accordingly no
ground whatever for questioning their physical reality. (1914b: 113–114)
When concluding Lecture III of OKEW, Russell takes his hypothetical construction
to have the advantage of dispensing with such arguments, and so rebutting arguments
against our knowledge of the external world:
Difficulties . . . [that arise] from the differences in the appearance which one
physical object presents to two people at the same time . . . have made people
doubtful how far objective reality could be known by sense at all, and have made
them suppose that there were positive arguments against the view that it can be
so known. Our hypothetical construction meets these arguments. (1914a: 104)
Russell acknowledges that the results of his reconstructive efforts are somewhat
meager, failing to achieve his ambitious foundationalist aims. But he finds his four
other results worth the price of admission. For his hypothetical construction
shows that the account of the world given by common sense and physical science
can be interpreted in a way which is logically unobjectionable, and finds a place
for all the data, both hard and soft. It is this hypothetical construction, with
its reconciliation of psychology and physics, which is the chief outcome of our
discussion. (1914a: 104)
NOTES
1. In preparing this chapter, I benefited in particular from reading Charles Fritz (1952),
Peter Hylton (1992), Chris Pincock (2006), Gregory Landini (2011), Gary Hatfield
(2013), and Scott Soames (2014). I am grateful to support from the Templeton
Foundation during a visit to Oxford University where I drafted this chapter. I am
also grateful to comments from the editors on the penultimate draft that led to
improvements.
2. On the origins of the phrase “sense-data,” see Hatfield (2014: 951–952). The first
time G.E. Moore used the phrase was in his 1905 paper “The Nature and Reality of
the Objects of Perception.”
3. The argument appears prominently in Hume, Enquiry Concerning Human
Understanding, 1758, sec. XII. For criticisms of the argument and related arguments,
see J.L. Austin (1962), Frank Jackson (1977), and Michael Huemer (2001).
4. Witness C.D. Broad: “All of the objects . . . that are directly perceived by anyone from
anywhere under any conditions must certainly exist and have the qualities that we
perceive them to have at least as long as we perceive them” (1914–1915: 232).
464 SKEPTICISM
5. Even dreaming involves veridical sense-perception in Russell’s sense. “Objects of

sense, even when they occur in dreams, are the most indubitably real objects known
to us” (1914a: 85–86). The error we make in dreaming is not that we perceive or
sense things that are not there, but that we make false predictions about other sense-
data we are likely to perceive. Dreams, unlike waking experience, are irregular and
unpredictable from experience. Echoes of Berkeley.
6. The distinction between the act and the object of sensation was advanced by
Samuel Alexander (1859–1938) and was fundamental to G.E. Moore’s paper,
“The Refutation of Idealism” (1903). In his later “neutral monism” phase, Russell
abandons the distinction, preferring to construct even mental acts out of sense-data
(what he later calls percepts).
7. For discussion of Russell on acquaintance, see Roderick Chisholm (1977) and David
Pears (1981).
8. Russell is of course aware of the other senses. And so our private worlds are even more
complex: “there are several spaces for each person according to the different senses
which can be called spatial” (1914a: 109). Russell emphasizes the space and the role
experience plays in correlating locations in visual space with locations in tactile space.
9. Russell acknowledges Frege’s 1884 priority (1959: 70). Russell’s prime motive was
ontological, to “get rid of numbers as metaphysical entities” (1959: 71).
10. Russell had a related semantic motive:
in so far [as] . . . common sense is verifiable, it must be capable of interpretation
in terms of actual sense-data alone. The reason for this is simple. Verification
consists always in the occurrence of an expected sense-datum. . . . Now if an
expected sense-datum constitutes a verification, what was asserted must have
been about sense-data; or, at any rate, if part of what was asserted was not
about sense-data, then only the other part has been verified. (1914a: 88–89,
emphasis added)
Here Russell seems to endorse the verifiability theory of meaning. Soames (2014)
makes a big deal out of this in his interpretation of Russell. And it seems to follow
from Russell’s doctrine that understanding requires acquaintance. However, Russell
himself doesn’t make much of this in Our Knowledge of the External World or even
in RSDP or UCM. Russell seems more driven by ontological considerations. Soames
also thinks this supports reading Russell as a phenomenalist, with counterfactual
analyses of material object statements. See Pincock (2006) for some critical discussion
of Soames. I remain agnostic. For recent scholarly discussion of Russell on logical
construction, see also Bernard Linsky (2014).
11. Following Russell, for the sake of simplicity I shall restrict the discussion in what
follows to visual sense-data: colors, shapes, sizes, and distances.
12. How do we know that sensible objects across perspectives are similar? We cannot
sense sensible objects outside of our own private perspectives, and so we cannot
know this by sense-perception. But, Russell thinks, we can know it by testimony.
I shall return to this, for it raises obvious difficulties for his epistemological project.
13. I table two important issues that Russell treats at some length. The first involves
unusual viewing circumstances, such as wearing blue-tinted glasses or physiological
abnormalities, like being sick (1914a: 85–88). The second involves sense-data
of bodily motion: when I verify a table by walking up to it, not only my visual
sense-data change, but so do my tactile sense-data and my sensations of my bodily

movement. Thus to say “that is a table” is not only to talk about visual sense-data,
but correlated sense-data of muscular and bodily movement (1914a: 84–85).
14. This is how Russell constructs material objects in Lecture III of Our Knowledge
of the External World. But in Lecture IV he notes that resemblance is clearly not
sufficient for two aspects being members of the same thing: “as the Comedy of
Errors illustrates, we may be led astray if we judge by resemblance. This shows
that something more is involved for two different things may have any degree of
likeness up to exact similarity” (1914a: 113). He then argues that continuity won’t
do either. He then seems to settle on a different criterion altogether: conformity
to the laws of physics. “Things are those series of aspects which obey the laws of
physics” (1914a: 115–116).
15. For discussion, see Gary Hatfield (2014: 959–960).
16. Russell relied on these two locations to construct a six-dimensional space of
perspectives: each aspect has a three-dimensional location as the place from
which the aspect is perceived, and another three-dimensional location at which
the aspect resides. Aspects then have six-dimensional locations within the system
of perspectives.
17. Russell believed it would be “pathological” to doubt our sense-data (1914a: 78).
18. Without knowledge of other minds, “testimony heard or read is reduced to noises
and shapes, and cannot be regarded as evidence of the facts which it reports”
(1914b: 89–90).
19. “A complete application of the method which substitutes constructions for inferences
would exhibit matter wholly in terms of sense-data, and even, we may add, of the
sense-data of a single person, since sense-data of others cannot be known without
some element of inference. This, however, must remain for the present an ideal. .
. It would give me the greatest satisfaction to be able to dispense with it, and thus
establish physics on a solipsistic basis” (1914b: 116).
20. Stout (1908–1909: 238). This debate is discussed at length in Nassim (2008).
See also Hatfield (2013).
21. Russell was partly inspired here by T. Percy Nunn (1909–1910).
REFERENCES
Austin, J. L. 1962. Sense and Sensibilia. Oxford: Clarendon Press.
Broad, C. D. 1914–1915. “Phenomenalism,” Proceedings of the Aristotelian Society, 15:
227–251.
Carnap, Rudolf. 1928. Der Logische Aufbau der Welt. Leipzig: Felix Meiner Verlag.
Chalmers, David. 2012. Constructing the World. New York: Oxford University Press.
Chisholm, Roderick. 1977. “On the Nature of Acquaintance: A Discussion of Russell’s
Theory of Knowledge.” In G. Nakhnikian (ed.), Bertrand Russell’s Philosophy, 47–56.
London: Duckworth.
Hatfield, Gary. 2013. “Psychology, Epistemology, and the Problem of the External World:
Russell and Before.” In E. Reck (ed.), The Historical Turn in Analytic Philosophy, 171–
200. New York: Palgrave MacMillan Press.
466 SKEPTICISM
Hatfield, Gary. 2014. “Perception and Sense-Data.” In M. Beaney (ed.), The Oxford
Handbook of the History of Analytic Philosophy, 948–973. Oxford: Oxford University
Press.
Huemer, Michael. 2001. Skepticism and the Veil of Perception. Lanham, MD: Rowman
and Littlefield.
Hume, David. 1758. Enquiry Concerning Human Understanding. Edinburgh.
Hylton, Peter. 1992. Russell, Idealism, and the Emergence of Analytic Philosophy. Oxford:
Clarendon Press.
Jackson, Frank. 1977. Perception. Cambridge: Cambridge University Press.
Landini, Gary. 2011. Russell. New York: Routledge.
Linsky, Bernard. 2014. “Russell’s Theory of Descriptions and the Idea of Logical
Construction.” In M. Beaney (ed.), The Oxford Handbook of the History of Analytic
Philosophy, 407–429. Oxford: Oxford University Press.
Moore, G. E. 1903. “The Refutation of Idealism,” Mind 12: 433–453.
Moore, G. E. 1905–1906. “The Nature of Reality and the Objects of Perception”
Proceedings of the Aristotelian Society 6: 68–127.
Nassim, Omar. 2008. Bertrand Russell and the Edwardian Philosophers: Constructing the
World. New York: Palgrave MacMillan Press.
Nunn, T. Percy. 1909–1910. “Are Secondary Qualities Independent of Perception?”
Pears, David. 1981. “The Function of Acquaintance in Russell’s Philosophy,” Synthese 46:
149–166.
Pincock, Chris. 2006. “Rejoinder to Soames,” Russell 26: 77–86.
Quine, W. V. O. 1969. “Epistemology Naturalized.” In his Ontological Relativity and
Other Essays, 69–90. New York: Columbia University Press.
Russell, Bertrand. 1912. Problems of Philosophy. London: Williams and Norgate.
Russell, Bertrand. 1914a. Our Knowledge of the External World as a Field for Scientific
Method in Philosophy. Chicago and London: Open Court.
Russell, Bertrand. 1914b. “The Relation of Sense-Data to Physics,” Scientia 16. Reprinted
in Russell 1918.
Russell, Bertrand. 1915. “The Ultimate Constituents of Matter,” The Monist 25: 399–417.
Reprinted in Russell 1918.
Russell, Bertrand. 1918. Mysticism and Logic and Other Essays. London: Allen & Unwin.
Russell, Bertrand. 1959. My Philosophical Development. London: Allen & Unwin.
Russell, Bertrand and Alfred North Whitehead. 1910–1913. Principia Mathematica,
3 volumes. Cambridge: Cambridge University Press.
Soames, Scott. 2014. The Analytic Tradition in Philosophy, volume 1: The Founding Giants.
Princeton: Princeton University Press.
Stout, G. F. (1908–1909). “Are Presentations Mental or Physical? A Reply to Professor
Alexander,” Proceedings of the Aristotelian Society 9: 226–247.
CHAPTER THIRTY-FOUR
Moore and Mooreanism

ANNALISA COLIVA
1 INTRODUCTION
In this chapter I concentrate on G. E. Moore’s “Proof of an External World,”1 which
first appeared in 1939. Since then, it has been the object of different and contrasting
interpretations and is nowadays at the core of a large debate in epistemology. In
Section 2, I present the paper to place it in its proper context. In Sections 3–5,
I consider some influential interpretations of it. In so doing I assess all of them
from a historical point of view, pointing out how they are all somewhat wanting
as renditions of Moore’s strategy. Finally, in Section 6, I put forward my own
interpretation of it.
2 MOORE’S PROOF OF AN EXTERNAL WORLD

Moore’s proof of an external world is often presented without mentioning its
original context and as if it was directed against skepticism about the external
world. However, “Proof of an External World” (PEW) is a long essay divided into
two parts. In the first one, Moore takes his lead from Kant’s famous observation, in
the Critique of Pure Reason:
It still remains a scandal to philosophy . . . that the existence of things outside of
us . . . must be accepted merely on faith, and that, if anyone thinks good to doubt
their existence, we are unable to counter his doubts by any satisfactory proof.
(Kant 1929: B xxxix )
Moore claims that Kant was not able to give a successful proof of the existence of
things outside of us and that his own proof will remedy the situation. However,
before presenting it, he introduces a series of terminological distinctions, meant to
clarify the meaning of the expression ‘things outside of us’. According to Moore, the
philosophical tradition in general, and Kant in particular, erroneously believe that
the following expressions are equivalent:
A. ‘Things outside of us’.
B. ‘External things’.
468 SKEPTICISM
C. ‘Things which are external to our minds’.

D. ‘Things which can be met in space’.
E. ‘Things presented in space’.
According to Kant, all these locutions are synonymous because they make reference
to phenomena as opposed to noumena. The former are necessarily presented in
space—the pure form of sensibility that allows us to perceive outer things. In
contrast, according to Moore, these expressions are not equivalent because he
does not subscribe to Kant’s transcendentalism, either about empirical objects, or
about space.
According to Moore, there may be things that are presented in space and yet
cannot be met within it. For instance, pains or itches are presented in a part of one’s
body, yet cannot be met in space. Moreover, according to Moore, animals’ pains
are external to our minds, yet cannot be met in space. Since, for Moore, ‘physical
object’ means “an object which exists independently of being perceived by us (human
beings),” he thinks that by giving a proof of the existence of physical objects he will
ipso facto prove that there are things that can be met in space and that are external to
us. However, this latter claim is prima facie problematical. For Moore himself points
out that animals’ pains are external to human minds as we cannot perceive them.
Still, it does not follow from this that they are what we intuitively regard as physical
objects. Thus, the correct claim is that a physical object is everything that we could
perceive, which, however, exists independently of the fact that we actually perceive
it. With this clarification in hand, let us now turn to the proof itself.
By holding up his hands in front of himself and in clear view, Moore makes a
gesture with the right hand and says:
(1) Here is a hand.
Then, making the same gesture with the left hand, he says:
(2) Here is another.
He then concludes:
(3) There are at present two human hands.
Since the conclusion concerns the existence of objects which can be met in space,
Moore claims that (3) entails
(4) There are physical objects.
and hence, that he has proved
(5) There is an external world.
Still, it is clear that an idealist could concede both premises—(1) and (2)—and the
intermediate conclusion—(3)—and yet deny that that entails that there are physical
objects—(4)—, if by ‘physical object’ one meant objects existing independently of
being perceived by us. Hence, an idealist would not take Moore’s performance to
show the truth of (5).2
MOORE AND MOOREANISM 469
Moreover, it must be noticed that up to this point nothing has been done to show
that the premises are known to be true and are not merely assumed to be such; nor
to show that the conclusion of the argument is known. Hence, up to now, Moore’s
proof has no bearing whatsoever against skepticism. In fact, some years later, Moore
himself maintained, in response to his critics, that his proof was directed merely
at an idealist and not against a skeptic (1942: 668). For, in his opinion, in order
to take issue with a skeptic he should have proved that he knew its premises, and
therefore, that he was not dreaming. Yet, Moore himself acknowledged (1959a:
149) that he could not have proved such a thing. For all his evidence would have
been compatible with the hypothesis that he might be dreaming of it.
Why is it that most readers of Moore’s paper have taken his proof as directed
at a skeptic?3 Furthermore, given Moore’s explicit pronouncements, is this
reading legitimate? In order to answer both these questions, we have to take into
consideration the sequel of Moore’s paper, where he claims that his proof is a
rigorous one because:
(a) The premises are different from the conclusion;
(b) they are known to be true and are not merely believed to be true;
(c) the conclusion really follows from the premises.
For, given (b) and the fact that the inference is valid, it follows that also the
conclusion of the argument is known.4 Hence, if it is true that Moore knew that
there were two hands, it follows that he also knew that there was an external world,
and this is clearly an anti-skeptical thesis.
This, however, raises the following issue: how could Moore maintain that he
knew that his premises were true, from which it follows that he also knew the
conclusion of his argument, while holding that he was unable to prove that he knew
them, and that that was necessary in order convincingly to oppose skepticism?
The most charitable interpretation of Moore’s claim, which can also explain
the interest Moore’s work stirred in other philosophers such as Wittgenstein, is as
follows. If one is a philosopher of common sense, it does not matter how much a
skeptic can press one to give a justification for one’s claims to knowledge. Hence,
it does not matter if one does not know how one knows that here there are two
human hands, or, more precisely (cf. Section 5), if one cannot prove that one
knows it. For such ignorance is entirely consistent with the fact that one does know
such a thing. In support of this interpretation, consider what Moore in effect writes
in PEW:
I certainly did at the moment [in which the proof was given] know that which I
expressed by the combination of certain gestures with saying the words “There is
one hand and here is another.” . . . How absurd it would be to suggest that I did
not know it, but only believed it, and that perhaps it was not the case! (1959a: 146)
I can know things, which I cannot prove; and among things which I certainly
did know, even if (as I think) I could not prove them, were the premises of
my . . . [proof]. (1959a: 150)
470 SKEPTICISM
We will come back to this issue in Section 6. Before doing so, however, let us
now turn to some influential interpretations and assessments of Moore’s proof.
3 THOMPSON CLARKE AND BARRY STROUD: AT

THE ORIGINS OF CONTEXTUALISM
Thompson Clarke (1972) introduces a distinction between plain talk and
philosophical talk. In his view, the former is what is produced within all our usual
linguistic practices, with their characteristic embedment in non-linguistic activities.
The latter, in contrast, is what is produced while doing philosophy. Philosophical
talk extrapolates from any ordinary practice and from non-linguistic activities, to
consider language in its own right. While in plain talk the conditions of meaningful
discourse are subject to pragmatic constraints—such as relevance, and other forms
of appropriateness (see Grice 1957)—in philosophical talk these limitations are
removed and words are considered as such. Any well-formed sentence of natural
language can be subject to philosophical analysis. A philosophical question, like the
skeptical question about the foundations of our knowledge, is formulated within
philosophical talk and, according to Clarke, is perfectly legitimate, since it satisfies
a deep intellectual need that is not fulfilled by any of its counterparts in plain talk
(1972: 292).
Clarke then points out that if Moore’s use is taken to be part of plain talk, it
reveals a philosophical lobotomy—that is, a sort of split-brain in one’s philosophical
pronouncements—since Moore means to oppose philosophical theses. If, in
contrast, it is taken as part of philosophical talk, it is dogmatic, since it does not face
the skeptical challenge of explaining how he knows that there are two hands where
he seems to see them (and consequently, that there is an external world) and simply
counters “Because I do.”
Barry Stroud’s interpretation of Moore’s proof closely resembles Clarke’s.
However, according to Stroud, Moore’s proof is given within plain talk. If so, it is
a good proof, but, obviously, it cannot have any anti-skeptical bearing, since it does
not even face the skeptical challenge. From the inside, it is a good proof because
it appeals to the greater degree of certainty possessed by the premises of Moore’s
argument over the degree of certainty possessed by those premises, which, within
plain talk, would be necessary in order to maintain that he may only be dreaming of
having two hands. Still, none of this shows that a doubt about Moore’s knowledge
of his premises is impossible. Hence, the proof fails to address the philosophical
issue and, for this reason, when it is considered from the outside, in the way in
which confronting the skeptical challenge requires, it cannot be successful.
Hence, it is possible to doubt from the inside only when there are actual reasons
to doubt that p, or when there are stronger reasons to doubt it than the ones one can
produce in favor of one’s claim to know that p. Since, however, with respect to the
premises of Moore’s proof there are no such reasons—or at any rate, they are not
stronger than the ones in favor of holding those premises—any form of skepticism
appears, from the inside, totally misguided.
From the outside, in contrast, any doubt is legitimate, inasmuch as it is possible

or conceivable. Thus, it is perfectly right to doubt that the premises of Moore’s
argument are known because one can raise hypotheses, such as the one from
dreaming, that would call into question any supposed instance of sensory-based
knowledge regarding physical objects. Therefore, Moore, by failing to show why
a merely possible doubt—like the Cartesian one—is illegitimate or, in effect, no
doubt at all, does not address the skeptic, whose challenge is raised at the purely
philosophical level. According to Stroud, Moore’s strategy is wanting because it
provides no account of why our knowledge from the inside is legitimate and because
it offers no diagnosis and solution (or dissolution) of what, in his view, the skeptical
mistake would amount to.
This way, Stroud introduces a sort of evaluative lobotomy that, however, can
help clarify the relationship between ordinary and philosophical discourse. This in
turn is, in Stroud’s view, what is really at stake in the debate over skepticism and
common sense. From such a perspective, Moore’s proof is interesting because—its
failure notwithstanding—it contradicts skepticism while being compatible with it.
It contradicts it because it claims that we do have knowledge both of its premises
and of its conclusion. Still, it is compatible with it because such a claim is made
in a context other than the skeptical one. Similar considerations would apply in
the other direction too—that is, skepticism contradicts common sense, yet it is
compatible with it.5
Hence, on Stroud’s view, Moore’s mistake consists in failing to appreciate that, if
this is the right description of the relationship between ordinary and philosophical
discourse, then his negation of the skeptical thesis cannot be a confutation of
skepticism. Thus, “the price of philosophical skepticism’s immunity . . . would be
the corresponding immunity of all our ordinary assertions to philosophical attack”
(Stroud 1984: 127).
Let us assess Clarke’s and Stroud’s interpretations. In an important letter to
Malcolm (1993: 214), Moore claims that he is using the verb ‘to know’ according to
its ordinary meaning. Furthermore, he maintains that his use of ‘I know’ in relation
to the premises of his proof is meant to engage with his philosophical opponents,
contrary to what Stroud says.6 Hence, on the one hand, according to Moore, it
is with an ordinary conceptual repertoire that one enters philosophical discourse,
contrary to what Clarke claims. On the other, in his view, there is no separation
between philosophy and common sense, in such a way that statements respectively
made in those different contexts could contradict each other yet somehow be
compatible with one another, as opposed to what Stroud holds. These two claims
together amount to the view that, according to Moore, the concept of knowledge is
not context-sensitive.7
Finally, and most importantly, according to Moore’s own understanding of
skepticism, the latter arises when the issue of showing how one can know what
one claims to know, and thus prove that one does indeed know it, is raised. To ask
such a question does not produce a change of context according to Moore. Nor
does it show the context-sensitivity of knowledge. On the contrary, it depends on
making use of an invariant concept and yet asking a different question—not whether
472 SKEPTICISM
it is known that p, but, rather, how one can prove that p is known (if indeed it is).
Moore, moreover, agrees that he is not able to answer such a question. Yet, as we
have repeatedly seen, he refuses to agree with the skeptic that, because one cannot
answer it, it follows that one does not know that p. This, however, is not simply a
dogmatic and unphilosophical position, as Clarke claims.8 Rather, it depends on a
specific conception of the relationship between the conditions of knowledge and
their obtainment and the possibility of proving that they are in fact satisfied.
4 MOORE AND HUMEAN SKEPTICISM: WRIGHT’S

INTERPRETATION OF THE PROOF
Crispin Wright (1985) puts forward another reading of Moore’s proof. According
to him, it is important to make explicit the grounds on which Moore claims to know
the premises of his proof in the circumstances in which the proof was offered. As
is apparent and as Moore himself made clear, in those circumstances, the assertion
“Here is a hand” was based on his sensory evidence.9
According to Wright, if Moore’s grounds are made explicit, it becomes
immediately evident why the proof fails. He reconstructs it as follows:
(i) A given proposition describes the salient aspects of my experience at the
time in question;
(ii) I have a hand;
(iii) Therefore, there is an external world.
In fact (I) amounts to saying that there is a proposition that correctly describes the
relevant aspects of Moore’s experience in the circumstances in which his proof was
given, for example, “I am perceiving (what I take to be) my hand.” According to
Wright, (II) is then inferred from (I), and (III) follows from (II) since a hand is a
physical object. Moreover, given that the premises are known, according to Moore,
so would be the conclusion.
However, on Wright’s view, it is clear that (I) can ground (II) only if one can
already take it for granted that one’s experience is being produced by causal
interaction with physical objects. Hence, any sensory experience can warrant a
belief about empirical objects only if it is already assumed that there is an external
world. However, in order justifiably to go from (I) to (II), one needs already to have
a warrant for (III). Hence, the proof is epistemically circular (or question-begging);
antecedent and independent warrant for (III) is needed in order to have warrant for
(II) in the first place,10 given one’s current sensory experience, as described in (I).
Ironically enough, then, Moore’s proof—on Wright’s understanding of it—rather
than being a response to skepticism, instantiates the template of a powerful form of
skepticism, that Wright calls “Humean,” as opposed to “Cartesian.” The difference
between these two forms of skepticism resides in the fact that while the latter makes
play with uncongenial scenarios, such as the hypothesis that one may be the victim
of a sustained and lucid dream, whereby one would be systematically unable to tell
whether one is dreaming or not, the former does not. From such a starting point,
the Cartesian skeptic then claims that for any specific empirical proposition we
take ourselves to know on the basis of our sensory experience, it is metaphysically
possible that it be produced in a non-standard way. From the impossibility of
excluding that this is the case, he takes it to follow that we do not know any such
empirical proposition and, as a consequence, that we do not know (III)—that there
is an external world (cf. Wright 1985, 2004; Coliva 2008, 2010b).
Humean skepticism, in contrast, merely draws on an epistemic gap between having
a certain kind of evidence and warrantedly forming a belief about a domain that
goes beyond what one’s experience immediately testifies to—a gap that is bridged
by inferences such as the (I)–(II)–(III) argument just offered, or indeed by inductive
inferences—whence the title of “Humean” for this form of skepticism. For, in order
warrantedly to go from the first premise, which is about one’s sensory experience,
to the second, which is about an object whose existence is independent of one’s
experience, warrant for the conclusion of the argument—that there is an external
world—must be independently available. Since, however, by skeptical lights, there
is no way of getting such an independent warrant, the argument fails to provide
warrant for its conclusion. Wright calls this phenomenon “failure of transmission
of warrant.” For the need of an antecedent warrant for the conclusion in order
warrantedly to go from (I) to (II) prevents the warrant one may after all have for
(I)—if somehow the skeptic was wrong in claiming that independent warrant for
(III) could not be attained—from being transmitted to (II) and thus to (III). That
is to say, Moore’s proof cannot give one either a first warrant to believe (III), or
further epistemic support for the warrant one might already have to believe it.11
If Wright’s reading of Moore’s proof were correct, it would be devastating. For,
Moore could insist that despite being unable to prove that he knew his premises
he knew them, and—given the principle of closure—he would indeed know
the conclusion. Yet, he would not be in a position to acquire knowledge of the
conclusion, or indeed somehow enhance it, by running his proof. Interpreted thus,
Moore’s proof would simply be no proof whatever of (III)—namely, that there is
an external world. For, characteristically, proofs are means that allow us to extend
our knowledge from their premises to their conclusions and thus provide us with
reasons for first believing them, or else with reasons that enhance our epistemic
support for believing them. On Wright’s reading of it, in contrast, Moore’s proof
would dramatically fail to do so.
Now, two things are worth pointing out. First, things might be different if Wright’s
attack were meant to impugn merely Moore’s ability to redeem his knowledge of (III),
as Wright’s more recent discussions of Moore’s proof seem sometimes to suggest
(2004: 167, 210–211; 2007), and if this de-coupling could be matched by endorsing
an externalist notion of knowledge (and/or warrant). In such a case Moore’s proof
could establish that its conclusion is known—since its premises would be—and yet,
just as Moore held, fail at proving, against skepticism, that either the premises or the
conclusion be known. We will come back to this issue in Section 6.
Second, Jim Pryor has recently challenged the correctness of Wright’s views upon the
structure of empirical warrants. If Pryor were right, it would certainly come as a relief
for the proof’s prospects of success. Let us now turn to a discussion of this rejoinder.
474 SKEPTICISM
5 MOORE’S COMEBACK: PRYOR’S

DOGMATIST INTERPRETATION
Jim Pryor (2004) claims that the proof is fine, from an epistemological point of
view. In particular, it does not exhibit failure of warrant transmission. For, on
Pryor’s understanding of the structure of perceptual warrants, it suffices, in order to
possess such a warrant, merely to have a certain course of experience while lacking
any reason to doubt that the experience is veridical, at least when perceptually basic
beliefs are at stake.12 Hence, there is no need, on his view, to possess an independent
warrant for the conclusion of that argument, namely, that there is an external world.
Pryor, however, thinks that although Moore’s proof is perfectly in order from
an epistemic point of view, it is not successful against a skeptic. In particular, it is
dialectically ineffective for, according to him, a skeptic will think it is (likely) false13
that there is an external world. For such a reason he will not consider Moore’s
experience as of a hand in front of him sufficient for warrant of the corresponding
empirical belief. Starting with an unwarranted premise, the proof is not able to
confer warrant upon its conclusion and hence to convince a skeptic that there is
an external world. Yet, according to Pryor, skeptical doubts are not legitimate. He
holds that skepticism is a “disease” we should not catch, or else cure ourselves of
(Pryor 2004: 368).
Is this interpretation correct? In fact, Moore never explicitly argued for the
view about perceptual warrants that Pryor sees as the key to the proof’s epistemic
success. No doubt he made explicit the grounds of his proof—that is, the fact
that he believed its premises on the basis of his perceptual evidence. However, he
never said that that would be sufficient, by itself, to give one a warrant—or indeed
knowledge—of certain propositions about physical objects such as “Here is a hand.”
On the contrary, one of the main tenets of Moore’s philosophy of perception is his
appeal to sense data. Now, it merits emphasis that any such account would entail
a conception of perceptual warrant (and therefore knowledge) as dependent on
some extra element beside the occurrence of the sense-datum itself—at least unless
the latter were taken to be identical with some part or other of a physical object (a
view considered and rejected in Moore 1959a: §iv). Such an extra element would
presumably be the (warranted) assumption that the experience one is having be
produced by causal interaction with physical objects.
Indeed, in “Four Forms of Scepticism” (FFS), Moore explicitly said that he agreed
with Russell that propositions about specific material objects in one’s surroundings
are not known immediately, but always on the basis of some “analogical or
inductive argument” (1959b: 226). Thus, it is almost certain that he did not endorse
in PEW14 the conception of the structure of empirical warrants (or of knowledge)
that Pryor proposes.
Moreover, it should be kept in mind that Moore raised the issue of having to
prove that he was not dreaming because he realized that that was what stood in
the way of his opponent’s recognition that the premises of his proof were known.
If so, however, Pryor’s and Moore’s understanding of what would be needed to
confront a skeptic would dramatically diverge. For, according to Pryor’s account,
it would suffice simply to remind him that, in order to possess a perceptual warrant
for “Here is a hand,” it is enough to have a certain course of experience when there
is in fact no reason to doubt that there is an external world and the conditions are
such that no doubt of that kind is reasonable. In Moore’s view, in contrast, it should
be proved that, in those circumstances, such a doubt would be unreasonable. Indeed
Moore thought he could make some gesture in that direction, and accordingly said
that the grounds for thinking that he might have been dreaming of his hand were
weaker than the ones available to him to claim knowledge of his premises, or even
totally non-existent.
True, this move is unsuccessful, once coupled with other things Moore says. For
assume it is legitimate to distinguish between, on the one hand, knowledge and the
conditions of its obtainment and, on the other, the ability to prove that they are
satisfied, as Moore claims. Assume, further, as he argues, that the skeptical challenge
concerns such a proof. Then, merely to insist that one does know that there are
hands and that, given one’s evidence, the hypothesis that one may be dreaming is less
supported, or even unsupported by it, can’t exclude the metaphysical possibility that
it be the case. Hence, Moore did not succeed in proving that he was not dreaming.
Accordingly, he also failed to prove that he knew the premises of his proof, contrary
to what he thought he should have done in order to counter a skeptic. Still, it is
clear that he characterized the skeptical position not as a “disease” but as a genuine
challenge that arises out of asking a kind of question that cannot be answered by
simply exposing the structure of empirical warrants. It is to such a question and to
its legitimacy that we shall now turn.
6 HAVING KNOWLEDGE AND BEING ABLE

TO PROVE THAT ONE DOES
In order to assess Moore’s proof from a historical point of view, we need to consider
the issue of the relationship between, on the one hand, having knowledge and the
conditions of its obtainment and, on the other, the possibility of proving that those
conditions do in fact obtain and thus that one really has that knowledge one takes
oneself to have, in such a way as to be able rationally to claim or redeem it. The
“How do you know?” kind of question that the skeptic typically asks is meant to raise
that issue. Hence, that question, when voiced by a skeptic, is not a simple request of
exhibiting the grounds of one’s knowledge. Rather, it is meant as a request of giving
a proof of what one claims to know. To exemplify once more with the premises of
Moore’s proof, Moore can say that his ground for holding that there is a hand in front
of him is his current visual experience. A skeptic, however, is precisely questioning
Moore’s ability to prove that his experience is indeed veridical in those circumstances.
Hence, according to Moore, the skeptic holds the following view of the
relationship between having knowledge and being able rationally to redeem it:
(1) If you cannot show how you know that p—that is, prove that you do know
it—and cannot, therefore, rationally redeem your claim to knowledge, you
do not know that p.
476 SKEPTICISM
Moore, however, thinks that (1) does not hold because he was in fact endorsing a
somewhat externalist conception of knowledge.15 The caveat is apposite because he
never proposed anything that would suggest his leaning toward one of the views
that, later on, would have been qualified as externalist—for example, reliabilism,
relevant alternative theories, counterfactual analyses, etc. Still, he introduced a kind
of move, which would then become the typical externalist maneuver, that is, the
contention that one can know that p even if one is unable to prove that one does.
The evidence in Moore’s writings that he would deny (1) is plenty.16 For instance,
in PEW, he quite explicitly said that neither he nor anyone else may have been able
to prove the truth of his premises:
[If what is required is a general proof for the existence of physical objects], (t)his,
of course, I haven’t given; and I do not believe it can be given: if that is what is
meant by proof of the existence of external things, I do not believe that any proof
of the existence of external things is possible. (1959a: 149)
Yet, he was adamant that such an impossibility would not have impaired the fact
that he did know them. To repeat the relevant quotation:
I can know things which I cannot prove; and among the things which I certainly
did know, even if (as I think) I could not prove them, were the premises of
my . . . [proof]. (1959a: 150)
In the recent literature on skepticism, it is often remarked that if one makes this
kind of move, one will then face some “crisis of intellectual conscience” or “angst.”17
That is to say, it is conceded that one may know that p, even if one is unable to
prove it. Yet, one would not be able to reassure oneself that one does. Skepticism, on
this reading, would not directly challenge one’s knowledge but, rather, one’s ability
rationally to redeem it. Hence, skepticism would not be characterized by (1), but by:
(2) If you cannot show how you know that p—that is prove that you know it—
and cannot therefore, rationally redeem your claim to knowledge, you may
know that p, but you cannot reassure yourself of it.
Thus, skepticism brings about an “intellectual crisis,” at least prima facie, because,
even supposing that through a gift of nature, as it were, we knew that p, we would
also feel the intellectual need to be able to prove that we really do.
The problem then becomes that of explaining the sources and legitimacy of
this intellectual need. The issue is subtle, yet of the utmost epistemological and
metaepistemological significance. In particular, a supporter of this view will have
to claim that we have just an externalist notion of knowledge—which makes us
hold that we know that p even if we cannot say how—and then a sort of self-
reflective spontaneous attitude which forces us to look for an explanation of how
that knowledge may have come about. Skepticism would thus be due to such a
deeply rooted human attitude. Yet, it would have no bearing on whether we possess
knowledge. One may then try to explain why, adaptively, it would be useful for us
to have such an attitude; by saying, for instance, that it forces us to make inquiries,
which sometimes show that we do not really know what we thought we knew.
Hence, that attitude prevents us from indulging in mistakes. Moreover, such

inquiries into the sources of our knowledge may actually increase it, by providing an
explanation of how it comes about. A case in point, which exemplifies both aspects,
is that of chicken sexers. For, by inquiring into their ability, it was discovered that
it is based on the operations of the sense of smell and, thereby, the incorrect belief
that it depended on sight and touch was removed. One may then say that the price
to pay, its advantages notwithstanding, is that at times such a self-reflective attitude
gets in the way of our recognition of our real epistemic status and makes us worry
when there is no need to. The circumstances in which Moore claimed to know
that there was a hand where he seemed to see it would just be an example of this
downside of our self-reflective attitude. For clearly he knew that there was a hand
in what appeared to be cognitively optimal conditions. Yet, if our self-reflective
attitude kicks in and demands that we provide a proof of the fact that those are
indeed optimal conditions, how could we accomplish such a task? We could only set
into operation the same cognitive faculties in the same kind of optimal conditions
for which the problems of how we could prove them to be reliable and optimal,
respectively, were raised. Hence, we would be caught up in a circle that would
prevent us from being able to prove that we know what we take ourselves to know.18
Faced with this charge, an externalist could either say that, provided one does
know that here is one’s hand, one would also be able to prove, via (successive steps
of something like) Moore’s proof, that one does know that one’s sense organs are
working reliably, that one is not dreaming, and so forth. Surely, we will not be able
to settle the issue against an opponent who doubts that our sense organs are working
reliably, that we are not victims of lucid and sustained dreams, etc., because such an
opponent would not concede that we know that there is a hand in the first place. Yet,
this is simply a dialectical failure that does not impugn our knowledge. Furthermore, an
externalist should notice that the skeptic would raise a doubt that, on closer inspection,
is illegitimate because it would be grounded in a mistaken conception of knowledge.
Alternatively, an externalist could concede that we do have knowledge but
cannot actually prove that we know that our sense organs are working reliably
and that we are not victims of lucid and sustained dreams. Moreover, he could
recommend that we had better tame our self-reflective attitude whenever its setting
into motion would get in our way by raising challenges that cannot, in principle, be
met. Nevertheless, both kinds of explanation would also have a metaepistemological
consequence: that of casting doubt—albeit for different reasons—on the legitimacy
of the skeptical challenge.
In my view, Moore somehow anticipated, in many respects, the kind of
approach that epistemic externalists have developed after him. He did not offer
a diagnosis of why the skeptical challenge would be illegitimate by the lights of an
externalist epistemologist, though. This may explain both why he was criticized by
his contemporaries, who were rooted in an internalist conception of knowledge,
and why, also nowadays, quite independently of one’s epistemological preferences,
which may even go in that very same direction, one may find Moore’s strategy
somewhat unconvincing. Yet, it is obvious that he had the great merit of initiating
the kind of epistemology that is nowadays dominant.
478 SKEPTICISM
NOTES
1. It is impossible to do justice to all of G. E. Moore’s epistemological writings in the
space allotted. For a fuller analysis, see Coliva (2010a: ch. 1).
2. Alternatively, if by “physical objects” one meant something compatible with idealist
theses, such as objects that afford the possibility of occurrent perceptions, (4)
would follow, but not (5), if by “external world” one meant a world populated by
objects that exist independently of being perceived by us. Obviously, since Moore
has painstakingly defined “physical object” and “external world,” the alternative
reading just presented is not the intended reading of Moore’s proof, as he himself
made clear in Moore (1942: 669–70).
3. A notable exception is Sosa (2007: 52).
4. Unless one denied the principle of closure for knowledge—according to which, if
you know that p and you know that p entails q, you know that q. Dretske (1970)
and Nozick (1981) denied closure, but Moore never proposed such a thing.
5. Stroud’s claims are dubious. For, if “knowledge” is context-sensitive in the way
proposed, there would be no real contradiction between skepticism and common
sense because, in order to have a contradiction, p should be both known and not
known in the same context.
6. For a similar objection to Stroud’s interpretation of Moore, see McGinn (1989: ch. 3).
7. Travis (1989: 165–166), who is sympathetic to Moore’s claim that he knows
the premises of his proof, and who supports contextualism about knowledge
ascriptions, recognizes this point and deems it the source of the failure of Moore’s
anti-skeptical strategy.
8. McGinn (1989: 49–53) rightly notes that Moore is directly engaging with skepticism
and is not confining his claims to knowledge merely to ordinary contexts. Yet, she
herself shares the view that he is dogmatic and somewhat unphilosophical because
he fails to diagnose what is wrong with skepticism.
9. That Moore was basing his claim on his available evidence may appear obvious, but
it is not within the historiography on the proof. For it has been denied that Moore
made explicit the criterion on the basis of which he could assert his premises. See
Malcolm (1949), Stroll (1994: 50–52), and Sosa (2007).
10. Since warrant is a necessary condition for having knowledge, on the tripartite
conception of knowledge, and nowhere does Moore impugn the tripartite account,
lack of such a warrant would impugn one’s alleged knowledge by Moore’s lights.
Thus, the fact that Wright is talking about warrant while Moore talks of knowledge,
though certainly inaccurate from a historical point of view, makes no substantial
difference from a conceptual one.
11. These qualifications are important in order to clarify the difference between
“transmission failure” and the failure of the principle of closure for epistemic
operators.
12. “Here is a hand” would be such a perceptually basic belief. If one found this
claim odd, Pryor would allow substituting it with “Here is a pinkish expanse.”
I am not sure whether Moore himself would be happy with that substitution.
Nevertheless, the important point (for both) is that the premises of the proof be
about physical objects.
13. This is indeed a contentious rendition of the skeptical position, for a skeptic is
no idealist! Rather, on the basis of philosophical arguments he holds an agnostic
position and, in particular, that it cannot be warrantedly believed either that there is
an external world, or that that there is not. See Coliva (2008; 2010b).
14. But it should be kept in mind that Moore wrote FFS in between 1940 and 1944,
thus after PEW.
15. As Baron Reed has insightfully remarked, traditional foundationalists would reject
(1), too, but that would not make them externalists. Basic beliefs grounded in
the given or in self-presenting states cannot be proven, though they still count as
knowledge. In reply, it should be pointed out that we are concerned with beliefs
about physical objects here and that the basic beliefs foundationalists would normally
appeal to would not be of this kind. Skeptics would then challenge the passage from
having knowledge of these basic elements to having knowledge of physical objects,
based on the former piece of knowledge.
16. Sosa (2007: 50) concurs with this appraisal.
17. The former phrase is Wright’s (2004: 167, 210–211); the latter is Pritchard’s
(2005); similar remarks have been made by Stroud (1994).
18. For a discussion and different appraisals of these forms of “bootstrapping”
arguments, see Sosa (1994), Cohen (2002; 2005), and Vogel (2000; 2008).
REFERENCES
Clarke, Thompson. 1972. “The Legacy of Skepticism,” The Journal of Philosophy 69:
754–769. Reprinted in M. Williams (ed.), Scepticism, 287–302. Aldershot: Dartmouth,
1993.
Cohen, Stewart. 2002. “Basic Knowledge and the Problem of Easy Knowledge,” Philosophy
Cohen, Stewart. 2005. “Why Basic Knowledge is Easy Knowledge,” Philosophy and
Phenomenological Research 70: 417–430.
Coliva, Annalisa. 2008. “The Paradox of Moore’s Proof,” The Philosophical Quarterly 58:
234–243.
Coliva, Annalisa. 2010a. Moore and Wittgenstein: Scepticism, Certainty and Common
Sense. Basingstoke: Palgrave Macmillan.
Coliva, Annalisa. 2010b. “Moore’s Proof and Martin Davies’ Epistemic Projects,”
Australasian Journal of Philosophy 88: 101–116.
Dretske, Fred. 1970. “Epistemic Operators,” The Journal of Philosophy 67: 1007–1023.
Grice, Paul. 1957. “Meaning,” The Philosophical Review 66: 377–388.
Kant, Immanuel. 1929. Critique of Pure Reason, translated by N. K. Smith, London:
Macmillan.
McGinn, Marie. 1989. Sense and Certainty. A Dissolution of Scepticism. Oxford: Blackwell.
Moore, George E. 1942. “A Reply to My Critics.” In P. A. Schilpp (ed.), The Philosophy of
G. E. Moore. Evanston and Chicago: Northwestern University.
Moore, George E. 1949. “Letter to Malcolm.” In his Selected Writings. London and
New York: Routledge, 1993: 213–216.
480 SKEPTICISM
Moore, George E. 1959a. “Proof of an External World,” Proceedings of the British

Academy 25, 1939: 273–300. Reprinted in Philosophical Papers. London: George Allen
and Unwin, 1959: 127–150.
Moore, George E. 1959b. “Four Forms of Scepticism.” In his Philosophical Papers. London:
George Allen & Unwin: 196–226.
Nozick, Robert. 1981. Philosophical Explanations. Cambridge, MA: Harvard University
Press.
Nuccetelli, Susana and Gary Seay. 2007 (eds.), Themes from G. E. Moore. New Essays in
Epistemology and Ethics. Oxford: Oxford University Press.
Pritchard, Duncan. 2005. “Scepticism, Epistemic Luck and Epistemic Angst,” Australasian
Pryor, James. 2004. “What’s Wrong with Moore’s Argument,” Philosophical Issues 14:
349–378.
Sosa, Ernest. 1994. “Philosophical Skepticism and Epistemic Circularity,” Proceedings of
the Aristotelian Society, Supplementary Volume 68: 263–290.
Sosa, Ernest. 2007 “Moore’s Proof.” In S. Nuccetelli and G. Seay (eds.), Themes from G. E.
Moore. New Essays in Epistemology and Ethics, 49–61. Oxford: Oxford University
Press.
Stroll, Avrum. 1994. Moore and Wittgenstein on Certainty. New York: Oxford University
Press.
Stroud, Barry. 1984. The Significance of Philosophical Scepticism. Oxford: Clarendon Press.
Stroud, Barry. 1994. “Scepticism, ‘Externalism,’ and the Goal of Epistemology,” Proceedings
of the Aristotelian Society, Supplementary Volume 68: 291–307.
Travis, Charles. 1989. The Uses of Sense. Oxford: Oxford University Press.
Vogel, Jonathan. 2000. “Reliabilism Leveled,” The Journal of Philosophy 97: 602–623.
Vogel, Jonathan. 2008. “Epistemic Bootstrapping,” The Journal of Philosophy 105:
518–539.
Wright, Crispin. 1985. “Facts and Certainty,” Proceedings of the British Academy 71:
429–472.
Wright, Crispin. 2004. “Warrant for Nothing (and Foundations for Free)?” Proceedings of
the Aristotelian Society, Supplementary Volume 78: 167–212.
CHAPTER THIRTY-FIVE
Wittgenstein
and Skepticism:
Illusory Doubts
MICHAEL WILLIAMS
1 WITTGENSTEIN: LIFE AND PHILOSOPHY

Ludwig Wittgenstein (1889–1951) is widely regarded as one of the pre-eminent
philosophers of the twentieth century. He was born in Vienna, the youngest of eight
children. His father’s family was of Jewish descent but had converted to Protestant
Christianity. His mother had a Jewish father but followed her mother’s Catholic faith,
and the Wittgenstein children were brought up as Catholics. Although Wittgenstein
lost his faith as a schoolboy, questions of religion and personal morality preoccupied
him throughout his life.
Wittgenstein’s father was an immensely wealthy self-made industrialist, and the
family was highly cultivated, with music at the center of its cultural life: Wittgenstein
himself was deeply musical. However, for all its advantages the family was troubled.
Two of Wittgenstein’s brothers committed suicide, a likely factor being the clash
of their musical or literary proclivities with their father’s determination to see
them follow a business career, and Ludwig Wittgenstein had a lifelong struggle
with depression.1 But although often difficult to get along with, he was capable of
forming deep friendships and was a compelling, if unorthodox, teacher.
As a child, Wittgenstein showed a practical bent and was sent to technical high
schools rather than a classical gymnasium. But his sister, Grete, introduced him to
the philosophy of Schopenhauer as well as to contemporary currents in Viennese
intellectual life, and he was uncertain of his suitability for a career in engineering.
Nevertheless, in 1908, he enrolled at Manchester University to pursue research in
aeronautical engineering. However, his interest in mathematics led him to questions
concerning the nature of mathematical truth, the peculiar certainty of mathematical
propositions, and the relation of pure mathematics to logic. Abandoning engineering
for philosophy, in 1911 he moved to Cambridge, where he studied with Bertrand
482 SKEPTICISM
Russell. He also discussed philosophical questions with G. E. Moore and attended

some of Moore’s lectures.
Wittgenstein spent only five terms at Cambridge. At the outbreak of the
First World War in 1914, he left the university without a degree and enlisted in
the Austrian army. Despite hazardous frontline duty, he continued his work in
logic and, by the war’s end, had a close-to-final version of his Tractatus Logico-
Philosophicus (henceforth, TLP). After some difficulties with publishers, the book
appeared in 1922.2 Apart from a short paper, it is the only work he published in
his lifetime.
Following the lead of Frege, Russell had been engaged in an attempt to reduce
mathematics to logic, defining basic mathematical terms in the vocabulary of
pure logic and developing a system of formal proof. But what is logic? What is
a proposition? When does a proposition express a logical truth? Wittgenstein’s
Tractatus offers answers. At the base of language, there are elementary propositions,
saying how things stand with some particular “objects.” Complex propositions are
formed from elementary propositions with the aid of “logical constants” (“not,”
“and,” “or”) and devices of quantification (“all” and “some”). All complex statements
are “truth-functional,” in the sense that the truth-value of any complex statement is
strictly determined by the truth-values of the elementary propositions of which it is
composed.3 The only necessary (or logical) truths are truth-functional tautologies,
such as “Either Baltimore is in Maryland or it is not the case that Baltimore is in
Maryland.” The only necessary falsehoods are truth-functional contradictions, such
as “MW was born in England and it is not the case that MW was born in England.”
Since they are true no matter what or false no matter what, tautologies and
contradictions take no stand on how things are and so fail to express propositions.
But while senseless (sinnlos), they are not nonsensical (unsinnig): they say nothing
but they show—graphically illustrate—the truth-functional characteristics of the
logical constants.
This is bad news for philosophy. Metaphysics seeks truths that are both
substantive and necessary. For example, the Ontological Argument purports to
show that the proposition that God exists is necessarily true. There are no such
truths: the “statements” of metaphysics are nonsensical. The same goes for the
sentences of the Tractatus itself, which are neither tautologies nor statements about
how things stand in the world. Wittgenstein “resolves” the paradox by invoking a
metaphor that Sextus Empiricus uses to characterize skepticism: his “propositions”
are a ladder we climb but throw away once we “see the world aright”; and this is
something we show by avoiding philosophical talk. So too for ethics, religion, and
aesthetics. Ethical, religious, and aesthetic talk is idle chatter: a person’s ethical,
religious, or aesthetic sensibility is shown in her actions and choices. “Whereof one
cannot speak, thereof one must be silent” (TLP 7).
Having put an end to philosophy, Wittgenstein left academic life, spending some
years in the Austrian countryside as an elementary school teacher. However, he
kept some philosophical contacts and began to question central doctrines of the
Tractatus. In 1929, he returned to Cambridge, first as a fellow of Trinity College and
later as professor.4 There he began the inquiries that culminated in his Philosophical
WITTGENSTEIN AND SKEPTICISM: ILLUSORY DOUBTS 483
Investigations (henceforth, PI), published in 1953, two years after his death.5 The
book is a devastating critique of the Tractatus and the approach to philosophy
it exemplifies.6
In the Tractatus, language has a single, fundamental purpose: saying how things
are. Reference—the relation between a name and the object for which it stands—
is the root of all meaning. There is no vagueness. Wittgenstein abandons all this,
offering a new conception of meaning but not a general theory. We must avoid
presuming that language must work a certain way: “don’t think but look” (PI 66).
We then find that language is used in many ways: to issue commands, express hopes
and fears, tell stories, speculate about what might be, and so on. Instead of basing
meaning on one fundamental semantic relation—naming or reference—we should
look at the different ways words are used. Vagueness is inevitable. Language is, or
can be made, precise enough for our undertakings, practical or theoretical, but what
we ought to say in any possible circumstances is not fixed in advance.
Wittgenstein analogizes using words to playing a game. He also invents simple
“language-games” to make vivid his new approach to meaning. Like games, language
is subject to rules: some “moves” are permitted and others not. In the builder’s
game (PI 2), a builder calls for building-materials, and an assistant brings them.
If the builder calls “Slab” and gets a pillar, the assistant has made a wrong move.
However, if the rule were not generally followed, there would be no “game” at
all. This does not mean that the rules of a language-game can be formulated so as
to dictate what to do in every imaginable situation. Perhaps (not Wittgenstein’s
example) the builders come across an item that is rather like a pillar but that could
also be used as a beam: in that case, they would have to decide what to call it.
We should not think that the meaning of a word can always, or even ever, be
captured by a definition. Take “game” itself. Many things are called games: “board
games, card games, ball games, athletic games, and so on” (PI 66). Some games
are competitive, some not. Some you play with other people, some by yourself.
Games are united by “family resemblance,” “a complicated network of similarities
overlapping and criss-crossing” (PI 66).7 Nor can we think of ourselves as everywhere
guided by rules, since an explicit rule would be open to interpretation. Using words
correctly is a kind of know-how: understanding is shown by how a person uses
words in practice.
Even in making statements, words are used in different ways. Suppose the
builders have words for numbers. If we are under the spell of the idea that meaning
is always based on naming, we might think that just as ‘slab’ refers to slabs, number-
words refer to numbers. Since numbers don’t exist in the way that slabs do, we
may be tempted to suppose that they exist in some more ethereal way. But this
metaphysical temptation withers when we see that numerical symbols do not get
their meaning from naming things, as ‘Michael Williams’ names me, but from how
they are used in counting and calculation. We can say that ‘slab’ refers to slabs
and “1” to the number one, but we should not let this blind us to the different
ways in which these expressions function. The Tractatus and the Investigations have
this much in common: in both works, questions about meaning and the “logic” of
language lead to the view that philosophy presents confusions to be dispelled, not
484 SKEPTICISM
problems requiring theoretical solution. This is a kind of skepticism: skepticism

about philosophy as a theoretical discipline.
2 THE ORIGINS OF ON CERTAINTY:

WITTGENSTEIN AND MOORE
Toward the end of Wittgenstein’s life, his former student Norman Malcolm, whom
he was visiting at Cornell, suggested that they meet to discuss issues raised in two
papers by G.E. Moore: “A Defence of Common Sense”8 and “Proof of an External
World.”9 These discussions led Wittgenstein to fill the four notebooks that are the
basis of On Certainty (henceforth, OC).10
While Moore’s two papers are generally seen as directed against skepticism, a
view Wittgenstein shares to some extent, Moore’s immediate concern is to combat
Idealism: Hegelian Idealism in Defence and subjective Idealism in Proof.
In Defence, Moore gives a list of propositions that belong to common sense.
Examples are that he has a body, that he has never been very far from the Earth, and
that the Earth has existed for many years past. Not only are these propositions (or
where personal pronouns are involved, their counterparts) accepted by everyone,
they are wholly true. In making the latter claim, Moore means to contest a central
claim of neo-Hegelian Idealists like his contemporary, the Oxford philosopher F. H.
Bradley: that our present beliefs are only partially true.11 This doctrine of degrees
of truth reflects the Idealist’s explaining both justification and truth in terms of
coherence. A belief is justified if it belongs to a coherent belief system, and truth is
idealized justification: coherence in the limit. Since our belief system is incomplete
and can be extended in various ways, current beliefs, though justified, are compatible
with propositions that are incompatible with each other and thus not wholly true. As
our belief system becomes more comprehensive and integrated, our beliefs become
more true, though only beliefs at the limit of coherence are wholly true. Moore will
have no truck with this. Ever the Realist, he insists that his commonsense beliefs,
here and now, are wholly true.12
In Proof, Moore responds to Kant’s claim that “It still remains a scandal to
philosophy . . . that the existence of things outside of us . . . must be accepted
merely on faith, and that, if anyone thinks good to doubt their existence, we are
unable to counter his doubts by any satisfactory proof”13; and he means to end the
scandal by producing the missing proof. However, Kant says that ‘external object’
is ambiguous, referring either to what exists as a thing in itself, with no logical
relation to human cognition, or to something that “belongs to external appearance.”
Taking the latter phrase to mean “things that are presented in space,” Moore argues
that Kant has already gone wrong. External things, as ordinarily conceived, are not
just presented in space: they are to be met with in space. Such things are essentially
public, and they exist whether or not anyone meets or (given the circumstances)
could meet them. Their existence is not tied to their ever being presented to anyone.
By contrast, after-images are presented in space, in that they seem to float before
our eyes. But they can’t be approached, viewed by different people or from different
angles, or re-encountered at some other time; and as “private,” they exist only
through being presented to someone. Being presented in space is neither necessary
nor sufficient for being a thing to be met with in space. Since Kant fails to make this
vital distinction, his proof fails utterly.
Having clarified what needs to be proved, Moore holds up his hands and, while
making a certain gesture, announces “Here is one hand, and here is another.”
Presumably, Moore’s gesture is intended to make manifest his hand’s character as
a thing to be met with in space, rather than something merely presented in space.
For example, it can be brought closer and viewed (by both Moore and his audience)
from different angles. He concludes that at least two external objects have been
proved to exist. The existence of indefinitely many such things could be proved in
the same way.
Moore argues that his performance really is a proof: his conclusion is different
from his premises, his premises logically imply his conclusion, and he knows his
premises to be true. However, he admits that he has not proved that he knows.
For example, he has not proved, and does not know how to prove, that, while
seeming to hold up his hand, he is not in fact dreaming. So while he has refuted
subjective idealism, succeeding where Kant failed, he has not given a complete
answer to skepticism.14
Rightly, Wittgenstein sees Moore’s Proof and Defence as raising different issues.
In so far as Moore’s Proof engages skepticism, it is the Cartesian problem of our
knowledge of the external world: can we know that external things even exist? The
question raised by his Defence, which extends the bounds of certainty far beyond
the everyday knowledge of things around us that the Cartesian skeptic claims to
doubt, is what justifies the extension. On both topics, Wittgenstein thinks that
Moore goes badly wrong. Cartesian skepticism is not to be met with a proof but
exposed as an illusion. Moore’s blithe assurances that his commonsense certainties
belong to everyone are unfounded. Deep—sometimes intractable—differences of
opinion are a fact of life.
We see, then, that Wittgenstein has two principal aims. The first is to dissolve
Cartesian skepticism. The second is to investigate what to say about certainties
beyond those that a dissolution of the Cartesian problem would secure. We also
see that although Defence was written before Proof, for Wittgenstein the Cartesian
skeptical problem has philosophical priority. Accordingly, he begins his investigation
with the later paper. We shall follow him.
3 SKEPTICISM AND IDEALISM: MOORE’S PROOF 15

In the Tractatus, Wittgenstein dismisses skepticism as obviously nonsensical.
In a way, he retains this view. He also retains the conviction that considerations
pertaining to linguistic meaning hold the key to coming to terms with skepticism,
though he now draws on the resources of the Investigations. The major change is
that he no longer takes skepticism’s illusory character to be obvious. It has to be
shown. Understanding how Moore’s proof misfires will offer essential clues.
486 SKEPTICISM
Wittgenstein is less interested in the “proof” itself than in Moore’s argument in

favor of its being a proof, in particular his claim that he knows that his premises
are true.
1. If you do know that here is one hand, we’ll grant you all the rest.
When one says that such and such a proposition can’t be proved, of course that
does not mean that it can’t be derived from other propositions: any proposition
can be derived from other ones. But they may be no more certain than it is itself.
2. From its seeming to me—or to everyone—to be so, it doesn’t follow that it
is so.
What we can ask is whether it can make sense to doubt it. (OC)
This opening remark is mildly misleading. Moore’s “proof” turns out to fail even if he
does know that here is one hand. Nor is it simply wrong to say that this is something
Moore knows. The problem is that Moore appeals to everyday knowledge in the
context of responding to the skeptic. The skeptic questions whether anyone knows
anything whatsoever about the external world.16 Such a doubt cannot be assuaged
by Moore’s assuring us that he knows that here is one hand. We must take a less
direct approach and ask when doubts make sense.
The hyper-generality of skeptical doubts makes their meaning problematic: “If
e.g. someone says ‘I don’t know if there’s a hand here’ he might be told ‘Look
closer.’—This possibility of satisfying oneself is part of the language-game. Is one
of its essential features” (OC 3). In the absence of any idea of how I might be going
wrong, an expression of doubt would mean nothing. Further, in many everyday
situations, no answer I could give to the question “How do you know?” would be
more certain than the proposition supposedly in doubt. How do I know that here is
one hand? Well, I can see my hand. Is that really an answer? Or is it just a dismissal
of the question?
This line of thought suggests that genuine doubts have grounds: a reasonable
suspicion that I may have made a particular kind of mistake. However, if doubts need
grounds, a general doubt is impossible. The skeptic will demur. I invite his questions
by representing myself as not merely having an opinion but as knowing. If I can’t give
an answer that is any more certain than the proposition whose epistemic credentials
I have been asked to produce, this is grist to his mill. On the other hand, in everyday
life, this is not how the language-game of knowing and doubting is played.
4. What about such a proposition as “I know I have a brain”? Can I doubt it?
Grounds for doubt are lacking! Everything speaks in its favour, nothing against
it. Nevertheless it is imaginable that my skull should turn out empty when it was
operated on.
5. Whether a proposition can turn out false after all depends on what I make
count as determinants for that proposition. (OC)
There are two points here. First, skeptics assume that “How do you know?” is a
permissible question anywhere and everywhere. But if doubts need grounds to make
sense, there is no right to ask epistemic questions no matter what has been claimed,
by whom, or in what circumstances. Second, grounds for doubt are harder to come
by than the skeptic thinks. We can imagine (vorstellen) all sorts of extraordinary
happenings: we can picture them. (Wittgenstein would surely put Demon-deception
and brain-in-a-vat tales in this category.) But ordinarily, far-fetched imaginings are
not treated as grounds for doubt. Nor, for all we have seen, is there any reason why
they should be. In our everyday epistemic practices, we (as everyday practitioners)
set the rules for what counts as a reasonable or even an intelligible doubt. The
question arises, has the skeptic found a paradox in our ordinary ways of talking
about knowledge and certainty? Or is he subtly changing the rules of the language-
game, and if so why?
If the skeptic’s doubts make sense, there must be some possibility of my being
mistaken, even with respect to the most banal and obvious assertions about objects
and events in the world around me. Is this possible?
17. Suppose now I say “I’m incapable of being wrong about this: that is a book”
while I point to an object. What would a mistake here be like? And have I any
clear idea of it?
18. “I know” often means: I have the proper grounds for my statement. So if the
other person is acquainted with the language-game, he would admit that I know.
The other, if he is acquainted with the language-game, must be able to imagine
how one may know something of the kind.
19. The statement “I know that here is a hand” may then be continued: “for
it’s my hand that I’m looking at.” Then a reasonable man will not doubt that
I know. . . . (OC)
If I have a clear idea of how I might be mistaken, I also understand what it would
take to know. Notice that the topic here is Cartesian skepticism, so we are talking
about mistakes as they arise with respect to simple empirical judgments. We are not
talking about just any beliefs that we might judge to be false. For example, we are
not talking about religious beliefs. We might indeed judge such beliefs to be false
and explain how our other person came to hold them, without supposing that there
is any way they could (reasonably) be shown to be true. However in cases like this,
we would not think of the other person as simply making a mistake, in the sense at
issue here. These are matters for later.
Now according to the skeptic, I can be mistaken in thinking “Here is one
hand,” even if I have no idea about how I could come to correct my error. If
Wittgenstein is right to connect reasonable doubt, as ordinarily understood, with
specific ways of going wrong, it would seem to follow that the skeptic’s doubts fail
to be reasonable and perhaps less than fully intelligible. But Wittgenstein is in no
hurry to draw this conclusion. Granted, the reasonable man will not doubt that
I know. However,
19 Nor will the idealist; rather he will say that he was not dealing with the
practical doubt which is being dismissed, but there is a further doubt behind that
one.—That this is an illusion has to be shewn in a different way.
488 SKEPTICISM
Just because the skeptic’s doubts are out of the ordinary, we can’t conclude that
they are senseless or even wholly unreasonable. But neither can we take their
intelligibility for granted.
24. The idealist’s question would be something like: “What right have I got not
to doubt the existence of my hands? (And to that the answer can’t be: I know that
they exist.) But someone who asks such a question is overlooking the fact that
a doubt about existence only works in a language-game. Hence, that we should
first have to ask: what would such a doubt be like?, and don’t understand this
straight off. (OC)
This pivotal thought makes it clear why Wittgenstein has been emphasizing the
difference between everyday and skeptical doubt. Supposedly, skepticism is an
“intuitive” paradox, arising from reflection on ordinary ways of thinking about
knowledge and justification. However, the extraordinary character of skeptical
doubting suggests that it is a language-game of its own, with its own enabling
presuppositions. They must be brought into the open. Perhaps skeptical doubting
will turn out to be a natural extension of ordinary doubting. But then again,
perhaps not.
Wittgenstein works toward an initial diagnosis by addressing the question of my
right not to doubt the existence of my hands:
25. One may be wrong even about “there being a hand here.” Only in particular
circumstances is it impossible.—“Even in a calculation one can be wrong—only
in certain circumstances one can’t.”
26. But can it be seen from a rule what circumstances logically exclude a mistake
in the employment of rules of calculation?
What use is a rule to us here? Mightn’t we (in turn) go wrong in applying it?
(OC)
These remarks introduce three important themes.
The first theme is that, in certain circumstances, a particular claim cannot be
doubted at all because expressions of doubt fail to make sense. In many ordinary
situations, certainty is the default condition. This should be no surprise: unless
most moves were uncontroversially correct, there would be no language-game
in which to make a wrong move. This “transcendental” point stands in sharp
contrast to the skeptic’s claim that while radically skeptical doubts are ignored
for everyday practical purposes, they are always lurking in the shadows, ready
to emerge when we retire to the study. As Wittgenstein sees things, even in my
study I cannot really doubt the existence of my hands. Admitting that skeptical
doubts come up only when we set aside practical concerns does not suffice to
make them intelligible.
The second theme is that contextually sensitive immunity to doubt is not tied
to particular types of propositional content. Claims about “external” objects are
not as such especially vulnerable to skeptical questioning. In arguing for this point,
Wittgenstein challenges the traditional association of certainty with claims that are
“a priori,” “necessary,” or self-evident. If he is right, traditional epistemological

distinctions may be less aids to solving skeptical problems than their source.
There is an important methodological lesson here. We must not let philosophical
preconceptions distort our view of how our language-games actually work: “Don’t
think but look.”
The third theme is that there is no strict rule determining when and where doubt
is impossible, hence where knowledge resides.17 To master a language-game is to
acquire a kind of know-how, a vital component of which is the ability to recognize
when and where doubts and questions can properly be raised. Even when we can
give a rule of thumb, it will involve a phrase like “in normal circumstances,” and
we “recognize normal circumstances but cannot precisely describe them” (OC 27).
This is a further hint that Wittgenstein’s response to epistemological skepticism
will involve a healthy measure of skepticism about “epistemology,” understood as a
general theory of knowledge.
Wittgenstein is now ready to ask: does external-world skepticism make sense?
Reflecting his Moorean inspiration, he confronts Cartesian doubt in its most radical
form. The issue is not just our knowing things about external things but our knowing
that such things, understood as physical objects, even exist.
As noted, fairy-tale imaginings are not normally regarded as grounds for doubt.
Now Wittgenstein goes a step further: can I even imagine there being no physical
world at all?
35. But can’t it be imagined that there should be no physical objects? I don’t
know. And yet “There are physical objects” is nonsense. Is it supposed to be an
empirical proposition?—
And is this an empirical proposition: “There seem to be physical objects”? (OC)
It doesn’t matter what feats of imagination I can pull off. The proposition that
there are physical objects, as asserted by some metaphysicians and denied by others,
involves conceptual confusion.
36. “A is a physical object” is a piece of instruction which we give only to someone
who doesn’t yet understand either what “A” means or what “physical object”
means. Thus it is instruction about the use of words, and “physical object” is a
logical concept. (Like colour, quantity . . .) And that is why no such proposition
as: “There are physical objects” can be formulated.
Yet we encounter such unsuccessful shots at every turn. (OC)
The conclusion of Moore’s “proof” is one such. Moore’s contrast between “things
to be met with in space” and “things presented in space” has to do with questions
of individuation and re-identification and is therefore “logical” in Wittgenstein’s
sense. Accordingly, the proof amounts to no more than “instruction in the use of
words.” Using his hands as exemplars, Moore is teaching his audience the use of
the technical term ‘external thing’. To think of this exercise as an existence proof is
sheer confusion.
490 SKEPTICISM
4 CERTAINTY, MEANING, AND TRUTH

While exposing Moore’s confusion deepens our understanding of the failure of his
proof, it does not provide a full diagnosis of the illusions besetting the skeptic (or
idealist) and his realist opponent:
37. But is it an adequate answer to the scepticism of the idealist, or the assurances
of the realist, to say that “There are physical objects” is nonsense? For them after
all it is not nonsense. It would, however, be an answer to say: this assertion, or its
opposite is a misfiring attempt to express what can’t be expressed like that. And that
it does misfire can be shewn; but that isn’t the end of the matter. We need to realize
that what presents itself to us as the first expression of a difficulty, or of its solution,
may as yet not be correctly expressed at all. Just as one who has a just censure of a
picture to make will often at first offer the censure where it does not belong, and an
investigation is needed in order to find the right point of attack for the critic. (OC)
The persistence of the philosophical dispute shows that the participants can
make some sense of it. We must look more deeply into the language-game of
skeptical doubting.
Wittgenstein returns to the themes he introduced at OC 25–26: that in certain
circumstances, mistakes are impossible; that this circumstantial certainty is content-
indifferent; and that it is not governed by strict rules. But this time around he begins
with knowledge in mathematics. Philosophers have often traced mathematical
knowledge to a special “inner” process, such as the exercise of rational insight. But
the “inner process” or “state” is unimportant: “What is interesting is how we use
mathematical propositions” (OC 38). An essential aspect of the use of mathematical
propositions is that, in certain circumstances, “a calculation is treated as absolutely
reliable, as certainly correct” (OC 39). With respect to “12 x 12 = 144,” there is
no question of my saying “I see what you mean, but I’m not sure I agree.” If I think
you might be mistaken, I haven’t learned my multiplication tables. That is why an
expression of doubt fails to make sense. Without certainty, it is not just knowledge
that goes missing but meaning and understanding.
Wittgenstein reinforces the point that there is no need for a special faculty, giving
us insight into “necessary” truths, by taking mystery out of modality:
43. What sort of proposition is this: “We cannot have miscalculated in 12 x 12 =
144”? It must surely be a proposition of logic.—But now, is it not the same, or
doesn’t it come to the same, as the statement 12 x 12 = 144?
44. If you demand a rule from which it follows that there can’t have been a
miscalculation here, the answer is that we did not learn this through a rule, but
by learning to calculate. (OC)
Exempting some propositions from doubt is built into how we play the game. As
calculations become more complicated, the possibility of making a mistake creeps in,
and the need for checking may arise. When does it arise? It depends. No rule covers
all cases, and we don’t need one. “Nothing is lacking. We do calculate according to
a rule, and that is enough” (OC 46).
With respect to propositions about things in the world, the logico-linguistic

situation is essentially the same. In some situations, the notion of making a mistake
gets no grip, in which case neither does doubt.
52. This situation is thus not the same for a proposition like “At this distance
from the sun there is a planet” and “Here is a hand” (viz., my own hand).
The second can’t be called a hypothesis. But there isn’t a sharp boundary line
between them.
53. So one might grant that Moore was right, if he is interpreted like this: a
proposition saying that here is a physical object may have the same logical status
as one saying that here is a red patch.
54. For it is not true that a mistake merely gets more and more improbable as
we pass from the planet to my own hand. No: at some point it has ceased to be
conceivable. (OC)
This is the point of attack we have been looking for. The external world skeptic
assumes that certainty is tied to propositional content in a way that is independent
of circumstances.18 He is wrong on two counts.
First, he thinks that elementary mathematical propositions are self-evident: their
truth is grasped by rational insight. Of course, people make mistakes in calculation,
but potentially, and in some cases actually, mathematical propositions are absolutely
certain. By contrast, for propositions about things around us certainty is never more
than a point that we may approach asymptotically but can never reach. Being always
haunted by a shadow of a doubt, all such propositions are hypotheses. This is not how
our language of epistemic assessment works: “Are we to say that certainty is merely
a constructed point to which some things approximate more, some less closely? No.
Doubt gradually loses its sense. This language-game just is like that” (OC 56). “Here is
my hand” is every bit as certain as “2 + 3 = 5” and certain in the same way.19
The skeptic will reply that Wittgenstein has overlooked the principal reason for
regarding propositions about things around us as hypotheses. No such proposition
is immediately evident because all such propositions are epistemically dependent on
perceptual experience. This is why skeptical hypotheses are central to the external-
world problem. If I were the victim of a Cartesian Demon, or a brain in a vat, my
perceptual experience would be just what it would have been had I been living a
“normal” life. This is why such possibilities, however remote, cannot be ruled out
and why every external world proposition is haunted by a shadow of a doubt.
This reply also ties certainty to content, but in a different way. This time, the
contrast is not between mathematical and empirical propositions but between two
kinds of empirical proposition. Propositions concerning perceptual appearances, or
what used to be called “sense-data,” are intrinsically certain in a way that propositions
about things in the external world never can be. This is because propositions about
how things appear to me express knowledge that I would have even if I were a
victim of Demon deception or a brain in a vat.20 However, if we recognize that
certainty is tied not only to content but to circumstances, then we see that “Here is
a hand” can be every bit as certain as “Here is a red patch” (or “It seems to me as if
492 SKEPTICISM
there were a hand here”) and, again, certain in the same way. Wittgenstein returns
to his point in the third notebook:
250. My having two hands is, in normal circumstances, as certain as anything
that I could produce in evidence for it.
That is why I am not in a position to take the sight of my hand as evidence for
it. (OC)
The real thrust of the skeptic’s claim that philosophical reflection on knowledge
requires setting aside practical interests is to detach certainty altogether from
circumstances, making it a property of abstract “propositions” rather than claims
made and beliefs held by particular people in particular historical situations.21 This
is his fundamental mistake.
In the continuation of OC 54, Wittgenstein brings his points together in a general
claim about the relation between meaning and truth. That doubt about propositions
concerning things around us must, at some point, cease to be conceivable “is already
suggested by the following: if it were not so, it would be conceivable that we should
be wrong in every statement about physical objects; that any we ever make are
mistaken.” Having expressed agnosticism with respect to the claim that the total
nonexistence of physical objects is imaginable, Wittgenstein now turns to whether it
is possible (möglich): “So is the hypothesis possible that all the things around us don’t
exist? Would that not be like the hypothesis of our having miscalculated in all our
calculations?” (OC 55). In the second notebook, a suggestion becomes a conclusion:
80. The truth of my statements is the test of my understanding of these statements.
81. That is to say: if I make certain false statements, it becomes uncertain whether
I understand them.
The connection between truth and understanding is pervasive and essential. If
we didn’t get elementary calculations mostly right, there would be no practice of
calculation. By parity of reasoning, if all our thoughts and sayings about things
around us were false, they wouldn’t be about things around us. They wouldn’t be
about anything, since meaning would have evaporated. Grounding meaning in use
reveals the circumstantial nature of certainty, which in turn exposes the skeptic’s
context-independent doubt for what it is: an illusion.
5 CERTAINTY INTERNALIZED: MOORE’S MISTAKE

For many philosophers, the claim that the certainty of a proposition about things
around us differs only superficially from that of a simple mathematical proposition
is a hard pill to swallow. Don’t the cases just feel different?
Wittgenstein has another card to play. He introduces the central idea early in the
first notebook of OC:
6. Now, can one enumerate what one knows (like Moore)? Straight off like that,
I believe not.—For otherwise the expression “I know” gets misused. And through
this misuse a queer and extremely important mental state seems to be revealed.
12. —For “I know” seems to describe a state of affairs which guarantees what
is known, guarantees it as a fact. One always forgets the expression “I thought I
knew.”
Wittgenstein does not object to the idea that knowledge is “infallible” in the sense
that knowing excludes the possibility of my being mistaken. An important—perhaps
the main—use of “I know” is to imply that I can’t be wrong, either because I have
compelling grounds (OC 243) or because I have said something to which, in the
circumstances, the notion of doubt doesn’t apply. However, there is a route to this
“infallibilist” conception of knowledge that must be blocked.
21. Moore’s view really comes down to this: the concept “know” is analogous to
the concepts “believe,” “surmise,” “doubt,” “be convinced” in that the statement
“I know . . .” can’t be a mistake. And if that is so, then there can be an inference
from such an utterance to the truth of an assertion. And here the form “I thought
I knew” is being overlooked.—But if this latter is inadmissible, then a mistake
in the assertion must be logically impossible too. And anyone who is acquainted
with the language-game must realize this—an assurance from a reliable man that
he knows cannot contribute anything.
With respect to reporting my mental state, I have first-person authority. At least in

normal circumstances, my sincere assertion that I believe, surmise, or am convinced
that such and such is sufficient to establish that this is how things are with me. I can’t
take myself to believe that p when I merely surmise that p, or to be convinced when
I am dubious. In assuring his audience that he knows that here is one hand, Moore
shows that he thinks that the “mental state” of knowing is like this.22 It isn’t. If
I say “I thought I knew,” I will be readily understood: I was making some kind of
mistake, or new evidence has led me to change my mind. By contrast, “I thought
I believed” would take some explaining, and “I thought I surmised” has no use at all.
As implying the impossibility of error, knowledge has nothing to do with
feeling sure.
42. One can say “He believes it, but it isn’t so,” but not “He knows it, but it isn’t
so.” Does this stem from the difference between the mental states of belief and of
knowledge? No.—One may for example call “mental state” what is expressed by
tone of voice in speaking, by gestures, etc. It would thus be possible to speak of
a mental state of conviction, and that may be the same whether it is knowledge
or false belief. (OC)
The principle that if I know, I know that I know must not be misunderstood. “‘If
I know something, then I also know that I know it, etc.’ amounts to: ‘I know that’
means ‘I am incapable of being wrong about that.’ But whether I am so needs to be
established objectively” (OC 16). My knowing depends on the objective certainty
of what I assert or believe: that is, on the impossibility, in the circumstances, of my
going wrong in saying that.
The thought that the difference between knowing and merely believing is one
of “mental state” is a disaster. If we combine the triviality that knowledge is factive
494 SKEPTICISM
with the thought that knowing is a mental state subject to first-person authority,
we are led to conclude that my awareness that my mental state is one of knowing,
which is immediate and transparent to introspection, guarantees the truth of what I
know. Knowledge of the world around us becomes instantly problematic. How can
I guarantee the truth of propositions about things around me merely by reflecting on
my mental states? I can’t. It follows that what I know “immediately” must itself be
subjective: how it is with my sense-data, or how things appear to me.
In the second notebook, Wittgenstein suggests that the temptation to internalize
certainty can also arise from the “primitive meaning” of “I know.”
90. “I know” has a primitive meaning similar to and related to “I see” (“wissen,”
“videre”). And “I knew he was in the room, but he wasn’t in the room” is like
“I saw him in the room, but he wasn’t there.” “I know” is supposed to express
a relation, not between me and the sense of a proposition (like “I believe”) but
between me and a fact. So that the fact is taken into my consciousness. (Here is the
reason why one wants to say that nothing that goes on in the outer world is really
known, but only what happens in the domain of what are called sense-data.) This
would give us a picture of knowing as the perception of an outer event through
visual rays which project it as it is into the eye and the consciousness. Only then
the question at once arises whether one can be certain of this projection. And
this picture does indeed show how our imagination presents knowledge, but not
what lies at the bottom of this presentation.
This is a complex passage, but the central idea is clear enough. Assimilating
knowing facts to seeing objects leads to a picture of perceptual knowledge in
which experiential content is “given”—present to the mind—in a way that facts
involving “external” events and objects never can be. Experience thus constitutes or
underwrites a kind of knowledge that we could have even if all our beliefs about the
world around us were false. As with Moore’s mistake, a temptation arises to think
that knowledge properly so called is a mental state that cannot be confused with
merely believing. But internalizing certainty makes knowledge of the world become
incurably problematic. Diagnosing the sources of this fatal temptation is essential if
the illusion of Cartesian doubt is to be dispelled.
6 THE LIMITS OF ARGUMENT

Certain knowledge of the world around us goes with the very possibility of rational
thought. However, this certainty does not extend automatically to the commonsense
propositions enumerated in Moore’s Defence. These propositions may be certain for
us, but they do not all belong to the human form of life.
Take the proposition that the Earth has existed for many years past. It is not
something we question, if we think of it at all. It is implied by anything and everything
we believe about history, paleontology, or astrophysics. But must everyone share
such commitments?
85. What goes into someone’s knowing . . . history, say? He must know what it
means to say: the earth has already existed for such and such a length of time. For
not any intelligent adult must know that. We see men building and demolishing
houses, and are led to ask: “How long has this house been here?” But how does
one come on the idea of asking this about a mountain, for example? And have
all men the notion of the earth as a body, which may come into being and pass
away? (OC)
Knowing history involves certainties that are not vouchsafed simply by our being
competent judgers of things in the world around us. Even if Moorean propositions
are widely shared, this may be because they are abstractions from more specific
commitments: the contemporary scientist and the person who thinks that the Earth
has always existed both believe that the Earth has existed for many years past, but their
agreement is superficial. However, even this superficial agreement is not guaranteed.
Wittgenstein imagines a king brought up to believe that the world began with him.
Moore might be able to persuade the king otherwise, but could he prove him wrong?
There is one way that we can’t prove that the Earth has been around for a while:
“‘It is certain that after the battle of Austerlitz Napoleon. . . . Well, in that case it’s
surely also certain that the earth existed then’” (OC 183). The “proof” is hopeless
because everything that is known about history depends on documents and other
things that have survived from the past. That the Earth has existed for many years
past is taken for granted in any kind of historical thinking. The proof tacitly assumes
what it sets out to prove.
The apparent impossibility of proving that the Earth has existed for many years
past is no reason to doubt its longevity. Wittgenstein: “What we call historical
evidence points to the existence of the Earth a long time before my birth;—the
opposite hypothesis has nothing on its side” (OC 190). As we have seen doubts need
grounds, and there are none, at least none that we would ordinarily consider. But
what about Russell’s “possibility” that the Earth came into existence five minutes
ago, replete with what we take to be evidence of great antiquity?23 With respect to
skepticism concerning knowledge of the past, Russell’s hypothesis plays the role
played by Demon-deception and brains in vats in arguments for external world
skepticism, and it invites the same response: it’s a fairy tale, not a ground for doubt.
Fairy tale or not, does Russell’s “hypothesis” at least show that the Moorean
proposition might be false? Moore has no such worries. He insists that his
commonsense propositions, though contingent, are indisputably true. This cuts no
ice with Wittgenstein.
191. Well, if everything speaks for an hypothesis and nothing against it—is it
then certainly true? One may designate it as such.—But does it certainly agree
with reality, with the facts?—With this question you are already going round in
a circle. (OC)
Wittgenstein does not deny our right to call Moorean propositions certain. His point
is that our entitlement to treat such propositions as certain is not to be explained
by an appeal to their truth. Saying that p is true, or corresponds to the facts, is just
another way of insisting that p.24
Wittgenstein has excluded two ways of supporting Moorean certainties. How,
then, should we think of them, given that they look like empirical propositions? This
496 SKEPTICISM
question leads us to what many readers see as Wittgenstein’s master thought. This
thought is hinted at early in the second notebook: “It may be . . . that all enquiry
on our part is set so as to exempt certain propositions from doubt, if they are ever
formulated” (OC 88). The phrase “on our part” is important. Epistemic practices—
inquiring, justifying, doubting—take place in the context of an interlocking and
culturally transmitted network of deeply sedimented convictions. As Wittgenstein
remarks, “I did not get my picture of the world by satisfying myself of its correctness;
nor do I have it because I am satisfied of its correctness. No: it is the inherited
background against which I distinguish between true and false” (OC 94). That said,
the commitments that constitute this world picture do not derive their certainty
from tradition but from their distinctive logical role. Wittgenstein explains what he
has in mind with the aid of a metaphor:
95. The propositions describing this world-picture might be part of a kind of
mythology. And their role is like that of rules of a game; and the game can be
learned purely practically, without learning any explicit rules.
96. It might be imagined that some propositions, of the form of empirical
propositions, were hardened and functioned as channels for such empirical
propositions as were not hardened but fluid; and that this relation altered with
time, in that fluid propositions hardened, and hard ones became fluid. (OC)
In later passages, he introduces another striking figure: that of “hinge” propositions.
341. The questions that we raise and our doubts depend on the fact that some
propositions are exempt from doubt, are as it were like hinges on which
those turn.
342. That is to say, it belongs to the logic of our scientific investigations that
certain things are in deed not doubted.
343. But it isn’t that the situation is like this: We just can’t investigate everything,
and for that reason we are forced to rest content with assumption. If I want the
door to turn, the hinges must stay put. (OC)
Philosophers have often read these remarks as the basis of Wittgenstein’s response
to skepticism. They may be, but not in the way that is generally thought.
It is widely assumed that “hinge” propositions, because they structure our
epistemic practices, are not even candidates for being known to be true. If this is
what Wittgenstein intends, he faces what Duncan Pritchard identifies as a “core
problem”: that “from a skeptical view it is hard to see just what is so anti-skeptical
about the claim that the structure of rational evaluation has at its core arational
commitments. Isn’t that just what the radical skeptic claims?”25 However, it is far
from clear that this is Wittgenstein’s view. The “hinge” remark (OC 341) begins
“That is to say . . .” and glosses Wittgenstein’s previous observation (OC 340): “We
know, with the same certainty with which we believe any mathematical proposition,
how the letters A and B are pronounced, what the colour of human blood is called,
that human beings have blood and call it ‘blood.’” Apparently, Wittgenstein doesn’t
hesitate to treat hinge propositions as known.
That said, Wittgenstein often seems to take the opposite view. Is he of two
minds? However, when Wittgenstein declines to say that everyday certainties are
things he knows, it is almost invariably to distance himself from Moore’s use of
“I know” to combat skepticism. “I know,” Wittgenstein observes, may be taken to
express “comfortable certainty, not the certainty that is still struggling” (OC 357).
This use of “I know” is unexceptionable in everyday contexts. But in a discussion
of skepticism it is pointless, if not actively dangerous: remember Moore’s mistake.
Still, what gives us the right to this comfortable certainty when it comes to culturally
specific, thus potentially controversial, commitments, which are not things that just
anyone must know? Since the certainties in question extend far beyond anything
secured by his response to Cartesian skepticism, Wittgenstein must introduce some
new considerations or perhaps make explicit some further implications of ideas
already on the table. He does both; and as before his response to the skeptical
problem is rooted in his views about meaning and mind.
The first point to note is that Wittgenstein’s conception of meaning-as-use involves
a form of semantic holism: “When we first begin to believe anything, what we
believe is not a single proposition, it is a whole system of propositions. (Light dawns
gradually over the whole.)” (OC 141). Having a world picture is a precondition
of having any beliefs whatsoever. However rudimentary, this world picture will
have characteristic certainties that shape our practices of inquiry and epistemic
assessment. These will not be Moorean abstractions but specific articulations of
them: not just the certainty that the Earth has existed for many years but (if it had a
beginning) for how long and how it came into being.
An immediate consequence of Wittgenstein’s conception of extended certainty
is that there is no universal, trans-historical standard of rationality: or better, of
what is reasonable in the way of belief or doubt. Being reasonable is not a matter
of conforming to abstract “epistemic principles” but of constraint by bedrock
certainties: “I want to say: We use judgments as principles of judgment” (OC 124).
This is one of Wittgenstein’s deepest insights. It lies at the heart of his profound
skepticism concerning epistemology as a general theory of knowledge. Epistemology
represents a hankering after an all-purpose recipe for “rationally” resolving disputes.
Given that we use judgments as principles of judgment, there can be no such recipe:
what is reasonable cannot be detached from substantive commitments, which may
reach down to bedrock.
This will strike many philosophers as a case of out of the frying pan, into the fire.
In accordance with the role of judgments as principles of judgment, Wittgenstein
notes that “If Moore were to pronounce the opposite of those propositions he
declares certain, we should not just not share his opinion: we should regard him
as demented” (OC 155). But would we regard anyone who challenged elements
of Moorean common sense as out of his mind? Not necessarily. One of Moore’s
commonsense certainties is that he has never been far from the surface of the Earth.
Wittgenstein (OC 106) imagines someone telling a child that he had been to the
Moon. Wittgenstein tells the child that it was only a joke, that it is impossible to
get there; and ordinarily the child will be convinced. (Of course, Wittgenstein is
writing long before the Apollo missions.) But suppose there were a tribe whose
498 SKEPTICISM
members believe that people sometimes go to the Moon: perhaps this is how they
interpret their dreams. What could we say to show them the error of their ways, if
justification and doubt are “hinge-relative”?
Does Wittgenstein’s conception of certainty lead to cultural relativism and, if so,
what becomes of objective truth? Wittgenstein anticipates this question:
108. “Is there then no objective truth? Isn’t it true, or false, that someone has
been on the moon?” If we are thinking within our system, then it is certain
that no one has ever been on the moon. Not merely is nothing of the sort ever
seriously reported to us by reasonable people, but our whole system of physics
forbids us to believe it. For this demands answers to the questions “How did he
overcome the force of gravity?” “How could he live without an atmosphere?”
and a thousand others which could not be answered. But suppose that instead of
all these answers we met the reply: “We don’t know how one gets to the moon,
but those who get there know at once that they are there; and even you can’t
explain everything.” We should feel ourselves intellectually very distant from
someone who said this.
We should not be surprised at the last sentence, which seems to ignore the
question about truth. As we have already seen, Wittgenstein’s view of the concept
is deflationary. To ask whether it is true or false that someone has been on the
Moon is just to ask whether someone has been on the Moon. The answer (at the
time) is “No.” Period. Nor should we be overly perturbed by the reference to “our
system.” It is a given that they do not think as we do. The question is whether we
are epistemically entitled to take our views to be superior. Wittgenstein’s answer is
unequivocal: we are.
286. What we believe depends on what we learn. We all believe that it isn’t
possible to get to the moon; but there might be people who believe that that is
possible and that it sometimes happens. We say: these people do not know a lot
that we know. And, let them be never so sure of their belief—they are wrong and
we know it.
If we compare our system of knowledge with theirs then theirs is evidently the
poorer by far.
The Moon-visitors’ “system of knowledge”—that is, what they think they know
but which we know isn’t true—is so much poorer than ours because they don’t
even understand our questions, much less have answers for them. (While there is
no reason to suspect influence on Wittgenstein, this is a worthwhile element in
Bradley’s doctrine of partial truth and one that Moore simply failed to appreciate.)
What can we say to the Moon-visitors themselves? We don’t share enough
common ground for a straightforward argument. To get them to understand our
questions, we would have to teach them physics, and teaching isn’t a matter of
arguing. So suppose we managed to teach them physics: then what? There are two
possibilities. If they learned physics in a sense that implies taking physics to heart,
we would already have already got them to “look at the world in a different way”
(OC 92), so there would be no need for an argument.
Changing the shape of their intellectual riverbed would reconstruct their sense
of what is possible. With respect to lunar voyaging, Wittgenstein’s view is not that,
as it happens, no one has been on the Moon. He is certain of that because, at the
time of writing, flying to the Moon was impossible; and if the Moon-visitors learned
physics, they would agree.26 Alternatively, even if they learn physics but remain
attached to their lunar escapades, they might try to exempt Moon-visiting from
physical constraints, as in the thought that they go to the Moon in their dreams. A
response like this muddies the waters, in that it is no longer clear that they think of
themselves as traveling to the Moon in a sense that implies physical transportation.
A response along these lines is analogous to reconciling science with the Bible by
abandoning Biblical literalism. Wittgenstein might well be sympathetic to this.
Wittgenstein respected science but had no time for scientism: that is, he did not think
that everything worth taking seriously was subject to scientific explanation. But in the
example of the Moon-visitors, he envisages a less interesting response: that they don’t
know how they get to the Moon, but they know when they are there, “and even you
can’t explain everything.” Unlike the attempt to remove the conflict between physics
and Moon-visiting by treating the latter as in some way non-physical, this generic
epistemological remark is just an evasion: a simple refusal to engage. No one can
explain everything, but some people can explain a lot more than others.
Has Wittgenstein escaped skepticism, or relativism, only at the cost of embracing
dogmatism? No. Nothing in Wittgenstein’s account of certainty implies that, when
we meet people who think very differently from us, we will automatically favor own
views. Belief-revision can go both ways. There is nothing to learn from the Moon-
visitors, as Wittgenstein imagines them. However, more sophisticated interlocutors
could lead us to rethink some of our own erstwhile certainties. Our world picture—
though stable—is liable to change as we encounter new facts and new ideas.
This aspect of Wittgenstein’s thought, not prominent in the much-quoted
“hinge” passages, is insufficiently emphasized. To be sure, some things don’t change.
Elementary mathematical propositions are fossilized; and without certainties
concerning banal facts about things in our everyday environment we could have no
beliefs at all. But from the outset, Wittgenstein’s picture of knowledge and certainty
beyond this anti-Cartesian core is essentially dynamic:
97. The river-bed of thoughts may shift. But I distinguish between the movement
of the waters on the river-bed and the shift of the bed itself, though there is not
a sharp division of the one from the other.
In flowing through their channels, the streams may change the shape of the river.
Although this can happen slowly, through a kind of epistemic-semantic erosion, or
more dramatically, as in the Scientific Revolution, in neither case will such change
result from routine forms of argument: “Not all corrections of our views are on the
same level” (OC 300). But corrections get made: our descendants may think of us
(or we may come to think of our earlier selves) as Wittgenstein invites us to think
of the Moon-visitors. While there must be bedrock certainties if our inquiries are
to have their characteristic—indeed any—direction, such commitments are not an
intellectual straightjacket.
500 SKEPTICISM
The skeptic will argue that allowing for change to the “scaffolding” (OC 211) of
our inquiries compromises their claim to be objectively certain. In an important late
passage, Wittgenstein considers the skeptic’s argument:
599. One could describe the certainty of the proposition that water boils at circa
100o C. . . . The proposition is a very elementary one in our text-books, which
are to be trusted in matters like this because . . .—Now one can offer counter-
examples to all this, which show that human beings have held this or that to be
certain which later, according to our opinion, proved false. But the argument
is worthless. [Here Wittgenstein adds a marginal note: May it not also happen
that we believe we recognize a mistake of earlier times and later come to the
conclusion that the first opinion was the right one? etc.] To say: in the end, we
can only adduce such grounds as we hold to be grounds, is to say nothing at all.
I believe that at the bottom of this is a misunderstanding of the nature of our
language-games.27
Wittgenstein makes three points.
One—not profound but not to be overlooked either—is that correction is not a
one-way ratchet: rejection does not preclude all possibility of revival.
A second point cuts deeper. Of course, we can only adduce such grounds as we
hold to be grounds. But that is not the end of the matter: grounds are what we hold
to be such, in the way that the rules of football are what we recognize as the rules.
To be sure, when we recognize new grounds, we may take a critical view of our
former practice, treating the new grounds as factors we overlooked: that is, that we
should have considered. (That said, even with games like football, we may think of
modifications to the rules as changes we should have made before now.) The point
here is one that Wittgenstein introduced at the very beginning in his discussion
of grounds for doubt: the rules of language-games are immanent in our practices,
so that we are not merely responsible to them but also responsible for them. The
inferential structure of our practices of inquiry and epistemic assessment cannot be
“read off” from nature.28
This takes us to a third and decisive point: that without fixed points to
determine what are and are not proper inferential moves, reasonable doubts, or
relevant evidence, there is no inferential structure and thus no language-games at
all, primitive or sophisticated. There is no thought. In consequence, there is no
external perspective, outside of all language-games, from which we can judge that
our present certainties fall short of some epistemic ideal. If they need correction,
this will emerge through working with them: a retail business, though one that may
have far-reaching implications.
Where does all this leave us? Wittgenstein wants us to take seriously our
epistemic limitations. At the same, he wants to insist that recognizing our limitations
is not even the first step on the road to any form of radical skepticism or relativism:
the thought that all views are equal or that no view is really “justified,” much less
objectively certain.
In the first place, there are limits to argument, if argument is supposed to be
anything like Moorean proof. When disagreements engage bedrock certainties,
routine argument based on agreed evidence is impossible. This is true as much in

our disagreements with our ancestors or former selves as with contemporaries.
Faced with such deep disagreement, we can try teaching, as in the case of the
Moon-visitors. Of course, teaching depends on a willingness to listen and learn,
which might or might not be there. Or we could try looser, more coherentist
kinds of argument: persuasion, Wittgenstein would say. Again, we might
succeed or we might not: there is no recipe for getting someone to change
aspects of his or her world-picture. Everything depends on what we are talking
about, with whom, and the persuasive resources at our disposal. Deadlock is an
ineliminable possibility.
So why isn’t this a radically skeptical conclusion? Two reasons. First, deadlock does
not imply epistemic equality. Our inability to get anywhere with the Moon-visitors
gives us no reason to treat them as epistemic peers, so no reason to draw a skeptical
or relativist conclusion. This isn’t a recoil to dogmatism, since more sophisticated
interlocutors might give us reason to re-examine commitments of our own. Second,
suppose that we reach deadlock: while this happens, it points to no general skeptical
conclusion. Dialectical deadlock amounts only to undecidability with respect to some
particular dispute, in our present epistemic situation. With further evidence or new
ideas, seemingly intractable disagreements may be resolved, sometimes in favor of
one party but sometimes because we come to see the dispute as misconceived. Our
contingent limitations do not imply the impossibility of knowledge, any more than
fallibilism means that nothing is ever certain and properly regarded as such.
This point applies even at the level of hinge commitments. Pritchard’s “core
problem” with respect to the apparently “arational” character of hinge commitments
depends on treating Wittgenstein’s picture of knowledge, certainty, and justification
as primarily a view about “the structure of rational evaluation,” giving insufficient
weight to its fundamentally dynamic character.29 If we take a time-slice of our belief
system, hinge commitments may seem to float in the air. But over time, they either
continue to pay their way or become subject to revision. They are grounded in
practices of inquiry: in our life.
As Philosophical Investigations did with respect to meaning, On Certainty brings
knowledge and certainty down to Earth. Wittgenstein shows us knowledge as it is
for human beings, not disembodied intellects or thinkers wholly detached from any
concrete, historical situation. For philosophers who find the certainty Wittgenstein
describes as somehow less than we might wish for, he offers this advice: “Forget this
transcendent certainty, which is connected with your concept of spirit” (OC 47).
We should take it.
NOTES
1. A third brother, the only one to go into commerce, shot himself at the end of the First
World War when the men under his command refused to obey his orders.
2. Wittgenstein (1961). Wittgenstein called his manuscript Logische-Philosophische
Abhandlung. The Latin title, by which the book is generally known, was suggested by
Moore.
502 SKEPTICISM
3. Wittgenstein exploits the fact that the truth-functional logical constants can be
defined in terms of a single logical operation of joint negation, which he generalizes
to apply to any number of propositions. He attempts to extend his truth-functional
conception of logic to quantified propositions by treating “all”- and “some”-
statements as, respectively, conjunctions and disjunctions, possibly infinite. So “All
men are mortal” is analyzed as “Socrates is mortal, Plato is mortal, Michael Williams
is mortal . . . and so on.” “Some philosophers are Greek” comes out as “Socrates is
a philosopher and Socrates is Greek, or Plato is a philosopher and Plato is Greek,
or Descartes is a philosopher and Descartes is Greek, . . . and so on.” The adequacy
of this analysis of quantification has been disputed, and Wittgenstein himself had
second thoughts about it.
4. To make up for his lack of a degree, he was allowed to submit the Tractatus as a
Ph.D. thesis.
5. Wittgenstein (2009). Wittgenstein had hoped to publish his two books in a single
volume, so readers would have a clear understanding of the later work’s targets.
6. Though containing only first draft material, the notebooks are not just books of
notes: each has considerable thematic unity, and the order in which they were
written is significant.
7. It has been argued that Wittgenstein is wrong about “game.” See Suits (2014).
8. Moore (1970: 32–59).
9. Moore (1970: 127–150).
10. Wittgenstein (1969). Malcolm recounts his relationship with Wittgenstein, including
the discussion that led to On Certainty, in Malcolm (1984).
11. Bradley (1914).
12. Moore also takes issue with philosophers who hold that while his commonsense
propositions as ordinarily understood are false, there are related propositions that are
philosophically more respectable and wholly true. Moore (1970: 115) surely has in
mind Russell, who takes the ordinary notion of a thing, as an entity logically distinct
from its perceptual appearances, to be an invention of “prehistoric metaphysicians”
and advocates replacing such “inferred entities” with “logical constructions” out of
sense-data. This position is also Idealist in spirit, though the Idealism is Berkeley’s
rather than Hegel’s.
13. Kant, Critique of Pure Reason, B xl, note, quoted in Moore (1970: 127). Moore
follows the translation of Kemp Smith (1929).
14. Moore thought that critics had misunderstood his intentions. “I have sometime
distinguished between two different propositions, each of which has been made by
some philosophers, namely (1) the proposition ‘There are no material things’ and (2)
the proposition ‘Nobody knows for certain that there are any material things.’ And in
my latest published writing, my British Academy lecture called ‘Proof of an External
World,’ . . . I implied with regard to the first of these propositions that it could be
proved to be false in such a way as this; namely, by holding up one of your hands and
saying ‘This hand is a material thing; therefore there is at least one material thing.’
But with regard to the second of those two propositions, which has, I think, been far
more commonly asserted than the first, I do not think that I have ever implied that it
could be proved to be false in any such simple way.” See Moore (1968: 668).
15. I offer a more extended treatment of Wittgenstein’s reaction to Moore’s

Proof in Williams (2004a). However, the current discussion corrects some
details concerning Wittgenstein’s “infallibilist” account of knowledge and
“Moore’s mistake.”
16. For more detailed discussion of the peculiar generality of skeptical doubts, see
Williams (1996: ch. 1).
17. Sextus would call such a rule a criterion and would argue that no criterion can be
defended in a non-circular way. For Wittgenstein, the skeptical problem just shows
what’s wrong with the idea of a criterion.
18. In Williams (1996) I call this doctrine a version of “epistemological realism.” My
diagnosis of skepticism is much indebted to Wittgenstein.
19. In an important exploration of Wittgenstein’s response to skepticism, Marie
McGinn (1989: ch. 7–8) stresses the point that Wittgenstein treats the certainty of
“Moorean” propositions such as “Here is one hand” as essentially similar to that of
elementary mathematical propositions.
20. Judgments about appearances, though contingent, have often been thought to be
like mathematical propositions in that, if formulated with a clear understanding
of their content, they cannot be false. For a classic deconstruction of this idea, see
Sellars (1963: 164–167, para. 32–34).
21. Cf. Austin (1962: 118). “[T]he idea that statements about ‘material things’ as such
need to be verified is just as wrong as, and wrong in just the same way as, the idea
that statements about ‘material things’ as such must be based on evidence.” For
Austin (1961: 89), statements are the primary bearers of truth, and “A statement is
made and its making is an historic event, the utterance by a certain speaker or writer
of certain words (a sentence) to an audience with reference to an historic situation,
event or what not.”
22. Although Wittgenstein may be hard on Moore here, the view he attributes to
Moore was certainly held by some of Moore’s contemporaries. Thus Cook Wilson
writes that knowledge cannot be one of “two states of mind . . . the correct and the
erroneous one . . . quite indistinguishable to the man himself,” the reason being
that “as the man does not know in the erroneous state, neither can he know in the
other state” [Quoted by Charles Travis (2008: 293)]. Travis’s paper is an important
defense of a context-sensitive account of knowledge, though his inspiration is Austin
rather than Wittgenstein.
23. Russell (1921: Lecture XI).
24. In PI Wittgenstein favors a redundancy conception of truth. See PI 136. However,
although in my view Wittgenstein’s view of truth remains fundamentally deflationary,
there are passages in OC (e.g., 200) suggesting sympathy for an epistemic account of
truth as verifiability. I discuss this apparent tension in Williams (2004b).
25. Pritchard (2015: 71). The concern that Pritchard identifies lies at the heart of
disputes over how to read On Certainty. Strawson thinks of framework certainties
as analogous to Hume’s natural beliefs (1985: ch. 1). McGinn holds that we stand
to these certainties in a “non-epistemic” relation (1989: ch. 7). Crispin Wright holds
that because they enable inquiry, such certainties are the objects of a distinctive
rational but non-evidentially based epistemic attitude that he calls “entitlement”
504 SKEPTICISM
(2004). Danièle Moyal-Sharrock defends her “enactivist” view that such certainties
are not arational beliefs because they are not even propositional (2016). Pritchard
thinks that, while Moyal-Sharrock’s interpretation has the most textual support, it
makes better sense of Wittgenstein to read him as suggesting that hinge commitments
are not beliefs, insofar as beliefs are supposed to be “knowledge-apt” (2016: 90).
In my view, all these views fail because they start from the false presupposition
that Wittgenstein has an all-purpose response to skepticism, based pretty much
entirely on the idea of “hinge” propositions. All these commentators ignore the
special diagnostic treatment accorded Cartesian skepticism in the first notebook.
In Williams (2004a), I call this approach “the Framework Reading.” I hope in the
future to offer a detailed account of its manifold inadequacies, both exegetical
and philosophical.
26. The possibility at issue here is real possibility: the sense of “possibility” in which
there is no possibility of my abandoning academic life to play center forward for
Liverpool. Fans of skeptical fairy tales—the Earth’s having sprung into existence
five minutes ago or my being a brain in a vat—legitimate such “hypotheses” by
distinguishing senses of “possible”: while not physically or technologically possible,
these eventualities are imaginable and so possible “logically” or “metaphysically.”
From Wittgenstein’s standpoint, such distinctions serve only to defend a
methodology that has no independent rationale.
27. A version of this argument can be found in Sextus Empiricus (PH I 34): “Before the
founder of the school to which you adhere was born, the argument of the school,
which is no doubt sound, was not yet apparent, although it was really there in
nature. In the same way, it is possible that the argument opposing the one you have
just propounded is really there in nature but is not yet apparent to us; so we should
not yet assent to what is now thought to be a powerful argument.”
28. See Williams (1996: ch. 3).
29. Pritchard (2016: 94) recognizes that what he calls “personal hinges” are subject to
change. However, he fails to realize the extent to which Wittgenstein’s dynamic
vision of knowledge and certainty works against the supposed “core problem,” thus
removing the rationale for Pritchard’s own “nonbelief” account of hinge propositions.
REFERENCES
Austin, J. L. 1961. “Truth.” In his Philosophical Papers, 85–101. Oxford: Clarendon Press.
Austin, J. L. 1962. Sense and Sensibilia. Oxford: Oxford University Press.
Bradley, F. H. 1914. “Coherence and Contradiction.” In his Essays on Truth and Reality,
219–244. Oxford: Clarendon Press.
Kemp Smith, Norman. 1929. Immanuel Kant’s Critique of Pure Reason. London: Macmillan.
Malcolm, Norman. 1984. Ludwig Wittgenstein: A Memoir, 2nd edn. Oxford: Oxford
University Press.
McGinn, Marie. 1989. Sense and Certainty. Oxford: Blackwell.
Moore, G. E. 1970. Philosophical Papers, 3rd edn. London: Allen and Unwin.
Moyal-Sharrock, Danièle. 2016. “The Animal in Epistemology: Wittgenstein’s Enactivist
Solution to the Problem of Regress,” International Journal for the Study of Skepticism
6: 97–119.
Pritchard, Duncan. 2016. Epistemic Angst. Princeton, NJ: Princeton University Press.
Russell, Bertrand. 1970. “The Relation of Sense-Data to Physics.” In his Mysticism and
Logic, 108–131. London: Unwin Books.
Russell, Bertrand. 1921. The Analysis of Mind. London: Allen and Unwin.
Schilpp, Paul Arthur (ed.). 1968. The Philosophy of G.E. Moore, Vol. 2, 3rd edn. Lasalle, IL:
Open Court.
Sellars, Wilfrid. 1963. “Empiricism and the Philosophy of Mind.” In his Science, Perception
and Reality, 127–96. London: Routledge.
Strawson, Peter. 1985. Skepticism and Naturalism: Some Varieties. New York: Columbia
University Press.
Suits, Bernard. 2014. The Grasshopper: Games, Life and Utopia, 3rd edn. Calgary:
Broadview Press.
Travis, Charles. 2008. “A Sense of Occasion.” In his Occasion-Sensitivity, 290–315.
Williams, Michael. 1996. Unnatural Doubts. Princeton, NJ: Princeton University Press.
Williams, Michael. 2004a. “Wittgenstein’s Refutation of Idealism.” In D. McManus (ed.),
Wittgenstein and Scepticism, 76–96. Abingdon: Routledge.
Williams, Michael. 2004b. “Wittgenstein, Truth and Certainty.” In M. Kölbel and B. Weiss
(eds.), Wittgenstein’s Lasting Significance, 247–281. London: Routledge.
Williams, Michael. 2007. “Why Wittgensteinian Contextualism Isn’t Relativism,” Episteme
4: 93–114.
Wittgenstein, Ludwig. 1961. Tractatus Logico-Philosphicus, translated by D. F. Pears and
B. F. McGuiness. London: Routledge.
Wittgenstein, Ludwig. 1969. On Certainty, edited by G. E. M. Anscombe and G. H.
von Wright. New York: Harper.
Wittgenstein, Ludwig. 2009. Philosophical Investigations, revised 4th edn., edited by
P. M. S. Hacker and J. Schulte Chichester: Wiley-Blackwell.
Wright, Crispin. 2004. “Warrant for Nothing (and Foundations for Free)?,” Proceedings of
the Aristotelian Society supp. vol. 78: 167–212.
CHAPTER THIRTY-SIX
Richard Popkin on the

History of Skepticism
JOSÉ R. MAIA NETO
1 POPKIN’S RÉPUBLIQUE DES LETTRES 1

Richard H. Popkin was born in New York in 1923 and died in Los Angeles in
2005. He taught at the universities of Connecticut, Iowa, UC San Diego (whose
Philosophy Department he founded), Washington University in Saint Louis, and
UCLA. He worked in the archives of the major research libraries in the United States
and Europe (among many others, the Bibliothèque Nationale in Paris, the British
Museum in London, the Laurenziana in Florence, and the Clark Library in Los
Angeles). He played a major role in the development of the history of philosophy in
the United States (he was a co-founder of the Journal of the History of Philosophy),
in particular in the field of the early modern history of ideas and philosophy, in
which he is a major world reference. He founded, with Paul Dibon, the Archives
Internationales d’Histoire des Idées/International Archives of the History of Ideas, a
series that has thus far published more than 200 volumes. Popkin was also important
as a teacher and/or supervisor of major North American early modern historians of
philosophy such as Harry Bracken, Richard Watson, Philip Cummins, David Norton,
Donald Livingston, and James Force. He was probably the first North American
early modern scholar to establish intellectual connections with European scholars.2
Popkin established an international network, collaborating with scholars from the
United States, Canada, England, France, Italy, Germany, The Netherlands, Poland,
Argentina, Mexico, Brazil, and Israel. He organized international conferences in
Europe and the United States and published the proceedings.3 Popkin published
more than 200 articles. Some of those concerning early modern philosophers were
collected by Richard Watson and James Force in Popkin (1980). Others, mainly
on early modern millenarianism, were collected by Popkin (1992) himself. With
Paul Edwards, he edited the Readings in the History of Philosophy Series.4 He
also edited The Columbia History of Western Philosophy (1999) and contributed
to the diffusion of seventeenth-century skeptical views, translating and publishing
selections of Pascal (1989) and Bayle (1991).
RICHARD POPKIN ON THE HISTORY OF SKEPTICISM 507
2 THE HISTORY OF SCEPTICISM

Popkin’s most important and influential work is his book The History of Scepticism,
the first edition covering from Erasmus to Descartes (Popkin 1960), the second
from Erasmus to Spinoza (Popkin 1979), and the third from Savonarola to Bayle
(Popkin 2003). The second edition was translated into Spanish, Italian, French, and
Portuguese, and has stimulated original research on early modern skepticism in all
of these languages—often critical of Popkin’s own views.
As an undergraduate student at Columbia in 1941–1942, Popkin became
acquainted with Sextus Empiricus’s Outlines of Pyrrhonism in a survey course on
the history of philosophy.5 In 1945–1946, in a graduate seminar on Hume given
by Charles Hendel at Yale,6 Popkin perceived a great similarity between Sextus
and Hume and wondered how the skepticism of the former may have influenced
the latter.7 This question, which he discussed with his former teacher at Columbia
and thereafter lifelong collaborator, the great Renaissance scholar Paul Oskar
Kristeller, was the origin of the project that led to the History of Scepticism. In
its first plan, Hume would be the subject of the concluding chapter. In a research
project he wrote at the time, Popkin claimed that his aim was “to discover the
historical antecedents of the sceptical philosophy of David Hume. . . . My hope
is to publish my results in the form of a book on the history and influence of
Pyrrhonism from Montaigne to Hume.”8 But as the research developed, Popkin
realized that the material to be covered was too vast for just one volume. When
he published the first edition of his History of Scepticism, up to Descartes, he
envisaged three others. A second volume would deal with post-Cartesian
seventeenth-century skepticism: Simon Foucher, Pierre-Daniel Huet, and Pierre
Bayle on the skeptical side, and Baruch Spinoza, Nicolas Malebranche, Gottfried
Leibniz, and John Locke on the side of those whose anti-skeptical reaction led to
new forms of skepticism. A third volume would deal with the religious skepticism
characteristic of the Enlightenment: Hume and the French philosophes. Finally,
“if time permits, I hope to follow the sceptical trail from Hume to Kant, and to the
reconversion of scepticism into fideism in Hamann, Kierkegaard and Lamennais”
(Popkin 1960: xiii). Popkin then set to work on Bayle and Huet, for both were
central figures in connecting Hume to the earlier modern skeptical tradition.9 But
this was not the material incorporated in the second edition of the History of
Scepticism, whose additional chapters on Isaac La Peyrère and Spinoza (Popkin
1979: 214–228 and 229–248, respectively) rather reflect Popkin’s new major
interest at the time: Jewish intellectual history. His research on post-Cartesian
skepticism (Pascal, Huet, Foucher, Malebranche, Leibniz, Glanvill, Boyle, Locke,
and Bayle) was finally incorporated in the last edition of The History of Scepticism:
From Savonarola to Bayle.10
Popkin’s History of Scepticism has provided a research program for historians of
early modern philosophy and ideas all over the world. The following four sections
(3–6) deal briefly with four theses that have been much debated since the first edition
of the book.
508 SKEPTICISM
3 THE SIXTEENTH-CENTURY RELIGIOUS

CONTROVERSY AND THE PROBLEM
OF THE CRITERION
The first thesis in question was called by Popkin “the Popkin-Schmitt thesis,”
since it was also held by Charles Schmitt (1972: 163–170). In his autobiography,
Popkin (1988: 115) tells us that the thesis was formulated after he became aware
of the vast presence of skeptical views among sixteenth- and seventeenth-century
philosophers. The thesis is that the great impact of ancient skepticism on modern
thought resulted from the historical coincidence of the rediscovery of the works of
Sextus Empiricus and the religious controversy brought about by the Reformation
in the sixteenth century. Although there was a Latin fourteenth-century translation
and three copies of Sextus’s Outlines (Floridi 2002: 23–24), we have no knowledge
of any philosopher noticing it in Western Europe before the mid-fifteenth century
when Italian humanists became interested in his works, mainly for philological and
historical reasons. Manuscript copies of Sextus’s works increased rapidly through
the sixteenth century (Floridi 2002: 95). The diffusion was naturally much larger
when Henri Estienne translated into Latin and published Sextus’s Outlines in 1562
and Gentien Hervet translated what is known as Sextus’s Adversus Mathematicos
and published the two works in 1569.11
Popkin noticed that by denying the authority of the Pope and the councils to
decide religious controversies and by arguing that the direct examination of Scripture
and each individual’s persuasion is the rule of faith, the Reformers unintentionally
brought about the Pyrrhonian challenge of how to justify a criterion of truth without
reasoning in a circle or falling into an infinite regress.12 Because Christians no
longer shared a common criterion for the solution of religious controversies, what
counted as evidence for one party was not accepted by the other, making a solution
extremely difficult to find and therefore opening a “high road to Pyrrhonism,” as
Bayle describes the problem in the Dictionary (article Nicole, remark C). Against
the Reformers’ criterion, the Roman Catholics argued that there were many obscure
passages in Scripture that required an external, traditional institution (the Roman
Catholic Church) to interpret them. The Reformers replied that the Catholic Church
could not claim to be the infallible rule for interpreting the Bible since its supposed
authority derived from the Bible. Catholics fought back, claiming that since different
persons interpret the Bible quite differently, the Reformers’ rule of faith allows a
multiplicity of religious views, rendering a common orthodoxy impossible.13 If the
rule is personal persuasion and there are many different persuasions, how can the
true one be discerned? Luther cannot say it is his own since he is a party to the
disagreement. Reformers turned the same problem against the Catholics. If tradition
is the rule, which is the true one? The Roman Catholic or the Greek Orthodox? The
Roman Catholic cannot claim to be the legitimate judge of the true tradition since it
is one of the conflicting parties. Catholics claimed that the Reformers’ rule of faith
is subjective, compromising the objectivity of religious knowledge. The Reformers
replied that, since Catholics claim that human beings are fallible, needing the help of
the church to decide religious matters, how can Christians recognize the true church
and the true Pope? Only the Pope himself, supposedly infallible on such matters,
could know that he is the Pope (Popkin 2003: 13–15).
The tremendous repercussion of the religious controversy at the time explains,
according to Popkin, the great importance of skepticism in the period and why
Pyrrhonism, which poses the criterion problem, and not Academic skepticism,
played the major role in the skeptical crisis that led, still according to him, to
the major developments in early modern philosophy. This is why he called such
a crisis “crise pyrrhonienne.” If the criterion problem were not the crucial issue
in the religious controversy that was excruciating to hearts, minds, and bodies,
Sextus’s works would not have had the impact they had in different areas of culture
(philosophy, theology, literature, politics).14
The Popkin-Schmitt thesis has been much challenged. Sylvia Giocanti claims that
Popkin’s focus on the epistemological problems involved in the controversies of the
time is very limited and “leads to the destruction of the skeptical position” (Giocanti
2001: 37). Ayers (2004) denies that there was a skeptical crisis in the period.
According to him, Montaigne’s “skepticism” in the “Apology for Raymond Sebond”
(Essays II 12)—considered by Popkin the major manifestation of the crisis—is a side
effect of his overall Platonism. Maia Neto (2012) argues that Pyrrhonism in the
“Apology” is just a strategic weapon to resist Calvinist proselytism (a view partially
derived from Popkin’s) and therefore cannot be attributed to Montaigne himself.
Notwithstanding the importance of Sextus’s works for Montaigne and others,15 new
studies have shown that the role of Academic skepticism did not diminish after the
availability of Sextus’s Pyrrhonian works,16 including in Montaigne’s own case.17
Pierre Charron develops an Academic skeptical view of wisdom that was extremely
influential in the seventeenth century (see Maia Neto 2014). Academic skepticism
becomes even stronger in post-Cartesian skepticism at the end of the seventeenth
and first half of the eighteenth century. With Cartesian doubt, skepticism stops
being a moral philosophy (as it was among the ancient and Renaissance skeptics)
and becomes a philosophical method. In the last edition of his History of Scepticism,
Popkin takes into account the part of this research that was available to him to
recognize that philosophers such as Foucher and Huet (Bayle and Hume could also
be added to the list) “still admired the methodological contribution of Descartes,
although they strove to destroy the ontological dogmatism he had presented. They
thought the skeptical method, which Descartes started with, was an important
part of philosophical study” (Popkin 2003: 374–375 n. 23). Foucher and Hume
identified this philosophical method as Academic.
4 SKEPTICISM AND ANTI-SKEPTICISM FROM

RELIGION TO PHILOSOPHY
The second thesis, an implication of the first, is that the three main reactions to
the skeptical challenge in the sixteenth century were anticipated in the religious
controversies. The first is a skeptical defense of Catholicism in Erasmus’s On
510 SKEPTICISM
Free Will (1524). Erasmus holds that the theological problems concerning the
compatibility of grace and free will are too obscure to limited and fallible human
reason so that Christians should accept (even if not understanding as true) the views
of the Catholic Church. This position leads, according to Popkin, to the skeptical
fideism of Montaigne, Charron, La Mothe Le Vayer, and others, for whom the
human faculties are incapable of knowing even ordinary and natural things. The
second position was taken by Luther in his reply to Erasmus, On the Bondage of
the Will (1525). Luther claims that a skeptical attitude is unacceptable in religion
where what is at stake is of such importance (eternal salvation or damnation) that
any psychological state different from the firmest conviction is unnatural. This
position anticipates, in the religious field, the philosophical attempts to provide a
complete refutation of skepticism, of which the best known and most influential is
Descartes’s. The third position was held by the liberal reformer Sebastian Castellio
in his Concerning Heretics and De Arte Dubitandi. For Castellio, the rule of faith is
neither tradition nor subjective certainty but reason and the senses, which, because
limited, do not lead to certain knowledge. Given that certainty is unavailable in
religious knowledge, Christians must look for what is probable and reasonable.
Castellio’s position is thus a via media between the Catholic skeptical position
and Luther’s dogmatic position that led, in the philosophical field, to what Popkin
calls the “constructive or mitigated skepticism” (Popkin 2003: 112–127) of Marin
Mersenne and Pierre Gassendi.18 We have no knowledge of the essence of things
but can give probabilistic, experimental, and hypothetical explanations of the
phenomena.
Although Popkin’s view of the kinds of, and reactions to, skepticism in the
different early modern philosophies he examined have been much debated (see
below), and his interpretation of Erasmus’s and Castellio’s relation to skepticism has
been challenged by Backus (2009), most scholars today recognize that any accurate
historical treatment of sixteenth- and seventeenth-century philosophy cannot ignore
its religious and/or theological background. In the 2003 edition, Popkin enlarged
and diversified his presentation of the religious background of the skepticism of
the time, recognizing the role of fourteenth-century scholastic skepticism derived
from God’s omnipotence and the contingency of the world,19 pointing out different
religious motivations in the use of skepticism by Blaise Pascal, Henry More, and
the Quietists (Popkin 2003: 174–188), and expanding his analysis of Gianfrancesco
Pico della Mirandola in the context of Savonarola’s influence in the Florence of the
time (Popkin 2003: 19–28).
The remaining two theses expounded below concern, respectively, the skeptical
and Descartes’s reactions to the skeptical challenge.
5 SKEPTICISM AND FIDEISM

Montaigne, Charron, La Mothe Le Vayer, and most Christian skeptics of the time
claim that skepticism can be compatible with Christian faith and that it is even useful
in Christian apologetics. Popkin disputes the standard view that these claims, La
Mothe’s in particular, are mere lip service to religious and political authorities.20
Popkin argues that skepticism understood as an epistemological position is
compatible both with the suspension of judgment about religious claims and with
the acceptance of such claims on a non-rational basis, a religious position usually
called “fideism.” Popkin raises the following challenge to the standard irreligious
interpretation: how is it possible to distinguish, without appealing to emotional
pitch or relationships (as Pintard does), between (i) a use of skepticism by Charron,
La Mothe Le Vayer, and Bayle to undermine Christianity and (ii) a use of skepticism
by Pascal, Kierkegaard, and Hamann to support Christianity, given that they all
attack natural theology and apparently defend a blind leap of faith?
I have argued elsewhere (Maia Neto 1995) that the positions are quite different.
So-called skeptical fideists such as La Mothe Le Vayer argue for a skeptical position
on only natural or rational terms and then claim that this position is the most
compatible with Christianity since it recognizes the incapacity of human reason
to attain truth, indirectly supporting the necessity of an alternative way to it. This
alternative way can be, depending on the particular Christian skeptic in question,
supernatural grace, or the “blind” acceptance of the truths contained in the Bible as
the Revealed word of God, or the recognition that we cannot give up what has been
established by tradition given our incapacity to know on our own, or a combination
of some or all of these views. Pascal and Kierkegaard—and also Bayle apud Bracken
(1993), diverging from Popkin—never accept skepticism as a mental state previous
to faith but interpret it as a consequence of Scripture and/or just as a philosophical
tool to clarify specific features of Christian faith. The major problem that renders
skeptical fideism suspect according to the view of most scholars is that skepticism—
in particular Pyrrhonian skepticism, the branch of skepticism Popkin considered
the most important in the period—would be intrinsically incompatible with any
dogmatic affirmation whatsoever, even if not based on rational arguments.21
6 DESCARTES’S PHILOSOPHICAL PROJECT

AS A REFUTATION OF SKEPTICISM
Because Descartes is the central seventeenth-century philosopher, Popkin’s thesis
about him is probably the most influential. Lennon (2004: 123) says the following
about the relevance of Popkin’s reading of Descartes in Cartesian scholarship:
A variation on Whitehead’s old saw about the history of philosophy as a series

of footnotes to Plato—one that is more accurate than Whitehead’s original
version—is that all histories of early modern philosophy since 1960 are footnotes
to Popkin, certainly as far as Descartes is concerned.
Popkin’s view of Descartes as endeavoring to “conquer skepticism” but ending as “a

skeptic malgré lui” (the titles of, respectively, chapters 9 and 10 of the 2003 edition
of The History of Scepticism) became standard, at least among English-speaking
Descartes scholars.22 Before Popkin, Cartesian doubt was little studied and almost
never—certainly not in English commentaries—in connection to the reappraisal
512 SKEPTICISM
of skepticism in the sixteenth and early seventeenth centuries. Popkin argued that
Descartes’s quest for certain knowledge was a reaction to a widespread skepticism
diffused mainly in France through Montaigne, Charron, Gassendi, La Mothe Le
Vayer, and others. He also showed that Descartes’s anti-skepticism differed from
most others in the period, which failed to meet the criterion problem and vainly
attempted to defeat the skeptical threat either by reaffirming or reformulating
Aristotle’s type of response23 or by dogmatically assuming human access to the
truth.24 Still more influential—for it goes much beyond early modern philosophical
scholarship—is Popkin’s view that Descartes’s attempt to defeat skepticism,
though the best available in the seventeenth century, failed and that his strategy to
radicalize doubt (by putting in doubt the existence of an external, material world)
in order to overcome it with the certainty of the cogito, resulted in a skepticism still
more formidable than the previous one. More recently, the view that Descartes’s
metaphysics was to a significant extent a reaction to the skepticism of his time has
been challenged. Scholars argue that Descartes’s real target was the Aristotelians,
not the skeptics.25 Lennon (2008: 62–77) provides a very detailed analysis of the
passages in which the skeptics are mentioned by Descartes and concludes that
Descartes despised them—at least the Pyrrhonians—referring to their views almost
always only defensively, that is, when accused of bringing about skepticism.
Popkin’s view on Descartes’s relation to skepticism certainly needs revision. It is
not plausible that refuting skepticism was a major goal of Descartes’s, even though he
did argue that this was one of the benefits of his philosophy (see CSM I 127, 181–182;
II 12, 374–375, 408). Paganini (2008; 2008b; 2009), though denying that refuting
skepticism was a major goal of Descartes’s, presents a very compelling case for
Popkin’s interpretation by showing, in more detail and depth than Popkin did,
that not only was Descartes’s position partially anticipated by Charron,26 Francisco
Sanches, Tommaso Campanella, and La Mothe Le Vayer, but also that he was
genuinely concerned with confronting atheistic skepticism of the type, according to
Paganini, upheld by La Mothe Le Vayer.
Popkin’s most insightful view about Descartes’s connection to the skeptics of
his time is his view that Charron anticipated Descartes in using doubt as a method
to eliminate all acquired beliefs, establishing a provisional morality compatible
with this elimination, and highlighting the need of an active role of the will in
this process. The main common ground is the will to eliminate all acquired beliefs,
in Charron’s case according to Popkin, in order to clean the mind so as to better
dispose it to receive supernaturally revealed truth (through grace and revelation)
and, in Descartes’s case, in order to preserve in the mind only the innate truths of
the cogito, God, and extension, which are the foundations of the new science.
In his comparison between Descartes, on the one hand, and the “noveaux
Pyrrhoniens” (Charron in particular) on the other, Popkin points out two major
differences, namely, the intensity of Cartesian doubt and the fact that it does not
result in a tabula rasa:
The process of doubting compels one to recognize the awareness of oneself,
compels one to see that one is doubting or thinking, and that one is here, is
in existence. The discovery of true knowledge is not miraculous, not a special

act of Divine Grace. Instead the method of doubt is the cause rather than the
occasion of the acquisition of knowledge. Its truth . . . is the result of Divine
intervention—not of sudden, new intervention, but rather a continuous and
permanent act of grace that sustains our mind with its innate ideas, and with its
natural light that compels us to accept as true that which we are unable to doubt.
Thus, the method of doubt leads naturally to the cogito, and not supernaturally
to truth as the nouveaux pyrrhoniens claimed. (Popkin, 2003: 151–152)
Popkin’s analysis of the cogito places it in its philosophical context. I argue in Maia
Neto (2014: 98) that Popkin is right in claiming that the Cartesian doubt is much
more radical than the doubt of his contemporary skeptics, whose skepticism did
not include the existence of the external world, remaining, like the ancient one,27 a
skepticism about properties. However, as has also been remarked by other scholars
(e.g., Giocanti 2001), the épochē of the French skeptics is not a blank tablet waiting
for a miraculous reception of the truth. Far from it, they hold a critical rationalism
(or rationalism without dogmas), that is characteristic of the Academic skeptical
tradition and that is, in the specific case of Charron, the starting point of Descartes’s
metaphysics. Descartes’s first principle, supposedly immune to skeptical doubt, is
mainly based on the moral Academic self-assurance of Charron’s wise man. This
reverses Popkin’s view of “Descartes conqueror of skepticism,” since the relevant
kind of skepticism for Descartes is not the object of his refutation but a major
inspiration for his own philosophy. However, Charron’s skeptical Academic wisdom
is only the point of departure, since through hyperbolic doubt Descartes transforms
the self-assurance of the Academic wise man into the metaphysical certainty of the
cogito. Although Descartes does not see himself as going against—but only beyond—
Charron,28 he completely reverses the latter’s Academic skepticism, inaugurating,
as Popkin rightly claims, a new form of dogmatism, in a certain way immune to
skepticism, since the moral skeptical position itself becomes—through hyperbolic
doubt—the first principle of the new philosophy.29
7 POPKIN’S SKEPTICISM
Unlike most philosophers examined in this book, Popkin did not advance skeptical
arguments of any kind. But also unlike most of them, Popkin did consider himself
a skeptic. His skepticism appeared in his view of the history of philosophy as a
perennial and fruitless effort made by dogmatic philosophers to refute skepticism. At
least this is how he sees the history of early modern philosophy, whose major events
are the skeptical attack on Aristotelianism and Descartes’s attack on skepticism
and establishment of a new dogmatism, only to give rise to the stronger skepticism
of Bayle and Hume.30 Besides a skeptic, Popkin (1988: 117) considered himself a
believing Jew. How so? I conclude this chapter with two brief considerations on
this matter. The first is that this double identity is consistent with Popkin’s belief
in the coherence of the position of the Christian skeptics he studied. The second
is the following passage on the skeptics he quoted from Kant’s preface to the first
514 SKEPTICISM
edition of the Critique of Pure Reason: “the sceptics, a species of nomads, despising
all settled modes of life, broke up from time to time all civil society. Happily they
were few in number. . .” He comments:
The sceptics as nomads, having no home, are like the Jews were, wandering across
the world, and like the unaffiliated, non-Zionist Jews are still. The intellectual
nomads (though few in number) like myself, find no home except in the broad
humanistic values of Socrates, Montaigne, Spinoza, Tolstoy, Orwell, and Victor
Serge, among others. This may make one a permanent outsider, one who sees the
faults and failings and dangers of the insiders. (Popkin 1988: 147)
NOTES
1. One of Popkin’s heroes, the skeptic Pierre Bayle, edited and wrote from 1684 to
1687 one of the first journals of reviews, the Nouvelles de la République des Lettres.
The title expresses Bayle’s ideal of a “republic” of philosophers and savants in which
nationality and religion would not compromise the free interchange and criticism
of ideas.
2. Popkin collaborated closely with Paul Henry, Alexander Koyré, Robert Lenoble,
Bernard Rochot, Henri Gouhier, and Elisabeth Labrousse, among many others. He
hired Herbert Marcuse when he was chair of the Philosophy Department of the
University of California at San Diego.
3. One of these publications (Popkin 1996) was, to my knowledge, the first volume
covering, like the present one, the complete history of skepticism, from the ancient
Academic skeptics to Wittgenstein. Because the book was the publication of a
conference to which Popkin invited scholars to present their ongoing research, it
does not have the completeness of the present volume.
4. The series, published in 1966 by Macmillan, comprised eight volumes with
major philosophical texts in English from Thales to the twentieth-century
analytic tradition.
5. “I took the Loeb Library edition out of the library, and on some subway ride from
Columbia to our house in the Bronx, or vice-versa, I encountered a philosophical
author that I could make sense of, who spoke to me” (Popkin 1988: 106).
6. Popkin got a PhD in Philosophy from Columbia University in 1950. His thesis was
on the philosophy of mathematics and logic.
7. The relevance of Popkin’s insight to contemporary Hume scholarship is attested by
Don Baxter’s chapter on Hume in the present book.
8. Undated fellowship application cited by Jeremy Popkin (2009: 23). Popkin’s first
published article on skepticism was on Hume (Popkin 1950).
9. Bayle is acknowledged today as a major source of Hume’s skepticism. Huet is
cited by Hume in Part I of the Dialogues Concerning Natural Religion (Hume
1980: 10).
10. With respect to the history of skepticism from Bayle to Hume to Kant, Popkin’s
initial view that skepticism did not play a major role in the Enlightenment was
probably one of the reasons he did not carry through the initial project. However,
when he became acquainted with the work of the Italian scholar Giorgio Tonelli,
he changed his view and worked on Jean-Pierre Brissot and Condorcet. See his
two articles on the role of skepticism in the Enlightenment—the first dismissing
and the second recognizing its importance—in Popkin et al. (1997). Skepticism is
today a well-established topic of research in the history of the Enlightenment—see
Paganini et al. (2002), Bernier et al. (2005), and Charles et al. (2013). One of the
conferences Popkin organized was on the wide role played by skepticism around
1800, in particular in Germany just before and after Kant with Gottlob Ernest
Schulze and Friedrich Stäudlin. The proceedings were published in Van der Zande
et al. (1998).
11. The edition also included Diogenes Laertius’s Life of Pyrrho and Erasmus’s
translation of Galen’s De Optimo Docendi Genere.
12. For the problem of the criterion, see the chapter on Sextus in the present book.
13. The division into sects in the Protestant camp was already a tendency by the time
Sextus’s works were published.
14. For the relevance of the skepticism of the period in political thought, see Battista
(1966), Bianchi (1988), Laursen (1992), and Lessa (2009). For its importance in
English literature, see Hamlin (2005).
15. For the centrality of Sextus’s Pyrrhonism to Montaigne’s skepticism in the Essays,
see Paganini’s chapter in this book.
16. For the presence of Academic skepticism in general during the early modern period,
see Maia Neto (1997; 2014); in Pedro de Valencia, see Levy (2001) and Laursen
(2009); in Grotius, Tuck (1983); in Descartes, Lennon (2008: 242–244; 2011) and
Giocanti (2013).
17. See Limbrick (1977), Panichi (2009), and Eva (2013). In probably the first critical
reading of the manuscript of the first draft of the History of Scepticism, Kristeller
remarks on Popkin’s dismissal of Academic skepticism (see Jeremy Popkin 2009: 26).
18. In the 2003 edition, Popkin attributes this position also to the “Philosophers of the
Royal Society: Wilkins, Boyle, and Glanvill” (2003: 208–218).
19. Popkin was influenced mostly by Funkenstein (1987). However, Popkin (2003: 50)
believed that this scholastic and theology-based skepticism was less influential in
the sixteenth and seventeenth centuries than the skepticism derived from ancient
sources (mainly Sextus). For fourteenth-century skepticism, see Lagerlund’s and
Grellard’s chapters in this book.
20. Popkin (2003: 34–35, 55, 85–88) criticizes two major French scholars who support
the view that the French sixteenth- and seventeenth-century skeptics were irreligious,
namely, Busson (1922; 1933) and Pintard (1943). This traditional view, although
no longer hegemonic in the cases of Montaigne and Charron, is still accepted in the
case of La Mothe Le Vayer (e.g., Cavaillé 2002: 141–191) and has been supported
against Popkin’s view by, among others, Moreau (2007: 85–105) and Gros (2009:
540–542, 551–552).
21. Many early modern scholars have argued for this view: see Penelhum (1983),
Gregory (1992), and Smith (2009). See also Giocanti’s chapter on La Mothe Le
Vayer in this book. Naya (2000; 2009) and McKenna (2003) argue that Academic
skepticism—but not Pyrrhonism—can be used in Christian apologetics.
22. See, for instance, Curley (1978). See also Lennon’s (2008: 55–62) review of the
influence of Popkin’s “standard interpretation” of Descartes in the literature.
516 SKEPTICISM
23. See Popkin (2003: 105–111). Popkin finds this kind of “Aristotelian” reply to
skepticism “in some comments of Sir Francis Bacon” (2003: 110). For criticism
of Popkin’s view of Bacon as an anti-skeptic, see Oliveira (2002: 74–84) and
Manzo (2009).
24. According to Popkin (2003: 128–142), this was the position held by Herbert of
Cherbury and Jean de Silhon.
25. See, for instance, Gaukroger (1995: 181–186) and Secada (2000: 38–41).
26. See also Paganini’s chapter on Charron in this book.
27. But maybe unlike fourteenth-century skepticism—see the chapters by Christophe
Grellard and Henrik Lagerlund in the present book.
28. The text that shows the crucial role of Charron’s De la Sagesse in Descartes’s
philosophical project is the opening paragraph of Descartes’s unfinished dialogue,
The Search for Truth (CSM II 400). CSM’s translation of the passage, however,
conceals its Charronian source. See my analysis of the passage and criticism of
CSM’s translation in Maia Neto (2014: 117–120).
29. For a different view of Descartes’s relation to Academic skepticism, see
Lennon (2011).
30. See Laursen’s chapter on Bayle and Baxter’s chapter on Hume in this book.
REFERENCES
Ayers, Michael. 2004. “Popkin’s Revised Scepticism,” British Journal for the History of
Backus, Irena. 2009. “The Issue of Reformation Scepticism Revisited: What Erasmus and
Sebastian Castellio did or did not know.” In J. Maia Neto, G. Paganini, and J. Laursen
(eds.), Skepticism in the Modern Age: Building on the work of Richard Popkin, 61–89.
Leiden: Brill.
Battista, Ana M. 1966. Alle Origini del pensiero politico libertino: Montaigne e Charron.
Milano: A. Giuffrè.
Bayle, Pierre. 1991. Historical and Critical Dictionary: Selections. Translated, with an
introduction and notes, by R. Popkin. Indianapolis: Hackett.
Bernier, Marc and Sébastien Charles (eds.). 2005. Scepticisme & Modernité. Saint-Étienne:
Publications de l’Université de Saint-Étienne.
Bianchi, Lorenzo. 1988. Tradizione liberina e critica storica. Milano: Franco Angeli.
Bracken, Harry. 1993. “Bayle’s Attack on Natural Theology: The Case of Christian
Pyrrhonism.” In R. Popkin and A. Vanderjagt (eds.), Scepticism and Irreligion in the
Seventeenth and Eighteenth Centuries, 254–266. Leiden: Brill.
Busson, Henri. 1922. Les sources et le développement du rationalisme dans la littérature
française de la Renaissance (1533–1601). Paris: Letouzey et Ané.
Busson, Henri. 1933. La pensée religieuse française de Charron à Pascal. Paris: Vrin.
Cavaillé, Jean-Pierre. 2002. Dis/simulations: Jules-César Vanini, François La Mothe Le
Vayer, Gabriel Naudé, Louis Machon et Torquato Accetto: religion, morale et politique
au XVIIe siècle. Paris: Honoré Champion.
Charles, Sébastien and Plínio Smith (eds.). 2013. Scepticism in the Eighteenth Century:
Enlightenment, Lumières, Aufklärung. Dordrecht: Springer.
Curley, Edwin. 1978. Descartes against the Skeptics. Cambridge, MA: Harvard University
Press.
Descartes, René. 1984. The Philosophical Writings. 2 vols, translated by J. Cottingham,
R. Stoothoff, and D. Murdoch. Cambridge: Cambridge University Press.
Eva, Luiz. 2013. “Montaigne et les Academica de Cicéron,” Astérion. Philosophie, histoire
des idées, pensée politique 11 (online).
Funkenstein, Amos. 1987. “Scholasticism, Scepticism and Secular Theology.” In R. Popkin
and C. Schmitt (eds.), Scepticism from the Renaissance to the Enlightenment, 45–54.
Wiesbaden: Harrassowitz.
Gaukroger, Stephen. 1995. Descartes: An Intellectual Biography. Oxford: Clarendon Press.
Giocanti, Sylvia. 2001. Penser l’irrésolution. Montaigne, Pascal, La Mothe Le Vayer: Trois
Giocanti, Sylvia. 2013. “Comment traiter ce qui n’est pas ‘entièrement certain et
indubitable.’ Descartes héritier des Académiques de Cicéron,” Astérion. Philosophie,
histoire des idées, pensée politique 11 (online).
Gregory, Tulio. 1992. “Pierre Charron’s ‘Scandalous Book.’” In M. Hunter and D.
Wootton (eds.), Atheism from the Reformation to the Enlightenment, 87–109. Oxford:
Gros, Jean-Michel. 2009. Les dissidences de la philosophie à l’âge classique. Paris: Honoré
Champion.
Hamlin, William. 2005. Tragedy and Scepticism in Shakespeare’s England. Hampshire:
Plagrave/St. Martin’s Press.
Hume, David. 1980. Dialogues concerning Natural Religion and the Posthumous Essays.
Edited, with an introduction, by R. Popkin. Indianapolis: Hackett.
Laursen, John Christian. 1992. The Politics of Skepticism in the Ancients, Montaigne, Hume
and Kant. Leiden: Brill.
Laursen, John Christian. 2009. “Pedro de Valencia’s Academica and Scepticism in Late
Renaissance Spain.” In J. Maia Neto, G. Paganini, and J. Laursen (eds.), Skepticism in
the Modern Age: Building on the Work of Richard Popkin, 111–123. Leiden: Brill.
Lennon, Thomas M. 2002. “What Kind of Skeptic was Bayle?” Midwest Studies in
Lennon, Thomas M. 2004. “Huet, Descartes, and the Objection of Objections.” In J. Maia
Neto and R. Popkin (eds.), Skepticism in Renaissance and Post-Renaissance Thought:
New Interpretations, 123–142. New York: Humanity Books.
Lennon, Thomas M. 2008. The Plain Truth: Descartes, Huet and Skepticism. Leiden: Brill.
Lennon, Thomas M. 2011. “Descartes, Arcesilau e a estrutura da epokhé,” Filosofia e
educação 25: 37–62.
Lessa, Renato. 2009. “Montaigne’s and Bayle’s Variations: The Philosophical Form of
Skepticism in Politics.” In J. Maia Neto, G. Paganini, and J. Laursen (eds.), Skepticism
in the Modern Age: Building on the Work of Richard Popkin, 211–228. Leiden: Brill.
Lévy, Carlos. 2001. “Pierre de Valence, historien de l’Académie ou académicien?” In P-F.
Moreau (ed.), Le scepticisme au XVIe et au XVIIe sècle, 174–187. Paris: Albin Michel.
Limbrick, Elaine. 1977. “Was Montaigne Really a Pyrrhonian?” Bibliothèque d’Humanisme
et Renaissance 28: 67–80.
518 SKEPTICISM
Maia Neto, José R. 1995. The Christianization of Pyrrhonism: Scepticism and Faith in
Pascal, Kierkegaard, and Shestov. Dordrecht: Kluwer.
Maia Neto, José R. 1999. “Bayle’s Academic Scepticism.” In J. Force and D. Katz (eds.),
Everything Connects, 264–276. Leiden: Brill.
Maia Neto, José R. 2012. “O Contexto Religioso-Político da Contraposição entre
Pirronismo e Academia na ‘Apologia de Raymond Sebond,’” Kriterion 126: 351–374.
Maia Neto, José R. 2014. The Charronian Legacy: Academic Skepticism in Seventeenth-
Century French Philosophy (1601–1662). Dordrecht: Springer.
Maia Neto, José and Elene Pereira Maia. 2002. “Boyle’s Carneades,” Ambix 49: 97–111.
Maia Neto, José, Gianni Paganini, and John Laursen (eds.). 2009. Skepticism in the Modern
Age: Building on the work of Richard Popkin. Leiden: Brill.
Manzo, Silvia. 2009. “Probability, Certainty, and Facts in Francis Bacon’s Natural Histories:
A Double Attitude Towards Skepticism.” In J. Maia Neto, G. Paganini, and J. Laursen
(eds.), Skepticism in the Modern Age: Building on the work of Richard Popkin, 123–137.
Leiden: Brill.
McKenna, Antony. 2003. “Scepticism at Port-Royal: The Perversion of Pyrrhonian Doubt.”
In G. Paganini (ed.), The Return of Scepticism: From Hobbes and Descartes to Bayle,
249–265. Dordrecht: Kluwer.
Moreau, Isabelle. 2007. « Guérir du sot »: Les stratégies d’écriture des libertins à l’âge
classique. Paris: Honoré Champion.
Naya, Emmanuel. 2000. Le Phénomène pyrrhonien: lire le scepticisme au XVIe siècle, Ph.D.
dissertation, Université de Grenoble III.
Naya, Emmanuel. 2009. “Renaissance Pyrrhonism: A Relative Phenomenon.” In J. Maia
Neto, G. Paganini, and J. Laursen (eds.), Skepticism in the Modern Age: Building on the
work of Richard Popkin, 13–32. Leiden: Brill.
Oliveira, Bernardo. 2002. Francis Bacon e a Fundamentação da Ciência como Tecnologia.
Belo Horizonte: Editora da UFMG.
Paganini, Gianni. 2008. Skepsis. Le débat des modernes sur le scepticisme: Montaigne—Le
Vayer—Campanella—Hobbes—Descartes—Bayle. Paris: J. Vrin.
Paganini, Gianni. 2008b. “Charron et le scepticisme des modernes,” Corpus: revue de
philosophie 55: 231–249.
Paganini, Gianni. 2009. “Descartes and Renaissance Skepticism: The Sanches Case.” In J.
Maia Neto, G. Paganini, and J. Laursen (eds.), Skepticism in the Modern Age: Building
on the work of Richard Popkin, 249–267. Leiden: Brill.
Paganini, Gianni, Miguel Benítez, and James Dybikowski (eds.). 2002. Scepticisme,
Clandestinité et Libre Pensée. Paris: Honoré Champion.
Paganini, Gianni and José Maia Neto (eds.). 2009. Renaissance Scepticisms. Dordrecht:
Springer.
Panichi, Nicola. 2009. “Montaigne and Plutarch: A Scepticism that Conquers the Mind.”
In J. Maia Neto, G. Paganini, and J. Laursen (eds.), Skepticism in the Modern Age:
Building on the Work of Richard Popkin, 183–211. Leiden: Brill.
Pascal, Blaise. 1989. Selections, edited by R. Popkin. New York: Macmillan.
Penelhum, Terence. 1983. God and Skepticism. Dordrecht: Kluwer.
Pintard, René. 1943. Le Libertinage érudit dans la première moitié du XVIIème siècle. Paris:
Boivin.
Popkin, Jeremy. 2009. “Richard Popkin and his History of Scepticism.” In J. Maia Neto,
G. Paganini, and J. Laursen (eds.), Skepticism in the Modern Age: Building on the work
of Richard Popkin, 15–34. Leiden: Brill
Popkin, Richard. 1950. “David Hume: His Pyrrhonism and His Critique of Pyrrhonism,”
Popkin, Richard. 1960. The History of Scepticism: From Erasmus to Descartes. Assen: Van
Gorcum.
Popkin, Richard. 1979. The History of Scepticism: From Erasmus to Spinoza. Berkeley:
Popkin, Richard. 1980. The High Road to Pyrrhonism. Papers collected by R. Watson and
J. Force. San Diego: Austin Hill.
Popkin, Richard. 1988. “Intellectual Autobiography: Warts and All.” In R. Watson and J.
Force (eds.), The Sceptical Mode in Modern Philosophy: Essays in Honor of Richard H.
Popkin, 103–149. Dordrecht: Martinus Nijhoff.
Popkin, Richard. 1992. The Third Force in Seventeenth Century Philosophy. Leiden: Brill.
Popkin, Richard (ed.). 1996. Scepticism in the History of Philosophy: A Pan-American
Dialogue. Dordrecht: Kluwer.
Popkin, Richard. 2003. The History of Scepticism: From Erasmus to Savonarola. Oxford:
Popkin, Richard, Ezequiel de Olaso, and Giorgio Tonelli (eds.). 1997. Scepticism in the
Enlightenment. Dordrecht: Kluwer.
Schmitt, Charles. 1972. Cicero Scepticus: A Study of the Influence of the Academica in the
Renaissance. The Hague: Martinus Nijhoff.
Secada, Jorge. 2000. Cartesian Metaphysics: The Late Scholastic Origins of Modern
Philosophy. Cambridge: Cambridge University Press.
Smith, Plínio. 2009. “Skepticism, Belief, and Justification.” In J. Maia Neto, G. Paganini,
and J. Laursen (eds.), Skepticism in the Modern Age: Building on the work of Richard
Popkin, 171–190. Leiden: Brill.
Tuck, Richard. 1983. “Grotius, Carneades and Hobbes,” Grotiana (NS) 4: 43–62.
Van der Zande, Johan and Richard Popkin (eds.). 1998. The Skeptical Tradition Around
1800: Skepticism in Philosophy, Science and Society. Dordrecht: Kluwer.
Watson, Richard. 1966. The Downfall of Cartesianism, 1673–1712: A Study of
Epistemological Issues in Late 17th Century Cartesianism. The Hague: Martinus Nijhoff.
Wright, John. 1986. “Hume’s Academic Scepticism: A Reappraisal of His Philosophy of
Human Understanding,” Canadian Journal of Philosophy 16: 407–436.
PART FOUR
Contemporary Skepticism
Introduction
Contemporary Skepticism
BARON REED
Skepticism has been a powerful spur to innovation in epistemology throughout the

history of philosophy. This is especially true in the present era, as many of the most
significant developments in recent epistemology have been motivated, at least in
part, as responses to global and perceptual skeptical challenges.
The main argument for global skepticism has its origin in the Modes of Agrippa,
presented by Sextus Empiricus in the Outlines of Pyrrhonism (PH I 164–177).
Take any one of your beliefs. The skeptic will ask, why do you think it is true? A
natural response might be to offer another of your beliefs in support of the first
one. But this second belief can also be called into question, and it will need to be
defended in turn. A third belief might then be offered in support of the second, but
it, too, can be challenged and is in need of defense. As further beliefs are offered
and challenged in turn, it becomes clear that there are three possible patterns your
reasoning might take. (i) At every step, you could offer a previously unmentioned
belief in response to the skeptic’s relentless challenges. But the skeptic will object
that this amounts to a vicious infinite regress—it’s vicious because at no point do you
ever succeed in offering a complete defense of any of the beliefs up for discussion.
(ii) Alternatively, at some point you could offer again a belief that has already been
used and challenged in the dispute. But the skeptic will object that this is circular
reasoning and fails because it, too, does not permit a complete defense of any of
the beliefs you have mentioned. (iii) Finally, you could offer a belief and refuse to
defend it further. But the skeptic will say this is an arbitrary place to stop, subject to
challenge itself, and incapable of putting to rest the questions that have been raised
to your other beliefs. Because every response to the skeptic’s questioning must fit
one of these patterns, and because the skeptic can raise this sort of challenge to any
belief, it looks as though there is no way to successfully defend any of your beliefs.1
The Pyrrhonists thought that the natural outcome of this sort of questioning would
be universal suspension of judgment. Other skeptics have thought that we are
incapable of living without beliefs, and so the skeptical challenge does not lead us
to abandon them altogether. Global skepticism, understood in this way, amounts
to the claim that our beliefs lack a desirable normative status: they fail to count as
knowledge, or they are irrational.
The considerations underwriting perceptual skepticism stem largely from early
modern philosophy. One major strand is the dream argument from Descartes’s First
Meditation (CSM II 12–13; AT VII 18–19).2 The meditator says that there is no
sure sign by means of which one can tell whether one is awake or dreaming. So,
524 SKEPTICISM
when you seem to see, for example, the pages of a book, you can’t be sure that
you are actually reading it or are only dreaming that you are reading. Because the
experience you are having is consistent with both possibilities, it is insufficient to
justify your belief that you are actually reading.3
Another strand has to do with the indirect realist account of perception that
became popular in the early modern period. Although it may seem obvious that
perception makes us directly aware of everyday objects like trees and flowers, Hume
says that this opinion “is soon destroyed by the slightest philosophy” (EHU XII
118). Our experiences vary with changes in our perceptual system—for example, the
same water can feel cold to one person and hot to another, depending on whether
they are hot or cold themselves—and a little reflection makes clear that sentient
beings with different perceptual systems will perceive objects very differently than
we do. Moreover, some of our experiences, like dreams, hallucinations, double-
vision, etc., seem to present to us objects that aren’t really there. In those cases, we
simply cannot be directly aware of mind-independent objects. What we are directly
aware of, then, are experiences that are misrepresentations of reality. But ordinary
cases of perception are just like these delusions, inasmuch as we often can’t tell,
just from the experience itself, whether it is veridical or not. If so, this means that
every case of perceptual experience involves a mental representation of which we
are directly aware and which makes us indirectly aware of what it represents when
the perception is veridical.4 The early modern philosophers sometimes combined
this sort of indirect realism with the distinction between primary and secondary
qualities.5 According to this distinction, our ideas of the secondary qualities—
colors, sounds, smells, etc.—do not resemble the qualities in the object that cause
them. Descartes, Locke, and others tried to maintain that our ideas of the primary
qualities—spatial properties, number, and motion—do resemble the qualities that
cause them, but later philosophers were not always persuaded.6 On the resulting
account of perception, then, we are aware of everyday objects only indirectly, by
means of a direct awareness of some representation that may be quite unlike the
object it purports to represent. Many philosophers began to worry that, on such
a view, perception seems to be incapable of providing us knowledge of the world
outside our own minds. We are, instead, trapped behind a “veil of perceptions.”7
In response to these two central forms of skepticism, contemporary
epistemologists have offered a rich variety of competing theories. One of the first
to become prominent was foundationalism. According to this view, we can solve
the regress problem posed by the Modes of Agrippa if we can identify a class of
basic beliefs, which are capable of justifying other beliefs without needing to be
justified themselves.8 That class was usually thought to comprise beliefs about
the “given” element in experience (Lewis 1946: 203)—that aspect of perception
that arises independently of, and prior to, our forming beliefs through the use of
perceptual concepts. Because we are directly aware of the given, it was thought that
beliefs arising immediately from our cognitive access to it would provide the secure
foundation from which all of our knowledge could arise.
Some philosophers have rejected the possibility that there could be a class of
basic beliefs at all (BonJour 1985: 30–33). Others have argued that the different
INTRODUCTION 525
roles assigned to the given—as something independent of our conceptualization

and as capable of justifying other beliefs—cannot be satisfied by the same thing
(Sellars 1963). And many philosophers have worried that, even if we can find a
level of basic beliefs that are directly justified through contact with the given,
they will be too impoverished, cognitively speaking, to justify the vast majority of
our other beliefs. Suppose it is granted that each of us can know with certainty,
on some occasions, that we are having an experience of something red or
something hot; even so, it is a long way from there to deriving an understanding
of, say, the Great Red Spot on Jupiter. Motivated by these anti-foundationalist
arguments, some epistemologists have tried to find other ways of getting past
the Modes of Agrippa. Coherentists have articulated a theory of justification
and knowledge that aims to rehabilitate what might otherwise look like circular
reasoning.9 Infinitists have tried to defend the possibility that an infinite regress
can be virtuous rather than vicious, underwriting justification for a challenged
belief.10 Other philosophers have argued that there is no non-skeptical solution
to the Modes of Agrippa available to us.11 A troubling consequence, if that is so,
is that we may often lack a standard by which disagreements can be brought to a
shared resolution.12
Many of the other responses to skepticism in contemporary epistemology can
be understood as ways of relaxing, or abandoning, one (or more) of the elements
of foundationalism. Some philosophers have thought that skepticism arises because
we have set our aim too high. Instead of looking for absolutely secure foundations,
capable of grounding certainty in our beliefs, we should be content with a lesser
sort of knowledge. According to fallibilism, it is possible for a belief to count as
knowledge even when it is grounded in a justification that is less than perfect.13 So,
for example, the fact that I sometimes dream about rain falling doesn’t mean that
I am unable to know it is raining now, when I see ripples in the puddles and feel
drops on my face. That evidence is good enough for me to know it is raining, even
though it doesn’t guarantee that I am not mistaken. Although fallibilism ought to
make it easier to have knowledge, some philosophers have suggested that it brings
with it a new vulnerability to skepticism.14 Fallibilism makes it possible for truth
and justification to be combined in the same belief by accident; in that case, the
belief fails to count as knowledge (Gettier 1963). On that point, epistemologists
largely agree; the problem, however, is that it may not be possible to explain how
truth and justification are combined in a non-accidental way in beliefs that do count
as knowledge.
Fallibilism may be understood as a family of theories, which differ from one
another in the way they explain how imperfect justification and knowledge
are possible. The fallibilist view that is closest to classic foundationalism is
dogmatism. According to this view, much of our knowledge rests on perception,
but the beliefs at the foundations need not be about the perceptual experiences
themselves; rather, the dogmatist takes perception to put us directly in cognitive
contact with everyday objects.15 The justification perception provides us is prima
facie or defeasible, in that it can be refuted or undermined by counterevidence.16
Dogmatists then argue that this prima facie justification is usually good enough to
526 SKEPTICISM
allow us to have perceptual knowledge of the objects around us and to overcome

the dream argument.17 A skeptic might note, however, that the existence of
dreams, delusions, and deceivers provides an undercutting defeater for this prima
facie justification. If so, the dogmatist will need to provide a rational response to
that defeating consideration.
Disjunctivism provides a second way of broadening the class of foundational
beliefs. Like dogmatism, disjunctivism allows for beliefs about everyday objects
to count as basic, but it offers a different account of the nature of perceptual
experience. Where the dogmatist grounds justification in the way things seem to
the perceiver, the disjunctivist says that perceptual justification arises from the fact
that the perception includes the object as a constituent. For this reason, the dream
argument for skepticism fails: someone who is having a waking experience of a
pine tree is in a different sort of perceptual state than someone who is dreaming of
a pine tree. The view is disjunctive in virtue of rejecting the argument from illusion
mentioned above; disjunctivists reject the notion that all perceptual experience must
be understood as involving a mental representation of external reality. At most, that
holds only for delusory experiences. Veridical experience, by contrast, provides us
with a direct cognitive access to the world around us.18 A skeptic will object, though,
that it is impossible for the person having the experience in question to tell whether
it is veridical or not. The debate between the skeptic and the disjunctivist, then, will
turn on whether this is so and, if it is, whether it matters from an epistemological
point of view.19
Fallibilism has also been developed in a variety of ways that break the connection
between justification and knowledge, on the one hand, and what we are immediately
aware of in experience, on the other. These theories, which fall under the rubric
of externalism, take justification or knowledge to be grounded in a causal or
subjunctive relationship holding between the belief and the fact that it is about.20
That relationship can hold without the believer being aware that it does. All that
matters is that, in having the belief, the believer has the right sort of cognitive fit
with her environment.21 If that is the case, then there is presumably a simple solution
to both skeptical challenges mentioned above.22 First, there is no need for either
a regress or a circle of beliefs to be offered in response to a skeptic’s questioning;
as long as the target belief has the requisite externalist property, it is justified (and
typically, when true, counts as knowledge as well). Second, the believer doesn’t
have to antecedently know that she is not dreaming in order for her belief to count
as justified; it simply has to be the case that she is not dreaming (and is otherwise
appropriately related to her environment). Some externalists have then argued that
the believer can make use of this knowledge to infer that she is not dreaming or
a brain in a vat or the victim of an evil demon.23 Other philosophers object that
this makes it too easy to reply to skepticism.24 A further problem for externalists
has to do with the role that rationality plays in their theories. Like the dogmatists,
most externalists have allowed that justification can be defeated by counterevidence
that one has or should have.25 If the considerations that underlie skepticism can be
used as a defeater for our purportedly justified beliefs—so that, even if our beliefs
do properly fit with our surrounding environment, we nevertheless have reason
INTRODUCTION 527
to doubt that they do—then externalists will have to find a rational response to
skepticism, just as internalists do.26
Given this sort of difficulty, some philosophers have sought to insulate our
everyday knowledge from the threat of skepticism.27 One strategy for doing
so is to argue, as Wittgenstein did, that the doubts raised by the skeptic are in
some sense illegitimate. The challenges we raise to one another’s beliefs occur in
frameworks that are shaped by our practical and theoretical needs. You may be
told by both your uncle and a doctor that you have pneumonia, but the doctor will
need to provide more evidence to support her diagnosis, due to the fact that we
have higher expectations of accuracy for physicians. Even so, we do not expect the
doctor to be able to trace her diagnosis back to first principles or to show that she
is not dreaming.28 Some philosophers have replied either that skepticism embodies
epistemological principles that govern all sorts of inquiry or that it creates its own,
legitimate sort of inquiry.29 At the heart of this dispute is a deep disagreement over
the nature of philosophy and the role it might play in everyday life.
Another insulation strategy aims to protect from skepticism our everyday
attributions of knowledge. According to contextualism about “knows” and cognate
terms, they are context-sensitive: sentences that include them may express different
propositions in different conversational contexts because those terms determine
epistemic standards that can shift, depending on our conversational needs.30 Just
as “he is tall” may assert very different propositions when I affirm it of a toddler
and of a professional basketball player, so, too, “she knows it will rain” may
assert different propositions when I affirm it of a farmer and of a meteorologist.
Contextualists argue that shifts in epistemic standards are especially apparent when
the practical stakes relevant to a knowledge attribution change in a conversation. But
they also occur when a skeptical challenge is posed to an attribution of knowledge.
When that happens, the epistemic standards governing the conversation go up to
such an extent that very few beliefs will meet them. This is why skepticism often
seems impossible to refute, once we are considering it. But, the contextualist will
point out, most of our conversations are governed by much less stringent epistemic
standards, and so our everyday knowledge attributions do not face this difficulty.
When we are not engaged in conversation with the skeptic, there is apparently no
bar to our continuing to attribute knowledge to ourselves and to others. Although
contextualism has proven to be a fruitful way of theorizing about how we talk
about knowledge attributions, some epistemologists have remained doubtful that
it provides much illumination about knowledge itself.31 It may also be possible to
regiment the skeptical challenge, so that it can be brought to bear on everyday
beliefs without distortion from shifting epistemic standards.32 If so, contextualism
may need to be supplemented with one of the other responses to skepticism
considered above.
In addition to the general and perceptual arguments for skepticism, there are
also skeptical challenges to other modes and fields of inquiry. Skepticism about
inductive inference, spurred in large part by Hume, continues to be a difficult
problem in epistemology and philosophy of science.33 Skeptical challenges have
been raised both to the possibility of a priori knowledge and to there being a
528 SKEPTICISM
coherent conception of the a priori at all.34 Our ability to know about the mental
lives of others continues to be problematic.35 Moral skepticism remains a compelling
possibility in ethical theory.36 And, although religious matters are no longer at the
forefront of contemporary research on skepticism as they were in the early modern
period, religious skepticism continues to be an illuminating perspective on questions
of faith and doubt.37
The skeptical challenges of today have their origin in the great skeptical arguments
of the past. As it has throughout the history of philosophy, skepticism continues to
provoke new insights and to adapt itself to the changing problems and perspectives
that compel our attention. For that reason, understanding skepticism in its full
variety, both past and present, remains an essential part of a philosophical way of life.
NOTES
1. This form of argument has also been called the problem of the criterion, the regress
argument, and the problem of epistemic circularity. See the chapters by Richard
Fumerton, Andrew Cling, and Markus Lammenranta in this book. According to
Richard Popkin (2003: ch. 1, 4), both Protestants and Catholics in the early modern
period used the problem of the criterion to call into question each other’s preferred
way of forming beliefs about theological matters; this was a major impetus in the
revival of skepticism. This sort of problem can also be seen in the “Cartesian Circle”
objection raised by Antoine Arnauld in the Fourth Set of Objections to Descartes’s
Meditations (CSM II 150; AT VII 214). See Alston (1986; 1993) for more on epistemic
circularity. See Chisholm (1966: ch. 4), Amico (1993), and Andrew Cling’s chapter in
this book for a related, metaepistemological version of the problem of the criterion.
Another argument for global skepticism can be found in the First Meditation, where
the meditator considers the implications of not knowing whether he was created by a
benevolent God, by natural causes, or by chance. This is often called the “evil demon”
argument, though that is a misleading way to think of it. The force of the argument
does not stem from the presentation of a skeptical scenario the meditator cannot rule
out; that would simply be a failure of evidence. Rather, the point of the argument is
that the very intellectual capacities with which the meditator might try to make use of
evidence are themselves in need of vindication. See Reed (2015: 72–73).
2. An updated version of this argument substitutes for the evil demon’s victim a brain in
a vat, hooked up to a computer that is stimulating it to have misleading experiences;
see Putnam (1981: ch. 1).
3. See Brueckner (1994), Vogel (2004), and Ram Neta’s chapter in this book for the
importance of the underdetermination of evidence in skeptical arguments. See
Stroud (1984) for a recent, influential presentation of skepticism that turns on the
underdetermination of our beliefs by the available evidence.
4. See Fumerton (1985: 78–87) on the argument from the possibility of hallucination
(or, as it is sometimes called, the argument from illusion).
5. Locke (II.8.9–26). Although the name for the distinction is Locke’s, the distinction
itself can also be seen in the work of many other early modern thinkers, including
Galileo, Descartes, and Boyle.
INTRODUCTION 529
6. See Bayle (Dictionary, “Pyrrho,” rem. B; 1965: 197): “Heat, smells, colors, and the
like, are not in the objects of our senses. They are modifications of my soul. I know
that bodies are not at all as they appear to me. They [i.e., the Cartesians] would
have wished to exempt extension and motion, but they could not. For if the objects
of our senses appear colored, hot, cold, odoriferous, and yet they are not so, why
can they not appear extended and shaped, in rest and in motion, though they are
not so?”
7. See Bennett (1971: 68–70) for a characterization of Locke’s theory of perception
using the “veil” metaphor.
8. See Lewis (1946), Chisholm (1966, 1989), and Fumerton (1985, 1995), as well
as the chapters by Fumerton and Cling in this book. By “justification,” I mean to
pick out a positive epistemic property of belief. There are many different ways of
understanding justification; for my purposes, it is easiest to characterize it by way of
example. Guessing it will rain tomorrow is not justified, even if it is true; forecasting
it will rain tomorrow on the basis of a well-constructed model and careful gathering
of evidence is justified, even if it is false. Someone with normal eyesight will be
increasingly justified in believing that there is a goldfinch in the tree as she gets closer
and closer to it. Belief that is both justified and true usually counts as knowledge,
though not in cases where the truth and the justification are combined merely by
accident; see Gettier (1963). There is still, of course, widespread disagreement over
what it means for a justified belief to be accidentally true.
9. See, e.g., BonJour (1985).
10. See Klein (1999) and Turri and Klein (2014).
11. See Fogelin (1994). For more on Fogelin’s “Neo-Pyrrhonism,” see Sinnott-
Armstrong (2004) and the chapter by Lammenranta in this book. See Fumerton
(1995: ch. 7) on the prospects of avoiding skepticism from the framework of
classic foundationalism. For more on responses to the regress problem, see the
chapters by Fumerton and Cling in this book, DePaul (2001), and Turri and
Klein (2014).
12. For the relation between disagreement and skepticism, see Bryan Frances’s
chapter in this book, Frances (2005), Machuca (2012), and Christensen and
Lackey (2013).
13. See Hetherington (1999), Reed (2002), and Brown (forthcoming). See Hetherington
(2011), Sosa (2011), and Reed (2013) for an exploration of the idea that there
may be different grades or kinds of knowledge, where some of them fall short
of certainty.
14. See Hetherington (1996), Reed (2007, 2009), and Hetherington’s chapter in this
book. See Neta’s chapter in this book for a consideration of the prospects of using
fallibilism in a reply to skepticism about the external world.
15. See, e.g., Pryor (2000, 2004) and Huemer (2001).
16. Pollock (1986) distinguishes between rebutting defeaters, which indicate that the
target belief is false, and undercutting defeaters, which indicate that the target
belief was unreliably formed. Suppose, for example, I am justified in believing,
on the basis of visual perception, that a dog in the distance is a husky. That
justification can be rebutted by expert testimony that the animal is, in fact,
530 SKEPTICISM
a coyote; it can be undercut by expert testimony that huskies and coyotes are
indistinguishable from this distance. For more on different types of defeaters, see
Lackey (2008: 44–46).
17. See Pryor (2000, 2004). Davies (2004) and Wright (2004) agree that perception
provides a kind of prima facie justification for knowledge of everyday objects, but
they argue that this justification does not “transmit” to the conclusion that we know
we are not dreaming.
18. See McDowell (1982), Snowden (1990), Butchvarov (1998), Pritchard (2012), and
the chapter by Duncan Pritchard and Chris Ranalli in this book.
19. Williamson (2000) offers a view that is similar in some ways to disjunctivism, as
he takes there to be a significant difference between being in the “good case” of
veridical perception and being in the “bad case” of deception or dreaming. See
Schiffer (2009) for a critical appraisal of Williamson’s response to skepticism and
Williamson (2009) for his reply to Schiffer.
20. See Dretske (1970), Armstrong (1973), Goldman (1979), Nozick (1981), Williamson
(2000), Sosa (2007, 2011), and Matthias Steup’s chapter in this book.
21. There is considerable disagreement among externalists over how to best understand
what it means to have a proper fit with one’s environment. Armstrong (1973) offers
a fairly simple causal theory, while Sosa thinks that the belief needs to be produced
by an intellectual virtue (2007) or manifest a competence (2011) and, at least in
the case of reflective knowledge, must be the target of the believer’s epistemic
perspective (2009).
22. See Zalabardo (2012). See Steup’s chapter in this book for an argument that
the externalist is in fact in no better position to respond to the skeptic than is
an internalist.
23. Sosa (2009: ix) calls this “virtuous circularity”; the circularity arises because the
person is using the very intellectual capacities she is vindicating with the inference.
See also Van Cleve (1979).
24. Fumerton (1995: 177) argues that “the very ease with which externalists can deal
with the skeptical challenge . . . betrays the ultimate implausibility of externalism as
an attempt to explicate concepts that are of philosophical interest.” See also Cohen
(2002) on the “problem of easy knowledge” and Steup’s chapter in this book.
25. See, for example, Goldman (1979), Nozick (1981), Bergmann (2004), and Lackey
(2008). Externalists embraced defeaters as a way of responding to the objections
pressed by BonJour (1985: ch. 3).
26. See Reed (2006).
27. See Burnyeat (1984) on insulating responses to skepticism in ancient and early
modern philosophy. See also the chapter in Part I of this book on Sextus Empiricus
by Tad Brennan and Cliff Roberts.
28. See Michael Williams’s chapter on Wittgenstein in Part III of this book, as well
as Williams (1996). See also Bergmann (2004) and Ruth Weintraub’s chapter in
this book on skepticism about induction and the “functionalist” framework for
answering it.
29. See Stroud (1984) and Fumerton (1995: ch. 6), respectively.
INTRODUCTION 531
30. See Cohen (1988), DeRose (1995, 2009), Lewis (1996), and Michael Blome-
Tillmann’s chapter in this book, as well as Blome-Tillmann (2014).
31. See, e.g., Sosa (2000) and Kornblith (2000).
32. See Reed (2010: 225–27).
33. See Weintraub’s chapter in this book; see also Swinburne (1974).
34. See Otávio Bueno’s chapter in this book; see also Beebe (2011).
35. See Anil Gomes’s chapter in this book; see also Avramides (2001).
36. See Richard Joyce’s chapter in this book, Joyce (2012, 2014), and Sinnott-
Armstrong (2006).
37. See J. L. Schellenberg’s chapter in this book, as well as Schellenberg (2007, 2013).
REFERENCES
Alston, William. 1986. “Epistemic Circularity,” Philosophy and Phenomenological Research
47: 1–30.
Alston, William. 1993. The Reliability of Sense Perception. Ithaca, NY: Cornell University
Press.
Amico, Robert. 1993. The Problem of the Criterion. Lanham, MD: Rowman and Littlefield.
Armstrong, David. 1973. Belief, Truth and Knowledge. Cambridge: Cambridge University
Press.
Avramides, Anita. 2001. Other Minds. London: Routledge.
Bayle, Pierre. 1965. Historical and Critical Dictionary. translated by R. Popkin. Indianapolis,
IN: Bobbs-Merrill.
Beebe, James. 2011. “A Priori Skepticism,” Philosophy and Phenomenological Research 83:
583–602.
Bennett, Jonathan. 1971. Locke, Berkeley, Hume: Central Themes. Oxford: Clarendon
Press.
Bergmann, Michael. 2004. “Epistemic Circularity: Malignant and Benign,” Philosophy and
Blome-Tillmann, Michael. 2014. Knowledge & Presuppositions. Oxford: Oxford University
Press.
BonJour, Laurence. 1985. The Structure of Empirical Knowledge. Cambridge, MA: Harvard
University Press.
Brown, Jessica. Forthcoming. Knowledge, Evidence and Fallibility. Oxford: Oxford
University Press.
Brueckner, Anthony. 1994. “The Structure of the Skeptical Argument,” Philosophy and
Burnyeat, Myles. 1984. “The Sceptic in his Place and Time.” In R. Rorty, J. Schneewind,
and Q. Skinner (eds.), Philosophy in History, 225–254. Cambridge: Cambridge
University Press.
Butchvarov, Panayot. 1998. Skepticism about the External World. Oxford: Oxford
University Press.
532 SKEPTICISM
Chisholm, Roderick. 1966. Theory of Knowledge, 1st ed. Englewood Cliffs, NJ: Prentice-
Hall.
Chisholm, Roderick. 1989. Theory of Knowledge, 3rd ed. Englewood Cliffs, NJ: Prentice
Hall.
Christensen, David and Jennifer Lackey (eds.). 2013. The Epistemology of Disagreement.
Cohen, Stewart. 1988. “How to Be a Fallibilist,” Philosophical Perspectives 2: 91–123.
Davies, Martin. 2004. “Epistemic Entitlement, Warrant Transmission, and Easy
Knowledge,” Aristotelian Society Supplement 78: 213–245.
DePaul, Michael (ed.). 2001. Resurrecting Old-Fashioned Foundationalism. Lanham, MD:
Rowman and Littlefield.
DeRose, Keith. 2009. The Case for Contextualism. Oxford: Clarendon Press.
Descartes, René. 1984. The Philosophical Writings of Descartes, vol. II, translated by
Dretske, Fred. 1970. “Epistemic Operators,” Journal of Philosophy 67: 1003–1013.
Fogelin, Robert. 1994. Pyrrhonian Reflections on Knowledge and Justification. Oxford:
Frances, Bryan. 2005. Scepticism Comes Alive. Oxford: Clarendon Press.
Fumerton, Richard. 1985. Metaphysical and Epistemological Problems of Perception.
Lincoln, NE: University of Nebraska Press.
Fumerton, Richard. 1995. Metaepistemology and Skepticism. Lanham, MD: Rowman and
Littlefield.
Gettier, Edmund. 1963. “Is Justified True Belief Knowledge?” Analysis 23: 121–3.
Goldman, Alvin. 1979. “What Is Justified Belief?” In G. Pappas (ed.), Justification and
Knowledge, 1–23. Dordrecht: Reidel.
Hetherington, Stephen. 1996. “Gettieristic Scepticism,” Australasian Journal of Philosophy
74: 83–97.
Hetherington, Stephen. 1999. “Knowing Failably,” Journal of Philosophy 96: 565–587.
Hetherington, Stephen. 2011. How to Know: A Practicalist Conception of Knowledge.
Oxford: Wiley-Blackwell.
Huemer, Michael. 2001. Skepticism and the Veil of Perception. Lanham, MD: Rowman
and Littlefield.
Principles of Morals, edited by P. H. Nidditch. Oxford: Clarendon Press.
Joyce, Richard. 2012. “Metaethical Pluralism: How Both Moral Naturalism and Moral
Skepticism May Be Permissible Positions.” In S. Nuccetelli and G. Seay (eds.), Ethical
Naturalism: Current Debates, 89–109. Cambridge: Cambridge University Press.
Joyce, Richard. 2014. “Taking Moral Skepticism Seriously,” Philosophical Studies 168:
843–851.
Klein, Peter. 1999. “Human Knowledge and the Infinite Regress of Reasons,” Philosophical
Perspectives 13: 297–325.
INTRODUCTION 533
Kornblith, Hilary. 2000. “The Contextualist Evasion of Epistemology,” Philosophical

Issues 10: 24–32.
Lackey, Jennifer. 2008. Learning from Words: Testimony as a Source of Knowledge. Oxford:
Lewis, C.I. 1946. An Analysis of Knowledge and Valuation. La Salle, IL: Open Court.
Lewis, David. 1996. “Elusive Knowledge,” Australasian Journal of Philosophy 74: 549–567.
Locke, John. 1975. An Essay Concerning Human Understanding, edited by P. Nidditch.
Machuca, Diego (ed.). 2012. Disagreement and Skepticism. New York: Routledge.
McDowell, John. 1982. “Criteria, Defeasibility and Knowledge,” Proceedings of the British
Academy 68: 455–79.
Press.
Pollock, John. 1986. Contemporary Theories of Knowledge. Totowa, NJ: Rowman &
Littlefield.
Popkin, Richard. 2003. The History of Scepticism: From Savonarola to Bayle. Oxford:
Pritchard, Duncan. 2012. Epistemological Disjunctivism. Oxford: Oxford University Press.
Pryor, James. 2000. “The Skeptic and the Dogmatist,” Noûs 44: 517–549.
Pryor, James. 2004. “What’s Wrong with Moore’s Argument?” Philosophical Issues 14:
349–378.
Putnam, Hilary. 1981. Reason, Truth and History. Cambridge: Cambridge University Press.
Reed, Baron. 2002. “How to Think about Fallibilism,” Philosophical Studies 107: 143–157.
Reed, Baron. 2006. “Epistemic Circularity Squared? Skepticism about Common Sense,”
Philosophy and Phenomenological Research 73: 186–197.
Reed, Baron. 2007. “The Long Road to Skepticism,” Journal of Philosophy 104: 236–262.
Reed, Baron. 2009. “A New Argument for Skepticism,” Philosophical Studies 142: 91–104.
Reed, Baron. 2010. “A Defense of Stable Invariantism,” Noûs 44: 224–244.
Reed, Baron. 2013. “Fallibilism, Epistemic Possibility, and Epistemic Agency,” Philosophical
Issues 23: 40–69.
Reed, Baron. 2015. “Skepticism and Perception.” In M. Matthen (ed.), The Oxford
Handbook of Philosophy of Perception, 66–80. Oxford: Oxford University Press.
Schellenberg, J. L. 2007. The Wisdom to Doubt: A Justification of Religious Skepticism.
Ithaca, NY: Cornell University Press.
Schellenberg, J. L. 2013. Evolutionary Religion. Oxford: Oxford University Press.
Schiffer, Stephen. 2009. “Evidence = Knowledge: Williamson’s Solution to Skepticism.”
In P. Greenough and D. Pritchard (ed.), Williamson on Knowledge, 183–202. Oxford:
Sellars, Wilfrid. 1963. “Empiricism and the Philosophy of Mind.” In his Science, Perception,
and Reality, 127–196. Atascadero, CA: Ridgeview.
Sinnott-Armstrong, Walter (ed.). 2004. Pyrrhonian Skepticism. Oxford: Oxford University
Press.
534 SKEPTICISM
Sinnott-Armstrong, Walter (ed.). 2006. Moral Skepticisms. Oxford: Oxford University

Press.
Snowden, Paul. 1990. “The Objects of Perceptual Experience,” Aristotelian Society
Supplement 64: 121–150.
Sosa, Ernest. 2000. “Skepticism and Contextualism,” Philosophical Issues 10: 1–18.
Sosa, Ernest. 2007. A Virtue Epistemology: Apt Belief and Reflective Knowledge, volume I.
Sosa, Ernest. 2009. Reflective Knowledge: Apt Belief and Reflective Knowledge, volume II.
Sosa, Ernest. 2011 Knowing Full Well. Princeton, NJ: Princeton University Press.
Swinburne, Richard (ed.). 1974. The Justification of Induction. Oxford: Oxford University
Press.
Turri, John, and Peter Klein (eds.). 2014. Ad Infinitum: New Essays on Epistemological
Infinitism. Oxford: Oxford University Press.
Van Cleve, James. 1979. “Foundationalism, Epistemic Principles, and the Cartesian Circle,”
Vogel, Jonathan. 2004. “Skeptical Arguments,” Philosophical Issues 14: 426–455.
Williams, Michael. 1996. Unnatural Doubts: Epistemological Realism and the Basis of
Scepticism. Princeton, NJ: Princeton University Press.
Williamson, Timothy. 2000. Knowledge and Its Limits. Oxford: Oxford University Press.
Williamson, Timothy. 2009. “Replies to Critics.” In P. Greenough and D. Pritchard (eds.),
Williamson on Knowledge, 279–384. Oxford: Oxford University Press.
Wright, Crispin. 2004. “Warrant for Nothing (and Foundations for Free)?” Aristotelian
Society Supplement 78: 167–212.
Zalabardo, José 2012. Scepticism and Reliable Belief. Oxford: Oxford University Press.
CHAPTER THIRTY-SEVEN
Regress Arguments
and Skepticism
RICHARD FUMERTON
1 INTRODUCTION
The history of epistemology has been shaped to a very large extent by attempts
to meet the challenge of skepticism. One of the most powerful weapons in the
skeptic’s arsenal has always been the threat of problematic regress. Almost no
contemporary epistemologists accept skepticism, at least in its most extreme forms,
so epistemologists often conceive of their task as to block alleged problematic
regresses, or to allow them but to argue that they are harmless. In what follows, I’ll
examine a number of different potentially problematic regresses and evaluate the
ways in which various schools of epistemology respond to the challenge they pose.
2 DISTINCTIONS
Let’s begin by making some familiar distinctions. Skepticism itself comes in many
forms. It is useful to distinguish skepticism concerning knowledge from skepticism
concerning justified belief. It is further useful to distinguish, with respect to both
kinds of skepticism, global and local versions. As I’ll define the view, the global
skeptic concerning knowledge claims that no one knows anything (presumably, not
even that no one knows anything). The view might even be held with a strong modal
operator—it is impossible for anyone to know anything. The global skeptic certainly
makes a startling claim. The claim, if true, would render false countless statements
we make in ordinary discourse. But one can soften the claim a bit by talking about
what we know with absolute certainty, where one makes clear that one is employing
very high standards for achieving this sort of knowledge. There have been relatively
few global knowledge skeptics. Most are content to embrace skepticism with respect
to certain kinds of knowledge. So one might be a skeptic with respect to knowledge
of the external world, other minds, the past, the theoretical world of physics, the
future, God, global warming, and the existence of unicorns. The more consensus
there is that one doesn’t know something (like that there are unicorns), the more
536 SKEPTICISM
odd it sounds to characterize the view that we lack such knowledge as a form of
skepticism, but we won’t worry about that subtlety here.
We can also make a distinction between global and local skepticism with respect
to justified belief. And again we can hold the view with or without a modal operator.
The most extreme form of skepticism is the view that it is impossible for anyone to
have any justification at all for believing anything. Again, it trivially follows that,
if the view is true, one has no justification at all for believing it. Asserting the view
probably violates all sorts of “norms of assertion” (though odd assertions, and
assertions that violate norms of conversation, can still be true).1 As with knowledge-
skepticism, it is much more common to find epistemologists concerned only with
arguments designed to cast doubt on the possibility of justifying belief in certain
kinds of propositions.
Many of the regress arguments that drive epistemological views actually threaten
the possibility of global skepticism with respect to justified belief. Let us consider the
most familiar regress argument for foundationalism.
3 THE EPISTEMIC REGRESS ARGUMENT

FOR FOUNDATIONALISM
The foundationalist believes that we must distinguish two sorts of justification—
inferential and noninferential. Further, the foundationalist argues, we need
to recognize that all inferentially justified beliefs must be legitimately inferred
ultimately from propositions that we are noninferentially justified in believing.
That some beliefs are only inferentially justified, if justified at all, seems plausible
enough. I take myself to have good reason to believe that my wife is home after
noticing that her car is in the garage. The car’s being in the garage obviously doesn’t
entail that my wife is home, but that fact together with all sorts of background
information (that my wife doesn’t lend her car to others, that cars don’t find their
own way home, etc.) allows me to infer with relatively high probability that my
wife is home. While it might be easy to give examples of beliefs that are only
inferentially justified, if justified at all, it is much harder to give an uncontroversial
general characterization of what kinds of beliefs involve inference. So the radical
British empiricists all thought that to justifiably believe any proposition describing
the physical world one would need to justifiably infer that proposition from some
other proposition describing the character of one’s subjective experience. It also
used to be taken for granted by most epistemologists that inductively formed
beliefs, beliefs about other minds, beliefs about the (at least distant) past, and
beliefs based on testimony, all involve explicit or implicit inference. The notion
of implicit inference, however, becomes critical. I do believe that the next raven
I see will be black and I have that belief, at least in part, because of my experience
with past ravens. But I might be hard pressed to describe with any sort of accuracy
the specific occasions on which I have observed a black raven. I believe that the
Battle of Hastings took place in 1066, but I doubt I could describe the occasion on
which I was taught that or read that. In the end we probably can’t get clear about
REGRESS ARGUMENTS AND SKEPTICISM 537
what inferential justification is until we contrast it with another sort of justification

recognized by the foundationalist—noninferential justification.
We might begin by characterizing noninferential justification negatively: a belief
is noninferentially justified if it is justified but its justification is not constituted,
even in part, by the having of other different justified beliefs. The epistemic regress
argument for the existence of noninferential justification proceeds as follows:
(1) One is justified in believing P by inferring it from some other proposition E1
only if one is justified in believing E1 (Garbage in—garbage out).
(2) If the only way to justify a belief is through inference, then to be justified
in believing E1 one would need to have inferred it from some other
proposition E2 which one also justifiably believed, and, again, if the only
way to justify a belief is through inference, we could have a justified belief
in E2 only if we inferred it from something else E3 which one justifiably
believed, and so on ad infinitum.
(3) Finite minds cannot complete an infinitely long chain of reasoning. Arguably,
even infinite minds couldn’t complete such a chain.
So,
(4) If the only way to justify a belief is through inference, we have no
justification for believing anything (the most extreme of all skeptical
positions).
Notice that we haven’t got an argument yet for the existence of noninferentially
justified beliefs. But if we add as an additional premise that global skepticism
about justification is false, we get the foundationalist’s conclusion that there are
noninferentially justified beliefs. We haven’t given yet a positive analysis of what
noninferential justification is (it has been understood only negatively—as justification
that is not inferential), or made a claim about what one is noninferentially justified in
believing. But with an anti-skeptical commitment, we have reached the conclusion
that there are noninferentially justified beliefs.
Traditional epistemologists have often assumed, I think, that we need to block
not just one regress threatening the possibility of justified belief. Some (a decided
minority now) think that we are justified in believing P on the basis of E1 only
if in addition to being justified in believing E1, we are also justified in believing
that E1 makes probable P (where E1’s entailing P can be thought of as the upper
limit of making probable). Again if the only way to justify a belief is through
inference, inferential justification for believing P on the basis of E1 would require
us to legitimately infer that E1 makes probable P from some other proposition F1,
a proposition we would need to justifiably infer from some other proposition F2,
and so on (another regress begins). But we would also need to justifiably believe that
F1 does make likely that E1 makes likely P, a proposition we would need to infer
from some other proposition G1, which in turn would need to be inferred from G2,
and so on ad infinitum. And we would, of course, need to be justified in believing
that G1 makes likely that F1 makes likely that E1 makes likely that P, a proposition
we would need to infer. . . . On this view, if all justification were inferential, we
538 SKEPTICISM
would face not just one, but infinitely many regresses mushrooming out in infinitely
many directions.
The above arguments designed to force a choice between skepticism and
foundationalism are most naturally thought of as epistemic regress arguments. The
vicious nature of the regress is supposed to have something to do with the fact that
finite minds cannot come up with infinitely many arguments of the sort required by
the presupposition that all justification is inferential. There are at least two responses
to the argument. The coherentist famously rejects the idea that justification is
“linear.” As the label for the view implies, the coherentist claims that the justification
for believing a given proposition is a function of the way that proposition coheres
with other propositions believed (where we get different versions of the view
depending on whose beliefs coherence among which are supposed to constitute
epistemic justification). For reasons of space I won’t talk much about the coherence
theory here. Coherence seems to me neither necessary nor sufficient for the having
of justified beliefs. A single necessarily false proposition in my belief system renders
the whole system inconsistent, but does nothing to defeat the justification I have for
believing that I exist. And Foley (1979) has argued (successfully I think) that we can
even be aware that we have inconsistent beliefs each of which is still nevertheless
perfectly justified. Think about lotteries and the fact that we can justifiably believe
of each ticket that it will lose even as we justifiably believe that there will be a
winner. Or think about the proofreader who justifiably believes of each page that
it contains no errors even as she concludes quite reasonably that there is bound to
be a missed typo somewhere in the manuscript. So coherence doesn’t seem to be
necessary for justification. And all kinds of examples seem to suggest that coherence
without access to that coherence doesn’t generate epistemic justification. But a
coherentist can’t give a coherentist account of the relevant access without again
facing regress. The coherentist’s account of access to coherence would require a
coherentist account of access to what one believes. But all the coherentist can point
to in justifying a belief about what he believes is coherence with the metabelief about
beliefs and the rest of his beliefs. But that doesn’t do any good unless one has access
to that coherence, and off we go.
The infinitist (a term coined by Klein 1999) challenges the assumption that, if all
justification were inferential, finite minds couldn’t satisfy plausible conditions on
inferential justification, at least if the requirement for inferential justification insists
only on the capacity to come up with the relevant justifications. And we should be
cautious before we denigrate the wondrous capacities of the human mind. In some
sense, it seems that we can have not only infinitely many beliefs, but infinitely many
justified beliefs. I justifiably believe that 2 is greater than 1, that 3 is greater than 1,
and so on ad infinitum. I justifiably believe that I exist, that either there are unicorns
or I exist, that either there are warlocks or I exist, and so on, for (arguably) infinitely
many propositions I have the capacity to disjoin with the proposition that I exist. It
is not clear to me, however, that the availability of infinitely many justified beliefs
will help us to avoid another sort of regress that will still force us to choose between
foundationalism and either skepticism or the view that the very idea of justification
is incoherent.
4 CONCEPTUAL REGRESS
Let us return to the idea of inferential justification, and let us focus again on the
principle that one is inferentially justified in believing P on the basis of E only if one
is justified in believing E. The principle seems true, indeed necessarily true. But why
is it necessarily true? Perhaps the most straightforward answer is that the principle
unpacks the very idea of inferential justification. It is, plausibly, analytic that one is
justified in believing P on the basis of E only if one is justified in believing E. While
one might be able to define inferential justification this way, however, one obviously
can’t without circularity define justification in general as inferential justification.
Our characterization of inferential justification is parasitic upon understanding what
a justified belief is. The obvious solution is to recognize that one needs a recursive
analysis of epistemic justification. The recursive analysis requires a base clause. The
base clause will state a sufficient condition for noninferential justification, a condition
that does not itself invoke the concept of justification. Inferential justification will be
defined as justification that can be traced ultimately to noninferential justification.
Being a descendent of stands to being a son or daughter of, and being instrumentally
good stands to being intrinsically good, as being inferentially justified stands to
being noninferentially justified. In the end, the infinitist won’t be able to offer an
analysis of epistemic justification. We can talk all we want to about the capacity to
offer arguments for premises, but we need to add that the relevant arguments will
require premises that are justified. Those premises might be justified by further
argument, but only if those premises are justified. And the infinitist will never be
able to tell us what epistemic justification is without eventually relying on some
notion of noninferential justification (see Fumerton 2013).
5 FOUNDATIONS OR SKEPTICISM
The epistemic regress argument leaves one a choice between foundationalism
and skepticism. The conceptual regress argument leaves one a choice between
allowing that we have a concept of noninferential justification and rejecting the
very intelligibility of epistemic justification. It is easy enough to conclude that
foundationalism must be true (given what most take to be the implausibility of
extreme skepticism), but it is much more difficult to give a plausible account of
foundational justification. My own view relies heavily on unanalyzable concepts
of acquaintance and correspondence (fit) between thought and fact. Its defense
in large part involves arguments against alternative conceptions of noninferential
justification—in particular externalist foundationalisms. We are not about to canvas
competing accounts of noninferential justification here. If one is convinced that
neither infinitism nor the coherence theory of justification is plausible, however,
one might conclude that all of our energy as epistemologists should be devoted
to figuring out what the correct account of noninferential justification is. But,
before we reach that conclusion, we should probably look more carefully at reasons
philosophers have for worrying about just how one could even in principle end
a regress of justification in an intellectually satisfying way. Let’s look at two such
540 SKEPTICISM
reasons—BonJour’s influential argument against foundationalism (offered when he

was a coherentist about empirical justification) and the problem of easy justification.
6 BONJOUR AND REGRESS

BonJour (1985) offered a deceptively straightforward global argument against any
form of foundationalism, an argument that influenced a great many philosophers
(but which BonJour himself has since abandoned). Any version of foundationalism
will identify some property of a belief or a believer (call it J) as the property
exemplification of which renders a belief noninferentially justified. But surely, the
argument goes, the mere exemplification of J wouldn’t justify for the believer S the
relevant belief unless (1) the believer were aware that J is exemplified. Furthermore,
even if S were aware of J, without some reason to think that (2) J beliefs are likely to
be true, S would still lack any intellectually satisfying justification.
A great deal probably hinges on the notion of what is supposed to make a
justification intellectually satisfying. But the basic idea can be illustrated with
examples. So consider condition (1). Some foundationalists have argued that a
belief that P is noninferentially justified if it is caused by the fact that P (where the
causal chain leading from the relevant fact to the belief involves no beliefs as causal
intermediaries). The belief’s being caused by its truthmaker would certainly render
it true, but if I am trying to figure out whether or not I have any justification for
believing P, the belief’s causal ancestry will hardly help me answer the question when
I have no idea what that causal ancestry is. But condition (2) is no less plausible.
Suppose I notice that some of my beliefs are particularly vivid. I have no idea,
however, whether vivid beliefs are more likely to be true than those that are not. If
I am looking for assurance that what I believe vividly is true, how could I get such
assurance without establishing (or at least getting some justification) for believing
that vivid beliefs are likely to be true?
It takes but a little reflection to worry that BonJour’s two conditions spell doom
for foundationalism. On one interpretation, BonJour’s argument suggests that
condition J constitutes noninferential justification for believing P only if it doesn’t
really! It is only J conjoined with awareness of it and its connection to truth that could
constitute justification (call that J and A). But J and A couldn’t constitute justification
either—it’s only J and A conjoined with awareness of J and A and their connection
to truth, and so on ad infinitum. Alternatively, one could insist that noninferential
justification is exhausted by J, but J is such that necessarily one can’t be in state
J without being aware (or having the capacity to be aware) of J and its connection to
truth. This latter view is at least coherent, but still seems to require infinitely many
capacities to entertain more and more complex propositions, something that might
worry those of us who can’t keep thing straight past four or five metalevels. I can
probably form the thought that I have the three acts of acquaintance that yield first-
level noninferential justification. And I can probably get myself acquainted with that
second-level thought and its correspondence to the complex fact that yields first-
level justification. But as the levels move up, the thoughts and the facts that make
them true get ever more complex—so complex that it is hard to see how a mere
mortal might be able to keep them straight.
In the abstract, it is clear how the foundationalist must respond to BonJour’s
argument. If a foundationalist puts forth some property J that is supposed to be
genuinely sufficient for noninferential justification, then that foundationalist had better
remember what the view is. That foundationalist should not concede that something
other than the exemplification of J is required for noninferential justification. Indeed,
reflection on BonJour’s argument should make one realize that, if successful, it is a
“nuclear” objection that defeats any proposed account of justification (foundational
or not). To illustrate, BonJour himself defended a coherentist account of empirical
justification. Crudely, S’s belief that P is justified when P coheres in the “right” way
with the rest of what S believes. Call the relevant property of cohering C. But the
argument against foundationalism presumably applies equally to C. Coherence by
itself wouldn’t do S any good when it comes to S’s seeking philosophical assurance
for believing P. It would only be coherence coupled with S’s awareness of that
coherence and S’s awareness of how coherence renders likely the truth of beliefs
that cohere that would give S intellectually satisfying justification. But however the
coherentist tries to understand that awareness (presumably it would be in terms of
yet more coherence), it wouldn’t do the job unless we had awareness of that, and we
are off on another apparently vicious regress.
So if we are going to try to build a recursive analysis of justification on our
understanding of noninferential justification, we will succeed only if we find
some set of conditions that constitute noninferential justification without adding
to those conditions the requirement that we be aware of them. The externalists
are quite content to jettison the awareness requirements, but their externalist
analyses seem vulnerable to what Bergmann (2006) calls (reminding us of BonJour’s
argument) the subject’s perspective objection—the idea that having noninferential
justification seems strangely divorced from anything that gives one, from the first-
person perspective, assurance of truth. The internalist, by contrast, needs to find
an account of noninferential justification that connects having that justification to
assurance, but without requiring infinitely many levels of awareness. I think the
acquaintance theory I have developed elsewhere walks that fine line, but, if it
doesn’t succeed, skepticism looms once again on the horizon. BonJour’s argument
against foundationalism is heuristically valuable. It reminds us that our search for a
plausible account of epistemic justification cannot be divorced from our search for
intellectual assurance.
7 THE PROBLEM OF EASY JUSTIFICATION

Most epistemologists have converged on the idea that we must avoid regress in
both securing and understanding justification. But the impulse to always move in
the direction of inferential justification when seeking justification is undeniably
seductive. We can feel that impulse in yet another context when we consider the
problem of easy justification.
542 SKEPTICISM
The locution “the problem of easy knowledge” was coined by Stewart Cohen
(2002), but versions of the problem were also discussed in Fumerton (1995) and
Vogel (2000). Let’s briefly look at examples that illustrate the dilemma. Suppose,
as some do, that we have basic knowledge, or, for our purposes, noninferential
justification for believing that something is red when the belief is based on the fact
that the thing looks red. Consider a close variation of Cohen’s example. In his
hypothetical situation his son starts worrying that a given object that looks red isn’t
really red (but is, rather, some other color under distorting light). Cohen imagines
trying to respond this way to his son’s concern:
(1) That looks red (known through introspection)
(2) That is red (noninferentially justified)
(3) Since it is a red thing that looks red, it isn’t an object with some other color
distorted by light to look red.
We can relax. But the kid will surely realize that Dad pulled a fast one. If the kid
wanted some sort of assurance that the object is red, he didn’t get it. Vogel considers
a crude reliabilism that takes the reliability of a belief-forming process (i.e., suitably
belief-independent) to be a sufficient condition for the resulting belief’s being
justified. He then imagines a situation in which a person is absurdly trying to
check on the reliability of his car’s gas gauge through an inductive argument whose
premises rely on trusting the gauge. The reasoning proceeds this way: the gauge
reads half full so the car is half full; the gauge reads a quarter full, so the car is a
quarter full; the gauge reads an eighth full, and so on. Conclusion: this seems to be
a pretty reliable gas gauge! But surely this is no way to get yourself justification for
believing that the gas gauge is reliable.2 Finally, imagine someone trying to justify
reliance on testimony, and suppose, as some do, that testimony is a basic source
of justification:
(1) S1 testified that P, so (I conclude based on the testimony) P.
(2) S2 testified that Q, so (I conclude based on the testimony) Q.
(3) S3 testified that R, so (I conclude based on the testimony) R.
Repeat until you get tired and conclude (on inductive grounds)
(4) Testimony is generally reliable.
Again, self-respecting skeptics and those who take the problem of skepticism
seriously will conclude that this attempt to vindicate reliance on testimony is a
bad joke.
What does all of this have to do with regress? Well, there is an obvious solution
to each of the pathetic attempts described above that were designed to justify belief
in the reliability of various sources of belief. In the case of basic (noninferential)
justification, the solution is to conclude that the alleged basic justification supporting
the premises wasn’t basic at all. In concluding that the object is red, the father
needed as a separate empirical premise the claim that usually when an object
like this looks red it is red, and that premise needs to be justified for the father
before the father can reach conclusions about the color of the object. In the cases
of problematic attempts to “easily” use an inferential source to generate justified
beliefs in the source’s reliability, we can offer a similar diagnosis. When justifying
the premise asserting that the gas tank is half full by noticing that the gauge indicates
that the tank is half full, we need to rely on the unstated premise that the gauge
is reliable. But that premise is the very conclusion we are trying to establish with
our inductive argument. Because our conclusion appears as a premise justifying the
premise, we are blatantly begging the question. And to get ourselves justification for
believing the premises of our inductive argument for the reliability of testimony, we
need to have antecedent and independent justification that testimony is generally
reliable.3 When that critical premise needed to get justified belief in the premises
of our inductive argument is made explicit, we again realize that the conclusion of
our inductive argument appears as critical support for its premises—the argument
is viciously circular.
In fact, I think that this diagnosis of why the “easy” justification isn’t available
in our specific examples is quite correct. But can we generalize? Shall we always
conclude that before we form a justified belief based on some input (whether a belief
or some other stimulus), we need a justified belief in a connecting principle? That
doesn’t seem like a good idea. We clearly face regress.
8 THE CARROLL PUZZLE

Consider Lewis Carroll’s (1895) famous dialogue between the Tortoise and Achilles.
The two end up discussing the following proof from Euclid:
(A) Things that are equal to the same are equal to each other.
(B) The two sides of this Triangle are things that are equal to the same.
(Z) The two sides of this Triangle are equal to each other.
The Tortoise convinces Achilles that he wouldn’t be forced by reason to the conclusion
Z unless he accepted not only the premises (A) and (B) but the conditional: (C) If (A
and B) then Z. But given that this acceptance of (C) is critical, the Tortoise insists
that it be included as an additional premise yielding a different argument:
(A) Things that are equal to the same are equal to each other.
(B) The two sides of this Triangle are things that are equal to the same.
(C) If A and B are true, Z must be true.
(Z) The two sides of this Triangle are equal to each other.
But Achilles is quickly convinced that we can’t stop here. Reason won’t force us
to accept (Z) on the basis of the conjunction of (A), (B), and (C) unless we see
the connection between that conjunction and (Z), so we should add yet another
conditional (D), and off we go on a never-ending regress.
The regress with which the Tortoise tortures Achilles is illustrated with what
looks like a case of inferential justification. But as we have seen one could also
544 SKEPTICISM
imagine the Tortoise attempting to generate a similar regress with conditions

(J) that are supposed to generate noninferential justification for believing some
proposition P. When the Tortoise does, we have BonJour’s argument that we
considered earlier against foundationalism. Both the Tortoise and BonJour want
us to concede that J’s obtaining by itself won’t force us by reason to the conclusion
that P. We’ll also need to be committed (in some sense) to the position that the
belief is a fitting response to the obtaining of J.4 But if make such a commitment and
formulate it as a premise necessary to reach the conclusion that P, we’ve just lost our
noninferential justification.
9 THE PRINCIPLE OF INFERENTIAL JUSTIFICATION

AND THE THREAT OF REGRESS
We know we need to resist the Tortoise’s suggestion that whenever we reach
conclusions based on some state we are in or some justified beliefs we have, we
should add as a separate premise a proposition asserting that there is the appropriate
evidential connection between the basis of our belief and the propositions believed.
But at the same time, it seems hard to avoid the conclusion that we are sometimes
irresistibly moved to add the additional “connecting” principle as a premise. In the
case of the gas gauge discussed above, I argued that before you could rationally
trust the gas gauge to reach a conclusion about the level of fuel, you would need an
additional premise asserting the reliability of the gauge. In the case of testimony, I
argued for the much more controversial conclusion that you could not justify your
premises relying on testimony without a premise stating that the relevant testimony
(at least of this sort, under these conditions) is reliable. But the reason one can’t
reach a conclusion about the level of the gas based on observing the gauge, or a
conclusion about the truth of what somebody asserts based on their merely asserting
it is, I think, that the gas gauge display and the testimony are simply not evidence
at all supporting the respective beliefs about the level of the gas and the truth of
the testimony.
To be sure, this sounds really odd. Normally, if I’m asked what my evidence is
for thinking that my tank is half full, it would surely be appropriate to respond by
appealing to what the gas gauge indicates. It would be more than a bit pedantic to
start talking not only about the gas gauge but about whatever general evidence I
have for thinking that gas gauges are normally reliable. But that only shows that
in ordinary conversation we rely heavily on enthymematic descriptions of our
reasoning. The total body of evidence from which we infer even commonplace truths
about our environment is, I would argue, extraordinarily complex and involves all
sorts of presuppositions. If we fail to realize this, we might be seduced by examples
of enthymematic descriptions of reasoning to generalize to the conclusion that one
always needs to add to whatever premises one appeals to in inferring a conclusion
yet another premise asserting the relevant connection between one’s premises and
one’s conclusion. One can’t infer from litmus paper’s turning red in a solution that
the solution is acidic without the additional premise that the color of the litmus
paper is a reliable guide to the acidity of the solution, but that’s only because there
is no evidential probabilistic connection at all between the proposition that the

litmus paper turned red and the proposition that the solution is acidic. It is no
longer that popular a view, but I would argue that it is also impossible for me to
infer that your name is “Baron” from the fact that you told me it was, and that
fact alone, because there is also no evidential connection between the former and
the latter.
So the critical question we need to address in dealing with the puzzle posed by
the Tortoise is whether or not we need a premise describing the relevant connection
between premises and conclusion (or in the case of noninferential justification,
the basis of our belief and the proposition believed), when we have legitimate
reasoning where the premises (or the noninferential basis for our belief) do make
probable or entail the truth of what we believe. It is tempting to suggest at this
point that one should simply decline the Tortoise’s invitation to do that. In the case
of inferential justification, however, it still seems right to me that one needs to be
able to see the connection between one’s premises and one’s conclusion if one is
to have intellectually satisfying justification for believing the conclusion based on
the premises. Even when the premises of an argument entail the conclusion of the
argument, we surely want to distinguish the person who is aware of the entailment
from the person who is not. Both may be caused to believe the conclusion by belief
in the premises, but only the former is fully justified in believing the conclusion
based on the premises. It is precisely this thought that leads me to endorse the
following (highly controversial) principle of inferential justification:
PIJ
One is justified in believing P on the basis of E only if one is (1) justified in believing
P and (2) justified in believing that E makes probable P (where entailment can be
thought of as the upper limit of making probable).
Won’t insistence on such a principle lead to vicious regress? It is crucial that in
responding to this concern one makes clear that the suggestion is not that one adds
the proposition that E makes probable P as an additional premise from which one
infers P. That is to fall into the very trap that the Tortoise set for Achilles. But to
insist on justified belief in that connection (call it C) as a necessary condition for
being inferentially justified in believing P on the basis of E is not the same thing as
to insist that the inference in question is really an inference from the conjunction E
and E makes probable P to P.
My former student, Sam Taylor (2013), insists that this response is too clever by
half. If a justified belief in a connection between one’s premises and one’s conclusion
is part of the story of what justifies one in believing one’s conclusion, why isn’t the
proposition describing that connection functioning just as a premise does? And,
if that’s right, then why aren’t we vulnerable to the Tortoise’s puzzle? Taylor’s
suggestion is that we distinguish justified belief in the critical connection from non-
propositional awareness of the connection, and insist that only the latter is necessary
for intellectually satisfying justification. Indeed, he suggests this would make an
account of inferential justification analogous to what an acquaintance theorist
should say about noninferential justification. On my own view, for example, one is
546 SKEPTICISM
noninferentially justified in believing P when one is directly acquainted with the fact
that P and a “fit” between the thought that P and the fact that P. I specifically insist
that one does not need to know the distinct proposition that the thought fits the
fact in order to have the noninferential justification—I’d better insist on that lest the
noninferential justification disappear! Insisting that we need only awareness of fit
in the case of noninferential justification, and awareness of probabilistic connection
in the case of inferential justification, not only avoids Carroll-style regress, but also
allows for those who lack such sophisticated concepts as fit and probability to have
justified beliefs.
But suppose that one insists that for inferential justification one needs to be
justified in believing a proposition describing the relevant probabilistic connection
between premises and conclusion. One also insists that the relevant “connecting”
proposition not be considered a premise among other premises. Still, how do we
avoid regress, at least in the case of probabilistic connection? If E makes probable
P (where E is one’s total evidence), won’t it be an empirical truth that E makes
probable P and won’t one need to infer such a truth from some other proposition E2,
beginning yet another inference pattern that involves regress? I think regress does,
in fact, loom if one accepts the principle of inferential justification and holds the
view that propositions describing evidential connections are empirical propositions
justifiable only a posteriori. If there is a solution, then, it is to argue that when E
makes probable P one can in principle know that a priori. Or if one accepts Taylor’s
view, one would need to be able to argue that we can be directly aware of such
facts as that E makes probable P. But isn’t it obvious that we can’t know a priori
such truths, nor can we be directly acquainted with their truth makers? Didn’t we
just suggest that whether or not one truth makes probable another is surely an
empirical matter?
That probably is the conclusion that most epistemologists reach—certainly most
externalists—but it is not the only view that has been held in the history of philosophy.
Keynes (1921) argued that making probable is an internal relation holding between
propositions (where an internal relation is one that holds necessarily when its relata
exist). The view isn’t as strange as it initially seems. I suspect the main reason we
think that evidential connections hold only contingently is that we confuse genuine
evidential connections with the premises and conclusions of arguments described
enthymematically. It is almost obvious that the litmus paper’s turning red is only
a contingent indicator of the solution’s acidity, but if we were right earlier, litmus
paper’s turning red is not something from which we can legitimately infer that the
solution is acidic. There are no legitimate arguments with that form.5 By contrast,
consider the question whether it is true and, indeed, necessarily true, that my
seeming to remember having experience E makes prima facie probable for me that
I had experience E. Or consider the question whether it is true, indeed, necessarily
true, that the conclusion of an enumerative inductive argument is made prima
facie probable by that argument’s premises. Don’t confuse these questions with the
question whether it is possible for apparent memories to mislead, or possible for the
conclusion of an inductive argument to be false even when the premises are true. That
clearly is possible. Nor should we confuse the question with the question whether
one can imagine a world in which apparent memory usually leads one astray, or
inductive arguments usually have false conclusions. Even that seems possible. But it
is far less obvious that one can even conceive of a world in which apparent memory
wouldn’t make prima facie likely for one who has the apparent memory that which
one seems to remember, or a world in which the conclusion of an enumerative
inductive argument wouldn’t be (epistemically) probable relative to its premises
(and those premises alone). If one does conclude (as Keynes did) that we can make
sense of these “quasi-logical” probability relations holding between propositions,
we will need to make sense of a probability relation that cannot be reduced to
statistical generalizations—the probability to which this epistemologist appeals can’t
be the concept of probability defined in terms of contingent relative frequencies.
10 MODEST FOUNDATIONALISM AND YET

ANOTHER THREAT OF REGRESS
I said earlier that the key to avoiding the kind of regress that leads us to skepticism is
to find a plausible account of foundational justification—the kind of justification that
give us assurance of a truth P without our needing to have additional justification
for believing some proposition other than P. The goal of this chapter is to discuss
regress without evaluating all of the different proposals for how to understand
noninferential justification. These days there are important and interesting externalist
versions of foundationalism that compete with more traditional views seeking to
locate the source of noninferential justification in the mind’s awareness of its own
mental states (coupled perhaps with a similar sort of awareness of abstract objects
like properties or propositions and their relations). But a full critical evaluation of
these views is the topic for another paper.
I want to close, however, by adding one note of warning for foundationalists (of
all stripes) who are fully committed to cauterizing both epistemic and conceptual
regress. It is much more popular these days for foundationalists to adopt what is
sometimes called modest foundationalism. The classic foundationalist (inspired
often by Descartes) was convinced that one locates genuine foundational justification
only when one finds justification so strong that it eliminates the possibility of error.
Worried that foundations meeting this test will be too impoverished to support
the huge edifice of knowledge and justified belief we seek, some foundationalists
wanted to relax this criterion for foundational justification and allow that a belief
could be noninferentially justified and defeasible. Let’s say that my justification J for
believing P is defeasible when there might be something other than J which, if added
to J, would defeat my justification for believing P.
Let us illustrate the point with an example Goldman worried about in his
early piece introducing a reliabilist version of foundationalism. Like classic
foundationalists, Goldman wanted to understand epistemic justification recursively.
To do so, he needed to find a base clause—a non-epistemic condition upon which
justification supervenes (a sufficient condition for justified belief that was not itself
an epistemic condition). His idea was that a belief is noninferentially justified when
548 SKEPTICISM
it is produced by a belief-independent process that is unconditionally reliable. A

process is belief-independent just when the input to the belief-forming process is
something other than a belief. With the exception of introspection of belief states,
belief-forming processes that take as their input other beliefs will output justified
beliefs only if the input beliefs are themselves justified. To yield justified beliefs,
the belief-dependent processes need only be conditionally reliable, need only yield
mostly true beliefs when the input beliefs are true. But Goldman worried about the
following hypothetical situation. Let’s suppose that reliance on apparent memory
is a good example of a belief-independent, unconditionally reliable process. But
also imagine someone who seems to remember that he did X, but where he has
been given plausible evidence that his memory is completely unreliable. If such a
person goes on to believe that he did X, we would surely want to conclude that
the belief he formed is unjustified. How do we fix our account of noninferential
justification? Goldman suggests that we complicate it a bit. Let’s say that a belief
P is noninferentially justified when the belief is produced by a belief-independent
process that is unconditionally reliable and where there are no other reliable
processes available which, had they been used as well, would have resulted in a
different belief. It sounds OK initially. But if we think about the view, we realize
that the other processes to which the account refers are both belief-independent
and belief-dependent processes. But the only relevant belief-dependent processes are
those that are available and that take as input justified belief. If we say all that, we
have failed to give an account of noninferential justification that does not invoke a
concept of justification. Conceptual regress looms again.
The point is quite general. If foundationalism needs to avoid conceptual regress,
it needs to find a recursive analysis that invokes a concept of noninferential
justification that doesn’t itself appeal to our understanding of epistemic justification.
Any “moderate” foundationalist will need to protect the account of noninferential
justification by talking about the absence of defeaters. But the absence of relevant
defeaters can only be specified ultimately by talking about the absence of other
relevant justification. Our attempt to produce a non-circular account of justified
belief (our attempt to avoid conceptual regress) is now in danger of failing.
NOTES
1. So some would argue that there is some sort of implicit rule that one shouldn’t assert
what one doesn’t know to be true. A weaker suggestion is that one shouldn’t assert
what one has no epistemic reason to believe. Weaker still is the idea that one shouldn’t
assert what one doesn’t believe. It is not clear to me that philosophers follow any of
these norms in their philosophical writings.
2. The situation actually isn’t quite this bad, at least if one allows oneself certain
background assumptions as one initially trusts the gauge. So imagine that one looks at
the gauge as it reads full. A few seconds later one looks at the gauge and it indicates that
the tank is almost empty. A moment later, though, the gauge tells you once again that
the tank is full. At this point one can use one’s background evidence (that the levels
of fuel in the tank don’t vary in this way) to reach the conclusion that something is
wrong with the gauge. The fact that this didn’t happen doesn’t support the conclusion
that the gauge is accurate, but it is not completely worthless—something that could
have gone wrong, didn’t.
3. I’m not suggesting that it will be easy to get the critical justified belief without relying
on testimony. The threat of skepticism is no doubt part of what drives people to the
conclusion that testimony is in some sense a basic source of justification.
4. Sosa (2009) talks of such a commitment being present even in the case of what he
calls “animal-level” knowledge.
5. Think about the form of that argument: P; therefore, Q. If there are legitimate
nondeductive arguments, they will have a form just as surely as do deductively valid
arguments. So one sort of inductive argument has something like the following form:
(all) most observed F’s have been G; therefore, (all) most F’s are G. Reasoning from
apparent memory arguably has the form: S seems to remember X; therefore, X. The
form of reasoning to the best explanation is a matter of much dispute, but it is surely
what we need from explanationists if we are to take seriously their suggestions as to
how to understand commonplace inference.
REFERENCES
Bergmann, Michael. 2006. Justification without Awareness. Oxford: Oxford University
Press.
BonJour, Laurence. 1985. The Structure of Empirical Knowledge. Cambridge: Harvard
University Press.
Carroll, Lewis. 1895. “What the Tortoise Said to Achilles,” Mind 4: 278–280.
Foley, Richard. 1979. “Justified Inconsistent Beliefs,” American Philosophical Quarterly
16: 247–258.
Fumerton, Richard. 1995. Metaepistemology and Skepticism. Boston: Rowman and
Littlefield.
Fumerton, Richard. 2004. “Epistemic Probability,” Philosophical Issues 14: 149–164.
Fumerton, Richard. 2006. “Epistemic Internalism, Philosophical Assurance, and the
Skeptical Predicament.” In T. Crisp, M. Davidson, and D. Vander Laan (eds.), Knowledge
and Reality: Essays in Honor of Alvin Plantinga, 179–192. Dordrecht: Springer.
Fumerton, Richard. 2013. “Infinitism.” In P. Klein and J. Turri (eds.), Ad Infinitum: New
Essays on Epistemological Infinitism, 74–86. Oxford: Oxford University Press.
Goldman, Alvin. 1979. “What Is Justified Belief?” In G. Pappas (ed.), Justification and
Sosa, Ernest. 2009. Reflective Knowledge, volume II. Oxford: Oxford University Press.
Taylor, Samuel. 2013. The Problem of Easy Justification: An Investigation of Evidence,
Justification and Reliability. Dissertation, University of Iowa.
Vogel, Jonathan. 2000. “Reliabilism Leveled,” Journal of Philosophy 97: 602–623.
CHAPTER THIRTY-EIGHT
The Problem of
the Criterion
ANDREW D. CLING
1 INTRODUCTION
The problem of the criterion is the ancient skeptical paradox posed, roughly, by
this argument:
To know a proposition, we must first know a criterion of truth. To know a
criterion of truth, we must first know a proposition. Therefore we cannot know
any proposition or any criterion of truth.
The problem is nascent in Meno’s paradox and appears in the skeptical “modes”
of Sextus Empiricus, the problem of the Cartesian circle, the problem of induction,
Hegel’s Phenomenology of Spirit, and elsewhere.1 It plays an organizing role in
Chisholm’s influential epistemology and vigorous current debates are about issues
it raises.
The problem, properly understood, is not just about knowledge, justification,
or any other specific epistemic value. It is a problem for any epistemic value that
beliefs can have only if they are supported by the right kinds of reasons. I explain
the problem, discuss some anti-skeptical responses, and show how some important
contemporary debates are relevant to it. I conclude with a discussion of epistemic
pluralism, human flourishing, and skepticism.
2 CHISHOLM
Roderick Chisholm’s account has strongly influenced contemporary discussions of
the problem.2 Chisholm intends to pose the problem as an argument from plausible
premises to skeptical conclusions, a paradox. In fact, however, Chisholm presents
two arguments.
The first argument, paraphrased from Montaigne, concerns knowledge:
[1] To know whether things really are as they seem to be, we must have a
procedure for distinguishing appearances that are true from appearances that are
THE PROBLEM OF THE CRITERION 551
false. [2] But to know whether our procedure is a good procedure, we have to
know whether it really succeeds in distinguishing appearances that are true from
appearances that are false. And we cannot know whether it does really succeed
unless we already know which appearances are true and which ones are false.
And so [C] we are caught in a circle. [So [3a] we cannot know which appearances
are true and [3b] we cannot know whether any procedure for distinguishing true
from false appearances is good.] (Chisholm 1973: 3; here and below, material in
brackets has been supplied by me.)3
The second argument poses the problem in terms of two questions about knowledge:
[1] You cannot answer question A [“what do we know?”] until you have answered
question B [“how do we know?”]. And [2] you cannot answer question B until
you have answered question A. Therefore [3] you cannot answer either question.
[3a] You cannot know what, if anything, you know, and [3b] there is no possible
way for you to decide in any particular case. (1973: 14)4
Chisholm (1973: 12–15) takes these arguments to be the same and claims that there
are three responses: reject [1] (particularism), reject [2] (methodism), or accept [3a]
and [3b] (skepticism). He endorses particularism but claims that any response begs
the question against the others (1973: 37).
These arguments are different. The conclusion of the first—henceforth
Chisholm’s first argument—is that we cannot know any proposition and we cannot
know that any procedure for distinguishing true from false propositions is good.
The second argument is metaepistemological. Its conclusion is that we cannot
know any proposition about knowledge or any principle for distinguishing true
from false propositions about knowledge. Because propositions about knowledge
constitute a proper subset of propositions, and epistemic principles—principles
expressing necessary and/or sufficient conditions of knowledge—can be criteria for
distinguishing true from false propositions about knowledge, the problem posed
by the second argument is a special case of the problem posed by Chisholm’s first
argument (Cling 1994: 262–266).5
Chisholm’s first argument provides a good starting place for understanding
the skeptical argument but it obscures some important features of the problem.
First, Chisholm’s first argument does not make it explicit that the problem requires
critical assumptions about the reason-providing relationships between propositions
and criteria. What makes [1] and [2] plausible are their implicit claims that knowing
a proposition P requires knowing a criterion of truth that is a reason for P and vice
versa. A statement of the problem needs to make explicit reference to these reason-
providing relationships. Second, Chisholm’s temporal language—“already know,”
“until”—needs to be explained. The problem concerns the epistemic, not temporal,
priority that good criteria and good propositions seem to require. Chisholm’s first
argument indirectly refers to this independence requirement by stating that [1] and
[2] require a circle [C], but it does not display the principle that is required to rule
out circles of reasons. In fact, however, [1] and [2] do not require circles of reasons,
they jointly require endless regresses of reasons: sequences in which each component
552 SKEPTICISM
has a successor that is a good reason for it. Endless sequences can be circular, have
infinitely many components, or both. So we need to state the necessary injunctions
against all types of endless sequences of reasons. Finally, although Chisholm’s first
argument is about the possibility of knowledge, the problem arises for any epistemic
value propositions can have only if they are supported by the right kinds of reasons.
3 EPISTEMIC REASONS
There are epistemic and non-epistemic reasons for belief. An epistemic reason is a
factor that is taken to count in favor of believing a target proposition P because it is
taken to count in favor of P’s truth. A non-epistemic reason is a factor that is taken
to count in favor of believing a target proposition P because it is taken to count
in favor of that belief’s having other benefits. Knowing that most people with the
disease I have will die from it can be a good epistemic reason to believe that I will
die from the disease. I might, however, have a good pragmatic reason to believe that
I will not die from the disease if believing this would improve my chance of survival.
Reasons for belief must be mental states with content that we can take to count in
favor of the beliefs for which they are reasons.6 Epistemic reasons, therefore, must
be beliefs or belief-like mental states with propositional contents that can be taken
to count in favor of the truth of their targets.7
A good reason must (a) have content that stands in an appropriate relation to its
target and (b) have the normative standing that epistemic reasons must have. That
I have $1 million entails that I have at least $500 thousand and therefore stands
in a reason-relevant relation to the latter. It is not a good epistemic reason for me
to believe anything, however, because it lacks standing: the normative epistemic
authority reasons must have if they are to be good reasons for any targets. That I
have $1 million lacks the standing to be a good evidence-providing reason since,
alas, it seems to have no epistemic value: I have no good reason to believe it and
plenty of good reasons to reject it.
Two kinds of epistemic reasons are evidential reasons and procedural reasons
(Cling 2014b: 164). An evidential reason is a mental state with content P that is
taken to count in favor of a target Q by a person S because S takes P to (deductively
or inductively) imply that Q is true. P is a good evidential reason for a proposition Q
only if (a) S believes that P, (b) P does (deductively or inductively) imply Q, and (c)
S’s belief that P has the epistemic value required for standing. Hopes, for example,
are mental states with propositional content, but hopes are not, by their nature, the
sorts of things that can stand in the support relation required of good evidential
reasons. Beliefs are the right sorts of things to stand in the relevant support relations,
but beliefs lacking the epistemic value required for standing are not good reasons.
Beliefs that it is epistemically impermissible or irresponsible to hold are not good
evidential reasons for any proposition, for example. A critical question raised by the
problem of the criterion, therefore, is whether propositions or criteria of truth can
have epistemic value without being supported by good epistemic reasons.8
Procedural reasons involve criteria of truth. The word “criterion” derives from
the Indo-European word for sifter (Watkins 2000: 113). Criteria are principles by
means of which we may sift or sort things in order to identify what we seek: wheat,
sheep, Shinola™, truths, etc. A criterion of truth is a principle according to which
a detectable property D is positively relevant to the truth of propositions. An ideal
criterion of truth would be a true principle with this form: for any proposition
P, P has D if, and only if, P is true. We should, however, be liberal and count
other principles—principles that express only sufficient, necessary, or probabilistic
conditions of truth; fallible principles; principles restricted to a limited domain;
etc.—as criteria of truth.
A procedural reason is a belief that C where C is a criterion of truth. C is not
usually evidence for a proposition P that C identifies to be true, but a belief that C
can be a reason to believe that P is true by expressing and underwriting the process,
procedure, norm, rule, or method (henceforth “method”) by which a person
identifies P to be true.9 That (S) propositions about macroscopic objects identified
to be true in light of sensory experiences are usually true is not evidence that (E)
there is a written sentence in front of me now but a person S’s belief that (S) can
count in favor of the truth of (E) if (S) expresses and underwrites the method by
which S identifies P to be true. That (I) inductive generalizations with true premises
usually have true conclusions is not evidence that (R) the sun will rise tomorrow
but can count in favor of (R) if S’s belief that (I) expresses and underwrites the rule
of inference governing the method by which S identifies (R) to be true on the basis
of S’s other beliefs. A belief that C is a good procedural reason for a target P for a
person S only if (a) C is a criterion of truth that is implicit or explicit in the method
by which S identifies P to be true, (b) C picks out P, and (c) S’s belief that C has the
required epistemic standing.10 A good procedural reason would put us in a better
epistemic position with respect to the propositions that a method M identifies to be
true by making our commitment to M explicit and by being a good reason to trust
M. A good reason to trust a method must be, at least, a belief that it is reliable or is
in some other way epistemically trustworthy.
Criteria of truth are themselves propositions, but I assume that we can distinguish
them from propositions that are not criteria. Exactly how criteria of truth figure in
the methods we use to identify truths is a question I also leave to the side. Unless
criteria are believed, however, they cannot be epistemic reasons since such reasons
must be beliefs.11
4 THE PROBLEM SCHEMA

We are now in a position to state the problem of the criterion more carefully and
completely. I shall state the problem as a schema for skeptical arguments about an
otherwise unspecified epistemic value Ф. Let P be a variable for propositions that
do not express criteria of truth, and let C be a variable for principles that do.12 This
schema generalizes and completes Chisholm’s first argument:
(1) A proposition P can have epistemic value Ф only if some C has Ф and C is an
epistemic reason for P.
(2) A criterion C can have epistemic value Ф only if some P has Ф and P is an
epistemic reason for C.
554 SKEPTICISM
(NC) No P or C can have Ф if that requires a finite, circular sequence of epistemic

reasons that have Ф.
(NIR) No P or C can have Ф if that requires an infinite sequence of epistemic
reasons that have Ф.
(3a) ∴ No proposition can have Ф.
(3b) ∴ No criterion can have Ф.
The premises of the argument are plausible for many epistemic values Ф. (1)
says that a proposition P can have epistemic value only for persons who believe
a principle underwriting the method by which they have identified P to be true
and for whom that belief has that epistemic value. For example, it is plausible that
P is justified only for persons who are justified in believing that the method by
which they have identified P to be true is reliable. (2) says that a criterion can have
epistemic value only for persons who have a good epistemic reason to believe that
criterion. For example, it is plausible that a principle for distinguishing true from
false propositions can be known by a person S only if S knows evidence that supports
that principle. (NC)—for “No Circles”—states, plausibly, that no proposition can
have epistemic value if that requires a finite circle of epistemic reasons that have
that value. Likewise, (NIR)—for “No Infinite Regresses”—states, plausibly, that no
proposition can have epistemic value if that requires an infinite regress of good
epistemic reasons. (NC) and (NIR) are weak since they do not entail that circles
or infinite regresses of reasons cannot be relevant to Ф. In particular, they are
compatible with the idea that circles or infinite regresses of reasons might play a
significant role in enhancing the epistemic value of some propositions or criteria.13
(1) and (2) jointly entail that epistemic value Ф requires either a circle or an
infinite regress of good epistemic reasons. This, together with (NC) and (NIR),
entails (3a) and (3b). So either at least one of these plausible assumptions is false or
skepticism is true. The schema shows that there is a version of the problem of the
criterion for any epistemic value Ф for which (1), (2), (NC), and (NIR) are plausible.
5 THE RANGE OF RESPONSES

Following Chisholm, we can understand the types of responses to the problem of
the criterion in terms of (1), (2), (NC), (NIR), (3a), and (3b). To describe these types
of responses, I use Chisholm’s terminology as far as possible. Particularism holds
that (1) is false: a good proposition need not be supported by a good criterion of
truth. Perhaps, for example, one need not know that the procedure by which one
has identified a proposition to be true is reliable in order to know that proposition.
Methodism holds that (2) is false: a good criterion need not be supported by any
good evidential reason. Perhaps, for example, a person can be justified in believing
a criterion of truth without having any justified evidential reason for that criterion.
Skepticism about an epistemic value is true if the argument is sound.
This list leaves out responses that reject (NC) or (NIR). Coherentism, as we
may call it, is a family of epistemological theories that reject (NC) on the grounds
that propositions and criteria can have epistemic value Ф on the basis of finite but
circular sequences of reasons. Likewise infinitism, as we may call it, is a family

of epistemologies that reject (NIR) on the related grounds that propositions and
criteria can have value Ф on the basis of infinite regresses of good reasons.
This list still undercounts the types of responses. For one thing, there are anti-
skeptical epistemologies that reject more than one of the premises of the argument.
One variety of contextualism, for example, holds that propositions and criteria can
have epistemic value without being supported by good epistemic reasons provided
they are immunized against the need for reasons by the conventions governing the
relevant context of inquiry, so both (1) and (2) are false. Similarly, intuitionism can
hold that because we are able to apprehend the truth of some criterion and some
proposition without having epistemic reasons for them, both (1) and (2) are false.
So although there will be as many types of anti-skeptical response to the problem
of the criterion as there are premises in the skeptical argument—four—there is
no limit to the arguments that might be given against any premise on the basis
of different epistemological theories. The epistemological theories that underwrite
these arguments need share little else beyond the premises they reject.
I shall not discuss specific epistemologies in detail. Rather, I consider each
premise of the skeptical argument schema and call attention to problems for any
epistemology that rejects it.
6 DISSOLVING THE PROBLEM

Some claim that the problem is not genuine. Robert Amico, for example, argues
that skepticism is self-refuting if “the skeptic” proposes to support skepticism with
reasons and inert if “the skeptic” has no beliefs. So the problem dissolves (Amico
1993: 119–139).14
A problem for this response is that the problem does not require skeptics. The
problem exists because propositions about what epistemic value requires that are
individually plausible to anti-skeptics entail skepticism. Noting that it is constituted
by such propositions is to recognize, not to dissolve, the problem. An anti-skeptical
response must give an account that explains why one of the premises of the argument
is false.
7 PARTICULARISM: REJECTING (1)

Particularist epistemologies imply that (1) is false because they imply that
propositions can have epistemic value without being supported by any criterion
having that value. Assuming that propositions with epistemic value can be good
(though possibly fallible) epistemic reasons for the propositions they deductively
or inductively imply, a challenge for particularist epistemologies is to explain how
propositions that are not supported by good criteria can be good epistemic reasons.
In particular, particularist epistemologies must account for the conditions under
which such propositions can be good reasons to believe criteria of truth.
Some examples of particularist epistemologies are traditional foundationalism,
reliabilism, intuitionism, some versions of contextualism, and many externalist
556 SKEPTICISM
theories, such as reliabilism. Traditional foundationalism holds that propositions can

acquire epistemic value by being apprehended in a way that makes us directly aware
of their truth or by standing in a relevant evidential relationship to non-doxastic
mental states such as sensory experiences.15 So (1) is false. Reliabilism implies that
beliefs have epistemic value if they are produced by methods that produce sufficiently
more true than false beliefs. If so, then (1) is false since it is possible that some belief
is produced by a reliable process that does not include any epistemic reasons.
There are two major challenges for particularist epistemologies: arbitrariness and
easy knowledge. The claim that some beliefs can have epistemic value without being
supported by any criterion of truth implies that it can be good to trust a method
for identifying true propositions—the method by which one has formed those very
beliefs—without having any beliefs about that method. It seems, however, that there
are some methods that it would not be good to trust without having evidence for
their reliability. If you seek truth about the road to Larissa, it is not good to rely on
the testimony of a stranger you have no reason to trust even if the stranger happens
to be reliable. It seems arbitrary, however, to suppose that things are any different
for the more general methods we happen to trust: sense perception, rational insight,
memory, norms for deductive and inductive reasoning, etc. Second, if we claim that
beliefs that are not supported by criteria of truth can be good reasons to believe in
the reliability of the methods that produce them, we sanction a kind of epistemic
circularity and face the problem of easy knowledge.16
The problem of easy knowledge is about whether beliefs that are produced by a
method M can be good reasons to believe that M is reliable. If (1) is false, then beliefs
that are not supported by any criterion can have epistemic value. Hence, in the
absence of some principle preventing it, such beliefs can be good evidential reasons
for the criteria that underwrite the very methods that produce them (bootstrapping).
This seems too easy. If beliefs can be good evidential reasons for the reliability of the
very methods that produce them, then a kind of epistemic circularity is not vicious.17
A plausible constraint on good evidential reasons, however, is that beliefs that result
from a method cannot be good evidential reasons for the reliability of that method.
This constraint would explain why it is not good to believe, for example, that a gas
gauge is reliable on the basis of a belief about its track record—it has been correct
every time I have checked—that is itself based on beliefs acquired by reading that
gauge. Two responses to this problem that would allow us to maintain that (1)
is false have received significant attention in contemporary epistemology: denying
closure and biting the bullet.
Rejecting (1) implies that bootstrapping is legitimate only if the epistemic value
required to be a good evidential reason is closed under deductive and inductive
implication. To say that epistemic value Ф is closed under implication is to say that
if P has Ф, then P transmits Ф to any proposition Q for person S if P implies Q,
perhaps provided that S recognizes the implication, bases the belief that Q on the
belief that P, there are no relevant defeaters, and so on. If closure fails, therefore,
we can reject the view that readings of a gas gauge are good evidential reasons for its
reliability while still maintaining that some beliefs are good without being supported
by good procedural reasons. To do so, however, we must reject one or more of the
following sorts of claims: (R) the gas gauge reads ½ full is a good evidential reason
for (H) the gas tank is ½ full; (R) & (H) is a good evidential reason for (T) the gauge
is correct this time; (T) is a good evidential reason for (C) the gauge has been correct
every time I have checked; or (C) is a good evidential reason for the gauge is reliable.
It is hard to see how any of these claims is mistaken (Barnett 2014: 181–187). This
complex issue is vigorously debated in contemporary epistemology.18
Another response to the problem of easy knowledge is to bite the bullet, endorse
bootstrapping, and accept that knowledge can be easy. On this view, beliefs
produced by method M and for which one has no procedural reason can be good
and can support M is reliable: this kind of epistemic circularity is not vicious. James
Van Cleve argues that this strategy provides the key to solving both the problem
of the Cartesian circle and the problem of induction, and is the only alternative to
skepticism.19 This approach, however, faces the significant challenge of making a
principled distinction between when it is and when it is not good to trust a method
without having evidence for its reliability.
8 METHODISM: REJECTING (2)

Methodist epistemologies imply that (2) is false because they hold that it is possible
that a criterion can have epistemic value without being supported by a good
evidential reason. A challenge for any methodist epistemology is to explain how this
is not unacceptably risky and arbitrary (Chisholm 1973: 17).
Two examples of methodist epistemologies are criteriological intuitionism
and criteriological contextualism. Criteriological intuitionism holds that we can
apprehend the correctness of some criterion without having a good epistemic
reason for it. So criteriological intuitionism implies that (2) is false since a belief
in an intuited criterion can have epistemic value without being supported by any
epistemic reason.20 Likewise, a contextualist can hold that the rules governing what is
appropriate in a context of inquiry immunize some criteria against any requirement
that we have epistemic reasons for them.21 So (2) is false because unsupported
criteria can have epistemic value in contexts where support is not required.
Methodist epistemologies face a problem similar to one of the problems for
particularist theories: it appears to be arbitrary to believe a criterion without a
good epistemic reason. Just as particularists claim that it can be reasonable to trust
an unsupported method—whether or not we explicitly believe it to be reliable—
methodist epistemologies imply that it can be reasonable to believe that a method
is reliable without having any good epistemic reason for that belief. Methodist
epistemologies must, therefore, explain why it is reasonable to believe some criteria
without having reasons for them, but not reasonable to believe others without having
good reasons. An intutionist might claim that some criteria are special because we
can simply apprehend their correctness without good epistemic reasons. This does
not solve the problem, however, unless there is an explanation for why we should
trust the products of intuition without evidence for the reliability of intuition itself.
Responsible inquiry seems to require that we have good evidential reasons for some
of the criteria that express and underwrite our belief-forming methods. Methodists
558 SKEPTICISM
must explain why this does not license arbitrariness and, specifically, why it is true
in some cases but not others.
9 COHERENTISM AND INFINITISM:

REJECTING (NC) OR (NIR)
Coherentist epistemologies reject (NC) on the grounds that even though epistemic
value always requires a circle of good epistemic reasons, beliefs can have epistemic
value because not all circularity is vicious.22 Infinitist epistemologies reject (NIR)
on the grounds that even though epistemic value always requires an infinite regress
of epistemic reasons, beliefs can have epistemic value because we can have infinite
regresses of good reasons.
An attractive feature of both coherentist and infinitist epistemologies is that they
imply that epistemic value requires good epistemic reasons. Therefore they have the
resources to address the arbitrariness problems for particularism and methodism.
Coherentism faces the challenge of explaining why circles of reasons can be
good. If all complete sequences of good reasons are finite circles, then every good
belief must appear both sooner and later in a sequence of reasons in which every
component has a successor that supports it. This appears to imply that it can be
good to assume the very proposition for which we need a reason. Coherentist
epistemologies must also explain why circularity is good in some cases but not in
others. Like particularism, coherentism also faces the problem of easy knowledge.
For if the beliefs in a circular system can be good, it seems that such beliefs could
be good reasons to believe in the reliability of the very methods that produce them.
Though they occupy a position in the space of possible responses to the problem
of the criterion, infinitist epistemologies have not usually been taken seriously.
Due to work by Peter Klein and others, however, infinitist epistemologies have
been defended by some contemporary philosophers, though they have not usually
been advanced as solutions to the problem of the criterion.23 As a response to that
problem, infinitism has a number of attractive features. Because they imply that
good beliefs must be supported by good reasons, infinitist epistemologies are not
subject to the arbitrariness problems for particularism and methodism. Unlike
coherentist epistemologies, they need not accept circles of reasons. Nor need they
face the problem of easy knowledge, provided that beliefs further “upstream” in an
infinite sequence of good epistemic reasons are not produced by the same methods
that produce the “downstream” beliefs. A major challenge for infinitist responses to
the problem of the criterion is to explain how it is possible to have infinitely many
procedural reasons, hence infinitely many methods of belief formation.24
10 CONCLUSION: VALUES, PLURALISM,

AND SKEPTICISM
Perhaps there are different kinds of epistemic goodness and good epistemic reasons.
If so, then different epistemologies might identify different kinds of good beliefs
and reasons. A common practice in epistemology is to debate the conditions on
epistemic goodness by appealing to intuitions about whether or not persons would

have knowledge or justification in test cases involving the number of coins in one’s
pocket, painted mules, and so on.25 A plausible explanation for the logjam of
intuitions about whether there is knowledge or justified belief in such cases is that
different epistemologies have identified different kinds of knowledge and justified
belief. For example, if our goal is simply to have true beliefs, reliably produced
beliefs might be good whether or not we have any reasons for them. If, however, our
goal is to be responsible inquirers or to be rationally autonomous agents, then such
beliefs might not be good. Pluralism about epistemic values and kinds of epistemic
reasons supports the idea that we broaden epistemological inquiry by investigating
the conceptions of human flourishing that inform our intuitions about whether or
not persons have good beliefs in the cases that constitute epistemological examples
and counterexamples. In particular, it supports the idea that we consider the kinds
of reasons we need in order to satisfy the ideals expressed by different conceptions
of a meaningful life. If the kinds of good beliefs and reasons we need according to
a given conception of flourishing need not satisfy all of the premises that pose the
problem of the criterion, we have the basis for an anti-skeptical response to the
problem for that kind of epistemic value. I fear, however, that there are attractive
conceptions of flourishing that do require the kinds of good beliefs and reasons
that satisfy (1), (2), (NC), and (NIR): being intellectually responsible and being
rationally autonomous agents, for example.26 If so, skepticism about such reasons
and the epistemic goods that require them is true. Should we abandon the ideals that
require such reasons? Might we, instead, be better persons by accepting ideals that
we recognize to be impossible? I hope so, but making this plausible and figuring out
what it implies for how we ought to conduct our intellectual lives requires a lot of
explaining. I end on this skeptical note hoping that these questions and the problem
of the criterion will be the focus of continued research in epistemology and ethics.27
NOTES
1. Sextus (1996: 104 = PH I, 114–117; 105 = PH I, 121–123; 128–129 = PH II,
19–20). The problem of the Cartesian circle is posed in Arnauld’s objections
to Descartes’s Meditations in Descartes (1984: 150 = AT 214). Descartes’s reply,
about which there remains deep disagreement concerning both its nature and success,
is in his replies to Arnauld’s objection (Descartes 1984: 171 = AT 245–246). Van
Cleve (1979) presents an exceptionally clear explanation of the Cartesian circle, a
criticism of many proposed solutions, and an original proposed solution. Tlumak
(1978) proposes a solution to the problem that is deeply informed, philosophically
astute, and underappreciated (see also Tlumak 2007: 52–59). The problem of
induction is raised in Hume (1977: section 4, 15–25). Weintraub (1995) shows that
the problem of induction is a version of the problem of the criterion. The problem of
the criterion is also raised by Hegel (1977: 52), who uses it to support his dialectical
method.
2. Chisholm says, “I am tempted to say that one has not begun to philosophise [sic] until
one has faced this problem and has recognized how unappealing, in the end, each of
the possible solutions is” (1973: 1).
560 SKEPTICISM
3. Chisholm also poses the problem, among other places, in Chisholm (1957: 30–39;
1977: 119–134).
4. For critical discussion, see Cling (1994; 1997).
5. It seems that most contemporary epistemologists have identified the problem of the
criterion with the narrower metaepistemological problem. See, for example, Steup
(1992; 1995). Weisberg (2010: 598) addresses the problem in connection with
bootstrapping, discussed below, but calls it “the regress problem.” For whatever
reason, many discussions of issues concerning the most general version of the
problem of the criterion have not been identified to be relevant to it.
6. For arguments for this, see Turri (2009), Gibbons (2010), and Cling (2014: 59–60).
7. I leave to the side the difficult question of whether non-doxastic states—with or
without propositional content—can be epistemic reasons. In Cling (2008: 408–
411), I argue that reasons must be beliefs or belief-like mental states.
8. Strictly speaking, epistemic reasons must be de dicto beliefs. Since such beliefs
have propositional content, I shall sometimes call propositions instead of beliefs
“reasons,” either as shorthand or in contexts where it is the propositional content
of the belief that plays the essential role.
9. As the literature on reliabilism—the family of views according to which positive
epistemic values are to be understood in terms of the reliability of belief-forming
methods—shows, it is non-trivial to individuate methods. See, e.g., Feldman (1985).
10. Giving only necessary conditions on good epistemic reasons as I have leaves it open
that additional conditions must be satisfied.
11. In Cling (2014: 60–61), I make use of an attractive concept of norms due to
Haugeland (1982: 16–17) that identifies norms with complex dispositions, not
propositions. I conjecture that it is norms, in Haugeland’s sense, that play the
fundamental role in most of the methods by which we identify propositions to be
true and that a crucial role for procedural reasons is to make explicit the principles
that are implicit in such norms and to give us reasons to trust those norms.
12. Epistemic propositions also need to be indexed to persons and times. Nothing in
my discussion turns on this, so I abbreviate. When the epistemological chips are
down, however, we can spell out principles in full. Premise (1) of the argument, for
example, is short for a proposition P can have epistemic value Ф for a person S at a
time t only if some C has Ф for S at t and C is an epistemic reason for P for S at t.
13. For an argument that a proposition can enhance the justification we have for it by
being in its own reason ancestry, see Cling (2002).
14. Amico seems to be addressing only the metaepistemological version of the
problem, but my criticism still applies. Amico has more to say on dissolution and
pseudoproblems in Amico (1993: 13).
15. Moser (1985) offers a detailed survey and critique of a wide range of theories
of justification, in addition to defending a powerful version of internalist
foundationalism according to which non-doxastic sensory appearances provide
evidence for basic beliefs.
16. On the problem of easy knowledge and bootstrapping see, e.g., Fumerton (1995:
173–181), Vogel (2000; 2008), Cohen (2002)—who explicitly notes the relevance
of the problem of easy knowledge to the problem of the criterion—and Barnett

(2014). Weisberg (2012) gives a useful overview of the issues.
17. Excellent discussions of epistemic circularity are given in Alston (1986; 1993). I have
attempted to characterize and apply the notion in Cling (2002: 254–256; 2003:
281–287). It will be clear from my discussion that problems about bootstrapping
and epistemic circularity are intimately related.
18. For a good introduction to some of the issues about closure, see the debate between
Dretske (2005a,b) and Hawthorne (2005). Barnett (2014) is especially helpful on
the connections between bootstrapping, epistemic circularity, and closure, and his
paper has greatly influenced my account of these issues.
19. Van Cleve has developed this view in connection with these problems in Van
Cleve (1979; 1984; 2003). For a criticism of Van Cleve’s proposed solution to the
problem of the Cartesian circle, see Cling (1997).
20. Van Cleve (2003: 50–51) cites Reid (1969: 617, 622, 625, 630) as holding
this position.
21. Williams (2001: 169–170) outlines this type of contextualist response to the
problem of the criterion.
22. DePaul (1988) defends coherentism as a response to the problem of the criterion
in ethics. Elgin (1996) defends a coherentist epistemology for all beliefs, though
it contains a particularist element by allowing that beliefs have a default positive
epistemic status—initial tenability (Elgin 1996: 101–02)—until we have epistemic
reasons to revise them. Initial tenability, however, may depend upon having non-
epistemic reasons: “That the sentences we accept do not in general frustrate our
efforts is some reason to accept them” (Elgin 1996: 102). If so, then Elgin’s system
resembles Lynch’s, which also claims that we can have good non-epistemic reasons
for our methods (Lynch 2012: 78; 88–108).
23. Defenses of infinitism are in Klein (1999) and Aikin (2010). Infinitism is discussed
from a wide range of perspectives in Klein and Turri (2014). I criticize infinitism
in Cling (2004; 2014a). I discuss the relationship between the epistemic regress
problem and the problem of the criterion in Cling (2014b), though the statement of
the problem in this chapter is, I think, superior to the one I gave in that one.
24. Klein (1999) responds to the related finite minds problem about evidential reasons
by weakening the conditions on having a reason: reasons are merely second-order
dispositions to be disposed to form the relevant beliefs in the required contexts. Can
a similar sort of response give us access to infinitely many belief-forming methods?
For an argument that the kinds of reasons that Klein provides to solve the finite
minds objection are not the kinds of reasons we ought to want if we are to attain
certain kinds of human flourishing, see Cling (2014a: 62–71).
25. An important boon to this approach was provided by Gettier (1963). For a detailed
account of a range of attempts to solve Gettier’s problem, see Shope (1983).
26. I argue that there are such reasons in Cling (2009: 339–340; 2014a: 62–71;
2014b: 165–170).
27. I am very grateful to my colleague Nicholaos Jones and to Diego Machuca and
Baron Reed for very useful comments on an earlier draft of this chapter. The errors
that remain are mine alone.
562 SKEPTICISM
REFERENCES
Aikin, Scott. 2010. Epistemology and the Regress Problem. New York: Routledge.
Alston, William P. 1986. “Epistemic Circularity,” Philosophical Studies 47: 1–28.
Alston, William P. 1993. The Reliability of Sense Perception. Ithaca, NY: Cornell University
Press.
Amico, Robert. 1993. The Problem of the Criterion. Lanham, MD: Rowman & Littlefield.
Barnett, David. 2014. “What’s the Matter with Epistemic Circularity?” Philosophical
Studies 171: 177–205.
Chisholm, Roderick. 1957. Perceiving. Ithaca, NY: Cornell University Press.
Chisholm, Roderick. 1973. The Problem of the Criterion. Milwaukee: Marquette University
Press.
Chisholm, Roderick. 1977. Theory of Knowledge, 2nd ed. Englewood Cliffs, NJ: Prentice-
Hall.
Cling, Andrew. 1994. “Posing the Problem of the Criterion,” Philosophical Studies 75:
261–292.
Cling, Andrew. 1997. “Epistemic Levels and the Problem of the Criterion,” Philosophical
Studies 88: 109–140.
Cling, Andrew. 2002. “Justification-affording Circular Arguments,” Philosophical Studies
111: 251–275.
Cling, Andrew. 2003. “Self-supporting Arguments,” Philosophy and Phenomenological
Research 66: 279–303.
Cling, Andrew. 2004. “The Trouble with Infinitism.” Synthese 138: 101–123.
Cling, Andrew. 2008. “The Epistemic Regress Problem,” Philosophical Studies 140: 401–
421.
Cling, Andrew. 2009. “Reasons, Regresses, and Tragedy,” American Philosophical
Cling, Andrew. 2014a. “Reasons Require Reasons.” In P. Klein and J. Turri (eds.), Ad
Infinitum: New Essays on Epistemological Infinitism, 55–74. New York: Oxford
University Press.
Cling, Andrew. 2014b. “The Epistemic Regress Problem, the Problem of the Criterion, and
the Value of Reasons,” Metaphilosophy 45: 161–171.
DePaul, Michael. 1988. “The Problem of the Criterion and Coherence Methods in Ethics,”
Canadian Journal of Philosophy 18: 67–86.
Descartes, René. 1984. The Philosophical Writings of Descartes, vol. 2, translated by
J. Cottingham, R. Stoothoff, and D. Murdoch. New York: Cambridge University Press.
Dretske, Fred. 2005a. “The Case against Closure.” In M. Steup, and E. Sosa (eds.),
Contemporary Debates in Epistemology, 13–25. Malden, MA: Blackwell.
Dretske, Fred. 2005b. “Reply to Hawthorne.” In M. Steup, and E. Sosa (eds.), Contemporary
Debates in Epistemology, 43–46. Malden, MA: Blackwell.
Elgin, Catherine. 1996. Considered Judgment. Princeton, NJ: Princeton University Press.
Feldman, Richard. 1985. “Reliability and Justification,” The Monist 68: 159–174.
Fumerton, Richard. 1995. Metaepistemology and Skepticism. Lanham, MD: Rowman &
Littlefield.
Gibbons, John. 2010. “Things that Make Things Reasonable,” Philosophy and
Haugeland, John. 1982. “Heidegger on Being a Person,” Noûs 16: 15–26.
Hawthorne, John. 2005. “The Case for Closure.” In M. Steup and E. Sosa (eds.),
Contemporary Debates in Epistemology, 26–42. Malden, MA: Blackwell.
Hegel, G.W.F. (1977). Hegel’s Phenomenology of Spirit, translated by A. Miller. New York:
Hume, David. 1977. An Enquiry Concerning Human Understanding, edited by E. Steinberg.
Klein, Peter and John Turri (eds.). 2014. Ad Infinitum: New Essays on Epistemological
Infinitism. New York: Oxford University Press.
Lynch, Michael. 2012. Praise of Reason. Cambridge, MA: MIT Press.
Moser, Paul K. 1985. Empirical Justification. Boston: D. Reidel.
Reid, Thomas. 1969. Essays on the Intellectual Powers of Man, edited by B. Brody.
Cambridge, MA: MIT Press.
Sextus Empiricus. 1996. The Skeptic Way: Sextus Empiricus’s Outlines of Pyrrhonism,
translated by B. Mates. New York: Oxford University Press.
Shope, Robert. 1983. The Analysis of Knowing: A Decade of Research. Princeton, NJ:
Princeton University Press.
Steup, Matthias. 1992. “Problem of the Criterion.” In J. Dancy and E. Sosa (eds.),
A Companion to Epistemology, 378–381. Cambridge, MA: Blackwell.
Steup, Matthias. 1995. “Problem of the Criterion.” In R. Audi (ed.), The Cambridge
Dictionary of Philosophy, 653. New York: Cambridge University Press.
Steup, Matthias and Ernest Sosa (eds.). 2005. Contemporary Debates in Epistemology.
Malden, MA: Blackwell.
Tlumak, Jeffrey. 1978. “Squaring the Cartesian Circle,” The Southern Journal of Philosophy
16: 247–257.
Tlumak, Jeffrey. 2007. Classical Modern Philosophy: A Contemporary Introduction.
Turri, John. 2009. “The Ontology of Epistemic Reasons,” Noûs 43: 490–512.
The Philosophical Review 88: 55–91.
Van Cleve, James. 1984. “Reliability, Justification, and the Problem of Induction,” Midwest
Studies in Philosophy 9: 555–567.
Van Cleve, James. 2003. “Is Knowledge Easy—or Impossible? Externalism as the only
Alternative to Skepticism.” In S. Luper (ed.), The Skeptics: Contemporary Essays, 45–59.
Aldershot: Ashgate.
Vogel, Jonathan. 2008. “Epistemic Bootstrapping,” The Journal of Philosophy 105: 518–539.
564 SKEPTICISM
Watkins, Calvert (ed.). 2000. The American Heritage Dictionary of Indo-European Roots.
Boston: Houghton Mifflin.
Weintraub, Ruth. 1995. “What was Hume’s Contribution to the Problem of Induction?”
Weisberg, Jonathan. 2010. “Bootstrapping in General,” Philosophy and Phenomenological
Research 81: 525–548.
Weisberg, Jonathan. 2012. “The Bootstrapping Problem,” Philosophy Compass 9: 597–610.
Williams, Michael. 2001. Problems of Knowledge: A Critical Introduction to Epistemology.
CHAPTER THIRTY-NINE
Neo-Pyrrhonism
MARKUS LAMMENRANTA
1 INTRODUCTION
Classical Pyrrhonian skepticism had a practical side: the skeptics tried to attain
tranquility by suspending beliefs. David Hume (1975: 160) famously attacks this
form of skepticism thus: (1) life without beliefs is psychologically impossible, and
(2) even if it were possible, life would be very short because intentional action is not
possible without beliefs. Assuming Hume is right, it is no surprise that there have
not been many Pyrrhonian skeptics.
It is also a common interpretation that classical Pyrrhonism is based on the
equal plausibility of conflicting or contrary propositions. When we find such sets
of propositions, it is psychologically impossible for us to believe one of them rather
than the other, and we are forced to suspend belief. However, if Hume is right, there
cannot be many such conflicts, and therefore classical Pyrrhonism does not offer a
serious skeptical challenge to our beliefs.
This seems to be a widespread diagnosis of classical Pyrrhonism among
contemporary epistemologists. It explains why they have been more interested in
Cartesian skepticism, which appears to raise more interesting skeptical problems.
However, there are a few philosophers who think that we should take Pyrrhonism
more seriously, and that it should even be accepted. When it is properly understood,
we can see that there have even been actual and, indeed, very influential Pyrrhonian
skeptics such as Michel de Montaigne, Hume himself, Ludwig Wittgenstein, and,
among contemporary philosophers, Robert Fogelin and Walter Sinnott-Armstrong.
We cannot say the same of Cartesian skepticism (Sinnott-Armstrong 2004a: 4).
How, then, should we understand Pyrrhonism? What constitutes updated
Pyrrhonism or neo-Pyrrhonism? Fogelin (1994: 5–10), who is the main
contemporary advocate, restricts the scope of skepticism to philosophical and
perhaps other theoretical matters. Here he follows Michael Frede’s interpretation of
classical Pyrrhonism: it is directed at the dogmatic beliefs of professors rather than
the common beliefs of ordinary people. In the taxonomy proposed by Jonathan
Barnes (1997: 60–62), Fogelin’s neo-Pyrrhonism is a form of urbane Pyrrhonism
that suspends beliefs about philosophy, rather than a form of rustic Pyrrhonism that
566 SKEPTICISM
suspends all beliefs. This sort of Pyrrhonism avoids Hume’s worries: we can surely
live without philosophical beliefs.
Another important feature of Fogelin’s neo-Pyrrhonism is that it arises from
philosophy: it is skepticism about philosophy, based on philosophical arguments.
Philosophical reflection on ordinary epistemic practices leads us to deny the
possibility of knowledge and justified belief. However, a neo-Pyrrhonian skeptic does
not conclude that we cannot have knowledge or justified beliefs, which would be
negative dogmatism. Neither does she reject ordinary epistemic practices: she rejects
the philosophical reflection that leads to the impasse (Fogelin 1994: 3–4). So Fogelin’s
neo-Pyrrhonism is based on a reductio ad absurdum of epistemology: because the
epistemological enterprise leads to dogmatic skepticism, it should be rejected.
This sort of neo-Pyrrhonism avoids the problems raised by Hume. Indeed, in
Fogelin’s view, Hume himself was such a Pyrrhonist. In this chapter, I concentrate
on Fogelin’s neo-Pyrrhonism and evaluate its plausibility. Does epistemology, when
seriously pursued, destroy knowledge, as Fogelin (2004: 164) argues? This depends
very much on the philosophical argument that is supposed to show that knowledge
is impossible and that he takes to be a reductio of its philosophical premises.
Unfortunately, it is not so clear what the argument is. Fogelin never formulates it
explicitly. He suggests that it derives from the so-called Five Modes of Agrippa,
but it is not easy to interpret these modes. I will formulate two different versions
of the Agrippan argument and consider whether either of them supports the neo-
Pyrrhonian moral. But let us first take a look at the moral itself.
2 FOGELIN’S NEO-PYRRHONISM
Fogelin argues in the second part of his book Pyrrhonian Reflection on Knowledge
and Justification that Agrippa’s problem arises from philosophical reflection on our
ordinary epistemic practices and that the theories of justification he considers cannot
solve it. The result is a kind of urbane Pyrrhonism that combines “philosophical
skepticism with skepticism about philosophy, that is, to have doubts about
philosophy on the basis of philosophical arguments” (1994: 3).
In order to explain how Agrippa’s problem arises from our ordinary epistemic
practices, Fogelin gives the following account of knowledge attributions:
“S knows that p” means that “S justifiably believes that p on the grounds that
establish the truth of p.” (Fogelin 1994: 94; 1997: 417)
Although this is what we say when we make knowledge attributions, as Fogelin

(1994: 91–93) points out, we often make serious knowledge attributions in the
face of error-possibilities that the grounds do not exclude. We simply ignore these
possibilities, or “rely on the grace of nature not to defeat us.” However, if we dwell
on them, the level of scrutiny will rise and we will no longer be ready to attribute
knowledge—indeed we will be inclined to deny it.1
Thus, our ordinary practice of knowledge attribution involves procedures for
invoking higher levels of scrutiny that lead us to deny knowledge. When we allow
the level of scrutiny to rise without restriction, we are inclined to say that nothing is
NEO-PYRRHONISM 567
known. According to Fogelin’s interpretation of Pyrrhonism, a Pyrrhonian skeptic

will go along with these different levels and attribute knowledge when the level is
low and deny knowledge when it is high. She will just report how things strike her
without privileging any particular level.
However, in Fogelin’s view, epistemologists have not been happy with this sort
of Pyrrhonism. They have sought to propose theories of justification to remedy the
fragility of our ordinary epistemic practices:
I hold that the theory of knowledge, in its traditional form, has been an attempt
to find ways of establishing knowledge claims from a perspective where the level
of scrutiny has been heightened by reflection alone. (Fogelin 1994: 99)
Unfortunately, this attempt fails because it falls into the Modes of Agrippa. Indeed,
“the epistemological enterprise, when relentlessly pursued, not only fails in its
efforts, but also Samson-like, brings down the entire edifice of knowledge around
it” (Fogelin 2004: 164). Thus, traditional epistemology destroys knowledge: it leads
inevitably to the conclusion that nothing is known or that no beliefs are justified.
We can take this to be a reductio ad absurdum of the epistemological enterprise,
which should therefore be discarded. We should become Pyrrhonists—or neo-
Pyrrhonists—who go on making knowledge claims about ordinary matters but
suspend belief about epistemology.
Before deciding whether Fogelin succeeds in showing that the epistemological
enterprise fails because it leads to dogmatic skepticism, let us consider the question
of whether Fogelin’s neo-Pyrrhonism itself is coherent by examining Walter Sinnott-
Armstrong’s neo-Pyrrhonism.
3 SINNOTT-ARMSTRONG’S NEO-PYRRHONISM
Sinnott-Armstrong (2004b: 188–189) expresses puzzlement about Fogelin’s attitude
to ordinary knowledge attributions. On the one hand, Fogelin (1994: 10, 88)
supposes that Pyrrhonism leaves ordinary beliefs untouched, including epistemic
beliefs about whether somebody knows something, but on the other hand he states
that, “in making knowledge claims, we always (or almost always) assert more than
we have a right to assert” (1994: 94). When the cost of being wrong is low, we make
knowledge attributions in the face of uneliminated error-possibilities, but when the
cost becomes excessive or we dwell on those error-possibilities, we deny knowledge
and even assert that nobody knows anything.
Fogelin’s Pyrrhonist seems to be self-contradicting, saying both that somebody
knows something and that nobody knows anything. Both assertions cannot be true.
Sinnott-Armstrong tries to reconcile these claims by appealing to contrastivism.
According to Sinnott-Armstrong, justification is not a binary relation between a
subject and a proposition but a ternary relation between a subject, a proposition,
and a contrast class:
Someone, S, is justified in believing a proposition, P, out of a contrast class, C,
when and only when S has grounds that rule out any other member of C but do
not rule out P. (2004b: 190)
568 SKEPTICISM
A contrast class is composed of propositions that are contrary to each other. We

are justified in believing one of these propositions just in case it is the only member
of the set that is left uneliminated by our grounds. Suppose, for example, that a
father takes his daughter to the zoo and looks at a zebra in a cage. His sensory
experience rules out the possibility that the animal is an antelope but does not rule
out the possibility that it is a cleverly painted mule. The father is therefore justified
in believing that he sees a zebra rather than an antelope, but he is not justified in
believing that he sees a zebra rather than a painted mule.
In order to reconcile Fogelin’s apparently conflicting claims, Sinnott-Armstrong
(2004b: 190) distinguishes three contrast classes that correspond to Fogelin’s three
levels of doubt:
The unlimited contrast class for P = all propositions contrary to P, including
skeptical scenarios that are systematically ineliminable.
The extreme contrast class for P = all propositions contrary to P that could
be eliminated in some way, even if doing so is not needed in order to meet
normal standards.
The everyday contrast class for P = all propositions contrary to P that could be
eliminated and need to be eliminated in order to meet normal standards.
We can now make sense of Fogelin’s apparently conflicting attributions and denials
of knowledge. They can all be true if we understand him as saying that, although we
are typically justified in believing things out of everyday contrast classes, we are not
justified in believing anything out of extreme and unlimited contrast classes.
However, according to Sinnott-Armstrong (2004b: 193), it is problematic
to understand our unqualified knowledge attributions in this way. The linguistic
meaning (or Kaplan’s character) of the sentences of the form “S is justified in
believing that p” should be represented as:
S is justified out of the relevant contrast class in believing that p.
This is also what makes our attributions of justification normative: to call a contrast
class relevant is to say that the believer needs to rule out all other members of the class in
order to be justified in believing one of them. According to Sinnott-Armstrong (2004b:
193–202), such normative epistemic judgments are suspect because nobody has found
plausible rules for relevance. Indeed, he suggests that a Pyrrhonian skeptic suspends
judgment about all normative matters and instead makes judgments of the forms:
S is justified out of the everyday contrast class in believing that p.
Nobody is justified out of the extreme contrast class in believing that p.
These judgments are quite compatible and purely descriptive.2 According to Sinnott-
Armstrong’s version of neo-Pyrrhonism, a Pyrrhonist is allowed to make such
judgments and suspends judgment only about normative and philosophical matters.
Although Sinnott-Armstrong’s view reconciles the apparently contradictory
knowledge attributions and denials of Fogelin’s neo-Pyrrhonist, it is in conflict
with the spirit of Fogelin’s neo-Pyrrhonism, which leaves our ordinary epistemic
NEO-PYRRHONISM 569
practices untouched and allows the Pyrrhonist ordinary epistemic judgments.

According to Sinnott-Armstrong, our ordinary epistemic judgments presuppose
something that is false or at least suspect, namely, that there are relevant contrast
classes. This is why purely descriptive claims should replace them. Fogelin’s version
of neo-Pyrrhonism is preferable. It is much harder for Sinnott-Armstrong’s skeptic
to avoid all philosophical commitments because she needs some reasons to revise
our ordinary epistemic practices, and it is hard to see how such reasons could be
anything other than philosophical.
It seems that contextualism offers a better way of reconciling the neo-Pyrrhonist’s
apparently conflicting knowledge attributions. It saves our ordinary epistemic
judgments, as Sinnott-Armstrong understands them, by making the relevance of the
contrast class dependent on the attributor’s context. When the stakes are low for us
and we ignore all uneliminated error-possibilities, our knowledge attributions are
true, but when the stakes grow higher and we dwell on some uneliminated error-
possibilities, our knowledge denials are also true. There is no contradiction, because
our knowledge attributions and denials express relations to different contrast
classes. Although Fogelin denies contextualism in the book, he acknowledges in a
more recent paper, “The Skeptics are Coming! The Skeptics are Coming!” (2004)
that it would be one version of neo-Pyrrhonism.
4 THE STRUCTURAL REGRESS PROBLEM

OR AGRIPPA’S TRILEMMA
How, then, does epistemology destroy knowledge? Fogelin argues that the theories
of justification that are put forth to remedy the fragility of our ordinary epistemic
procedures disappoint us because they fail to solve Agrippa’s problem, and this
failure will “bring down the whole edifice of knowledge” (2004: 164). What,
exactly, is Agrippa’s problem?
Fogelin (1994: 114) considers the common reference to Agrippa’s problem as the
infinite regress problem misleading, because a philosophical theory of justification
must avoid involvement not only with infinite regress but also with vicious circularity
and an appeal to arbitrary assumption (or hypothesis). The problem is rather a
trilemma: how can one avoid all three bad options?
According to Fogelin (1994: 114–115, 120–121), Agrippa’s trilemma constitutes
a serious philosophical challenge only if we make two assumptions. The first of
these is the normative principle of epistemic justification, which W. K. Clifford
expresses thus: “It is wrong always, everywhere, and for anyone, to believe anything
upon insufficient evidence.” As Fogelin interprets it, the principle says roughly that
epistemic justification is a matter of responsibility, and that responsibility requires
that we base our beliefs on good reasons (or grounds). The second assumption is
that we have or could have knowledge and justified beliefs.
Together these assumptions entail that we have knowledge that conforms to
Cliffordian standards. The task of a theory of justification is to explain how this
is possible. Thus, it seems clear that Fogelin sees the problem as a paradox. In
570 SKEPTICISM
assuming the Cliffordian principle, we must face Agrippa’s trilemma with its three
bad options, none of which can make beliefs justified. Yet, we believe that we do
have justified beliefs.
To be more precise, the paradox arises through the formulation of a valid
skeptical argument leading to the conclusion that there are no justified beliefs, and
then denying the conclusion. The argument seems to be this (cf. Comesaña 2005):
(S1) If a belief is justified, it is either a basic belief (justified independently of

reasoning) or an inferentially justified belief.
(S2) There are no basic beliefs.
(S3) If a belief is justified, it is made justified either by (a) an infinite chain
of reasons, (b) a circular chain of reasons, or (c) a chain of reasons
terminating in an unjustified belief.
(S4) No belief is made justified by an infinite chain of reasons.
(S5) No belief is made justified by a circular chain of reasons.
(S6) No belief is made justified by a chain of reasons terminating in an
unjustified belief.
(S7) Thus there are no justified beliefs.
This is a valid argument: the subsidiary conclusion S3 follows from premises S1

and S2, and then the final conclusion S7 follows from premises S3, S4, S5, and
S6. Whereas premises S2 and S6 derive from the Agrippan mode of hypothesis,
premises S4 and S5 derive from the modes of infinity and circularity.
Theories of justification offer different responses to this argument: foundationalism
(including holistic coherentism3) rejects premise S2, infinitism rejects premise
S4, circular (or linear) coherentism rejects premise S5, and Wittgensteinian
contextualism rejects premise S6. Fogelin (1994: 117) lists the same options for the
dogmatic epistemologist, except infinitism, which he takes to be hopeless.
Why, then, should all these attempts at responding to the argument fail? One
would expect it to follow from Cliffordism that all premises of the argument are
true, which is why Cliffordism leads to dogmatic skepticism. This would happen
if the principle stated that only justified beliefs could serve as reasons or evidence.
In that case every justified belief must be based on other justified beliefs, which
in turn must be based on other justified beliefs, and so on. This would rule out
foundationalism, which the majority of epistemologists take to be the most plausible
solution to the paradox, and thus leaves us with the three bad options.
The problem with this interpretation of Cliffordism, as Michael Williams (1999:
147) points out, is that the argument merely shows that the Cliffordian principle
is too strong and should therefore be rejected. It does not show that the whole
epistemological enterprise fails. Indeed, Fogelin (1994: 116–117) does not think
that Cliffordism rules out foundationalism. He sees Paul Moser as a Cliffordist and
foundationalist, and conducts an extensive discussion about Roderick Chisholm’s
foundationalism. Thus, Cliffordism as he understands it seems to allow other than
justified beliefs as reasons or evidence, such as non-doxastic experiences or seemings.
NEO-PYRRHONISM 571
However, if we accept this more liberal interpretation of the Cliffordian

principle, it becomes unclear why we should reject foundationalism, according to
which the regress ends in basic beliefs that are justified even though they do not
derive their justification from other justified beliefs. Nevertheless, Fogelin (1994:
123–162) argues that the best existing theories of this sort, such as Chisholm’s
foundationalism and BonJour’s holistic coherentism, fail because they do not even
come close to satisfying the success conditions for a theory of justification.
Fogelin (117–119, 138–139) provides three success conditions for such a theory:
(1) it should help to determine which particular beliefs are justified; (2) it should
explain why these particular beliefs are justified; and finally (3) it should be defended
by arguments that do not beg the question against skepticism, in other words: the
arguments should not depend on the undefended assumption that we do have some
justified beliefs.
The first two conditions seem quite reasonable, but although Fogelin sees much
to be hoped for in Chisholm’s and BonJour’s theories regarding them, it does not
seem unreasonable to assume that many theories can satisfy them. It is the third
condition that is hard to satisfy. Indeed, it seems to be impossible.4 However, it
is strange that Fogelin should hang on to it. He claims elsewhere that we need to
assume that we do have justified beliefs in order to have Agrippa’s problem at all, but
here he states that we are not allowed to make this assumption if we want to solve
the problem. Indeed, it was this assumption that points to Fogelin’s understanding
of the problem as a paradox, and it seems that all we need to do in order to solve a
paradox is to determine which of the propositions comprising it is false and explain
why. This is exactly what we can do if we have a theory that satisfies Fogelin’s first
two conditions for success.
Perhaps the problem is that we have competing possible solutions and should
not just assume that we have the right one, thereby begging the question against
the alternatives. One of those possible solutions is the skeptical one supporting the
conclusion that we have no justified beliefs. Thus, we should not just assume that
we have justified beliefs and beg the question against the skeptical option. This is
exactly what epistemologists do, Fogelin (1994: 119) claims, when they first assume
that we have justified beliefs and then argue by means of disjunctive elimination that
only their own theory can explain this. For example, foundationalists argue that
because infinitely regressive reasoning, circular reasoning, and reasoning starting
from arbitrary premises cannot make beliefs justified, there must be basic beliefs
that are justified independently of reasoning. This argument merely assumes that
we have justified beliefs and thus begs the question against the skeptical resolution
of the paradox.
However, if we understand Agrippa’s problem in this way, we should ask where
the motivation for the third condition of success comes from. Why are we not
allowed to beg the question?5 This requirement does not come from the Cliffordian
principle or the considered theories of justification, which are non-dialectical.
They allow justified beliefs that we cannot defend without begging the question.
It therefore seems that Fogelin’s dialectical requirement comes from the modes of
Agrippa. If this is the case, they may not constitute an epistemically neutral skeptical
572 SKEPTICISM
strategy that can be directed against any theory of justification, as Fogelin (1997:
397) assumes. They presuppose in themselves some requirements for justification.
5 THE DIALECTICAL REGRESS

Although the contemporary discussion of Agrippa’s problem typically focuses on
three modes, the original problem, as Sextus Empiricus describes it in the Outlines
of Scepticism (PH I 164–177), comprises five modes. It is the remaining two
modes that bring on the dialectical requirements. Fogelin refers to them as the
challenging modes:
Two of Agrippa’s modes, discrepancy and relativity, trigger a demand for
justification by revealing that there are competing claims concerning the nature
of the world we perceive. Given this competition, it would be epistemically
irresponsible for the Cliffordian to choose without argument one of these
competing claims over others. Thus the modes of discrepancy and relativity
force anyone who makes claims beyond the modest expression of opinion to give
reasons in support of these claims. (Fogelin 1994: 116)
He refers to the rest of them as the dialectical modes:
The task of the remaining modes—those based on regress ad infinitum, circularity,
and (arbitrary) hypothesis—is to show that it is impossible to complete this reason-
giving process in a satisfactory way. If the Pyrrhonists are right, no argument,
once started, can avoid falling into one of the traps of circularity, infinite regress,
or arbitrary assumption. (1994: 116)
Fogelin thus understands the modes as showing that it is impossible to adequately
justify or defend our beliefs once they are properly challenged. He thus appears
to be suggesting that Agrippa’s problem does not—as we assumed—concern the
structure of justification: it rather concerns the process of justification, our ability to
defend our beliefs by giving reasons for them (see also Williams 2001: 61–63). This
way of understanding the modes makes it easier to understand why Fogelin believes
that existing theories of justification fail: because there is disagreement about the
conditions of justification, one should not accept any theory of justification without
defending it against competing theories, but no such theory can be defended without
begging the question. However, the way Fogelin describes the modes makes them
more widely applicable. They do not merely challenge epistemologists to defend
their theories of justification: “The modes of discrepancy and relativity force anyone
who makes claims beyond the modest expression of opinion to give reasons in
support of these claims” (Fogelin 1994: 116). Hence, the Agrippan modes apply to
any knowledge claim whatever the subject matter.
It is somewhat surprising that Fogelin very quickly starts considering whether
certain existing theories of justification meet the conditions of success and ignores
the question of whether these theories can explain how to stop the process of
reason-giving in the first place. Perhaps this is because the phrases Fogelin (2004:
162) uses—“validating knowledge claims” and “presenting reasons establishing
NEO-PYRRHONISM 573
their legitimacy”—are ambiguous between showing that a claim or belief is true

(justifying it) and showing that it is justified (justifying the claim that it is justified).
In the book he concentrates on the latter rendering, although it is the former that is
more fundamental—the latter is just a special case of it.
In any case, Fogelin is committed to the failure of existing theories of justification
to address the first-order problem: if some such theory, such as foundationalism,
can solve the first-order problem, there seems to be no reason to doubt its ability to
solve the second-order problem as well. For example, if a foundationalist can stop
the first-order regress by appealing to basic beliefs, why could he not do the same
with the second-order regress? He could simply deny the arbitrariness of the second-
order belief that a particular belief is justified. It is not arbitrary because it is itself
a basic belief, which can then be used as evidence for the theory of justification.6
Therefore, in order to block this move Fogelin needs to show that foundationalism
fails to explain how to stop the reason-giving regress.
6 CAN FOUNDATIONALISM STOP THE

DIALECTICAL REGRESS?
We can thus assume that Agrippa’s problem concerns the reason-giving or dialectical
regress. Because Fogelin’s neo-Pyrrhonism is based on the failure of existing theories
of justification to solve it, we are owed an explanation as to why they fail and,
indeed, why the failure leads to dogmatic skepticism. Unfortunately, Fogelin does
not discuss this explicitly. We have to turn to another philosopher, Peter Klein (2005:
131–135), who understands Agrippa’s problem in a similar way and argues in greater
detail that foundationalism cannot solve the problem. Because, as Klein points out,
holistic coherentism is just a special case of foundationalism, foundationalism covers
very well the theories Fogelin discusses in his book.
In explaining why foundationalism cannot solve Agrippa’s problem, Klein
(2005: 133) imagines a dialogue between Fred, the Foundationalist, and Doris, the
Doubter. Fred asserts proposition p. Doris asks Fred for his reasons for believing
that p, perhaps saying, “Why do you believe that p?” or “How do you know that p?”
Fred then gives his reason, r1, for p. Doris asks why r1 is true, and Fred gives another
reason, r2. This goes on until Fred arrives at reason b, which he takes to be a basic
proposition. Doris naturally asks Fred for his reasons for b, but Fred, being a self-
conscious foundationalist, tells Doris that b is a basic proposition that does not need
reasons in order to be justified. Thus Doris’s request for reasons is inappropriate.
In Klein’s view, Doris can concede for the sake of argument that foundationalism
is true and that b is basic in virtue of possessing some property F, such that any
proposition having F is noninferentially justified. Even so, she can say to Fred: “Do
you think that propositions having F are likely to be true?” Once Fred considers
the proposition that such propositions are likely to be true, he faces three options:
to (1) affirm it, (2) deny it, or (3) withhold it. If he affirms it, Doris will ask him
for his reasons, and the regress does not stop. If he either denies or withholds it, he
should concede that that p is not, after all, justified for him, presumably because in
both cases he would have a defeater for his justification for believing that p.7 Klein
574 SKEPTICISM
concludes that foundationalism does not solve Agrippa’s problem. It cannot explain
how the reason-giving regress can be stopped. Indeed, though Klein does not put it
in this way, a foundationalist should concede that if foundationalism is true, there
are no justified beliefs. This is exactly the result Fogelin wants.
However, there is a response that a foundationalist can give: to affirm that
propositions having F are likely to be true, and even to give reasons for this
generalization, but to insist that the justification of b does not depend on this
generalization, which just explains why b is basic. A foundationalist can and should
have an explanation for what makes basic propositions basic. Such an explanation
does not mean that the propositions are not basic, however. On the contrary, it
is assumed to explain why subjects without the epistemological sophistication to
defend their beliefs can have justified beliefs (Bergmann 2004: 164–165).
Doris may point out that this response changes the subject, which is justification
rather than explanation. The aim of the modes—within the present interpretation—
is to show that no proposition can be justified by giving reasons for them. The
challenging modes are supposed to raise doubts about the truth of a proposition,
and pointing out that the proposition is basic does not answer the challenge
because even the truth of basic propositions can be challenged (cf. Klein 2004:
169–170). Foundationalism thus appears unable to explain how the reason-giving
regress can be stopped. Yet it does not follow that there are no justified beliefs if
foundationalism is true. Justification must also depend on the ability to respond
to challenges by giving reasons. This is what the modes must assume if they are
to support skepticism. However, it is what contemporary foundationalists can
and typically do deny, insisting that justification does not require an ability to give
reasons. Before evaluating this response, let us see what Agrippa’s problem looks
like once this implicit assumption about the nature of justification is made explicit.
7 THE DIALECTICAL PROBLEM

As shown above, two of Agrippa’s modes challenge us to give reasons for our beliefs,
and the rest of them aim to show that the reason-giving process cannot succeed. If
they are to constitute a skeptical argument for concluding that there are no justified
beliefs, some epistemic principle connecting justification to reason-giving must be
presupposed. Let us suppose that the principle is this:
(D1) If S is justified in believing that p, S can defend p against appropriate challenges.
Assuming this principle, Agrippa’s modes become relevant to justification. The
challenging modes show that there are appropriate challenges to the truth of p,
and the dialectical modes show that the challenges cannot be answered. Hence the
second premise:
(D2) S cannot defend p against appropriate challenges.
The conclusion follows from these premises via modus tollens:
(D3) S is not justified in believing that p.
NEO-PYRRHONISM 575
In fact, the challenging modes and the dialectical modes work in tandem. When
S tries to defend p by giving reasons for p, the challenging modes show that these
reasons can be challenged, and when S gives further reasons for these reasons, the
challenging modes are applied to these reasons, and so on. The dialectical modes
show that this reason-giving regress can end only in three bad ways, none of which
can provide an adequate response to the challenges. D2 is thus true, and D3 follows
from D1 and D2.
Now, it is clear why the appeal to basic beliefs appears to fail. Because the truth of
even these beliefs can be challenged, they are not able to stop the regress. However,
this argument may not serve the neo-Pyrrhonist’s dialectical purposes as Fogelin
understands them: the neo-Pyrrhonist is supposed to use premises that her dogmatic
opponent, the foundationalist, accepts and then to show that there are no justified
beliefs. The problem is that the foundationalist need not accept premises D1 and
D2. He can deny either one or the other.
8 ARE WE ALL NEO-PYRRHONISTS NOW?

Fogelin seems to be aware of this problem. In Fogelin (2004: 164), he notes that
many contemporary epistemologists—he calls them New Epistemologists—“have
foresworn this large-scale attempt at validation through reason-giving.” They have
either severed the connection between knowledge and reason-giving altogether, or
dispersed reason-giving into a plurality of contexts.
According to externalists/reliabilists, beliefs are made justified by their reliable
connection to reality. Thus, we can know without being able to defend our beliefs
by giving reasons for their truth or for their being based on reliable sources. What
matters is that they are, in fact, based on reliable sources. Hence, the modes of
Agrippa are no threat to our knowledge.
Contextualists, inspired by Wittgenstein’s last manuscript On Certainty, concede
that knowledge and justification depend on the ability to respond to challenges,
but they insist that the question of which challenges are appropriate and need to be
responded to is context-dependent. In philosophical or skeptical contexts in which
ineliminable skeptical error-possibilities are under discussion, I know nothing about
the external world because I cannot show that I am not a brain in a vat. However,
in ordinary contexts, in which the challenge that I may be a brain in a vat is not
appropriate, I do have knowledge about the external world because I can defend
myself against the challenges that are appropriate in these contexts.
Both options deny one of the premises of the reconstructed Agrippan argument.
Externalist/reliabilists deny D1. They deny that justification is a matter of being
able to respond to challenges. This is a very popular position nowadays, accepted
not only by externalists but also by many internalists who typically think that non-
doxastic experiences can make beliefs justified. There are far fewer Wittgensteinian
contextualists who accept D1 but deny D2. However, both options avoid the
skeptical conclusion arising from Agrippa’s modes.
What should a neo-Pyrrhonist say to such New Epistemologists? According to
Fogelin (2004: 166–170), she should point out that their doctrines will leave you
576 SKEPTICISM
dissatisfied if you are looking for good reasons for believing that you are not a brain
in a vat. Saying that you may still know, because your beliefs are reliably connected
to the external world, does not give you what you are looking for: good reasons for
believing that your beliefs are so connected, and that you are not a brain in a vat.
Thus the alleged consolation that the externalist/reliabilist can give simply changes
the subject.
The contextualist’s consolation is also beside the point. When you are looking for
reasons to deny the skeptical hypothesis, you are actually in a philosophical context
in which, the contextualist concedes, you know nothing about the external world.
It does not console you to be told that knowledge attributions made in ordinary
contexts may still be true. You are now in a philosophical context in which you
know nothing.
Fogelin admits that he has interpreted the New Epistemologists as if they are
continuing the traditional epistemological project of trying to show that we have
knowledge without falling into the modes of circularity, infinite regress, and
arbitrary hypothesis. However, there is no reason to assume that this is what they try
to do. They do, or at least should, realize that this is a project that cannot succeed
for simple logical reasons, as Ernest Sosa (2009: 159) points out:
It is impossible to attain a legitimating account of absolutely all one’s own
knowledge; such an account admits only justifications provided by inference or
argument and, since it rules out circular or endlessly regressive inferences, such
an account must stop with premises that it supposes or “presupposes” that one is
justified in accepting, without explaining how one is justified in accepting them
in turn.
It should be clear by now that it is impossible to show that we have knowledge (or
justified beliefs) without circularity, infinite regress, or assuming some knowledge (or
justified beliefs). No epistemologist should seriously pursue the end of legitimating
all our knowledge at once.
As his final point, Fogelin (2004: 170–171) concedes that New Epistemologists
who have given up the legitimating project are no longer the targets of a Pyrrhonian
attack, but only because they have themselves become neo-Pyrrhonists. This is a
very disappointing result. It means that virtually all epistemologists nowadays are
neo-Pyrrhonists. Why should we even bother to argue for the position? What is
worse, it cannot be right. Neo-Pyrrhonists are skeptics about philosophy, but most
epistemologists who reject the legitimating project are not such skeptics. Once the
legitimating project and either premise of the Agrippan argument are rejected, we
lack a philosophical argument for skepticism about philosophy or other matters.
9 HOW TO BE A NEO-PYRRHONIST
Fogelin argues that neo-Pyrrhonism results naturally from reflection on our ordinary
epistemic procedures. Because these procedures do not eliminate all possibilities of
error, it is always possible to raise the level of scrutiny by paying attention to new
error-possibilities. When we allow the level of scrutiny to rise without restriction,
NEO-PYRRHONISM 577
we will conclude that nothing is known. Fogelin understands theories of justification

as attempting to repair this fragility in our ordinary procedures. Unfortunately,
these theories fail because they cannot solve Agrippa’s problem. Therefore, we
have no choice but to settle for the fragility of ordinary knowledge attributions and
become neo-Pyrrhonists.
I have argued that Fogelin fails to show that contemporary theories of justification
cannot solve Agrippa’s problem. The quandary is that, as a neo-Pyrrhonist, Fogelin
may use only premises that advocates of those theories accept in order to show
that they have skeptical consequences. I have given two versions of the argument
on which Agrippa’s problem seems to be based: the structural and the dialectical.
In neither version are most contemporary epistemologists committed to the truth
of the premises of the argument. Indeed, it follows from their theories that one or
another of them is false. Hence, neo-Pyrrhonism appears to lose its motivation.
However, there is a more direct way of defending neo-Pyrrhonism. The neo-
Pyrrhonist may argue that the premises of Agrippan argument are intuitively
plausible, and that neo-Pyrrhonism provides the best resolution of the resulting
paradox. Only neo-Pyrrhonism both explains the intuitive plausibility of the
premises and avoids dogmatic skepticism.
It seems that the paradox does arise from the ordinary practice of knowledge
attribution. It is plausible to assume that the purpose of this practice is to share
information (Craig 1990). When I attribute knowledge to someone else, I pick her
out as an informant. If she says that p, I believe that p. However, I cannot believe her
and attribute knowledge to her unless I believe what she says is true. If I believe it is
true, there may be no need to ask for reasons, but if I am not convinced of the truth
of what she says, it is quite appropriate to ask her: “How do you know?” or “What
reasons do you have for believing so?” If she can answer my question by giving
reasons that I take to be true and to support her belief, I take her to be justified in
her belief and to know what she is saying. If she cannot do this, I will not take her
as an informant and will deny her knowledge. Thus, it seems that the practice of
knowledge attribution supports the view that knowledge and justification depend on
an ability to give reasons when properly challenged. Premise D1 appears to be true.
When I attribute knowledge to myself I typically volunteer myself as an informant
for other people. Because I want to convince them—to share my information with
them—I am prepared to give reasons when properly challenged. If I cannot do
this, I understand that I am not a good informant and am ready to take back my
knowledge self-attribution—indeed, I am ready to deny my knowledge. As my
audience grows larger and larger, it becomes less and less likely that I can convince
all of its members. Indeed, it is the point of the mode of discrepancy or disagreement
to bring it to mind that there are always some people that I cannot convince with
my reasons and must therefore give up my knowledge claim. Hence premise D2 also
appears to be true.
However, in ordinary contexts we are ready to attribute knowledge even though
we are aware that there are people whom the subject cannot convince. Thus, in one
context we are inclined to attribute knowledge that we deny in another context.
The semantic theory that aims to explain these apparently inconsistent attributions
578 SKEPTICISM
and denials is contextualism. According to the version of it that is relevant here, the
content and truth of knowledge attribution depend on the attributor’s intentions and
interests—especially on the intended audience. If this is true, we can easily explain
our intuitions about the Agrippan argument. If the intended audience is a skeptic,
we cannot defend our beliefs against all the challenges she raises. Consequently, the
premises and the conclusion of the argument are indeed true in a context in which
we are trying to convince a persistent skeptic. However, knowledge attributions
are often true in ordinary contexts, because either the intended audience has no
challenges8 or the subject can answer them.
This contextualist resolution of Agrippa’s paradox closely resembles Fogelin’s
neo-Pyrrhonism: such a neo-Pyrrhonist is a skeptic about philosophy. There are so
many persistent disagreements in philosophy that knowledge attribution will almost
always be false in philosophical contexts. Furthermore, one of the main criticisms
of contextualism has been that it has no epistemological consequences. It does not
follow from the fact that people often speak the truth when they use sentences of
the form “S knows that p” in ordinary contexts that anybody knows anything (Sosa
2000). This is no problem for the neo-Pyrrhonists. It is rather the central feature
of their position that they go on making knowledge attributions about everyday
matters while suspending judgment about the traditional epistemological question
concerning the scope of knowledge.
An alternative way of explaining the plausibility of the Agrippan argument is to
assume that it is based on confusion between knowledge and assertion (Rescorla
2009; Gerken 2012; cf. Turri 2012). One might concede that a warranted assertion
requires defense against challenges, but it can be denied that knowledge requires this:
you may very well know something that you cannot defend and thus properly assert.
Agrippa’s problem concerns only what you may assert and is therefore epistemically
harmless. However, this view still provides a version of neo-Pyrrhonism: its target is
assertion rather than knowledge. Indeed, Duncan Pritchard (2005: 213–220) argues
that this is the right way of understanding ancient Pyrrhonism. A Pyrrhonist may
very well have beliefs and thus knowledge, but she makes no dogmatic assertions.
Of course, contextualized neo-Pyrrhonism may even allow assertions in contexts
where there are no appropriate challenges.
There are thus two initially quite plausible versions of neo-Pyrrhonism that both
seem to accord with practices of knowledge attribution. The choice between them
depends on whether we should prefer semantic or pragmatic resolutions of Agrippa’s
problem and other skeptical paradoxes, which is a big question in contemporary
epistemology and philosophy of language.9
NOTES
1. Fogelin seems to accept some sort of relevant alternative account here. Cf. Lewis
(1996: 551): “S knows that P iff S’s evidence eliminated every possibility in which
not-P—Psst!—except for those possibilities that we are properly ignoring.”
2. As Juan Comesaña (2005) points out, these judgments are still normative, being
attributions of justification.
NEO-PYRRHONISM 579
3. According to holistic coherentism, justification does not transfer from one belief
to another. It emerges from the whole system of beliefs when there are enough
crisscrossing inferential relations among them. It has been noted that this view makes
every justified belief a basic belief. So holistic coherentism is a form of foundationalism,
which Sosa (1991c: 180) calls formal foundationalism.
4. See Chisholm (1982: 61–75) on the problem of the criterion. Chisholm insists,
plausibly, that we cannot solve the problem of the criterion without begging the
question against skepticism.
5. Chisholm (1982: 75) acknowledges that “we can deal with the problem only by
begging the question” and insists that this is exactly what we should do. See also Sosa
(2009: 176).
6. Chisholm (1982: 68–69) calls this strategy particularism. It seems that a particularist
should think that there are basic epistemic beliefs about particular cases of justified
beliefs. See, for example, Sosa (1991b: 158) and Lemos (2007: 160).
7. At least foundationalists such as Bergmann (2005: 426) and Feldman (2006: 232–233)
accept that not only the denial of reliability but also the conscious suspension of judgment
about it would defeat one’s justification.
8. There are, of course, other conditions that must be satisfied if knowledge attributions
are to be true, such as truth, reliability, and perhaps safety.
9. I wish to thank Raul Hakli, Jaakko Hirvelä, Antti Kylänpää, Arto Laitinen, Luca
Moretti, and Tommaso Piazza as well as the editors of this book, Diego Machuca and
Baron Reed, for comments and discussion.
REFERENCES
Alston, William P. 1989. Epistemic Justification: Essays in the Theory of Knowledge. Ithaca:
Cornell University Press.
Barnes, Jonathan. 1997. “The Beliefs of a Pyrrhonist.” In M. Burnyeat and M. Frede (eds.),
The Original Sceptics: A Controversy, 58–91. Indianapolis: Hackett.
Bergmann, Michael. 2004. “What’s not Wrong with Foundationalism,” Philosophy and
Bergmann, Michael. 2005. “Defeaters and Higher-Level Requirements,” The Philosophical
Chisholm, Roderick. 1982. “The Problem of the Criterion.” In his The Foundations of
Knowing, 61–75. Brighton: The Harvester Press.
Comesaña, Juan. 2005. “Review of Pyrrhonian Skepticism,” Notre Dame Philosophical
Reviews 2005.06.8.
Craig, Edward. 1990. Knowledge and the State of Nature. Oxford: Clarendon Press.
Feldman, Richard. 2006. “Epistemological Puzzles About Disagreement.” In S. Hetherington
(ed.), Epistemology Futures, 216–236. Oxford: Clarendon Press.
Fogelin, Robert. 1994. Pyrrhonian Reflections on Knowledge and Justification. New York:
Fogelin, Robert. 1997. “Précis of Pyrrhonian Reflections on Knowledge and Justification.”
Philosophy and Phenomenological Research 57: 395–400.
580 SKEPTICISM
Fogelin, Robert. 2004. “The Skeptics are Coming! The Skeptics are Coming.” In W. Sinnott-
Armstrong (ed.), Pyrrhonian Skepticism, 161–173. Oxford: Oxford University Press.
Gerken, Mikkel. 2012. “Discursive Justification and Skepticism,” Synthese 189: 373–394.
Principles of Morals, (ed.), L. A. Selby-Bigge, rev. P. H. Nidditch. Oxford: Oxford
University Press.
Klein, Peter. 2004. “What Is Wrong with Foundationalism Is that It Cannot Solve the
Epistemic Regress Problem,” Philosophy and Phenomenological Research 68: 166–171.
Klein, Peter. 2005. “Infinitism is the Solution to the Regress Problem.” In M. Steup and
E. Sosa (eds.), Contemporary Debates in Epistemology, 131–140. Oxford: Blackwell.
Lemos, Noah. 2007. An Introduction to the Theory of Knowledge. Cambridge: Cambridge
University Press.
Lewis, David. 1996. “Elusive Knowledge,” The Australasian Journal of Philosophy 74:
549–567.
Pritchard, Duncan. 2005. Epistemic Luck. Oxford: Oxford University Press.
Rescorla, Michael. 2009. “Epistemic and Dialectical Regress,” Australasian Journal of
Sextus Empiricus. 2000. Outlines of Scepticism, translated by J. Annas and J. Barnes. 2nd
ed. Cambridge: Cambridge University Press.
Sinnott-Armstrong, Walter. 2004a. “Introduction to Pyrrhonian Skepticism.” In W. Sinnott-
Armstrong (ed.), Pyrrhonian Skepticism, 3–9. New York: Oxford University Press.
Sinnott-Armstrong, Walter. 2004b. “Classy Pyrrhonism.” In W. Sinnott-Armstrong (ed.),
Pyrrhonian Skepticism, 188–207. New York: Oxford University Press.
Sosa, Ernest. 1991a. Knowledge in Perspective: Selected Essays in Epistemology. Cambridge:
Sosa, Ernest. 1991b. “The Foundations of Foundationalism.” In his Knowledge in
Perspective: Selected Essays in Epistemology, 149–162. Cambridge: Cambridge
University Press.
Sosa, Ernest. 1991c. “The Raft and the Pyramid: Coherence Versus Foundations In The
Theory of Knowledge.” In his Knowledge in Perspective: Selected Essays in Epistemology,
Sosa, Ernest. 2009. Reflective Knowledge: Apt Belief and Reflective Knowledge, Volume II.
Turri, John. 2012. “Pyrrhonian Skepticism Meets Speech-Act Theory,” International
Journal for the Study of Skepticism 2: 83–98.
Williams, Michael. 1999. “Fogelin’s Neo-Pyrrhonism,” International Journal of
Philosophical Studies 7: 141–158.
Williams, Michael. 2001. Problems of Knowledge: A Critical Introduction to Epistemology.
CHAPTER FORTY
Disagreement
and Skepticism
BRYAN FRANCES
1 INTRODUCTION
Odds are, you have some beliefs that you know full well to be fairly controversial.
Perhaps you think God exists, we have free will, socialism is better than communism,
affirmative action is unjust, and global warming is going to be catastrophic. Or think
of beliefs that aren’t out there in the public eye but are controversial none the
less: your belief that father had an affair when you and your sister were ten years
old (you believe it, your sister denies it), or your opinion that your son lied about
stealing from the corner store (your spouse thinks he’s innocent, you disagree). In
each case, you are well aware that there are people who disagree with you—people
who aren’t foolish or evidentially challenged or any more biased than you are. And
yet you think you’re right and they’re wrong; and you’re holding fast to your view
even though you know that the people who disagree with you have heard your side
of the story, have understood it, and still think you’re wrong. You’re aware that your
reasons, the ones that convinced you of the belief in question, have not impressed
them enough to agree with you.
In such a scenario, your belief may well be the true one; good for you. But
will it amount to knowledge? Will it be wise for you to keep your belief in those
circumstances? The disagreement skeptic thinks that awareness of disagreement
provides one with a reason to think one’s belief is false—a reason often strong
enough to make one’s epistemic position impoverished in the event that one retains
one’s belief in the face of recognized disagreement. In what follows I address these
questions, starting with a comparison of traditional skepticism with disagreement
skepticism, moving on to some inadequate skeptical arguments that turn on
disagreement, and then closing with a presentation and analysis of a superior
disagreement argument for skepticism.
582 SKEPTICISM
2 TRADITIONAL SKEPTICISM VERSUS

DISAGREEMENT SKEPTICISM
Disagreement skepticism is importantly different from traditional skepticism, not
only with regard to the supporting argument, which will be made obvious below,
but with regard to what the conclusion comes to. To see this, ask yourself how much
we should worry about whether traditional philosophical skepticism is true.
Consider the following standard argument for one traditional kind of
philosophical skepticism:
1. If a person knows that she has hands, then she can know that she wasn’t
kidnapped by Santa and unwittingly made into a handless brain in a
vat (BIV).
2. If she can know that she wasn’t kidnapped by Santa and unwittingly made
into a handless BIV, then she must have evidence, broadly construed, that
rules out that BIV scenario.
3. But she doesn’t have any such evidence.
4. Thus, she doesn’t know that she has hands.
There are just five possibilities with respect to this argument.
Premise (1) is false. She knows she has hands but can’t know that she isn’t a
handless BIV via Santa’s lunacy. That’s pretty odd: if she knows she has hands, then
why can’t she then deduce that she isn’t a handless BIV?
Premise (2) is false. Our subject can know she isn’t a Santa-induced handless BIV
but she doesn’t have evidence that rules out that BIV possibility. That’s odd, too:
how can someone know that X doesn’t obtain unless she has some evidence that
rules out X’s obtaining?
Premise (3) is false. She has some evidence that rules out the BIV scenario.
That’s pretty odd, too. For instance, it seems clear that there’s nothing she can do
to rule it out, like checking her hands carefully, pinching herself, or researching
Santa Claus. Anything she can point to as evidence could easily be part of
Santa’s deception.
A fourth possibility is that the claims in the argument are ambiguous, different
disambiguations have different truth-values, and different disambiguations will be
“chosen” depending on subtle contextual factors (of the subject, the person asserting
(1)–(4), or the person evaluating those assertions).
The fifth possibility is that the argument (1)–(4) is sound; this is a type of
traditional philosophical skepticism.
What is interesting, and what makes much of this book worthwhile, is the
fact that, no matter which of the five theories is right about (1)–(4), the correct
one says something interesting about knowledge, evidence, and perhaps related
epistemic notions.
But once you understand the reasoning behind the traditional skeptical thesis,
that thesis gets robbed of a good portion of its oddness. All the skeptic is saying
with her (1)–(4) is that the notion knowledge is a lot more peculiar than we thought
DISAGREEMENT AND SKEPTICISM 583
it was: she is not saying that our ordinary beliefs are false or even that our overall
evidence goes strongly against our ordinary beliefs. With some more detail:
(a) The traditional skeptic is not saying that your belief B (e.g., “I have hands”)
is false.
(b) She is not saying that your overall evidence is strongly against B (so much so
that you should disbelieve B).
(c) She admits that everything non-skeptics might point to as evidence for B
really does exist and is at least prudentially relevant to your holding B.
(d) She admits that you haven’t overlooked some new and powerful evidence
against your belief.
(e) She just says that technically speaking, all that stuff non-skeptics might
call evidence for your belief B doesn’t actually support B enough to make
it knowledge or justified: that “evidence” (e.g., your sensory experiences)
provides much weaker support than non-skeptics think it does. That is,
according to her, non-skeptics have not so much missed out on some
powerful evidence against B (that’s (d)) as grossly inflated the strength of the
so-called evidence for B that you possess (that’s (e)).
(f) According to her, it is simply impossible to discover whether B is true; so
you might as well go on acting the way you did before, when you believed B.
(g) Her skepticism is limited: it doesn’t apply to your beliefs about your own
mind or about mathematics.
(h) Contextualism, pragmatic encroachment, relevant alternatives theory,
reliabilism, and Gricean theory each supply prima facie serious objections
to her skeptical argument (see Goldman 2011 and Rysiew 2011 for
explanations of these ideas).
This just doesn’t strike me as a terribly worrisome kind of skepticism. Even if I become
convinced that this traditional kind of skepticism is true, I won’t be worried that
your belief in your hands is false in any way that really matters. Suppose I believe
that global warming is going to raise sea levels on the east coast of the United States
by over eight feet by 2100. The disagreement skeptic argues that I know no such
thing, as there is a real chance that the scientists who say the increase will be only
three feet are right. The radical skeptic says I know no such thing because there is a
chance that there is no earth. It’s even tempting—we see undergraduates do this all
the time—to think that, even if the skeptic is right, your belief in your hands is true
“in your reality,” in the reality you are epistemically confined to. Sure, your belief
might not be true “in the grand scheme of things,” from “the God’s eye point of
view of real truth,” but that fact only matters to philosophers.
Of course, the bald claim, “You know virtually nothing,” sounds awful and
shocking, but as soon as one understands the view behind the claim—claims (a)–
(f) in particular—one is inclined to think that all that has really been said is that
“knowledge” fails to latch on to what matters in our doxastic lives. When someone
says, “You think you know P, but you don’t,” it sounds as though she’s saying
584 SKEPTICISM
that P is false, or what you claim as evidence doesn’t exist. But the radical skeptic
isn’t saying P is false or that your so-called evidence doesn’t exist. For comparison,
imagine watching some children play a baseball game. Then someone says “No,
there is no baseball game there.” At first, you would find this shocking, but once you
learn what she is really saying—the game being played isn’t really baseball because
they are allowing four strikes for an out, whereas in baseball one gets just three
strikes—you no longer find it at all shocking, merely peculiar even if true.
Part of what makes disagreement skepticism so interesting is that (d)–(h) are
false for it. The phenomenon of disagreement supplies a kind of skepticism worth
worrying about because it generates an argument for the conclusion that many
of your most precious beliefs are false in a way that can make a difference. The
traditional skeptic says that, even if your belief is false, no one will ever know that
it’s false and there isn’t a damn thing anyone can do about this lack of knowledge.
The disagreement skeptic says none of that. This is a much more worrisome kind
of skepticism.
In what follows, I examine the cases of disagreement that generate what I take to
be the best arguments for skeptical conclusions that have some bite in the sense that
they generate skepticism for a large portion of our actual beliefs—especially ones
we really care about.
3 WHEN DISAGREEMENT DOES NOT

GENERATE SKEPTICISM
Clearly, the mere fact that you know of a bunch of people who disagree with you
does not preclude your knowing that they are wrong and you are right. A group of
children might think that so-and-so is the most famous person alive, but you could
easily know that they are wrong. Disagreement generates a worthwhile skeptical
argument only when the disagreement meets demanding conditions. However, it
isn’t easy to come up with an impressive skeptical argument based on disagreement.1
We begin with some stipulations. Suppose you’re faced with the question,
“Is belief B true?” You have your view on the matter: you think B is true. If you
are convinced that a certain person is clearly lacking compared to you on many
epistemic factors when it comes to answering the question “Is B true?”—factors
such as general intelligence, relevant evidence, time spent thinking the evidence
through, amount of distractions encountered while thinking about the evidence,
and background knowledge—then you’ll probably say that you are more likely than
she is to answer the question correctly (provided there are no other factors that give
her an advantage over you). If you are convinced that a certain person definitely
surpasses you on many of those factors when it comes to answering “Is B true?,”
then you’ll probably say that you are less likely than she is to answer the question
correctly (provided you think there are no factors that give you an advantage over
her). In general, you can make judgments about how likely someone is compared
to you when it comes to answering “Is B true?” correctly. If you think she is more
likely (e.g., the odds that she will answer it correctly are 90%, whereas your odds
are just 80%), then you think she is your epistemic superior on that question; if you
think she is less likely, then you think she is your epistemic inferior on that question;
if you think she is about equally likely, then you think she is your epistemic peer on
that question.
With those notions in mind, here is a simple rule to start us off looking at skeptical
arguments from disagreement:
Rule 1
If before the discovery of disagreement you think that a certain person is your
epistemic peer/superior on the question whether belief B is true, then if upon
realizing that the two of you disagree you retain your belief B, your belief B won’t
be reasonable.
The reasoning behind Rule 1 in the case of peers is the following. Prior to the discovery
of disagreement, you think that the two of you are equally likely to judge B correctly.
You probably think there is about a 90 percent, say, chance that you’ll be right and,
thus, a 90 percent chance your peer will be right. When you find out she believes
not-B while you believe B, you can’t conclude that there is a 90 percent chance not-B
is true and a 90 percent chance B is true. Such a conclusion is incoherent. Instead, you
have to conclude that there’s just a 50 percent she’s right and a 50 percent chance
you’re right. So you’re saying to yourself that there’s just a 50 percent chance B is
true. So you should suspend judgment on B. The reasoning in the case of superiors is
similar. In both cases, peers and superiors, if you don’t give up your belief B, then it
won’t be reasonable. And if it’s not reasonable, then it won’t amount to knowledge.
By that reasoning, if you discovered you had disagreeing peers or superiors on a
great many of your beliefs, then a great many of your beliefs would fail to amount
to knowledge, even if they were true—or so says Rule 1. That would be a limited
but significant kind of skepticism.
But many cases show that Rule 1 is dubious. Suppose Viktoriya is very confident
that B is true, where B is “Smoking causes cancer.” Viktoriya is highly confident B
is true because she is well aware that virtually all of the many experts on the topic
have long thought that it’s true. She has excellent testimonial evidence backing up
her belief in B. She knows, of course, that there are many people who verbally deny
the smoking-cancer connection, but she thinks those people are either ignorant,
foolish, lying, or caught in self-deception. Viktoriya is also convinced that Willard
is her peer on knowing whether B is true. The basis for this belief in peerhood isn’t
anything fancy or involved: she just supposes that Willard is about as intelligent and
scientifically literate as she is, since he is a software engineer like she is and he seems
generally knowledgeable about many scientific things in the news (e.g., he knew
about the latest mission to Mars). Then she finds out that Willard genuinely thinks
smoking doesn’t cause cancer.
By my lights, when Viktoriya learns of Willard’s opinion, she will have acquired
little or no reason to give up her belief that smoking causes cancer. When she learned
of Willard’s opinion she probably thought to herself, “Oh Jeez. He is one of those
people, ones who have been brainwashed by the tobacco companies and their shills.”
586 SKEPTICISM
She already knew about “those people”; actually running into one won’t matter to
her overall evidence regarding smoking and cancer. And even if she didn’t already
know of those people, she did know that for almost any topic there’s almost never
100 percent agreement among intelligent and informed people. So Rule 1 looks
dubious. If we are going to find a powerful skeptical argument from disagreement,
we will need a better rule to use.
Things don’t change if the person you disagree with is a recognized superior
on B. Even if Viktoriya thought that Willard was her superior on the issue (e.g.,
he is a doctor or a medical researcher), she would be reasonable in keeping B. She
would just recall or perhaps conclude that even the experts are rarely 100 percent in
agreement. She might think to herself “He probably thinks he knows something all
the many thousands of other experts missed. But the odds are really against that.”
She might conclude that Willard must be biased or somehow emotionally invested
in denying the smoking-cancer connection—or maybe she will just think to herself,
“Well, something is definitely going on with that guy on this topic” (so she is open to
the possibility that he isn’t saying what he appears to be saying). She will implicitly
think that even though he’s smarter than her, knows all about her good testimonial
evidence for the smoking-cancer connection, has thought about the issue a lot more
than she has, etc., there must be some explanation of why he went wrong.
One thing we can do in order to find a better skeptical rule is focus on groups
instead of individual people:
Rule 2
If you think that the group you’re disagreeing with is in an equal/better position
than your group (the group of people who, like you, believe B) to judge belief B
correctly, then retaining B will be unreasonable and B won’t amount to knowledge.
This rule is better than Rule 1 because it better captures the worrisome disagreements
we find ourselves in: often the most intriguing disagreements hold among very large
groups of people, some of whom are experts. Focusing on one-on-one disagreements,
as Rule 1 does, is useful mainly for theorizing about disagreement, not for applying
to real-life disagreements.
The smoking-cancer case isn’t a counterexample to Rule 2 because Viktoriya
did not think that, when it came to judging B, the group of deniers of B was in an
equal or better position than the group of affirmers of B. Indeed, it was precisely
because her group contains almost all the relevant experts that she was reasonable
in sticking with B.
However, Rule 2 is doubtful as well. Suppose in a college class the students
break into two groups of ten students in order to independently investigate some
complicated matter. Student Stu thinks the groups are evenly matched, based on his
modest knowledge of their abilities (pretend this is a small college and Stu knows a
great many of his fellow students). After a week of work, representatives of the two
groups give oral progress reports to the whole class. The representative from the
other group says that her group has figured out several things regarding the topic of
investigation (college resources), including the idea that there are about 113 student
computers in the college’s main library. But this strikes Stu as way too low: he was
100 percent confident in his belief B “The number of computers in the main library
is at least 140.” He briefly wonders whether the student in question is accurately
presenting the verdict of her group, but when he sees the members of her group
nodding their heads he knows that his group definitely disagrees with their group.
Stu is extremely confident the number is quite a bit more than 113, as he and his
group members have scoured the library to count the computers. He keeps his belief
B and starts to doubt whether he was right to think the other group had done about
as much work on the project as his group did.
We can use the college student case to cast doubt on the “superior” part of Rule
2 as well. Suppose Stu started out, before hearing the other group’s view on B, with
mildly good evidence that the other group was mildly better positioned to answer
“How many student computers are in the main library?” Suppose further that he
learns that, although the other group believes not-B, they do not do so with extreme
conviction: they believe it, but not vehemently. For what it’s worth, in my judgment
he is reasonable if he retains his belief B and just concludes that, despite their mild
superiority, the other group just got things wrong this time around. Stu knows they
were more likely than his group to judge B correctly, but that doesn’t mean that they
were guaranteed to judge B correctly. The main thing that secures his reasonability is
his extremely powerful overall evidence for B coupled with his comparatively mild
evidence for the other group’s mild superiority and mild belief in not-B.
Considerations like the ones above show that it’s no small matter to construct a
skeptical argument that centers on disagreement. So what are the troublesome cases,
the ones that do generate a worrisome kind of skepticism?
4 CONTROVERSY AND KNOWLEDGE

When one knows that one’s belief is highly controversial, then one has available
a good if potentially defeasible reason for serious doubt. However, once again it’s
tricky to formulate the skeptical argument, as there may be flaws lying in wait.2
For the sake of argument, I will assume that controversial beliefs very often start
out epistemically rational and even overall justified (in any of several senses of those
terms). Roughly put, the disagreement skeptic thinks that even if a controversial
belief starts out as knowledge, once one appreciates the controversy, one’s belief
will no longer amount to knowledge. However, just because someone learns that his
belief is highly controversial doesn’t guarantee that his belief will no longer amount
to knowledge. For instance, he might be the only living soul who knows that Oswald
assassinated Kennedy, as he is Oswald. So we will have to work harder to develop a
controversy-based skeptical argument.
In ordinary life, the cases of disagreement that generate powerful arguments for
types of skepticism that have real bite typically satisfy the following conditions:
You know that the belief B in question has been investigated and debated (i)
for a very long time by (ii) a great many (iii) very smart people who (iv) are
your epistemic superiors on the matter and (v) have worked very hard (vi) under
588 SKEPTICISM
optimal circumstances to figure out if B is true. But you also know that (vii) these
experts have not come to any significant agreement on B and (viii) those who
agree with you are not, as a group, in a better position to judge B than those who
disagree with you.
We typically come to know (i)–(viii) through testimony; the truths known are just
brute facts about society that we are well aware of. For instance, I might have some
opinion regarding free will or capital punishment or affirmative action or spiritual
experience or the causes of the Second World War. I know full well that these
matters have been debated by an enormous number of really smart people for a very
long time—in some cases, for centuries. I also know that I’m no expert on any of
these topics. I also know that there are genuine experts on those topics—at least,
they have thought about those topics much longer than I have, with a great deal
more awareness of relevant considerations, etc. It’s no contest: I know I’m just an
amateur compared to them. Part of being reflective is coming to know about your
comparative epistemic status on controversial subjects.
The person who knows (i)–(viii) is robbed of the reasonableness of several
comforting responses to the discovery of disagreement. If she is reasonable, then
she realizes that she can’t make, at least with high confidence, anything like the
following remarks:
(a) Well, the people who agree with me are smarter than the people who
disagree with me.
(b) We have crucial evidence they don’t have.
(c) We have studied the key issue a great deal more than they have.
(d) They are a lot more biased than we are.
This phenomenon is particularly prevalent with regard to religion, politics, morality,
and philosophy. To see why, recall college student Stu. When he learned that the
other group disagreed with him, he could reasonably say to himself, “Well, I guess
they didn’t do as much work on the issue as I thought” or “Maybe they aren’t as
smart as I thought” or “Perhaps they just screwed up the count of the computers
or made some other slip, even though they’re just as generally smart and thorough
as we are.” But when it comes to debates about free will, capital punishment,
affirmative action, and many other standard controversial topics, a great many of
us—not all of us, but many—either know that such responses are false or, if we do
embrace them, we do so irrationally. If when it comes to the question of whether
we have free will you say to yourself regarding the experts who disagree with you,
“Those people just don’t understand the issues,” “They aren’t very smart,” “They
haven’t thought about it much,” etc., then you are doing so irrationally in the sense
that you should know better than to say that, at least if you’re honest with yourself.
Now for the disagreement skeptical argument that centers on controversial belief:
(1) Suppose you believe B.
(2) Then you become aware of the controversial nature of B: you know (i)–(viii)
from before.
(3) If (2) is true of you, then your retained belief in B is epistemically second-rate
in a significant manner—and this is true even if B is true, you previously (before
the discovery of disagreement) had a propositionally and doxastically justified
belief in B, and your belief was not Gettiered.
(4) Thus, for any belief that makes (2) true, that belief will be epistemically
second-rate in a significant manner.
This argument concludes that even if your belief B is optimal in key epistemic
respects (true, previously propositionally and doxastically justified, not Gettiered),
it still faces a hurdle in order to attain high levels of epistemic goodness—a hurdle
it does not clear.
The argument leaves open what the conclusion amounts to. By “epistemically
second-rate,” it might be meant that the controversial belief fails to amount to
knowledge. Then again, maybe it fails to be doxastically and/or propositionally
justified. Perhaps both. Then again, it might mean that although the retained belief
is justified and amounts to knowledge, it still suffers from a serious epistemic defect.
Perhaps we can be unwise in keeping a belief even when it amounts to knowledge;
this might be true if the standard for knowledge is significantly lower than most
epistemologists have thought.
5 TWO OBJECTIONS TO THE DISAGREEMENT

SKEPTICAL ARGUMENT
First, the skeptical argument appears to have a severely limited scope, as there are
quite a few conditions in (i)–(viii)! For instance, (viii) will be denied in many religious
disputes, thereby falsifying (2): people tend to think that they have evidence that the
other side doesn’t have, or that the other side suffers from some serious epistemic
mistake that their side has avoided. Theists tend to think that non-theists are just
missing out on experiences of God; non-theists tend to think that theists either don’t
understand science sufficiently well or have been unable to get over their childhood
indoctrination. If a disputant has come to this comparative opinion in a justified way,
then the argument from disagreement doesn’t touch him, even if this comparative
opinion of his happens to be false. And if he has come to his comparative opinion
without justification, then although he has made an epistemic error (by hypothesis),
it’s arguable that it’s epistemically rational for him to retain his religious view, pro or
con, in the face of disagreement given the beliefs he has, since he thinks the people who
disagree with his group are definitely epistemic inferiors on the topics in question.
Setting aside premise (2), the argument is vulnerable to objections to (3). Pretend
that Pat thinks little boys are naturally more violent than little girls. He thinks this
because he has run a day care for little kids for twenty-five years, and he has talked
about the issue with many other experienced daycare workers over many years. He
is not, however, a child psychologist or anything related; neither has he read any
studies on the issue. Suppose further that he’s exactly right: little boys are naturally
more violent than little girls, no matter how you construe “natural,” “little,”
“violent,” “boy,” or “girl.” Assume that the experts on child behavior—at least,
590 SKEPTICISM
the ones doing the research with PhDs—have for a very long time been stubbornly
divided on the question whether little boys are naturally more violent than little
girls. They are well aware of the observational evidence that has suggested that boys
are naturally more violent—they know all about the kinds of evidence Pat has—but
for a couple decades now they have had studies that collectively strongly suggest the
exact opposite. In this story, the question of natural violence and gender has long
been one of the outstanding questions in child psychology.
Might Pat’s belief amount to knowledge provided: he obtained and holds the
belief on the basis of his experiences, experiences that gave him plenty of justification
for his true belief; the evidence the experts have that suggests otherwise is unknown
to him; and although he is vaguely aware that there have been some studies that go
against his view, he just shrugged his shoulders and never really dwelt on this fact?
Would the epistemic sin involved in that shrug—if it is such—be so bad as to ruin
his chance at knowledge?
I don’t know. Perhaps it matters what knowledge comes to. Over the centuries
many philosophers have thought that knowledge is something pretty special,
something that has to be fought for with real cognitive power. Others have thought
that any two-year-old child has loads of knowledge: if she is minimally functioning,
then she will know an enormous number of mundane facts. It’s not entirely clear
that anyone is wrong in that debate. We use the words “know” and “knowledge”
quite a bit in life, but that doesn’t mean that our use has fastened on to an epistemic
joint in nature, as “water,” “gold,” and “shark” have marked objectively existing
joints in nature. Perhaps “knowledge” is polysemous, which would go some way
toward explaining why we can never agree on the analysis of “the” notion knowledge
(because there is no such singular notion).
In any case, the story involving Pat suggests that even when one is aware that (i)
one’s belief is controversial among the experts, (ii) one is an amateur in many respects
compared to them, and (iii) one has no evidence that they have missed, it isn’t clear
that this will mean that one’s belief fails to amount to knowledge. So (3) is contentious.
The disagreement skeptic may well protest that Pat was in an unusual situation:
he knew P due to extensive experience, the experts had generated large amounts of
misleading evidence, Pat had none of that misleading evidence, and, although Pat knew
the issue was controversial, he didn’t think about it much and it more or less slipped
from his consciousness. The skeptic’s point here is that even if Pat is in a position to
have knowledge, for the vast number of controversial beliefs this position is rare.
NOTES
1. For introductions to the epistemology of disagreement, see Christensen (2009) and
Frances (2010b; 2014).
2. Goldberg (2009), Frances (2010a; 2013), Fumerton (2010), Kornblith (2010),
Machuca (2013), and Christensen (2014) each address the epistemology of
controversial belief. Feldman (2007), Oppy (2010), Thune (2011), Kraft (2012), and
Frances (2015) each address the specific case of religious controversy. General issues
regarding the epistemology of disagreement are addressed in Feldman and Warfield
(2010) and Christensen and Lackey (2013).
REFERENCES
Christensen, David. 2009. “Disagreement as Evidence: The Epistemology of Controversy,”
Philosophy Compass 4: 756–767.
Christensen, David. 2014. “Disagreement and Public Controversy.” In J. Lackey (ed.),
Essays in Collective Epistemology, 142–163. Oxford: Oxford University Press.
Christensen, David and Jennifer Lackey (eds.). 2013. Disagreement: New Essays. Oxford:
Feldman, Richard. 2007. “Reasonable Religious Disagreements.” In L. Antony (ed.),
Philosophers Without Gods: Meditations on Atheism and the Secular Life, 194–214.
Feldman, Richard and Ted Warfield (eds.). 2010. Disagreement. Oxford: Oxford University
Press.
Frances, Bryan. 2010a. “The Reflective Epistemic Renegade,” Philosophy and
Frances, Bryan. 2010b. “Disagreement.” In S. Bernecker and D. Pritchard (eds.), Routledge
Companion to Epistemology, 68–74. New York and Abingdon: Routledge.
Frances, Bryan. 2013. “Philosophical Renegades.” In D. Christensen and J. Lackey (eds.),
Disagreement: New Essays, 121–166. Oxford: Oxford University Press.
Frances, Bryan. 2014. Disagreement. Cambridge, UK: Polity Press.
Frances, Bryan. 2015. “Religious Disagreement.” In G. Oppy (ed.), Routledge Handbook of
Contemporary Philosophy of Religion, 180–191. London: Routledge.
Fumerton, Richard. 2010. “You Can’t Trust a Philosopher.” In R. Feldman, and T. Warfield
(eds.), Disagreement, 91–110. Oxford: Oxford University Press.
Goldberg, Sandy. 2009. “Reliabilism in Philosophy,” Philosophical Studies 124: 105–117.
Goldman, Alvin. 2011. “Reliabilism.” In E. Zalta (ed.), The Stanford Encyclopedia of
Philosophy (Spring 2011 Edition), http://plato.stanford.edu/archives/spr2011/entries/
reliabilism/.
Kornblith, Hilary. 2010. “Belief in the Face of Controversy.” In R. Feldman and T. Warfield
(eds.), Disagreement, 29–52. Oxford: Oxford University Press.
Kraft, James. 2012. The Epistemology of Religious Disagreement: A Better Understanding.
New York: Palgrave-Macmillan.
Oppy, Graham. 2010. “Disagreement,” International Journal for Philosophy of Religion
68: 183–199.
Rysiew, Patrick. 2011. “Epistemic Contextualism.” In E. Zalta (ed.), The Stanford
Encyclopedia of Philosophy (Winter 2011 Edition), http://plato.stanford.edu/archives/
win2011/entries/contextualism-epistemology/.
Thune, Michael. 2011. “Religious Belief and the Epistemology of Disagreement,”
Philosophy Compass 6: 712–724.
CHAPTER FORTY-ONE
Skepticism and
the Internalism-
Externalism Debate
MATTHIAS STEUP
1 INTRODUCTION
Since the early 1980s, it has been common to classify theories of justification as
internalist or externalist.1 Advocates of internalism take justification, in some way
or another, to be internal; externalists deny that justification is internal in the way
internalists assert it is. What is at issue here is not only whether a belief’s being
justified but also whether a belief’s not being justified is internal. The term j-status
will serve to refer to either one of these statuses. Justification, or the lack of it,
is conferred upon a belief by j-determiners: the factors that fix a belief’s j-status.
Whatever internality is, it would seem that the internality of j-determiners is both
necessary and sufficient for the internality of j-status.
2 THREE THEORIES OF JUSTIFICATION

To discuss what the internality of justification is supposed to be, it will be helpful to
have at our disposal sample theories of justification that qualify as either internalist
or externalist. We will consider classical foundationalism and evidentialism as
examples of internalism, and reliabilism as an example of externalism.
According to classical foundationalism, a subject’s beliefs are divided into two
sets.2 The first comprises inferred beliefs about the external world, and the second
is made up of foundational beliefs about the subject’s perceptual experiences. All
beliefs in the first set are ultimately justified by suitable beliefs in the second set.
For example, an external world belief such as “This object is red” is inferentially
justified by the foundational belief “This object looks red.” The latter belief is non-
inferentially justified through a special cognitive relation such as acquaintance
or direct apprehension, which provides a kind of justification that does not itself
require any further justifying beliefs.
SKEPTICISM AND THE INTERNALISM-EXTERNALISM DEBATE 593
Evidentialism is the view that whether a subject’s belief is justified is determined

by the total body of the subject’s evidence.3 A subject’s evidence includes not
only beliefs about one’s perceptual experiences, but also perceptual experiences
themselves. As a consequence, what justifies the belief “This object is red” need not
be a belief such as “This object looks red to me” but can be the kind of experience a
subject describes when she says, “It looks to me as though the object is red.”
Reliabilism, in its most basic form, is the view that a belief is justified if and only
if it is reliably produced.4 In other words, a belief’s origin in a reliable source is both
necessary and sufficient for its status of being justified. Consider again the example
of believing that the object before you is red. Reliabilism says that your belief is
justified because color perception is a reliable process. So, according to reliabilism,
your belief’s justificational status is determined by the following two facts: (i) your
belief originates in color perception and (ii) color perception is reliable.
Why do classical foundationalism and evidentialism qualify as internalist, whereas
reliabilism is externalist? To answer this question, we need to examine the notion
of internality that is in play when epistemologists debate whether justification is
internal or external.
3 INTERNALITY
According to Roderick Chisholm, a main advocate of internalism in the second half
of the twentieth century to be internal is to be recognizable on reflection (Chisholm
1977: 17; 1989: 62). But what is the difference between being able, and not being
able, to know some proposition p through reflection? The key thought appears to be
that you can know p solely through reflection if you can come to know p on the basis
of what you already have to go on, that is, without having to engage in any further
observations. Suppose you wonder what your car’s license plate says. If you remember
it, reflecting on the question will provide you with an answer. If you forgot, you will
have to go and look at your license plate. Or suppose you ask yourself, after you have
had a headache all morning long, whether you still have it now. Presumably you need
not consult a neurologist to find out. Reflecting on the question will provide you with
an answer to it. But if you wonder whether your colleague Carl, who complained
earlier in the morning about a headache, is now feeling better, reflection will not
answer your question. You will have to ask him. Finally, consider the question whether
“~(p & q)” is equivalent to “~p & ~q.” If you arrive at the correct answer (which
would be “no”) by thinking about the truth conditions of negation and conjunction,
you have answered the question by reflecting on it. If you answer it by asking a logician
or looking it up in a textbook, you have not.
According to an alternative criterion of internality, what is internal is what is
located within one’s own cognitive perspective.5 What is located outside of one’s
perspective is external. The problem with this proposal is that of illuminating the
concept of one’s own cognitive perspective. If it turns out that what is within and
what is outside of one’s perspective is precisely what is, and what is not, recognizable
on reflection, then the cognitive perspective criterion is not a genuine alternative
to the recognizability-on-reflection criterion. According to yet another approach,
594 SKEPTICISM
referred to as mentalism, internalism is the view that j-status is determined by

mental states only (see Conee and Feldman 2001). As before, it is not clear that this
proposal qualifies as a genuine alternative. Clearly, not any brain state qualifies as a
mental state. Is it possible to explain how mental and non-mental brain states differ
without relying on a cognitive accessibility criterion like introspection? It is doubtful
that that is possible. We may reasonably suspect that what is definitive of the mental
is introspective access. If that suspicion is correct, then mentalism is not a genuine
competitor for the recognizability-on-reflection criterion.6 In what follows we will
proceed to work with the latter.
Having at our disposal a criterion of internality, the difference between internalist
and externalist theories can be defined as follows:
Internalism: T is an internalist theory of justification iff, if T is true, then
justification is always recognizable on reflection.
Externalism: T is an externalist theory of justification iff, if T is true, then
justification is not always recognizable on reflection.
In these definitions, the word “always” is essential.7 Without it, we would get the
uncontroversial claim that justification is sometimes recognizable on reflection.
This rather modest claim should not generate disagreement. Reliabilists need not,
and should not, advocate the rather implausible claim that justification can never
be recognized on reflection. Rather, as we shall see below, what is distinctive of
reliabilism is that justification is not always recognizable on reflection.
4 RELIABILISM AS A VERSION OF EXTERNALISM

If reliabilism is true, then, for two reasons, justification is not always recognizable
on reflection. First, there are possible cases in which the presence of reliabilist
justification—justification resulting from a belief’s originating in a reliable source—
is not recognizable on reflection. Second, there are scenarios in which the absence of
reliabilist justification is entirely hidden and thus not recognizable on reflection (nor
in any other way). An example of the former situation is Laurence BonJour’s case of
Norman the reliable clairvoyant (BonJour 1985: ch. 3). Suppose Norman acquired
the faculty of clairvoyance for no noticeable reason, say, because of imperceptible
cosmic radiation. Suppose further that Norman’s faculty of clairvoyance, which we
assume is as reliable as Norman’s other perceptual faculties, produces, for the very
first time after Norman acquired it, a belief, namely the belief that the president is
currently in New York. Since Norman’s clairvoyance is reliable, Norman’s belief is
justified. But through reflection, this is not recognizable to Norman. If he were to
reflect on his belief’s epistemic credentials, Norman would have to conclude that
his belief is not justified, for he has no reason to think, right then and there, (a) that
his belief was produced by a newly acquired faculty, namely, by clairvoyance, and
(b) that, contrary to what he would reasonably expect, clairvoyance is, at least in his
particular case, reliable.
Next, let us turn to the kind of case where the absence of reliabilist justification is
not recognizable on reflection.8 Consider your belief that you have hands. Classical
foundationalists, evidentialists, and reliabilists all agree that your belief is justified.
Moreover, they would (or at least should) agree that you can recognize on reflection
that your belief is justified. Now suppose you are a brain in a vat (a BIV).9 As a BIV,
your belief that you have hands is of course false. Moreover, the belief’s source is
unreliable since nearly all beliefs originating from this source are false. Reliabilists
would therefore have to say that, if you are a BIV, your belief that you have hands
is unjustified. But if, as a BIV, you were to consider your belief’s j-status, you would
arrive—and reasonably so—at the conclusion that you are entirely justified in
believing that you have hands. You would arrive at this conclusion because your
being a BIV is, ex hypothesi, undetectable. You do not, therefore, have any reason
to suspect that your belief originates in an unreliable faculty. Hence there is no way
you could, as a BIV, recognize, on reflection or otherwise, that you are not justified
in believing that you have hands.
To sum up, if reliabilism is true, neither the presence nor the absence of
justification is always recognizable on reflection. That is why reliabilism is an
externalist theory.
5 CLASSICAL FOUNDATIONALISM AND

EVIDENTIALISM AS VERSIONS OF INTERNALISM
According to internalist theories such as classical foundationalism and evidentialism,
j-status is determined through factors that can be recognized through introspection
and a priori intuition, factors such as how the external world appears to one in
one’s experiences and what support relations obtain between the content of one’s
experiences and one’s beliefs about the external world. It is plausible think that
such j-determiners are frequently, perhaps even typically, recognizable on reflection.
Why think they always are? It is helpful to approach this question by considering
the two cases in which, if reliabilism is true, justification is not recognizable on
reflection. First, consider Boris, who unfortunately has been envatted. According to
reliabilism, Boris’s external world beliefs are not justified because, since Boris is a
BIV, his perceptual processes are unreliable. But their unreliability is undetectable to
a BIV. That’s why, if reliabilism is true, it is undetectable to Boris that his external
world beliefs are unjustified. If evidentialism is true, we get a different outcome.
Boris’s external world beliefs are then (typically) justified, and Boris can tell on
reflection that they are justified. This is so for the following reasons. First, according
to evidentialism, the unreliability of a belief’s origin is not a necessary condition
of the belief’s being justified. Second, Boris has plenty of evidence for his external
world beliefs, namely, the kind of perceptual experiences BIVs and normal people
share. Third, Boris can tell on reflection that he has such experiences and that
they support his external world beliefs. A similar line of reasoning will support
the conclusion that, according to classical foundationalism, Boris can recognize on
reflection the justificational status of his external world beliefs.10
Next, consider Norman the reliable clairvoyant. According to reliabilism,
Norman’s belief about the president’s location is justified because it is reliably
596 SKEPTICISM
produced. But since Norman has no reason to think that his belief has its origin in
reliable clairvoyance, he cannot tell on reflection that his belief is justified. In contrast,
according to classical foundationalism and experiential evidentialism, Norman’s
belief about the president’s location is not justified. And since introspection and a
priori reasoning will reveal to Norman that he has no support for his belief, he can
tell on reflection, according to these theories, that his belief is not justified.
Such goes, in any case, the traditional story of how internalist justification
works. J-status is always recognizable on reflection because the j-determiners of our
external worlds beliefs are transparent to the mind’s eye. Through introspection
we can tell what our perceptual experiences are, and through a priori intuition we
can recognize what beliefs these experiences support. We might wonder whether
such optimism about the luminosity of both the mind and epistemic support
relations is well advised. Timothy Williamson has argued, famously, that the mind
is not luminous (Williamson 2000: ch. 4). If his anti-luminosity argument is sound,
then introspection cannot always reveal to us what our perceptual experiences
are. It is then possible to be in a situation in which, for example, one simply
cannot tell whether an object looks red to one or not. Williamson’s argument
puts considerable pressure on traditional, internalist epistemology. However,
the soundness of his anti-luminosity argument has been challenged.11 And even
granting its soundness, it is unclear how much damage it inflicts on the traditional
internalist position.12
6 MOTIVATING INTERNALISM
Internalists have rejected reliabilism on the ground that origination in a reliable
cognitive process is neither necessary nor sufficient for making beliefs justified.
Beliefs that are false due to undetectable deception are unreliably produced, but,
internalists have argued, they can nevertheless be justified; beliefs that are in fact
reliably produced, but undetectably so, need not be justified. These are serious
arguments that put considerable pressure on reliabilism. But do they show that
a theory’s being internalist is something recommending it? They do not. If these
arguments are successful, they show that reliabilism is false, without showing
that reliabilism is false because it is an externalist theory. Let us distinguish, then,
between merely attempting to show that a particular version of externalism is false
and arguing for or against internalism as such. Why should we think that a theory’s
being internalist is something recommending it?
The perhaps dominant way of motivating internalism is to argue that justification
is conceptually linked to deontological status.13 The argument’s first premise is a
claim about the meaning of the word “justification.” To be justified in believing
p is to believe p in a way that does not violate the duties one has qua being an
intellectual being who aims at truth and the avoidance of falsehood. Alternatively,
the point can be expressed by saying that to be justified in believing p is to believe p
in a way that manifests epistemic responsibility, or in a way that is not epistemically
blameworthy. Let us express the first premise in terms of the latter point: justified
belief is epistemically responsible belief. The argument’s second premise asserts that
whatever determines whether a belief is responsibly held must be recognizable on

reflection. It then follows that j-determiners must be recognizable on reflection.
Why think that the second premise is true? The basic idea is that facts not
recognizable on reflection cannot determine the deontological status of either one’s
actions or one’s beliefs. Let’s consider actions first. If you cannot know that the
capsules in the Tylenol bottle you recently purchased are laced with cyanide, you
will not act irresponsibly if you offer a couple to your colleague.14 The passengers on
Pan Am Flight 103 (which crashed over Lockerbie) did not act irresponsibly when
they boarded the plane.15 Sometimes what people do has unfortunate outcomes. If
these outcomes are not recognizable or foreseeable at the time, they do not render
such actions irresponsible or unreasonable. An analogous point can be made about
beliefs. Sometimes, our beliefs are false. If nothing that’s recognizable to us at the
time indicates that they are false, such beliefs are not held irresponsibly. When the
passengers on Flight 103 believed the plane would in due time safely land in New
York City, their belief was not irresponsible, for nothing they could have figured out
at the time would have given them a reason to suspect there was a bomb on board.
Likewise, nothing a BIV can tell on reflection indicates that its external world beliefs
are unreliably produced (and indeed mostly false). Therefore, a BIV’s beliefs about
the external world are not epistemically irresponsible. Examples like these suggest
that what determines deontological status must be recognizable on reflection.16
In response to the epistemic responsibility argument for internalism, externalists
can reject the deontological conception of justification.17 They could easily agree
that evaluating beliefs as responsible or irresponsible is an interesting and legitimate
mode of epistemic evaluation but, at the same time, insist that a belief’s deontological
status is not related to truth in the way justification is supposed to be. William Alston
captures this point by saying that deontological justification is not truth-conducive
and thus is not the same as epistemic justification. He argues that deontological status
is sensitive to personal and situational limitations to which epistemic justification is
not. Alston describes two cases that illustrate this point (Alston 1985: 67f). First,
consider a philosophy major of limited talent. His interpretation of John Locke
might be responsibly arrived at because he did his best to read the text carefully; but,
given the limits of his understanding, it is not justified. Second, consider an isolated
tribe of indigenous people in a remote region in Brazil. Some of their beliefs are such
that, given what the tribe members have to go on, they are not irresponsibly held.
But, given the way these beliefs are actually arrived at, ways that are not at all likely
to produce true beliefs, they do not qualify as epistemically justified. According to
Alston, such examples show that responsibly held belief is one thing, justified belief
is another. They are not to be equated with each other. If that is correct, then the
deontological attempt to motivate internalism fails.
Next, we will turn to the question of how the internalism-externalism debate
is related to the issue of skepticism. Specifically, we will consider the following
two questions: First, does internalism have skeptical consequences that cannot be
attributed to externalism? Second, are externalists better positioned than internalists
when it comes to responding to arguments for skepticism about knowledge of the
external world?
598 SKEPTICISM
7 INTERNALISM, CLASSICAL FOUNDATIONALISM,

AND SKEPTICISM
Let the word “skepticism” designate the claim that we do not have knowledge of the
external world.18 Advocates of externalism could argue that whereas internalism,
when closely inspected, reveals itself as a form of skepticism, externalism is not
vulnerable to this accusation.19 Using reliabilism as a representative of externalism
and classical foundationalism as a representative of internalism, the argument
proceeds in two stages. At the first stage, it is asserted that reliabilism does not have
skepticism as a consequence:
According to reliabilism, having justified beliefs about the external world does
not require reasons in support of these beliefs. Nor does it require any beliefs
about, or awareness of, the fact that beliefs about the external world are reliably
produced. For such beliefs to be justified, their being reliably produced is
sufficient. Therefore, if reliabilism is true, external world beliefs are typically
justified and, if true, instances of knowledge. Reliabilism, therefore, qualifies as a
paradigmatically non-skeptical theory.
At the second stage of the argument, it is asserted that classical foundationalism
cannot avoid the outcome of skepticism:
According to classical foundationalism, our beliefs about the external world
qualify as instances of knowledge only if they are supported by foundational
beliefs about how the external world appears to us. But ordinarily, we do not form
foundational beliefs about how the external world appears to us. Rather, we tend
to acquire beliefs about the objects in the external world spontaneously, without
inferring them from beliefs about how these objects appear to us. Therefore, if
classical foundationalism is true, our external world beliefs are under ordinary
circumstances unjustified and thus fail to be instances of knowledge. Hence,
classical foundationalism, and therefore internalism, produces an unacceptably
skeptical outcome.
Although the claim that classical foundationalism generates the outcome of
skepticism is quite plausible, this argument suffers from an obvious limitation.
Classical foundationalism is just one form of internalism among others. Like
classical foundationalism, evidentialism assigns an important role to foundational
beliefs. However, according to evidentialism, foundational beliefs need not be
about how the external world appears to us. They can be beliefs about objects in
the external world. As such, they will make up a large part of a normal person’s
belief system. Therefore, the criticism that we do not under ordinary circumstances
form foundational beliefs does not apply to experiential evidentialism. To make the
internalism-leads-to-skepticism objection stick, externalists need to show that any
form of internalism suffers from the defect of making justification too difficult to
attain. That is, they need to identify an essential feature of internalism, a feature that
is necessarily a part of every version of internalism, and they need to show that, by
virtue of this feature, internalism is bound to have a skeptical outcome.
8 INTERNALISM, EPISTEMIC RESPONSIBILITY,

AND SKEPTICISM
If externalists wish to reject internalism on the ground that, whatever version it
takes, the view either renders knowledge of the external world impossible or at least
unacceptably minimizes its extent, they must meet the challenge of identifying a
skepticism-breeding feature that is essential to internalism. This feature, externalists
might argue, is the deontological conception of epistemic justification. Without it,
internalism remains unmotivated. Therefore, part and parcel of every version of
internalism is the claim that justified belief is epistemically responsible belief. But,
so the argument continues, the standards of epistemic responsibility are excessively
high. Therefore, internalism is bound to be a version of skepticism.
Why think that the standards of epistemic responsibility are excessively high?
The worry is that the demand for responsibility generates a vicious infinite regress.20
Suppose you believe p on the basis of a reason R. Suppose also your belief is justified.
R might be a perceptual experience with p as its content, or a further justified belief
from which p can be inferred. Initially, it might look as though the two conditions—
(i) having R and (ii) believing p on the basis of R—are sufficient for being responsible
in believing p. The externalist, however, might argue that (i) and (ii) are not sufficient
for being responsible in believing p. For your belief to be responsibly held, you must
also believe (iii) that there is a suitable support relation between R and p; that is,
you must believe that R makes it likely that p is true. Without such a metabelief, you
would not be responsible in believing p.21
Now two problems arise. First, in ordinary circumstances, we do not form
metabeliefs about the truth-conduciveness of our reasons. This is simply a fact
about human psychology. Second, a vicious infinite regress ensues. Suppose you
actually form the metabelief in question. If it is not justified in turn, it will hardly
be of help in making you responsible in believing p. Now, for your metabelief to
be justified, you need a reason R* that supports it, and you need a metabelief*
to the effect that R* supports your first metabelief. But your metabelief* needs
to be justified in turn, which calls for a reason R**, supporting your second
metabelief, and so forth. This regress is vicious because its increasing complexity
makes it impossible for human beings to move beyond the first few steps. So, if
epistemic responsibility does require justifying metabeliefs, then it is impossible
to satisfy the demands of such responsibility. Hence, if justified belief is indeed
epistemically responsible belief, justified belief remains elusive. Internalism leads
to skepticism.
To this line of reasoning, internalists can respond by questioning the premise that
epistemic responsibility requires justifying metabeliefs. According to an alternative
proposal, for responsibly believing p, the following is sufficient: the belief’s being
based on a reason R, and the existence of a true proposition to the effect that R
makes the truth of p likely. The true-proposition requirement does not have the
skeptical consequences of the requirement for a metabelief. It is consistent with the
fact that, in ordinary circumstances, we do not form justifying metabeliefs, and it
does not generate a vicious infinite regress.
600 SKEPTICISM
Alas, is the mere existence of a true proposition of the form R makes it likely that
p is true enough to render it responsible for a subject to believe p on the basis of R?
Arguably, it is not. In the very least, the subject must, it might be argued, somehow
appreciate that R makes it likely that p is true. And how can such appreciation
be instantiated short of believing the meta-proposition about the support relation
between R and p? According to the objection to internalism presently under
consideration, the demands of epistemic responsibility can only be discharged by
believing such meta-propositions. Internalists could insist, however, that epistemic
responsibility is not as demanding as the externalist critic would have it. They could
argue that, if a belief p is based on a reason R, understanding and appreciating
the epistemic support relation between R and p is sufficient for the belief to be
epistemically responsible. Moreover, understanding and appreciating the support
relation does not require the subject to form a metabelief to the effect that R supports
p. According to this approach, part of what it is to understand a proposition p is
to understand what counts as a reason in support of p. For example, an ordinary,
competent speaker of English understands that an object’s looking red is a reason
to think it is red, and that an object’s looking like a tomato is a reason for taking it
to be a tomato. Such a subject manifests an understanding and appreciation of these
support relations through the patterns of her epistemic conduct: typically, when she
believes an object to be red, that objects looks red to her; when the object does not
look red to her, she does not form the belief that it is red. Such an understanding,
internalists could argue, need not manifest itself through any epistemic metabeliefs.
By itself, it is enough to make it epistemically responsible for a competent speaker
to take an object to be red when the object looks red to her.
9 CLOSURE DENIAL AS AN EXTERNALIST

RESPONSE TO SKEPTICISM
We have considered two arguments for the conclusion that internalism leads
to skepticism. According to the first, internalism is identified with classical
foundationalism, which arguably makes skepticism unavoidable. According to
the second, internalism inevitably produces skepticism because the deontological
conception of justification as epistemic responsibility is too demanding. It turned out
that neither of these arguments is an unqualified success. Alternatively, externalists
could argue that, when it comes to rebutting skeptical arguments, they are in a strong
position whereas internalists are in a weak position.22 To examine the plausibility of
this claim, let us consider the following argument:
The BIV Argument

(1) If I know that I have hands, then I know that I am not a BIV.
(2) I do not know that I am not a BIV.
Therefore:
(3) I do not know that I have hands.
The first premise is plausible because it’s difficult to deny that knowledge requires
the ability to discriminate. If I cannot discriminate between a goldfinch and a canary,
I cannot know that the bird before me is a goldfinch simply by looking at it. If I
cannot tell Mark from Matt, who are identical twins, then I cannot know that the
person before me is Mark just by looking at him. Likewise, if I cannot discriminate
between being a normal person with hands and being a handless BIV, I’m not in a
position to know that I’m not a BIV. The second premise is also plausible. The BIV
hypothesis involves the stipulation that, if you were a BIV, your evidence—the total
set of your memories, beliefs, and introspective and perceptual experiences—would
be exactly the same as it is in the normal situation. This stipulation, it seems, makes
it impossible for me to have any reason to think I am not a BIV.
The argument we will discuss in this section aims to establish the conclusion that
rebutting the BIV argument is easier for externalists than it is for internalists. To discuss
this claim, we must first get clear on what a successful rebuttal of the BIV argument is
supposed to accomplish. A benchmark that’s neither too easy nor too ambitious is the
following: a rebuttal of the BIV argument is successful if, and only if, it makes a strong,
prima facie plausible case that at least one of the argument’s two premises is false. On
this understanding of rebutting a skeptical argument, the claim under consideration is
that externalists are in a better position than internalists to show that at least one of the
two premises of the BIV argument is false. We will first examine the plausibility of this
claim using Fred Dretske’s conclusive reasons theory as an example of externalism, and
then move on to discuss how the claim holds up when we consider reliabilism.
The first premise of the BIV argument expresses commitment to the closure
principle, which says the following:
Closure
If one knows that p and knows that p entails q, then one is in a position to know
that q.23
Many epistemologists think that the closure principle is extremely plausible. Let’s
apply the principle to the goldfinch example. If p = the bird is a goldfinch and q =
the bird is not a canary, then, given I know that p entails q, it is hard to see how
I could know p without also knowing (or at least being in a position to know) q.
Likewise, if p=I have hands and q=I’m not a BIV, then, given I know that p entails
q, it is hard to see how I could know the former proposition without also knowing
(or at least being in a position to know) the latter.
Nevertheless, Dretske’s theory of knowledge recommends that we should reject
the first premise of the BIV argument. According to Dretske, having a conclusive
reason is sufficient for knowledge.24 R is a conclusive reason for a subject S to believe
p if, and only if, the following condition is met: in all close possible worlds in which p
is false, R is false or, alternatively, S does not have R.25 Now, my reason for thinking I
have hands is that I have hand-like experiences: I can see and feel my hands. In close
worlds in which I do not have hands, I do not have hand-like experiences. Instead,
I would have stump-like, or prostheses-like, experiences. Of course, in a world in
which I am a handless BIV, I nevertheless have hand-like experiences; but such a
602 SKEPTICISM
world is not close. My hand-like experiences, therefore, are a conclusive reason for
me to think I have hands. That’s why, on Dretske’s theory, I know that I have hands.
Alas, my belief that I am not a BIV is not supported by a conclusive reason. Whatever
my reasons are for believing I am not a BIV, I have these reasons in the closest
possible worlds in which I am a BIV. That’s simply how the BIV hypothesis is set up.
Thus, on Dretske’s theory, I do not know that I am not a BIV. Hence, if Dretske’s
conclusive reasons theory of knowledge is right, we get the following picture: (i) I
know I have hands and (ii) I do not know I’m not a BIV, even though (iii) I know
that having hands entails not being a BIV. If (i) and (ii) are conjoined with (iii), then
the closure principle is false. So according to Dretske, the proper response to the
BIV argument is to reject the first premise. If the first premise is false, then my being
unable to know I’m not a BIV is not an obstacle to my knowing that I have hands.
Why does Dretske’s closure-denial strategy qualify as an externalist response to
skepticism? It does because Dretske’s conclusive reasons theory, like reliabilism, is
externalist. Whether a reason is conclusive is not always recognizable on reflection.
If you are a BIV, then, for you, the actual world is such that, in it, you have hand-like
experiences even though you do not have hands. So, if you are a BIV, your hand-like
experiences are not conclusive but rather misleading reasons. Alas, you have no way
of recognizing this. The non-conclusiveness of your hand-like experiences is entirely
hidden from you.
The strategy of avoiding skepticism by denying closure runs into two
serious problems. The first is that of abominable conjunctions (see DeRose 1995:
27–29). Consider:
Conjunction A: (a) I know that this bird is a goldfinch, but (b) I do not know that
it is not a canary.
This sounds pretty awful. If somebody were to admit (b), we would be very reluctant
to agree to part (a) of the assertion. Now consider:
Conjunction B: (a) I know that I have hands, but (b) I do not know that I am not
a handless BIV.
This, too, sounds quite awful. It is difficult to see why we should agree to Conjunction
B if we reject Conjunction A.
The second problem is no less serious. Dretske’s conclusiveness requirement is
actually quite demanding. A lot of what we take ourselves to know is not based on
conclusive reasons but on what we might call inductive evidence: on beliefs about
how the world by and large works. Consider the proposition that there isn’t a nuclear
bomb in your basement. Call this proposition “No Bomb.” You have strong inductive
evidence to think that No Bomb is true. Nuclear devices are exclusively kept at military
installations that are extremely well guarded. They are never in the basements of
houses in which ordinary citizens live. Alas, such inductive evidence is not conclusive.
In close possible worlds in which No Bomb is false (i.e., in close worlds in which there
really is a nuclear bomb in your basement), you still have strong inductive evidence
to think that No Bomb is true.26 To acquire a conclusive reason for No Bomb, you
would have to go into your basement and look. So, if Dretske’s conclusive reasons
theory were right, you couldn’t know that there isn’t a nuclear bomb in your basement
without having to go and check. That seems odd. An analogous outcome follows for
similar propositions such as: there isn’t a million dollars hidden inside your mattress,
or the current secretary of defense does not secretly plan to steal the gold in Fort Knox.
It’s easy to make a long list of such propositions. It would seem, then, that Dretske’s
externalist closure-denial response to skepticism significantly reduces the scope of
what we know. It avoids the outcome that we have no knowledge of the external
world at all by giving up a good chunk of such knowledge.
10 RELIABILISM AND SHOWING THAT ONE ISN’T

A BRAIN IN A VAT
Nothing about externalism requires its advocates to respond to the BIV argument
by denying the closure principle. An externalist might as well rebut the argument
by asserting “I know I’m not a BIV,” that is, by denying the BIV argument’s second
premise. What we will examine in this section is the claim that, when it comes to the
challenge of proving that premise wrong, externalists are in a better position than
internalists. To do so, we will focus on reliabilism and evidentialism.27
According to reliabilism, the belief “I’m not a BIV,” if true and reliably
produced, is an instance of knowledge. What the belief’s origin is is not obvious.
Non-philosophers do not even consider the question of whether they are BIVs.
Philosophers disagree on how to justify the rejection of the second premise of
the BIV argument. Some reject it because they take themselves to know they have
hands.28 Others reject it because they take themselves to know that BIVs do not
exist.29 For our purposes here, we can easily accommodate a variety of reasons
why people believe they are not BIVs. What matters is that, whatever the source of
the belief “I’m not a BIV” is, if the belief is true, then, according to reliabilism, it
amounts to knowledge if that source is sufficiently reliable. The reliabilist response
to the BIV argument, then, can be stated as follows:
(R) I know that I am not a BIV because my belief that I’m not a BIV has its origin
in a reliable source.
In contrast, the evidentialist rejection of the BIV argument’s second premise

goes thus:
(E) I know that I am not a BIV because I have knowledge-grade evidence for the
belief that I am not a BIV.
In each case, we are looking at a mere recipe for articulating a fully developed
response. Obviously, for (R) to carry weight, the reliabilist would have to present
evidence in support of the claim that her belief not to be a BIV stems from a reliable
source. Likewise, for (E) to carry weight, the evidentialist would have explain what
knowledge-grade evidence justifies her belief that she is not a BIV.
Why should we think that presenting evidence in support of (R) is easier than
presenting evidence in support of (E)? The thought might be that defending (R)
604 SKEPTICISM
is easier than defending (E) because reliabilism makes the acquisition of reliability
knowledge easier than evidentialism does. This thought, however, runs into
a serious problem, for it is in fact very difficult to demonstrate—as opposed to
merely knowing—that a particular belief has its origin in a reliable source. In general
terms, the challenge is to establish the reliability of one’s faculties. Alston, himself
a reliabilist, has argued that, while knowing that one’s faculties are reliable is easy,
it is not only difficult but in fact impossible to demonstrate that one’s faculties are
reliable (Alston 1993: ch. 2).30 Knowing it is easy because, according to reliabilism,
if a faculty F is indeed reliable and the belief “F is reliable” is itself reliably produced,
then that belief is an instance of knowledge. However, demonstrating that one’s
faculties are reliable is not easy because one cannot argue for the reliability of one’s
faculties without using one’s faculties. To show that perception is reliable requires
the use of perception; to show that memory is reliable requires the use of memory.
Alston refers to the circularity exhibited here as epistemic circularity. An argument is
epistemically circular if at least one of its premises is such that a legitimate appeal to it
requires antecedent knowledge of the conclusion. Suppose I appeal to a proposition
p as a premise in an argument for the reliability of perception. If showing that p is
true requires antecedent knowledge of the reliability of perception, appealing to p
will hardly serve as a good reason to think that perception is reliable.
It would be a mistake, then, to think that defending (R) is easier than defending
(E). In fact, it might be thought that, if Alston is right, the opposite is true: defending
(R) is more difficult than defending (E). But that, too, would be a mistake. For
the evidentialist, the task of showing that she has knowledge-grade evidence for
believing she is not a BIV involves the subtask of showing that the adduced evidence
stems from a reliable source. For in response to any claim of the form “I know
I’m not a BIV because of fact F,” the skeptic can ask: “How do you know that the
faculty through which you think you know F is reliable?” If the evidentialist does
not answer this question, she can hardly claim to have a successful response to the
BIV argument. If the evidentialist attempts to answer the question, she faces the very
same challenge reliabilists do when they attempt to defend (R).31
So, part of demonstrating that one knows one is not a BIV is to show that the
faculties through which such knowledge is acquired are reliable. Showing that
without epistemic circularity is impossible. Therefore, a successful refutation of the
BIV argument’s second premise requires a sound argument for the conclusion that
epistemic circularity can be benign. This is a formidable challenge. Two conclusions
follow from our findings. First, rejecting the BIV argument’s second premise is
difficult. Second, in the attempt to meet this challenge, neither reliabilism nor
evidentialism has an advantage over the other.
NOTES
1. All occurrences of “justification” or “justified” refer to epistemic justification, the kind
that bears on a belief’s likelihood of being true and turns, under the right conditions,
a true belief into knowledge. For early work on the internalism-externalism issue, see
Alston (1986) and BonJour (1985: ch. 3).
2. For contemporary examples of classical foundationalism, see BonJour (2001) and

Fumerton (2001).
3. See Conee and Feldman (1995; 2004). Related views are dogmatism (Pryor 2000)
and phenomenal conservatism (Huemer 2000).
4. See Goldman (1979) for the initial, classical articulation of reliabilism. For the
sake of succinct presentation, we are considering the simplified version according
to which reliable belief formation is both necessary and sufficient for justification.
Fully developed versions of reliabilism distinguish between inferential and non-
inferential justification and attempt to address the phenomenon of defeat.
5. Alston (1986) distinguished between a “cognitive access” and a “perspectival”
version of internalism.
6. According to Michael Bergmann (2006; 2008), an opponent of internalism, the
distinctive mark of internalism is some kind of awareness requirement; see (2008:
504). Awareness, however, is naturally interpreted as a special kind of knowledge.
The requirement Bergmann has in mind, therefore, is very demanding. Because
of its obvious regress implications, internalists should not define internality in
this way.
7. The always-recognizable-on-reflection criterion of internality might seem
excessively demanding. The worry might be that there is no theory of justification
that makes justification something the presence or absence of which can always be
recognized on reflection. However, “can be recognized” should not be identified
with “will be recognized.” To say that justification is always recognizable on
reflection is not to say that an attempt to determine through reflection whether
a belief is justified will always be successful. Someone who is momentarily
confused might at the present moment be unable to recognize what is in principle
recognizable to him.
8. Cases of this kind have become known as the “new evil demon” problem. See Cohen
(1984) and Ginet (1985).
9. A BIV’s mental states are induced through computer-controlled stimulation of nerve
endings in the brain. The stimulation is so perfect that the experiences of a normal
human being are duplicated. To a BIV, everything seems perfectly “normal,” which
is why the deception that is actually going on is undetectable.
10. The main thought would be that, according to classical foundationalism, (i) the
unreliability of a belief-producing process does not block justification, and (ii) to
BIVs and normal subjects alike, the support relations between foundational and
non-foundational beliefs are recognizable on reflection.
11. See Berker (2008), Steup (2009), and also Williamson (2009).
12. See Hawthorne (2005), Reed (2006), and Fumerton (2009).
13. For a lucid account of the relationship between epistemic deontologism and
internalism, see Plantinga (1993: ch. 1).
14. See http://en.wikipedia.org/wiki/Tylenol_murders.
15. See http://en.wikipedia.org/wiki/Pan_Am_Flight_103.
16. The claim is fairly plausible but not immune to proposed counterexamples. Suppose
a subject S is culpably in a situation in which he cannot recognize on reflection that p
is the case. If S had conducted a proper investigation, his total evidence would have
606 SKEPTICISM
concluded E*, but since S failed to conduct a proper investigation, his total evidence
is missing E*. Had it included E*, S would have noticed on reflection that p is true.
Without E*, S’s total evidence indicates that p is false. Arguably, S is epistemically
irresponsible in disbelieving p, even though this is not recognizable to him on
reflection. Against this kind of objection, it could be argued that a case like this
involves two different dimensions of evaluation. The agent was irresponsible in the
way he conducted the investigation; however, he is not irresponsible in disbelieving
p, for the very reason that his evidence indicates that p is false.
17. See Alston (1985; 1988).
18. There are of course other kinds of skepticism. Of these, the most prominent are
claims to the effect that we lack knowledge of the past, the future, or other minds.
For reasons of space, here we can focus only on skepticism about our knowledge of
the external world.
19. For arguments to the effect that internalism breeds skepticism, see
Goldman (1999).
20. For an argument roughly along these lines, see Michael Bergmann’s “dilemma for
internalism” argument (Bergmann 2006: ch. 1).
21. An argument along these lines is embedded in BonJour’s primary objection
to foundationalism in his The Structure of Empirical Knowledge (8BonJour
1985: 30–33).
22. For an excellent review of externalist responses to skepticism, see Bergmann
(2008).
23. For a discussion of the issues involved in rejecting and defending the closure
principle, see Hawthorne (2013).
24. See Dretske (1970; 1971; 2013). Another externalist closure-denial approach to
skepticism is Robert Nozick’s (1981) tracking theory of knowledge.
25. Alternatively, the notion of a conclusive reason could be defined in terms of a
counterfactual: R is a conclusive reason for S to believe p =df if p were false, R would
be false or S would not have R.
26. The closest worlds in which there is a nuclear bomb in your basement are not worlds
in which the reasons mentioned are false, but rather worlds in which terrorists got
hold of a suitcase bomb and, due to highly coincidental and unlikely circumstances,
ended up hiding it in your basement. In a situation like that, you still have excellent
reason for thinking there is not a nuclear bomb in your basement.
27. Another externalist approach to rejecting the BIV argument’s second premise is
Timothy Williamson’s controversial equation E=K: one’s evidence equals what
one knows. What normal Boris and envatted Boris know is vastly different.
Normal Boris has lots of knowledge of the external world. Envatted Boris
does not. Now, the standard line for saying normal Boris cannot know he is
not a BIV is that, if he were a BIV, his evidence would be the same: normal
Boris’s evidence=envatted Boris’s evidence. This line of reasoning is blocked if
Williamson’s E=K is correct, and one main challenge to claiming knowledge of
not being a BIV has been removed. Williamson’s approach qualifies as externalist
because, according to Williamson, only beliefs that are reliably based qualify as
knowledge. Given E=K, it follows that only what is reliably based qualifies as
evidence. For reasons of space, Williamson’s response to the BIV argument cannot
be discussed in this chapter.
28. See Huemer (2000: ch. 8).
29. See Steup (2011).
30. Ironically, it has also been argued that reliabilism makes arguments for reliability
too easy to come by. See Fumerton (1995: 173–178) and Vogel (2000). Cohen
(2002) has argued, however, that the problem—the problem of easy knowledge—
arises for all theories, internalist or externalist, that allow for basic knowledge:
knowledge acquired without knowledge that its source is reliable.
31. For an attempt to show how evidentialists can argue for the reliability of their
faculties, epistemic circularity notwithstanding, see Steup (2013).
REFERENCES
Alston, William. 1985. “Concepts of Epistemic Justification,” The Monist 68: 57–89.
Alston, William. 1986. “Internalism and Externalism in Epistemology,” Philosophical
Topics 14: 179–221.
Alston, William. 1988. “The Deontological Conception of Epistemic Justification,”
Philosophical Perspectives 2: 257–299.
Alston, William. 1993. The Reliability of Sense Perception. Ithaca: Cornell University Press.
Bergmann, Michael. 2006. Justification without Awareness. New York: Oxford University
Press.
Bergmann, Michael. 2008. “Externalist Responses to Skepticism.” In J. Greco (ed.), The
Oxford Handbook to Skepticism, 504–538. New York: Oxford University Press.
Berker, Selim. 2008. “Luminosity Regained,” Philosophers’ Imprint 8: 1–22.
BonJour, Laurence. 1985. The Structure of Empirical Knowledge. Cambridge: Harvard
University Press.
BonJour, Laurence. 2001. “Toward a Defense of Empirical Foundationalism.” In M.
De Paul, (ed.), Resurrecting Old-Fashioned Foundationalism, 21–38. Lanham, MD:
Chisholm, Roderick. 1989. Theory of Knowledge, 3rd edn. Englewood Cliffs: Prentice Hall.
Cohen, Stewart. 1984. “Justification and Truth,” Philosophical Studies 46: 279–295.
Conee, Earl and Richard Feldman. 1985. “Evidentialism,” Philosophical Studies 48: 15–35.
Conee, Earl and Richard Feldman. 2001. “Internalism Defended.” In H. Kornblith (ed.),
Epistemology: Internalism and Externalism, 230–260. New York: Blackwell.
Conee, Earl and Richard Feldman. 2004. Evidentialism: Essays in Epistemology. New York:
De Paul, Michael (ed.). 2001. Resurrecting Old-Fashioned Foundationalism. Lanham, MD:
DeRose, Keith. 1995. “Solving the Skeptical Problem,” Philosophical Review 104:1–52.
Dretske, Fred. 1970. “Epistemic Operators,” The Journal of Philosophy 67: 1007–1023
Dretske, Fred. 1971. “Conclusive Reasons,” Australasian Journal of Philosophy 49: 1–22.
608 SKEPTICISM
Dretske, Fred. 2013. “The Case against Closure.” In M. Steup, J. Turri, and E. Sosa (eds.),
Contemporary Debates in Epistemology, 27–40. New York: Wiley Blackwell.
Fumerton, Richard. 1995. Metaepistemology and Skepticism. Lanham: Rowman and
Littlefield.
Fumerton, Richard. 2001. “Classical Foundationalism.” In M. De Paul, (ed.), Resurrecting
Old-Fashioned Foundationalism, 3–20. Lanham, MD: Rowman and Littlefield.
Fumerton, Richard. 2009. “Luminous Enough for a Cognitive Home,” Philosophical
Studies 142: 67–76.
Ginet, Carl. 1985. “Contra Reliabilism,” The Monist 68: 175–187.
Goldman, Alvin. 1979. “What is Justified Belief?” In G. S. Pappas (ed.), Justification and
Goldman, Alvin. 1999. “Internalism Exposed,” The Journal of Philosophy 96: 271–293.
Hawthorne, John. 2005. “Knowledge and Evidence,” Philosophy and Phenomenological
Research 70: 452–458.
Hawthorne, John. 2013. “The Case for Closure.” In Steup, M., J. Turri, and E. Sosa (eds.),
Contemporary Debates in Epistemology, 40–56. New York: Wiley Blackwell.
Huemer, Michael. 2000. Skepticism and the Veil of Perception. New York: Rowman and
Littlefield.
Press.
Plantinga, Alvin. 1993. Warrant: The Current Debate. New York: Oxford University Press.
Reed, Baron. 2006. “Shelter for the Cognitively Homeless,” Synthese 148: 303–308.
Steup, Matthias. 2009. “Are Mental States Luminous?” In Greenough, P. and D. Pritchard
(eds.), Williamson on Knowledge, 217–236. New York: Oxford University Press.
Steup, Matthias. 2011. “Evidentialist Anti-Skepticism.” In T. Dougherty (ed.), Evidentialism
and Its Discontents, 105–122. Oxford: Oxford University Press.
Steup, Matthias. 2013. “Is Epistemic Circularity Bad?” Res Philosophica 90: 1–20.
Steup, Matthias, John Turri, and Ernest Sosa (eds.). 2013. Contemporary Debates in
Epistemology, 2nd edn. New York: Wiley Blackwell.
Williamson, Timothy. 2009. “Reply to Steup.” In P. Greenough and D. Pritchard (eds.),
Williamson on Knowledge, 370–374. New York: Oxford University Press.
CHAPTER FORTY-TWO
Skepticism and Fallibilism

STEPHEN HETHERINGTON
1 KNOWLEDGE-SKEPTICISM
Epistemology harbors many shades of, and guises for, skeptical thinking. This
chapter will discuss perhaps the simplest of all—knowledge-skepticism:
No belief is ever knowledge. That is, for no belief is there even a single moment
when it is knowledge.
Let us understand this as a thesis about any belief belonging to any actual or
possible person.1
Is it a thesis we should accept as true? It is, if it is derivable from grounds we
independently regard as true—a potential concession which brings us to our chapter’s
main question. Can knowledge-skepticism be derived from a fallibilist conception of
people’s cognitive powers and attainments? These days, many epistemologists claim
to be fallibilists. Yet typically they regard this as saving them from being committed
to knowledge-skepticism. (The usual explanation of this escape is said to be that
accepting a lower standard for knowing makes it easier to attribute knowledge.) Are
fallibilists correct in that optimistic assessment, though? Or does even fallibilism
imply a skepticism about there being any knowledge?
2 FALLIBILISM AND FALLIBILITY

That depends upon what is meant by “fallibilism.” Here is a generic initial attempt
to understand it:
Somehow, in some way, fallibility attends or infuses all beliefs (all instances of
having a belief).
But what is it for a belief to be fallible?
The guiding idea in reflection upon fallibility’s nature has long been that of
the possibility of being mistaken, even when believing carefully, say. You could be
mistaken even in a belief resting on what seem to you to be good evidential grounds:
A belief is fallible when it might be false in spite of its apparently good grounds.
610 SKEPTICISM
And because knowledge was generally thought of as a true belief supported by good
grounds,2 one apparent implication of that guiding idea would have been this:
A belief is fallible when, in spite of apparently being knowledge, it might not be
knowledge (due to being false).
This use of “apparently” conveys that in any such case it would seem to us—and
not randomly or unjustifiedly—that we have knowledge. Yet the use here of “might
not be” allows that perhaps we would not really have knowledge. This possibility of
mistake—of falsity in our apparently well-grounded belief—would be a possibility
of our actually not knowing even when apparently we do.
This is an initial step toward knowledge-skepticism. It accords fallibility a structure
akin to probably the most venerable philosophical idea that has ever spurred people
to skeptical heights or depths. That venerable idea asks us to gaze or reflect upon
what is apparently real, before conceiving of it as possibly only apparently real—
possibly not really real. Once we imbue our experiences and thoughts with that
conceptual pallor, no longer need we take ourselves to be gaining knowledge of
what is real through those experiences and thoughts. If we might not know even
when we seem to know, do we ever know—really? Should we lower our confidence
that we are connected as knowers to what is real?
Let us deepen that potential sense of disconnectedness. I offered a candidate
formulation of what it is for a belief to be fallible, followed by an implication of that
formulation. The initial formulation talked of a belief’s possible falsity even when
supported by apparently good grounds. The implication noted that this would be a
possible failure by the belief to be knowledge even in a setting of apparently being
knowledge. That same implication could have been gained from this alternative
possible initial formulation of fallibility’s nature:
A belief is fallible when it might be false in spite of its actually good grounds.
This is a stronger statement of fallibility’s nature because it talks of actually good
grounds (not of apparently good ones). It tells us that, for any belief you have, even
if your evidence supporting the belief is good, the belief need not be true. This is a
fallibilist view of what even genuinely good justification or support gives a belief—
namely, nothing that guarantees the belief’s being knowledge. Are we thereby closer
still to having to accept knowledge-skepticism if we adopt this fallibilist view of
justificatory support?
Before conceding so, we should note another complexity involved in conceiving
of fallibility. What does “it might be false” mean in our possible formulations of
fallibility’s nature? This is a first suggestion:
The belief in question is false in some possible situation. (If that situation is only
possible, not actual, it is as much like this actual one as possible—for a start, by
including the same grounds for the belief.)
There is a problem with that suggestion. It never allows us to understand a belief
as fallible when the belief’s content is necessarily true (e.g., when a mathematical
or logical truth is believed). For no such belief is ever false in even a single possible
SKEPTICISM AND FALLIBILISM 611
situation.3 Yet we naturally say that a person could form or hold a fallible belief
with such a content. (Fallibilism tells us to expect nothing more.) How can we
understand her as doing this?
Many epistemologists would respond to that challenge by proposing a shift of
perspective in our conceiving of fallibility. In particular, should we talk of having the
(necessarily true) belief fallibly, not of having the (necessarily true) fallible belief?
This shift could encourage us to regard a belief’s fallibility as always included in the
means—the method, the capacity, or more generally the circumstances—by which
the belief has arisen and/or is being held. For short: The belief would arise and/or
be held in a fallible way.
We might allow that a belief-forming way could be fallible by allowing for the
production of false beliefs, either actually or in alternative (non-actual) possible
situations relevantly similar to the situation where the belief in question has been
formed. We might regard a belief-holding way as fallible when it allows for the
belief’s possible replacement by false beliefs, either actually or in alternative (non-
actual) possible situations relevantly similar to the situation where the belief is
being held.
And even a way of thinking that has produced a necessarily true mathematical
belief on a given occasion might—if fallible—have produced instead a different
and mistaken mathematical belief.4 So, even when a necessarily true belief arises,
there would be fallibility in how it has arisen.5 Are we thus able now to understand
fallibilism’s saying that all our beliefs—including necessarily true ones—are attended
by fallibility? This is what fallibilism needs to understand.
3 SKEPTICAL IMPLICATIONS?
Yet once we understand the nature of the fallibility that is said by fallibilism to attend
all of our beliefs, we must ask whether such rampant fallibility poses a skeptical
challenge to our beliefs ever being knowledge. We have described fallibility as a
respect in which knowledge might be absent when it appears to be present, even
when good support for it is actually present. Does (1) imply (2)?
(1) At any time, for any belief: knowledge might be absent due to fallibility’s
being realized (i.e., due to the belief’s being false in spite of actually and/or
apparently being well supported).
(2) At any time, for any belief: knowledge is absent due to fallibility’s presence
(regardless of whether fallibility is being realized).
(1) is fallibilism, while (2) is knowledge-skepticism’s being said to be grounded in
(1). The rest of this section presents a few arguments that could be thought to take
us from fallibilism to knowledge-skepticism.
3a Degrees or Grades of Fallibility

Might fallibilism threaten knowledge by allowing degrees or grades of fallibility?
We talk regularly of individual belief-forming capacities or methods as more fallible
612 SKEPTICISM
or less so. The more reliably a method produces true beliefs rather than false ones,
the less fallible it is (all else being equal). For it would be more likely to generate a
true belief on a particular occasion, even as it might have generated a false belief.
But consider the possibility, on any given occasion, of a belief having been formed
in a very fallible way rather than merely in a fallible way. This leads to skeptical
reasoning, as follows.
Even those who feel no anxiety at the idea of our beliefs being formed fallibly
might be less sanguine at the thought of those beliefs arising very fallibly. Of course,
in practice it would be rare to know how fallibly a given belief has arisen. But this
admission, too, harbors skeptical potential. Once you admit to not knowing how
fallibly you have formed a particular belief, you should concede the possibility of
having formed it very fallibly—certainly of having formed it much more fallibly
than you realized. But this amounts to allowing that the belief has arisen in a way
that could very easily have given you a belief which is false while seeming true.
And once you allow, not merely that your belief could have been formed so as
to be false, but that this could very easily be what has happened, you should not
continue regarding the belief as knowledge. This problem pertains to all of your
beliefs, though. Consequently, you should view none of them as knowledge.
That would be a knowledge-skeptical stance. Attempted replies to it will
encounter responses of a form familiar from traditional skeptical arguments. If
you assure yourself that on this occasion you have not formed a mistaken belief,
you achieve nothing if you are relying on that same belief-forming method—with
its unknown grade of fallibility. If you use a different belief-forming method to
assess your use of the first one, you face the question of the new method’s grade of
fallibility. If it might likewise be too fallible (for all that you know to the contrary),
the same skeptical implication persists.
3b Gettiered Beliefs
One way for a true belief to arise fallibly, perhaps very fallibly, is for it to be
Gettiered—the centerpiece of what epistemologists call a Gettier case.6 And most
epistemologists deem any Gettiered belief not to be knowledge. Let us see how that
view of Gettiered beliefs—this way for a belief to arise fallibly—might help us to
argue for knowledge-skepticism.
There is dispute as to what makes a situation a Gettier case.7 There are two
general kinds of Gettier case:8
Helpful Gettier cases
Someone forms a true belief on the basis of what would normally be good
evidence of the belief’s being true. The belief is true. But it is made true by a
further circumstance within the setting—a circumstance not noticed within the
person’s evidence. (For instance, from outside some field it looks to you—and
so you believe—that there is a sheep in the field. You are correct—not, though,
because what you are seeing is a sheep. It is a disguised dog. A real sheep is
hidden elsewhere in the field.)9
Dangerous Gettier cases

Someone forms a true belief on the basis of what would normally be good evidence
of the belief’s being true. The belief is true. And (in contrast to what happens in
Helpful Gettier cases) the belief is made true by a circumstance noticed within
the person’s evidence. But there is a further circumstance within the setting, one
which—if it had been noticed in that same way by the person’s evidence—would
have made the same belief false while leaving intact the person’s sense of what
evidence she is using. (For instance, from outside some field, it looks to you in
a normal way that there is a barn in the field. You are correct. Your evidence,
though, takes no account of the many only-apparently-real barns—fake barns—
elsewhere along that road.)10
In neither sort of case (we are standardly told) is the belief knowledge. It is true. It
is fallibly justified. But it is not knowledge.
Why so? This question has spawned many epistemological theories of what a true
belief also needs in order to be knowledge, if its being fallibly justified is not enough.
Gettier acknowledged the fallibility present within his imagined cases (1963: 121).
We may assume that the believer within each Gettier case could do likewise, if
asked. Nonetheless, any Gettiered belief is produced far more fallibly within its
Gettier case than the believer realizes. She is aware of her evidence. But standard
interpretations of any given Gettier case also mention part of the case’s odd-yet-
unnoticed-by-the-believer circumstance. That combination of her evidence and this
other part-circumstance amount to the actual pattern (even if not exactly what she
would say is the pattern) by which she has gained her true belief. This will clearly—
to us, reading the case—have been a very fallible way of gaining a belief with the
content in question. Here is how the sheep-in-the-field case, for example, would
standardly be interpreted along such lines:
Your apparently good evidence of the sheep, plus your not really looking at a
sheep, constitutes a very fallible way for you to have reached the belief that there
is a sheep in the field.11
Is that why (according to the standard interpretation) such a belief is not knowledge?
Epistemologists differ over this. But what is clear is the standard interpretation’s
denying that any Gettiered belief is knowledge, while needing to insist that the belief
is supported only fallibly. I have added that the belief’s support is perhaps very fallible.
Each Gettier case is an odd situation. And no epistemologist would argue that you
are in a Gettier situation for each true belief you have. Still, this skeptical challenge
hovers nearby:
How could you know that you are not in a Gettier case, for each of your beliefs?
You never know this at the time. After all, even if a belief is true, it could be
present as part of your being in a Gettier situation without realizing that you
are; in which event, the belief would not be knowledge. You would not notice
any difference between being in that setting and the belief’s arising in a normal
way within a normal setting. The odd circumstance, in particular, would escape
your notice.
614 SKEPTICISM
That argument embeds, within a familiar form of skeptical reasoning, the possibility
of a belief’s being Gettiered. This form is one in which many epistemologists have
embedded the dreaming possibility or the evil demon possibility, both memorably
described by Descartes in his “Meditation I.”12 The form is simple:
Here is a possibility, P, whose actualization would be incompatible with your
belief B’s being knowledge: [P is then described—your belief’s being Gettiered,
or your dreaming, etc.]. You do not know that P is not being actualized. (You
lack this knowledge because any belief that P-is-not-being-actualized might itself
be present as part of P’s being actualized.) But in that event B is not knowledge.
(For instance, if you were dreaming seeing a barn, you would not know that you
are seeing it; and any belief that you are not dreaming could itself be present as
part of your dreaming.)
Any odd circumstance within a Gettier case (the circumstance somehow precluding
the Gettiered belief’s being knowledge) would be hidden similarly—like her dreaming
or her being deceived by an evil demon—from the believer. So, if epistemologists
are correct in saying that a (fallibly formed) belief’s being Gettiered precludes its
being knowledge, the possibility of a (fallibly formed) belief’s being Gettiered can
be used by skeptical reasoning much as the dreaming possibility and the evil demon
possibility can be. No one need actually be in a Gettier situation, if that skeptical
reasoning is to succeed; just as the dreaming argument does not require that anyone
is dreaming at a given time. It is sufficient, for these skeptical purposes, if no one
ever knows that she is not in a Gettier situation, or that she is not dreaming.13
3c Quantity-Fallibilism
Section 3a alerted us to the idea of there being grades of fallibility. Section 3b’s
Gettier cases are illustrative reminders that a belief can be very fallible. Those
cases also instantiate a further fallibilist moral, we will now see, that might have
knowledge-skeptical implications.
The cases contain circumstances which are such that if the believer were to know
of them, the situation would no longer be a Gettier case. No longer would the
setting contain a justified true belief falling short of being knowledge:
Re the sheep-in-the-field case
If you were to know that what you are seeing is not really a sheep, then (other
things being equal) you would cease believing that there is a sheep in the field or
you would no longer be justified in having that belief.
Re the fake-barns case
If you were to know that there are fake barns nearby—visually indistinguishable
for you from what you are seeing—then (all else being equal) you would either
cease believing that you are seeing a barn or you would no longer be justified in
having that belief.
Section 3b asked whether one ever knows that one is not in a Gettier situation—such
knowledge possibly being needed if one is to have other knowledge. We may now
generalize that question. Does one ever know that one knows enough about one’s
belief-context—enough for having some particular knowledge within that context?
That is a question about the quantity of one’s attendant knowledge. That question
can arise naturally if we accept fallibilism. Let us distinguish between quality-
fallibilism and quantity-fallibilism. The former is what has been characterized so far
in this chapter. It is the traditional sense of fallibilism. It concerns a modal quality
or strength of any given belief. (e.g., it implies that even a belief which appears, by
being based on good evidence, to be knowledge need not be true and therefore need
not be knowledge.)14 But we may also describe a quantitative fallibility, possibly best
understood as applying primarily to a person, only secondarily to her belief:
A person is being fallible whenever, in spite of apparently having done enough to
have a particular item of knowledge, she might not actually have done so (due to
there being more she needs to know).
Quantity-fallibilism says that people are always fallible in this way. Quantity-
fallibilism admits of degrees or grades: one could have done more or done less, even
when one might not have done enough. And quantity-fallibilism may be used in the
following way to derive knowledge-skepticism.
This supposed link to lacking knowledge will be adduced:
A person lacks a particular item of knowledge because she is being fallible,

whenever, in spite of apparently having done enough to have that piece of
knowledge, she might not actually have done so (due to there being more she
needs to know).
Then this challenge may be issued:
How could a person ever know that she has done enough for her belief—which
apparently is knowledge—actually to be knowledge? (She need not be troubled
by this question. Still, she might be. And some will say that she should be.) For
example, has she ascertained how her belief is being made true? (Hence, has
she checked on whether she is in a Gettier case like the sheep-in-the-field one?)
Does she know how much there is to know, and therefore how much more she
might need to know, in knowing the particular truth? (Hence, has she checked
on whether she is in a Gettier case like the fake-barns one?)15
So, imagine accepting, in a quantity-fallibilist spirit, that there is always more that
could be known, even when seeking to know some particular truth. Here is an
example I have used previously (Hetherington 2011a: ch. 5):
In knowing the particular state of affairs that here is a hand, you know this aspect
of the world in more or less detail.16 You could know more, or you could know
less, of what is involved in the fine structural details of the state of affairs that
here is a hand. Maybe you do not know all of what a hand surgeon would know
in knowing that here is a hand.
Maybe even the surgeon would not know all of the pertinent details within the
world’s including this state of affairs that here is a hand. Perhaps in principle
616 SKEPTICISM
continued science could always reveal more such details. Yet perhaps even our best
possible finished science would not report all such details.
To accept that picture of perennial incompleteness in our knowledge of the
world’s constitutive details opens the door to accepting that there could always be
much of relevance that one does not know. Even in apparently knowing a seemingly
simple state of affairs, one might not know much of what is involved in the world’s
including that state of affairs. This could be so for any state of affairs. Maybe you
never know even how much you do not know of what is involved in a state of affairs
you claim to know.
When should this potential limitation make you wonder whether you have the
particular knowledge that would be supported by the further-but-maybe-unattainable
knowledge? Do you really know that here is a hand, if you do not know much of
what this hand even is? The skeptical conclusion possibly to be inferred is that you
never have a particular piece of knowledge. You never know whether you have done
enough to have the particular knowledge; so you lack that knowledge.
4 ONWARDS, EVER ONWARDS

The term “fallibilism” comes from C.S. Peirce, a founder of American pragmatism
(Peirce 1931–1958: vol. 1, para. 171). His fallibilist picture was this:
No inquirer can ever claim with full assurance to have reached the truth, for new
evidence or information may arise that will reverberate throughout one’s system
of beliefs affecting even those most entrenched. (Hauser 1992: xxii)
That picture combines quality-fallibilism—the lack of “full assurance”—with
quantity-fallibilism—the ever-present possibility of “new evidence or information”
arising with potentially dire effects upon the beliefs already present.
Fallibilism pervades our cognitive efforts, thought Peirce17:
fallibilism is the doctrine that our knowledge is never absolute but always swims . . .
in a continuum of uncertainty and indeterminacy. (Hauser 1992: xxii)
Was Peirce also a skeptic? No. He thought that we know, in fallible ways.18 Should all
fallibilists be skeptics, though? As section 1 noted, the usual epistemological view of
fallibilism is that, if we can formulate it coherently, we may evade knowledge-skepticism:
we would understand knowledge’s not needing to attain a standard as impossibly
demanding as an infallibilist one. The quest continues to find such a conception of
fallible knowledge. In the meantime, this chapter has described a few ways in which
fallibilism might be thought to lead us toward knowledge-skepticism, not away from it.19
NOTES
1. Or—if there is collective belief—any actual or possible group of people. For simplicity,
this chapter will focus upon beliefs being possessed by an individual person. The
chapter’s theses and arguments also apply, mutatis mutandis, to collective beliefs and
believers.
2. Since 1963, that thought has not been so epistemologically universal. Section 3b will
say more about this.
3. This is equally so if we seek to understand “it might be false” in terms of
probability, such as with “its truth is not probabilistically certain or guaranteed.”
The probability of a necessarily true belief’s being true, given . . . well,
whatever . . . is 1. The belief is certain, in that sense at any rate. It might be
fallible, even so.
4. “What is the way of thinking used on a given occasion? Is this always manifest?” That
question arises for all theories assessing a belief (its being justified or knowledge)
even partly in terms of how it has been formed and/or is being held. If the question
is not answerable accurately and informatively, those theories face the problem of
generality. For its original formulation, see Feldman (1985).
5. McDowell (2011) advocates conceiving of fallibility only as a property of a belief-
forming capacity: doing so rescues beliefs from skeptical doubts. A perceptual
belief-forming capacity is fallible—he argues—if it can misfire, giving its user a
false perceptual belief; whenever there is no misfiring, the resulting true belief is
knowledge. Indeed, McDowell regards that true belief as conclusive knowledge. So,
his strategy is beside our present aim of examining whether skeptical doubts afflict
beliefs that are fallible.
6. For the original Gettier cases, see Gettier (1963).
7. For overviews of this issue, see Shope (1983) and Hetherington (2011c).
8. This categorization comes from Hetherington (1999).
9. This case’s initial version is from Chisholm (1966: 23 n. 22; 1977: 105; 1989: 93).
10. For this case, see Goldman (1976).
11. “Wait a moment. The situation’s odd circumstance also includes your belief’s
being made true by the real sheep hidden in the field. Once we include this in our
description of how your belief is formed, no longer can we describe as so fallible the
means—the evidence plus the circumstance—by which your belief has arisen. For
now the means is described as generating a true belief.” Yet epistemologists do not
include this further part of the situation’s odd circumstance in their usual assessments
(of knowledge as absent from the situation). Including within their assessment a
description of this further part of the odd circumstance would undermine their usual
claim that the belief is not knowledge. Accordingly, section 3b ignores this further
aspect of the situation, presenting a skeptical argument derived from the standard
interpretation of Gettier cases. For an argument against that standard interpretation,
see Hetherington (2011a: ch. 3).
12. For contemporary formulations of the skeptical reasoning using the dreaming
possibility, see Stroud (1984: ch. 1) and Hetherington (2011b).
13. See Hetherington (1996) for this skeptical argument.
14. This does not imply that any beliefs are false. It is potentially misleading to say, as
Talisse and Aikin do, that “[f]allibilism is the view that any number of our beliefs
are false, and that one should run one’s cognitive life in a way that one is able to
find and revise those false beliefs” (2008: 48). This is misleading unless the number
in question could be zero.
15. These questions need not be understood in terms of Gettier cases.
618 SKEPTICISM
16. I write here of knowing as knowing the world and a state of affairs, maybe not as
knowing a true proposition. On that choice, see Sayre (1997). I assume that a state
of affairs within the world can include within it further states of affairs—maybe only
some of which one need know, in knowing the former state of affairs.
17. On how Peirce intended his fallibilism to encompass mathematical claims, see
Cooke (2006: ch. 3).
18. One of Peirce’s marks of such fallibility is our need as individual inquirers to work
with others’ views (1931–1958: vol. 5, para. 265):
We individually cannot reasonably hope to attain the ultimate philosophy
which we pursue; we can only seek it, therefore, for the community
of philosophers.
As Hookway (2012: 156) puts Peirce’s point:
So long as we are careful in our reasoning, we can trust fallible methods of
belief formation, probably in the hope that any errors will eventually surface
and be corrected; but we can do this responsibly only as members of a
community of inquirers.
Yet unless we define truth so that it is bound to be what the community of scientific
inquirers would eventually-and-then-lingeringly settle upon as true, this chapter’s
skeptical thoughts are no less applicable here than to individual inquirers. And
although Peirce characterized truth in that way, he apparently defined it in terms of
correspondence (1931–1958: vol. 1, para. 578; vol. 2, para. 652; vol. 7, para. 659),
albeit with a pragmatist twist (ibid.: vol. 5, paras. 553–4). On Peirce on fallibilism
and truth, see Almeder (1980: ch. I) and Cooke (2006: ch. 6).
19. For more on the nature of fallibilism, see Hetherington (2005), Dougherty (2011),
and Reed (2012).
REFERENCES
Almeder, Robert. 1980. The Philosophy of Charles S. Peirce: A Critical Introduction.
Oxford: Basil Blackwell.
Chisholm, Roderick M. 1966. Theory of Knowledge. Englewood Cliffs, NJ: Prentice-
Hall.
Chisholm, Roderick M. 1977. Theory of Knowledge, 2nd edn. Englewood Cliffs, NJ:
Prentice-Hall.
Chisholm, Roderick M. 1989. Theory of Knowledge, 3rd edn. Englewood Cliffs, NJ:
Prentice Hall.
Cooke, Elizabeth F. 2006. Peirce’s Pragmatic Theory of Inquiry: Fallibilism and
Indeterminacy. London: Continuum.
Dougherty, Trent. 2011. “Fallibilism.” In S. Bernecker and D. Pritchard (eds.), The
Routledge Companion to Epistemology, 131–143. New York: Routledge.
Feldman, Richard. 1985. “Reliability and Justification,” The Monist 68: 159–174.
Gettier, Edmund L. 1963. “Is Justified True Belief Knowledge?” Analysis 23: 121–123.
Goldman, Alvin. I. 1976. “Discrimination and Perceptual Knowledge,” The Journal of
Hauser, Nathan. 1992. “Introduction.” In N. Hauser and C. Kloesel (eds.), The Essential
Peirce: Selected Philosophical Writings, Vol. 1 (1867–1893), ix–xli. Bloomington and
Indianapolis: Indiana University Press.
Hetherington, Stephen. 1996. “Gettieristic Scepticism,” Australasian Journal of Philosophy
74: 83–97.
Hetherington, Stephen. 2005. “Fallibilism,” Internet Encyclopedia of Philosophy. Available
at http://www.iep.utm.edu/f/fallibil.htm.
Hetherington, Stephen. 2011a. How to Know: A Practicalist Conception of Knowledge.
Malden, MA: Wiley-Blackwell.
Hetherington, Stephen. 2011b. “The Cartesian Dreaming Argument for External-World
Skepticism.” In M. Bruce and S. Barbone (eds.), Just The Arguments: 100 of the Most
Important Arguments in Western Philosophy, 137–141. Malden, MA: Wiley-Blackwell.
(Corrected version available at http://justthearguments.com/?page_id=49.)
Hetherington, Stephen. 2011c. “The Gettier Problem.” In S. Bernecker and D. Pritchard
(eds.), The Routledge Companion to Epistemology, 119–130. New York: Routledge.
Hookway, Christopher. 2012. “American Pragmatism: Fallibilism and Cognitive Progress.”
In S. Hetherington (ed.), Epistemology: The Key Thinkers, 153–171. London:
Continuum.
McDowell, John. 2011. Perception as a Capacity for Knowledge. Milwaukee: Marquette
University Press.
Peirce, Charles Sanders. 1931–1958. Collected Papers of Charles Sanders Peirce, vols. I–VI,
(eds.) C. Hartshorne and P. Weiss, vols. VII–VIII, (ed.) A. W. Burks. Cambridge, MA:
Harvard University Press.
Reed, Baron. 2012. “Fallibilism,” Philosophy Compass 7/9: 585–596.
Sayre, Kenneth M. 1997. Belief and Knowledge: Mapping the Cognitive Landscape. Lanham,
MD: Rowman & Littlefield Publishers.
Shope, Robert K. 1983. The Analysis of Knowing: A Decade of Research. Princeton:
Princeton University Press.
Talisse, Robert, and Scott Aikin. 2008. Pragmatism: A Guide for the Perplexed. London:
Continuum.
CHAPTER FORTY-THREE
Skepticism
and Contextualism1
MICHAEL BLOME-TILLMANN
1 WHAT IS EPISTEMIC CONTEXTUALISM?

Let’s begin with an example.2 Imagine schoolteacher Jones in the zoo explaining to
her class that the animals in the pen are zebras. Tom is unconvinced and challenges
Jones: “Are you sure those aren’t antelopes?” After Jones has explained the
difference between antelopes and zebras, Tom assures his classmates:
(1) She knows that the animals in the pen are zebras.
Has Tom spoken truly? Surely, Jones’s epistemic position seems good enough for
satisfying the predicate ‘knows that the animals in the pen are zebras’ (henceforth
‘knows Z’): Jones has visual experiences of a black and white striped horse-like
animal, she can discriminate reliably between zebras and antelopes, she has read the
sign on the pen that reads “Zebra Pen,” etc. Thus, Tom’s utterance of (1) seems to
be a paradigm case of a true ‘knowledge’ attribution.
Next, imagine a couple, Bill and Kate, walking along. Bill, a would-be
postmodernist artist, gives details of his latest ideas: he envisions himself painting
mules with white stripes to look like zebras, putting them in the zebra pen of a zoo
and thereby fooling visitors. Our couple randomly considers Jones, and Kate claims,
at the same time as Tom is asserting (1):
(2) She doesn’t know that the animals in the pen are zebras.
In Kate’s mind, for Jones to “know Z,” she must have better evidence or reasons
in support of Z than are momentarily available to her. In particular, Kate has it
that Jones’s evidence must eliminate the possibility that the animals in the pen are
painted mules. As long as her evidence, however, is neutral with respect to whether
or not the animals are cleverly painted mules, Kate claims, Jones doesn’t qualify as
‘knowing Z’.
What is going on in our little example? According to our intuitions, an utterance
of (1) is true in the context of the school class, while an utterance of (2) is true in the
context of the artistic couple. Moreover, (2) is the negation of (1): it doesn’t differ
from (1) except for containing the verbal negation ‘doesn’t’. And since the personal
SKEPTICISM AND CONTEXTUALISM 621
pronoun ‘she’ refers in both contexts to Jones, it seems that the schoolteacher
satisfies the predicate ‘knows Z’ in the context of the school class but not so in the
context of the artists.
How are we to account for these phenomena? First, note again that Tom and
Kate are talking about one and the same person—Jones—at exactly the same time.
Thus, we cannot resolve the situation by claiming, for instance, that Jones “knows
Z” in one context but not the other because she believes Z in one context but not
in the other. Similarly, we cannot plausibly respond that Jones has certain visual
experiences in one context that she is lacking in the other, or that she has the ability
to discriminate reliably between certain scenarios in one context but not the other,
or, finally, that she has read the sign on the zebra pen in one context but not in
the other. All factors pertaining to the subject are identical with respect to both
contexts, as the speakers in both contexts—Tom and Kate—are talking about one
and the same subject at one and the same time.3
Thus, what the above example suggests is that the mentioned factors—Jones’s
visual experiences, her discriminatory abilities, etc.—are sufficient for her to satisfy
‘knows Z’ in one context, but not so in the other. And it is this view that epistemic
contextualism (EC) takes at face value: how strong one’s epistemic position toward p
must be for one to satisfy ‘know(s) p’ may vary with the context of utterance. In the
artists’ conversational context, Jones needs to be in a stronger epistemic position—
she needs more evidence in support of Z—than in the school class’s conversational
context in order for her to satisfy ‘knows Z’. In fact, some contextualists describe the
situation by claiming that contexts of utterance are governed by so-called epistemic
standards.4 Given this terminology, the epistemic standards in our example are lower
in the context of the school class than in the context of the artists’ conversation. In
fact, in the former context, the standards are low enough for Jones to satisfy ‘knows
Z’, while in the latter they are too high: Jones doesn’t, in the artists’ context, satisfy
‘knows Z’ but, rather, satisfies ‘doesn’t know Z’. Now, the notion of an epistemic
standard can be explicated in a number of different ways. On the most popular way,
which is inspired by relevant alternatives approaches to contextualism, epistemic
standards are said to be higher in the school class’s context because, as David Lewis
(1996) puts it, satisfying ‘knows Z’ in that context doesn’t require the elimination
of the possibility that the animals are painted mules, while this is required in the
context of the postmodernist artists: more alternatives must be eliminated in the
context with the higher standards than in the context with the lower standards.
Given the hypothesized context-sensitivity of the predicate ‘know(s) p’, it is in
general possible that a subject satisfies the predicate in one conversational context
but doesn’t do so in another, or, in other words, that somebody in a given context
speaks truly when uttering a sentence of the form ‘x knows p’ while somebody in a
different context speaks falsely when uttering the very same sentence—even though
both speakers are speaking about the same subject x at the same time of utterance
t. Epistemic contextualism is, as a consequence, a linguistic or a semantic view—
namely, the view that ‘knowledge’-ascriptions—sentences of the form ‘x knows
p’—may express different propositions in different contexts of utterance. According
to EC, ‘knowledge’-ascriptions are, as Stanley (2005: 17) puts it, context-sensitive
622 SKEPTICISM
in a distinctively epistemological way: the content of a sentence S containing the

predicate ‘know(s) p’ can change with context, independently of whether S contains
further indexicals, is ambiguous, or is context-sensitive in any other way.
In a first approximation, we can thus define ‘epistemic contextualism’ as follows:
(ECʹ) Knowledge ascriptions may express different propositions relative to
different contexts of utterance, where this difference is traceable to the occurrence
of ‘know(s) p’ and concerns a distinctively epistemic factor.
Given semantic compositionality—the view that the content or semantic value of
complex expressions is a function of its ultimate constituents and the way in which
they are combined—(ECʹ) entails (EC):
(EC) The content of the predicate ‘know(s) p’ may vary with the context of
utterance in a distinctly epistemic way.
According to EC, the predicate ‘know(s) p’ adds context-sensitivity to a sentence
it occurs in, and this context-sensitivity is distinctly epistemic—that is, it goes over
and above the context-sensitivity that the verb contributes to the sentence by virtue
of its tense.5
The exact details as to how to semantically model the context-sensitivity of
‘know(s) p’ shall not concern us in this chapter. However, it is worth noting
that EC is not a lexical ambiguity theory—that is, it doesn’t claim that ‘know(s)
p’ is assigned multiple conventional meanings in English, as are the lexically
ambiguous expressions ‘bank’ or ‘orange’. On the contrary, contextualists have
commonly compared ‘know(s) p’ to indexical expressions, such as ‘I’, ‘that’, and
‘today’ or to gradable adjectives such as ‘flat’ and ‘empty’: these expressions are
widely taken to have only one conventional meaning—what Kaplan (1989) calls
their “character”—but different contents or semantic values in suitably different
contexts of utterance.
While there are several distinct ways to semantically model the context-sensitivity
of ‘know(s) p’, contextualists have often stressed an analogy between the semantics
of ‘know(s) p’ and the semantics of gradable adjectives such as ‘flat’, ‘empty’, or
‘tall’: just as what counts as satisfying ‘flat’, ‘empty’, or ‘tall’ may vary with context,
contextualists have argued, what counts as satisfying ‘know(s) p’ may vary with
context, too.6 Given this analogy, EC can be construed as claiming that (1) and (2)
in our above example stand in a relation similar to the relation between a basketball
coach’s utterance of “MB-T isn’t tall” and a jockey coach’s utterance of “MB-T
is tall”: while the surface syntax of these sentences suggests a contradiction, the
propositions expressed are compatible, as the semantic value of ‘tall’ changes with
the context of utterance. Ordinary usage of ‘tall’ and ‘know’ seem to be similar:
both expressions seem to be context-sensitive.7
2 EVIDENCE FOR CONTEXTUALISM

The main evidence for EC derives from our intuitions about the truth-values of
‘knowledge’-ascriptions in examples such as the above Zebra Case. However, there
are further, more familiar examples that have been presented in support of EC in the
literature. Consider, for instance, Stewart Cohen’s (1999: 58) Airport Case:
Mary and John are at the L.A. airport contemplating taking a certain flight to New
York. They want to know whether the flight has a layover in Chicago. They overhear
someone ask a passenger, Smith, if he knows whether the flight stops in Chicago.
Smith looks at the flight itinerary he got from the travel agent and responds, “Yes
I know—it does stop in Chicago.” It turns out that Mary and John have a very
important business contact they have to make at the Chicago airport. Mary says,
“How reliable is that itinerary? It could contain a misprint. They could have changed
the schedule at the last minute.” Mary and John agree that Smith doesn’t really know
that the plane will stop in Chicago. They decide to check with the airline agent.
As Cohen’s example suggests, the sentence “Smith knows that the flight stops
over in Chicago” seems true as uttered in Smith’s context but false as uttered in
Mary’s and John’s context. Moreover, it seems as though the practical interests
and goals of the conversational participants or how high the stakes are with regard
to the proposition that the flight will stop over in Chicago influence the respective
contexts’ epistemic standards, and thus whether Smith satisfies ‘knows’: in Smith’s
own context he satisfies ‘knows that the flight will stop over in Chicago’, but in
Mary’s and John’s context, where the stakes are significantly higher, he doesn’t.
Here is another example reinforcing this point—namely, Keith DeRose’s (1992)
famous Bank Cases, as presented by Stanley (2005: 3–4):
Low Stakes
Hannah and her wife Sarah are driving home on a Friday afternoon. They
plan to stop at the bank on the way home to deposit their paychecks. It is not
important that they do so, as they have no impending bills. But as they drive
past the bank, they notice that the lines inside are very long, as they often are on
Friday afternoons. Realizing that it isn’t very important that their paychecks are
deposited right away, Hannah says, “I know the bank will be open tomorrow,
since I was there just two weeks ago on Saturday morning. So we can deposit our
paychecks tomorrow morning.”
High Stakes
Hannah and her wife Sarah are driving home on a Friday afternoon. They plan
to stop at the bank on the way home to deposit their paychecks. Since they have
an impending bill coming due, and very little in their account, it is very important
that they deposit their paychecks by Saturday. Hannah notes that she was at the
bank two weeks before on a Saturday morning, and it was open. But, as Sarah
points out, banks do change their hours. Hannah says, “I guess you’re right. I
don’t know that the bank will be open tomorrow.”
Similar to Cohen’s example, DeRose’s case suggests that it is more difficult to satisfy
‘knows’ in a context in which the stakes are higher: in Low Stakes Hannah satisfies
‘knows that the bank will be open tomorrow’, whereas in High Stakes she doesn’t,
even though she is in exactly the same epistemic position in both cases. And, again,
624 SKEPTICISM
the defender of EC argues that the difference in our intuitions in the two bank cases
is due to the fact that the relevant ‘knowledge’-ascriptions are made in different
conversational contexts: in the context of High Stakes, the argument goes, it is
considerably more difficult to satisfy ‘knows that the bank is open on Saturday’ than
it is in the context of Low Stakes.8
The above examples and others like them have attracted a large amount of
critical attention in recent years. Note, for instance, that the argument in support of
EC emerging from the above data amounts to an inference to the best explanation:
in evaluating the support EC receives from the examples, we must therefore
compare EC’s account of the data with competing explanations of rival theories.
Such comparisons of explanatory virtue in philosophy, however, are usually rather
difficult and complicated. A further challenge to EC’s account of the above data
pertains to the data’s evidential status: some practitioners of so-called experimental
philosophy have, in recent studies, aimed to undermine the contextualists’ case by
arguing that the professional philosopher’s intuitions about the above cases do not
coincide with the intuitions of the more general public. Now, while the above issues
present interesting and legitimate challenges to EC, we shall, in this chapter, leave
them to one side and focus our attention on a topic that is commonly taken to
provide an important philosophical motivation for EC: skeptical puzzles.9
3 CARTESIAN SKEPTICAL PUZZLES

We have seen so far that EC receives some prima facie support from certain linguistic
data—namely, the examples discussed in the previous sections. Traditionally,
however, contextualists have also claimed that besides the above empirical evidence
in favor of their views there is also an independent, philosophical motivation for EC.
More specifically, contextualists have argued that the context-sensitivity of ‘know(s)
p’ suggested by the above examples provides us with an attractive resolution of
skeptical puzzles. Now, even though the issue of skepticism has somewhat moved into
the background in recent discussions of EC, it is certainly worthwhile considering the
traditional contextualist treatment of skepticism in some detail.10
To begin our discussion of the contextualist response to skepticism, consider the
following argument:
Skeptical Argument
(i) If I know that I have hands, then I’m in a position to know that I’m not a
handless brain in a vat.
(ii) I’m not in a position to know that I’m not a handless brain in a vat.
(iii) I don’t know that I have hands.11
The above argument is valid: if we accept its premises, we must accept its conclusion,
too. Moreover, the above Skeptical Argument leaves us with a philosophical puzzle:
its premises are highly plausible while its conclusion is highly implausible. One way
to bring this out in more detail is to consider the negation of its conclusion:
(iv) I know that I have hands.
The propositions (i), (ii), and (iv) form an inconsistent set; and so at least one of
them has to be rejected. However, merely rejecting one of the members of our set
doesn’t amount to a satisfactory resolution of our puzzle. As Stewart Cohen (1988:
94) has pointed out, an intellectually satisfying resolution of the skeptical puzzle
doesn’t merely block the argument by identifying the culprit. Rather, a satisfactory
resolution of the skeptical puzzle must, in addition, offer us an explanation of why
the false member of the set appeared so plausible at first glance (see also Cohen
1999: 63). Why is it that our intuitions about the truth-values of at least one of the
propositions at issue are misguided? And what exactly is the mistake we have made
when we find ourselves puzzled by the Skeptical Argument?
4 THE CONTEXTUALIST SOLUTION TO

SKEPTICAL PUZZLES
The traditional contextualist’s response to the skeptical puzzle is to claim that the
Skeptical Argument is unsound in conversational contexts that are governed by our
moderate everyday epistemic standards but sound in contexts with artificially high,
skeptical epistemic standards.12 In everyday contexts, the argument goes, I satisfy
‘knows that he isn’t a handless brain in a vat’ and premise (ii) of the Skeptical
Argument expresses a falsehood: if I satisfy, in ordinary contexts, ‘knows that he
isn’t a handless brain in a vat’, then I also satisfy, in ordinary contexts, ‘is in a
position to know that he’s not a handless brain in a vat’. Consequently, the Skeptical
Argument is unsound in ordinary contexts, and its conclusion doesn’t follow:
relative to ordinary contexts, I satisfy ‘knows that he has hands’ and the conclusion
of the Skeptical Argument expresses a falsehood.
However, as indicated already, things are different in so-called skeptical contexts
in which we practice epistemology and consider and discuss skeptical scenarios such
as the brain-in-a-vat scenario. In such contexts, the argument goes, the epistemic
standards are considerably higher—in fact, outrageously high—to the effect that
premise (ii) expresses a truth in such contexts. For instance, contextualists have argued
that because skeptical possibilities are epistemically relevant in skeptical contexts,
premise (ii) of the Skeptical Argument expresses a truth: skeptical possibilities are,
after all, uneliminated by our evidence, and we therefore do not, in skeptical contexts
in which they are relevant, satisfy the predicate ‘is in a position to know that s/he is
not a handless brain in a vat’. Consequently, when the skeptic asserts, in her skeptical
context, “MB-T doesn’t know that he has hands,” she asserts a truth. However, it is
crucial to emphasize that the truth of the skeptic’s assertion does not affect the truth-
values of my positive ‘knowledge’-ascriptions in ordinary contexts.
To illustrate this view further, it is worth noting that traditional contextualists aim
to resolve the tensions between our anti-skeptical intuitions on the one hand and the
intuition that skeptical arguments are sound (and their conclusions, therefore, true)
on the other. We can represent these intuitions as follows:
Anti-Skeptical Intuition (ASI)
People often speak truly when they assert ‘I know p.’
626 SKEPTICISM
Skeptical Intuition (SI)

People sometimes speak truly when they assert ‘Nobody knows p’ in contexts in
which skeptical arguments are discussed.
The traditional contextualist claims that both of these intuitions are correct and
only seemingly contradictory: they are correct because the semantic value of
‘know(s) p’ varies with the context of utterance; so when we claim in everyday
contexts that we ‘know p’, our utterances are not in contradiction to our utterance
of ‘Nobody knows p’ in skeptical contexts. Thus, it is crucial to note that, according
to the traditional contextualist, it is not (iv) which is shown to be true or (iii)
which is shown to be false. Rather, the traditional contextualist emphasizes that
our skeptical and anti-skeptical intuitions are exclusively intuitions about the truth-
values of utterance tokens, which are by their very nature situated in particular
conversational contexts. Our intuitions are not about the truth-values of sentences
as considered more or less in the abstract in a philosophical essay or discussion.
Once we appreciate this point and take into account the context-sensitivity of
‘knowledge’-attributions, the skeptical puzzle is—the traditional contextualist
argues—easily dissolved: the argument is sound in contexts with exceedingly
high or skeptical epistemic standards, but unsound in contexts with ordinary or
everyday epistemic standards.
5 ERROR THEORY AND CONTEXTUAL SHIFTS

An important question arises at this point: if the semantic value of ‘know’ can in
fact change in a way allowing for both (ASI) and (SI) to be true, why, then, are
we initially puzzled by skeptical arguments? Shouldn’t we be somehow sensitive
to or aware of the context-sensitivity of ‘know(s) p’ and avoid making the mistake
the contextualist ascribes to us? If EC is true, why, in other words, are skeptical
arguments puzzling in the first place?
To account for the puzzling nature of skeptical arguments, the contextualist is
committed to the view that we sometimes lose sight of the context-sensitivity of
epistemic terms, and in particular that we do so when confronted with skeptical
arguments.13 Therefore, according to the traditional contextualist, we do not
recognize that the skeptical conclusion is true in the context of discussions
of skeptical arguments while false in everyday contexts. Thus, the traditional
contextualist argues that, when we are puzzled by skeptical arguments, we fail to
realize that the propositions expressed by the arguments’ conclusions are perfectly
compatible with the propositions expressed by our everyday ‘knowledge’-claims.
Contextualists accordingly pair their semantics of ‘know(s) p’ with the view that we
are sometimes unaware of or tend to overlook the context-sensitivity of ‘know(s) p’.
Stephen Schiffer, in an important paper criticizing this view, aptly calls this element
of standard contextualism its error theory (see Schiffer 1996). Ultimately, it is, of
course, entirely due to this error theory that EC can claim to be able to account
for both the plausibility of skeptical arguments and our intuition that our everyday
‘knowledge’-ascriptions express truths.
We shall return to the plausibility of EC’s error theory later on in this chapter
(in Section 7). In the meantime, however, note that the contextualist’s resolution of
skeptical puzzles makes crucial use of the intuitive notion of a shift in or variation of
epistemic standards between contexts. It is important to note at this point, however,
that arguing for the context-sensitivity of ‘know(s) p’ on the basis of the claim that
the Bank Case, the Airport Case, and our Zebra Case involve contextual variation of
epistemic standards is one thing, while claiming that epistemic standards are “shifty”
in precisely the way required for a resolution of skeptical puzzles is quite a different
proposition. In fact, for the contextualist’s resolution of skeptical puzzles to be
credible, we need to be told more about what exactly the mechanisms underlying
contextual shifts and variations are. In other words, we need to be told more about
what epistemic standards are and how they are determined by context: why is it that
epistemic standards are high in so-called skeptical contexts, and why are they lower
in everyday situations?
Different contextualists have different stories to tell about what epistemic
standards are and about how they shift and vary. But it is fair to say that the original
approaches defended by early contextualists (such as DeRose (1995) and Lewis
(1996)) are highly problematic and must be refined and amended substantially
before we can grant the contextualist that her view offers a resolution of skeptical
puzzles. However, a detailed discussion of the more recent literature on epistemic
standards is beyond the scope of this chapter.14
6 CLOSURE
Another important aspect of the contextualist resolution of skeptical puzzles to
be mentioned here is that the contextualist resolves skeptical puzzles while fully
respecting our intuition that one can extend one’s knowledge by competent
deduction. To see what I have in mind here, consider the following principle, which
is familiar under the label Single-Premise Closure:
(CL) If x knows p and x knows that p entails q, then x is in a position to know q.
Here is an instance of (CL) for illustration: if, first, I know that the animal outside
my window is a fox and if, secondly, I know that its being a fox entails that it’s not
a cat, then I am also in a position to know that the animal outside my window is
not a cat. Of course, I am in a position to know that latter proposition because I
can competently deduce it from (i) my knowledge that the animal is a fox and (ii)
my knowledge that its being a fox entails that it is not a cat. Single-premise closure
captures fairly precisely the intuition that one can extend one’s knowledge by means
of deductive reasoning.
Now, while some epistemologists have argued that giving a response to the
skeptic requires us to give up (CL), the contextualist resolution gets by without
any such move.15 How does the contextualist avoid rejecting closure? Note that,
according to EC, every semantic value that the verb ‘knows’ can express in a given
context is, loosely speaking, closed under “known” entailment. Here is a more
628 SKEPTICISM
precise and contextualized formulation of the single-premise principle to illustrate

the idea:
(CLC) If x satisfies ‘knows p’ in context C and satisfies ‘knows that p entails q’ in
C, then x is in a position to satisfy ‘knows q’ in C.
(CLC) is a meta-linguistic principle. Loosely speaking, (CLC) says that (CL), its
non-contextualized cousin, expresses a truth as long as the conversational context
is kept fixed.16 Thus, unlike Nozick (1981) and Dretske (1970), who reject (CL) in
giving a response to the skeptic, the contextualist, by ascending semantically, merely
modifies and clarifies (CL) in a way that respects our intuitions about the validity
of the closure principle. Contextualists have traditionally taken this to be a great
comparative advantage of their theories over epistemological theories that advocate
closure failure.
7 CRITICISMS OF CONTEXTUALISM
Epistemic contextualism has been criticized for a number of reasons. In this
chapter, we shall focus on two criticisms of EC that have figured most prominently
in the recent literature. First, Ernest Sosa has wondered what the epistemological
relevance of contextualism is, given that EC is a linguistic view—that is, a view about
the predicate ‘know(s) p’ and its content rather than about knowledge. Considering
the contextualist’s resolution of skeptical puzzles sketched above, this worry may
appear somewhat surprising. However, Sosa thinks that EC, even though true, has
only little epistemological relevance, if any at all:
The main thesis of [EC] has considerable plausibility as a thesis in linguistics or
in philosophy of language. In applying it to epistemology, however, it is possible
to overreach. (Sosa 2011: 98)
What, then, is Sosa’s objection? To see what Sosa has in mind, let us introduce some
technical language. Let ‘KE’ express the content of ‘know’ in everyday contexts and
let ‘KS’ express the content of ‘know’ in skeptical contexts. Now consider (3), which
we derive from (ASI) by disquotation:
(3) People often speak truly when they assert that they know p.
Depending on whether the epistemic standards of our present context are those of
everyday contexts or those of skeptical contexts (3) expresses the content of either
(4) or (5):
(4) People often speak truly when they assert that they KE p.
(5) People often speak truly when they assert that they KS p.
Since Sosa assumes that contexts of epistemological enquiry are inevitably skeptical
contexts, Sosa thinks that (3), in the context of this chapter, expresses the proposition
expressed by (5). Now, the alleged problem for EC is that (5) is clearly false, for it
suggests that people in quotidian contexts assert that they KS p. However, when
people in quotidian contexts use the word ‘know’, its semantic value is always KE
rather than KS. Thus, Sosa complains that contextualists convey a falsehood, when
they assert (3) in a context of epistemological discussion.
How serious an objection is this to EC? The obvious reply to Sosa’s objection
is, of course, that the contextualist ought to distinguish more carefully between the
mention and the use of ‘know’ and thus only assert (ASI) instead of the disquoted (3):
Sosa’s objection rests on a conflation of the use/mention-distinction.17 Moreover,
it is worthwhile noting that more recent contextualists have developed accounts
according to which contexts of philosophical and epistemological enquiry are by
no means automatically skeptical contexts.18 On these more moderate views (3), in
the context of this chapter, expresses the content of (4) rather than that of (5), and
I therefore speak truly when, in the context of this chapter and the epistemology
classroom more generally, asserting (3). As a consequence, there are several ways the
contextualist can respond to Sosa’s charge that EC is epistemologically irrelevant.
Another recently influential type of objection to EC proceeds by highlighting
disparities between certain linguistic properties of ‘know(s) p’ on the one hand and
more recognized context-sensitive expressions on the other. Remember that, for
instance, contextualists compare ‘know(s) p’ to gradable adjectives, such as ‘flat’,
‘empty’, and ‘tall’. However, as Jason Stanley (2005, ch. 2) has pointed out, “know(s)
p” has very different syntactic properties from gradable adjectives: as Stanley shows
in great detail, ‘know(s) p’ is not syntactically gradable and doesn’t accept a large
number of constructions that gradable adjectives can be felicitously combined with.
Similarly, it has been pointed out that our semantic blindness toward the context-
sensitivity of ‘know(s) p’—discussed above under the label of EC’s error theory—
is not observed in connection with recognized indexicals such as ‘I’, ‘here’, and
‘today’. This fact is illustrated by the following principle:
(6) If an English speaker E sincerely utters a sentence S of the form ‘A knows
that p’, and the sentence in the that-clause means that p and ‘A’ is a name
or indexical that refers to a, then E believes of a that a knows that p, and
expresses that belief by S.
As John Hawthorne (2004: 101) points out, (6) seems entirely natural. But, of
course, (6) is false if ‘know(s) p’ is context-sensitive: since E’s context might be
governed by epistemic standards that are different from those operative in this
chapter, the disquotation of ‘knows’ in (6) is illegitimate. Since the possibility of
an asymmetry between E’s and our own epistemic standards is largely hidden from
competent speakers, the contextualist must accept that the context-sensitivity of
‘knows’ is non-obvious: it is not as readily detected by competent speakers as the
context-sensitivity of core indexicals.
Interestingly, however, similar phenomena are not observed with respect to ‘I’.
Consider (7), a disquotation principle for ‘I’:
(7) If an English speaker E sincerely utters a sentence S of the form ‘I am
hungry’, then E believes that I am hungry, and expresses that belief by S.
Clearly, it is not the case that every English speaker who utters “I am hungry”
believes that I, MB-T, am hungry.
630 SKEPTICISM
What is worth emphasizing in response to Hawthorne’s worry, however, is that

the gradable adjectives ‘flat’ and ‘empty’ display somewhat similar behavior with
respect to disquotation. Consider the following disquotation principles for ‘flat’
and ‘empty’:
(8) If an English speaker E sincerely utters a sentence S of the form ‘A is flat’,

and ‘A’ is a name or indexical that refers to a, then E believes of a that a is
flat, and expresses that belief by S.
(9) If an English speaker E sincerely utters a sentence S of the form ‘A is empty’,
and ‘A’ is a name or indexical that refers to a, then E believes of a that a is
empty, and expresses that belief by S.
As the intuitive plausibility of (8) and (9) demonstrates, the context-sensitivity of

‘flat’ and ‘empty’ is just as non-obvious or hidden from competent speakers as the
context-sensitivity of ‘knows’. Thus, on the assumption that gradable adjectives
are in fact context-sensitive, the context-sensitivity of ‘know(s) p’ has been shown
to be no more puzzling or mysterious than the comparatively humdrum context-
sensitivity of ‘flat’ and ‘empty’. It is due to data such as these that there is a growing
consensus in the literature that the semantic blindness objection is not all that
damaging to contextualism.
Finally, it is worth emphasizing that, even if ‘know(s) p’ varies in certain linguistic
respects—whether semantic, syntactic, or pragmatic—from other recognized
context-sensitive expressions, it is not clear whether the contextualist should be
worried about such a discovery. For why shouldn’t we accept that ‘know(s) p’ is
linguistically exceptional? Let’s not forget, after all, that ‘know(s) p’ combines a
number of fairly interesting and unique properties: ‘know(s) p’ is a factive verb
that accepts a sentential complement, and its satisfaction at a context is, arguably,
the norm of assertion, practical reasoning, and belief at that context.19 Moreover,
‘know(s) p’ gives rise to skeptical puzzles, which surely makes the predicate rather
unique. This combination of properties is no doubt unique, and we should therefore
not expect ‘know(s) p’ to function in each and every linguistic respect exactly like
other context-sensitive expressions. Thus, if EC should in fact commit us to the
uniqueness of ‘know’, then this shouldn’t worry us too much, as long as a coherent,
illuminating, and systematic account of this uniqueness can be given.
8 CONCLUSION
In summary, epistemic contextualism not only offers an interesting approach to
skeptical puzzles but also is motivated by a large set of empirical data. To be sure,
the philosophical jury is still out on epistemic contextualism: the view is, after all,
still rather contentious and hotly debated in the literature. However, it is undeniable
that EC is nowadays rather popular, not only among epistemologists but also among
philosophers more generally.20 And as we begin to achieve an increasingly better
understanding of natural language semantics in general and linguistic context-
sensitivity in particular, contextualists may hope that EC will someday become one
of the progressively more standard views in philosophy. At least to the author’s

mind, the prospects are rather bright that EC will one day be considered as making
a lasting and important contribution to our understanding of skepticism and
skeptical puzzles.
NOTES
1. This work was supported by Marie Curie Actions (PIIF-GA-2012-328969
“Epistemic Vocabulary”).
2. The following example is derived from the Zebra Case in Dretske (1970).
3. Similarly, it follows that we cannot plausibly explain the phenomena by pointing
out that Jones is in a different practical situation with respect to the two contexts:
she isn’t. This fact about the above example provides problems for so-called Subject-
Sensitive Invariantist accounts of knowledge. See, for instance, Fantl and McGrath
(2002); Hawthorne (2004); Stanley (2005).
4. Cf. Cohen (1988); DeRose (1995); Lewis (1996).
5. Cf. Schaffer and Szabó (2014) for the above definition of EC.
6. For this analogy see Unger (1975); DeRose (1995); Lewis (1996); Cohen
(1999, 2004).
7. Not all contextualists endorse the analogy to gradable adjectives (see, for instance,
Schaffer et al. [forthcoming]), but it is at this point helpful in illustrating the general
concept of context-sensitivity underlying the view.
8. Note that since the relevant ‘knowledge’-ascriptions in DeRose’s bank cases as
presented above are made from the first-person perspective—they are so-called
self-ascriptions—the data from DeRose’s example are not suited to support EC
over certain rival accounts such as Subject-Sensitive Invariantism (SSI): according
to SSI, whether one knows p depends in part on one’s own and thus the subject’s
(as opposed to the ascriber’s) epistemic standards (see Hawthorne (2004); Stanley
(2005); Fantl and McGrath (2009) for versions of this view). A straightforward
subject-sensitive invariantist explanation, however, is not available for Cohen’s
Airport Case and the Zebra Case in Section 1 of this chapter.
9. See DeRose (2011) for a critical discussion of some experimental philosophy papers
in the area.
10. Ludlow (2005), for instance, defends EC purely on the basis of the linguistic data,
leaving aside the issue of skepticism entirely.
11. Here is a formalized version of the argument, where ‘sh’ is shorthand for
‘skeptical hypothesis’ and where ‘op’ ranges over ordinary propositions about the
external world:
(i) Kp → ◊K¬sh. - A
(ii) ¬◊K¬sh. - A
(iii) ¬Kp. i, ii MT
12. The following is a description of standard contextualist views on skeptical puzzles,
as it can be found—more or less explicitly—in all major writings of contextualists.
See, for instance, DeRose (1995) and Cohen (1999).
632 SKEPTICISM
13. Cf. Cohen (1988: 106); DeRose (1995: 40).

14. Lewis’s (1996) “Rule of Attention” and DeRose’s (1995) “Rule of Sensitivity” offer
accounts of contextual shifts that, on the face of it, are useful for the resolution of
skeptical puzzles, but that turn out to be problematic for independent reasons. For
criticism of DeRose’s approach, see Blome-Tillmann (2009a), and, for a refinement
of Lewis’s relevant alternatives approach to contextualism that avoids a number of
problems, see Blome-Tillmann (2009b).
15. Note also that the skeptic uses (CL) when motivating premise (i) of her argument:
she reasons from the assumption that I know that my having a hand entails that I am
not a handless brain in a vat to the conclusion that if I know that I have a hand, then
I am in a position to know that I am not a handless brain in a vat, i.e., to premise
(i) of the Skeptical Argument. The precise reasoning is as follows: Closure Argument
for (i)
(1) (Kp^K(p → ¬sh)) → ◊K¬sh. from CL
(2) K(p → ¬sh) → (Kp → ◊K¬sh). 1 Exp
(3) K(p → ¬sh). - A
(i) Kp → ◊K¬sh. 2, 3 MP
Note that there are other ways to motivate (i), but I shall ignore them in this chapter
(see, for instance, Brueckner (1994)).
16. See DeRose (1995); Lewis (1996); Cohen (1999).
17. See Blome-Tillmann (2007) for more details on this line of reasoning.
18. See Blome-Tillmann (2009b) and Ichikawa (2011a, b).
19. See Blome-Tillmann (2013) for a discussion of the knowledge norms in a
contextualist account.
20. See Chalmers et al. (2014), whose data suggest that EC is the most widely accepted
view in the semantics of ‘knowledge’-attributions.
REFERENCES
Blome-Tillmann, Michael. 2007. “Contextualism and the Epistemological Enterprise,”
Proceedings of the Aristotelian Society, New Series 107 (1/3): 387–394.
Blome-Tillmann, Michael. 2009a. “Contextualism, Safety, and Epistemic Relevance,”
Blome-Tillmann, Michael. 2009b. “Knowledge and Presuppositions,” Mind 118 (470):
241–294.
Blome-Tillmann, Michael. 2013. “Contextualism and the Knowledge Norms,” Pacific
Chalmers, David J., and David Bourget. 2014. “What Do Philosophers Believe?”
Cohen, Stewart. 1988. “How to Be a Fallibilist,” Philosophical Perspectives 2: 91–123.
Cohen, Stewart. 1999. “Contextualism, Skepticism, and the Structure of Reasons,”

Philosophical Perspectives 13: 57–89.
Cohen, Stewart. 2004. “Contextualism and Unhappy-Face Solutions: Reply to Schiffer,”
DeRose, Keith. 1992. “Contextualism and Knowledge Attributions,” Philosophy and
DeRose, Keith. 2011. “Contextualism, Contrastivism, and X-Phi Surveys,” Philosophical
Studies 156: 81–110.
Dretske, Fred. 1970. “Epistemic Operators,” Journal of Philosophy 67: 1007–1023.
Fantl, Jeremy, and Matthew McGrath. 2002. “Evidence, Pragmatics and Justification,”
Fantl, Jeremy and Matthew McGrath. 2009. Knowledge in an Uncertain World. Oxford:
Hawthorne, John. 2004. Knowledge and Lotteries. Oxford: Oxford University Press.
Ichikawa, Jonathan. 2011a. “Quantifiers and Epistemic Contextualism,” Philosophical
Studies 155: 383–398.
Ichikawa, Jonathan. 2011b. “Quantifiers, Knowledge, and Counterfactuals,” Philosophy
Kaplan, David. 1989. “Demonstratives.” In J. Almog, J. Perry, and H. Wettstein (eds.),
Themes from Kaplan, 481–563. Oxford and New York: Oxford University Press.
Ludlow, Peter. 2005. “Contextualism and the New Linguistic Turn in Epistemology.” In
G. Preyer and G. Peter (eds.), Contextualism in Philosophy: Knowledge, Meaning and
Truth, 11–50. Oxford: Oxford University Press.
Nozick, Robert. 1981. Philosophical Explanations. Oxford: Oxford University Press.
Schaffer, Jonathan and Zoltan G. Szabó. 2014. “Epistemic Comparativism: A Contextualist
Semantics for Knowledge Ascriptions,” Philosophical Studies 168: 491–543.
Schiffer, Stephen. 1996. “Contextualist Solutions to Scepticism,” Proceedings of the
Aristotelian Society 96: 317–333.
Sosa, Ernest. 2011. Kowing Full Well. Princeton: Princeton University Press.
Stanley, Jason. 2005. Knowledge and Practical Interests. Oxford: Oxford University Press.
Unger, Peter. 1975. Ignorance: A Case for Scepticism. Oxford: Clarendon.
CHAPTER FORTY-FOUR
External World Skepticism

RAM NETA
1 INTRODUCTION
Skepticism has a long history in philosophy. But skeptical concern with “the
external world” is a more recent phenomenon. The phenomenon is sometimes
thought to have originated in the early modern period, perhaps with Descartes
(1993) or Locke (1975). But it is not clear that this is true: to know whether
Descartes’s argument for the existence of what he called “extended substance,”
or whether Locke’s claim to have sensitive knowledge of what he called “material
substance,” constitute attempts to address skepticism about “the external world,”
we would have to understand not only what Descartes and Locke meant by those
phrases, but also what we mean by the phrase “the external world.” I begin with a
discussion of that issue.
2 WHAT IS THE “EXTERNAL WORLD”?

One way to approach the task of trying to understand what it is that early-twenty-
first-century philosophers who’ve been trained in the Anglophone tradition might
have in mind when we speak of the “external world” is to look at the texts that
have most directly influenced our thinking. We can look, for instance, at some of
Bertrand Russell’s writings on what he called “matter.”
Russell (1912) asks what, if anything, we can know about a particular table that
we perceive. After noting the various ways in which our perception of the table
varies independently of any variation in the table itself, he writes:
Let us give the name of “sense-data” to the things that are immediately known
in sensation: such things as colours, sounds, smells, hardnesses, roughnesses, and
so on. . . . It is plain that if we are to know anything about the table, it must be
by means of the sense-data—brown colour, oblong shape, smoothness, etc.—
which we associate with the table; but, for the reasons which have been given, we
cannot say that the table is the sense-data, or even that the sense-data are directly
properties of the table. Thus a problem arises as to the relation of the sense-data
to the real table, supposing there is such a thing.
EXTERNAL WORLD SKEPTICISM 635
The real table, if it exists, we will call a “physical object.” Thus we have
to consider the relation of sense-data to physical objects. The collection of all
physical objects is called “matter.” (Russell 1912: 4–5)
In this passage, Russell introduces the term “physical object” by means of a single
example (the table), and his only hint as to what it is about that example that
makes it an example of the category at issue is that our perceptions of it can vary
independently of variation in it, and so cannot be identical to it. But it’s not clear
that this fixes what category Russell has in mind: there are plenty of things that we
typically perceive when we look at (say) a table, and many of those perceived things
are not identical with our perceptions of them. For instance, when I look at a table,
I typically perceive a situation of one kind or another (e.g., people sitting at a table
having dinner) and I typically perceive a complex scene (e.g., a dining room in a
state of disarray) and these things are not identical to my perception of them.
Further constraints on the category of “physical object” are imposed elsewhere in
Russell’s writings from the same period. For instance, Russell (1957) treats “physical
objects” as logical constructions out of actual and possible sense-data. This has
some utility: the logical construction itself imposes structural constraints on what
can count as a physical object; for instance, it is a consequence of how Russell’s
construction works that no physical object can be completely contained in two
entirely distinct regions of physical space at the same time. But the “sense-data” that
serve as the materials for the construction are themselves explained only as those
things that are given in sense at one time and that might be singled out by attention.
Unfortunately, this account of “sense-data” leaves open precisely how that category
is understood: why not say, for instance, that a particular table is given in sense at
one time, and might be singled out by attention? In other words, why exclude tables
from the category of “sense-data”? Nothing that Russell says in his characterization
of “sense-data” answers this question.
An alternative conception of the “external world” is provided by G. E. Moore, in
his celebrated “Proof of an External World” (1959b). Moore devotes most of that
paper to characterizing the category of external things, or “things to be met with in
space.” This phrase, Moore writes, is to be understood in such a way that
many “things,” e.g., after-images, double images, bodily pains, which might be
said to be “presented in space,” are nevertheless not to be reckoned as “things
that are to be met with in space,” and . . . there is no contradiction in supposing
that there have been and are “to be met with in space” things which never have
been, are not now, and never will be perceived, nor in supposing that among
those of them which have at some time been perceived many existed at times at
which they were not being perceived. (1959b: 156)
For Moore, the category of “external things” is the category of space-occupying things
that may fail ever to be perceived. This is as close as Moore gets to characterizing
external things, and it is also as close as anyone in the contemporary discussion of
skepticism about the external world has got to characterizing them. Thankfully, his
characterization, rough as it is, will prove useful to us in understanding what gives
636 SKEPTICISM
rise to skeptical doubts concerning our epistemic relation to the external world
specifically.
3 THE VARIETIES OF SKEPTICISM

Before canvassing the considerations that have raised skeptical doubts concerning
our epistemic relation to the external world, I should first distinguish the various
forms that such skepticism has taken. Skepticisms can differ in their attitude, their
depth, and their scope.
A skeptic’s attitude may be one of doubt, one of denial, one of suspension, or one
of withholding. In other words, a skeptic may harbor doubts concerning whether
we possess a positive epistemic relation to external things, or she may outright deny
that we possess any such positive epistemic relation, or she may suspend belief
concerning whether we possess any such relation, or she may altogether withhold
any attitude concerning whether we possess any such relation: one dimension along
which forms of skepticism differ, then, is in attitude.
Another way in which different forms of skepticism differ from each other is
in their depth, that is, the range of epistemic positions concerning which they are
skeptical. Our common convictions are that we know that we live on the Earth, we
justifiably believe that we live on the Earth, we have justification for believing that
we live on the Earth, it is rational for us to believe that we live on the Earth, and we
have evidence that we live on the Earth. A relatively superficial skepticism may call
into question the first of these common convictions, without calling into question
any of the others. A much deeper form of skepticism will call into question all five
of these common convictions.
Finally, skepticisms differ from each other in the range of facts our epistemic
relation to which they call into question. Thus, there is skepticism about the
external world, skepticism about the future, skepticism about the past, skepticism
about other minds, skepticism about value, and so on. Of course, there are also
forms of skepticism that are completely general with respect to range, that is, they
call into question our epistemic relation to all facts of any kind. Such completely
general forms of skepticism include the forms that have been defended by appeal
to Agrippa’s Trilemma (see Sextus Empiricus (1996) or Williams (2001)), those
that have been defended by appeal to the Problem of the Criterion (see Chisholm
(1982)) and that which was defended by Unger (1975).
In the next section, I survey the various arguments that have been offered for one
or another version of skepticism concerning the external world.
4 THE CONSIDERATIONS THAT GENERATE THE

PROBLEM OF THE EXTERNAL WORLD
The skeptical problem of the external world has been developed in four ways, though
it is a matter of some controversy what relation these four skeptical arguments
bear to each other (Brueckner (1994) argues, for example, that the argument from
underdetermination and the argument from closure are not distinct). What all
four of them have in common is that they depend on the thought that perceptual
experience can vary independently of the external world: there is, in short, a gap
between appearance and reality. But this crude idea gets developed in different ways
in each of the following four lines of thought.
4a The Argument from Fallibility

Our beliefs about the external world are formed in ways that are fallible: beliefs
formed in those ways (i.e., through a series of events that involve the stimulation of
one’s sensory surfaces, the resulting transmission of a chemical signal through one’s
neural pathways, and consequent belief fixation) are liable to be false, since the parts
of the external world that initiate that series of events can be utterly different from
how they are represented as being by the beliefs that lie at the end of that series.
But if a belief is formed in a way that is liable to be false, then it is a matter of good
fortune if and when that belief is true. But, when a belief is knowledgeably held,
then it is not a matter of good fortune that the belief is true. Thus, our beliefs about
the external world are not knowledgeably held.
The argument is articulated by Albritton (2011):
The argument, here, is from natural possibility to epistemic possibility. Imagine
a philanthropic lottery in which almost everyone wins, but to keep things
exciting . . . there are a few losing tickets. I have a ticket, like all the others. I have
no concrete reason to suppose that it, in particular, as distinguished from others,
may lose. But this lottery is so set up that some tickets do lose. That’s a perfectly
concrete fact about the lottery, and is a compelling reason . . . to suppose that
this ticket I’ve bought may lose. I don’t know that it will, but I don’t know that
it won’t, either. (Albritton 2011: 7–8)
The “natural possibility” that Albritton here takes to establish a skeptical conclusion
is the possibility that our ways of forming beliefs lead us into error. The plausibility
of Albritton’s view here is indicated by the fact that, in developing his very influential
account of knowledge, Armstrong (1973) claims that a belief can be knowledgeably
held only if it was formed in a way that, by the laws of nature, guarantees the
truth of the belief. If nature allows the possibility that a particular belief, formed
in a particular way, is false, then that belief is not knowledgeably held. But when,
Albritton’s skeptic pointedly asks, does nature not allow that possibility?
4b The Argument from Underdetermination

Descartes thought he could demonstrate the existence of extended substance in
his Sixth Meditation by appeal to the non-deceptiveness of God and our natural
inclination to believe that our sensory ideas are caused by extended things. Berkeley
(1975) thought he could demonstrate that Descartes was wrong by calling into
question the possibility that our sensory ideas represent material substances. But
Russell (1959) thought both philosophers were wrong: the existence of matter, said
Russell, is a hypothesis, and we cannot achieve certainty concerning the truth or
638 SKEPTICISM
falsity of that hypothesis. The hypothesis, Russell thought, is credible because it

provides the simplest and most conservative explanation of our having the sense-
data that we have. While there may be disagreement concerning Russell’s claim
that this explanation is simplest (perhaps it is simpler to say that God is the direct
cause of everything) or most conservative (perhaps it is more conservative, at least
for children and other primitive thinkers, to say that spirits are the direct cause of
everything), it seems more plausible to say that the hypothesis in question provides
the best explanation of our having the sense-data that we have—whatever exactly
is packed into its being the “best.” And so this is precisely the view that a number
of contemporary philosophers have adopted. We can call this view “abductivism.”
Abductivism comes in two main varieties: externalist and internalist. According
to the externalist version of the view, defended by Lycan (1988) and Goldman
(1988), what makes us justified in our belief that there are material objects, and
in many of our more specific beliefs about the behavior of those objects and their
causal role in shaping our sensory experiences, is the very fact—whether or not we
are aware of it—that the contents of those beliefs do provide the best explanation
for our having the sensory experiences that we have. According to this externalist
version of the view, explanatory “bestness” is a relation that can obtain between our
sensory experiences (the explananda) and the hypotheses that we believe about the
external world (the explanans) whether we are aware that it obtains. The internalist
version of abductivism, defended in various forms by Harman (1973), Vogel (1990),
and BonJour (1985), claims that we are justified in believing in the existence of
material objects, and also believing various hypotheses about their behavior, because
we are justified in believing that these hypotheses provide the best explanation of
our sensory experiences.
Abductivism, in either of its forms, paves the way for the skeptical argument
from “underdetermination.” Why, this skeptic will ask, should we grant that our
hypotheses about material objects do in fact provide the uniquely best explanation
of our sensory experiences? If, in fact, another hypothesis does just as well at
explaining our sensory experiences, then, by the externalist abductivist’s own lights,
this implies that we are not justified in our beliefs about the external world. Or,
even if no other hypothesis does just as well at explaining our sensory experiences,
still, so long as we have no reason to believe that our hypotheses about the external
world provide the uniquely best explanation of our sensory experiences, then, by
the internalist abductivist’s lights, this implies that we are not justified in believing
those hypotheses about the external world.
4c The Argument from Closure

You take yourself to know that you have hands. But notice that, if you do have
hands, then you are not merely a brain floating in a vat of nutrient fluid and being
electrochemically stimulated to have the sensory experiences that you have now:
such a brain does not have hands, but you do. So if you know that you do have
hands, then you must also be in a position to know that you are not such a brain.
But how could you know that you are not such a brain? If you were such a brain,
everything would seem exactly as it does now; you would (by hypothesis) have all
the same sensory experiences that you’re having right now. Since your empirical
knowledge of the world around you must somehow be based upon your sensory
experiences, how could these experiences—the very same experiences that you
would have if you were a brain in a vat—furnish you with knowledge that you’re
not such a brain? And if you don’t know that you’re not such a brain, then you
cannot know that you have hands.
Although the skeptical argument just stated applies specifically to your putative
knowledge that you have hands, it needn’t have targeted that proposition in
particular: your putative knowledge that you have feet, or that you have a head, or
that you live on the planet Earth, or that Lincoln is the capital of Nebraska—all of
these empirical propositions about the external world would have done just as well
for the skeptic’s purposes. And so, if the skeptical argument just stated can call into
question your knowledge that you have hands, it can call into question all of our
empirical knowledge of the external world.
This skeptical argument is an instance of modus tollens on the following principle:

If S knows that p, and S knows that p implies q, then S is in a position to know
that q.
If we let p be any ordinary empirical proposition about the external world (e.g.,
“I have hands”), and q be the negation of some skeptical hypothesis that is logically
inconsistent with p (e.g., “I am a handless brain in a vat”), then we get the following
instance of the principle above:
If S knows that she has hands, and knows that her having hands implies that she
is not a handless brain in a vat, then S is in a position to know that she is not a
handless brain in a vat.
The argument from closure begins from the premise that the consequent of this
conditional is false, but the second conjunction of the antecedent is true. It concludes,
then, that the first conjunct of the antecedent must be false. More generally, it
concludes, empirical knowledge of the external world is impossible.
Some philosophers, following Moore’s (1959a,c) lead, have been tempted to do
a modus ponens where the skeptic does a modus tollens, and argue that, because
we do have empirical knowledge of the external world, it follows that we must be
in a position to know that the skeptic’s hypotheses are false. But to this the skeptic
can reply: “How can you know that my hypotheses are false? You cannot know it
empirically, since your sensory evidence would be just the same if my hypotheses
were true. And of course you cannot hope to achieve knowledge of contingencies
about the external world non-empirically. So you cannot know that my hypotheses
are false.” What this skeptic demands is an account of how it is possible to know that
we are not in a skeptical scenario in which, by hypothesis, our sensory experiences
are exactly the way that they are now, but in which our beliefs about the external
world are generally wrong. It is not obvious how to satisfy this demand, and thus the
argument from closure can generate skepticism about the external world.
640 SKEPTICISM
4d The Argument from Dogmatism

Suppose, for the sake of argument, that you are in possession of some particular
bit of empirical knowledge about the external world—for instance, that you have
hands. In what way is it better to have this knowledge than merely to hold a true
opinion to the effect that you have hands? This is an issue that Socrates addresses by
means of an analogy that he offers at the end of Plato (1976):
This is an illustration of the nature of true opinions: while they abide with us they
are beautiful and fruitful, but they run away out of the human soul, and do not
remain long, and therefore they are not of much value until they are fastened by
the tie of the cause . . . . But when they are bound, in the first place, they have the
nature of knowledge; and, in the second place, they are abiding. And this is why
knowledge is more honourable and excellent than true opinion, because fastened
by a chain. (Plato 1976, 31)
How should we understand Socrates’s claim that knowledge, unlike true opinions,
is “fastened by a chain” or “fastened by the tie of the cause”? And how does such
“fastening” prevent knowledge from running “away out of the human soul”?
Although there is considerable scholarly controversy concerning these questions,
here is one plausible sort of answer: to know that p, for Socrates, involves not
merely having the true opinion that p but moreover involves understanding “the
cause” of one’s taking it to be true that p. The cause that one understands is the
reason for which one takes it to be true that p. And so knowledge differs from
true opinion at least by virtue of one’s having a good reason—a reason that one
understands—for taking it to be true. But how does this difference help to make it
the case that knowledge does not “run away out of the human soul”? Is this merely
an empirical claim to the effect that one is less likely to forget that p if one knows
that p than if one merely has the true opinion that p? I think not: even if Socrates
took such an empirical claim to be true, he gives no evidence for its truth here.
Rather, I think we can more plausibly understand Socrates’s claim if we think about
the way in which knowing that p entitles us to dismiss apparent counterevidence to
p. So, given that you know that you have hands, if a stranger tells you that you do
not have any hands, then you are entitled to dismiss the stranger’s testimony: her
testimony should not affect your high degree of confidence that you have hands.
We can abbreviate this point by saying that, when you know that p, your knowledge
is “indefeasible”: no apparent counterevidence to p can make it rational for you to
lower your confidence in p. This requirement on knowledge was defended in recent
philosophy by McDowell 1982.
But if this is what knowledge requires, then the case for skepticism about the
external world may seem overwhelmingly plausible, since it seems that very few, if
any, of our empirical beliefs about the external world satisfy this stiff requirement.
Consider, for instance, your very confidently held belief that you have hands: is
there any apparent counterevidence that can make it rational for you to lower your
confidence that you have hands? It seems there is: imagine if, a moment from now,
you have an extremely vivid and lifelike experience as of waking up from a dream,
lying in a hospital bed, looking down at your arms and seeing that your hands
have been amputated. In such a case, would it not be rational for you to be at least
slightly less confident of your having hands than you now are? And if this “waking
up in a hospital bed” experience were the beginning of a long stream of coherent
conscious experience, throughout which it seemed clear to you that you had no
hands, would this not make it irrational for you to maintain your present level of
confidence that you have hands? It seems so. But, in that case, if knowledge makes
it rationally permissible for you to dismiss all apparent counterevidence to what you
know, then you do not know that you have hands. Since the proposition that you
have hands was chosen as an arbitrary example of an empirically grounded belief
about the external world, and the same considerations would apply to virtually
any other empirically grounded belief about the external world, it follows that you
have virtually no empirical knowledge of the external world. The indefeasibility
requirement on knowledge—if indeed it is a requirement—seems to make the world
safe for skepticism.
The four sorts of arguments I’ve canvassed in this section—arguments from
fallibility, from underdetermination, from closure, and from dogmatism—have been
the main ways in which skepticism about the external world has been dialectically
generated. In the next section, I will canvass the main ways in which epistemologists
have attempted to rebut these four sorts of argument.
5 RESPONSES TO THE SKEPTICAL ARGUMENTS

5a Responses to the Argument from Fallibility
By far the most common response to the argument from fallibility—the response
issued (either explicitly or implicitly) by Goldman 1986, Nozick 1981, Lehrer 1974,
Williamson 2000, Sosa 1999, and many, many others—is simply to deny that natural
possibility (“possible for something to be F”) implies epistemic possibility (“possible
that something is F”). According to these philosophers, the fact that it is possible
for you to form the beliefs that you currently have in just the way that you formed
them, and yet be wrong in those beliefs, does not imply that, as things actually are,
those beliefs are not knowledgeably held. That they could have been wrong does not
imply that they could be wrong.
While this “fallibilist” view of knowledge is widely held, it is not obviously
correct. If it’s possible for this very belief (say, that I am now typing on a keyboard),
formed in the way it has been formed, to be wrong, then why isn’t it possible that
this very belief is wrong? Albritton’s analogy with the philanthropic lottery can seem
compelling here: since it’s possible for a particular ticket to be a loser, it seems possible
that the ticket is a loser. The defenders of fallible knowledge would reply that, even
if the analogy to the philanthropic lottery is compelling, what this shows is not that
knowledge requires forming one’s belief in a way that could not have gone wrong,
but rather it requires forming one’s belief in a way that would not have gone wrong.
If one forms one’s belief in a way that is reliable (where this “reliability” condition
can be cashed out in a variety of ways that fall short of infallibility), then one’s belief
642 SKEPTICISM
can be knowledgeably held even if fallibly formed. The reason that we cannot know,
of the ticket in the philanthropic lottery, that it is a winner is that it is a fair lottery,
and so one could just as easily (even if not just as probably) lose as win. It’s not the
case (the fallibilist will say) that if one had bought a ticket one would have won: it’s
only true that if one had bought a ticket one would probably have won. And the
latter modal condition is weaker than the former. So—the fallibilist concludes—to
explain why the ticket holder in the philanthropic lottery cannot know of his ticket
that it is a winner, we need not impose any condition on knowledge that is stronger
than the former condition.
So the lottery example does not dispose of fallibilism. But there is another
consideration that may seem to tell against fallibilism. Suppose, for reductio, that S
has fallible knowledge that p. Suppose also that S’s belief that p is based upon some
grounds g, and S knows that this is so, and knows that g are good (albeit fallible)
grounds for believing that p. In that case, S is in a position to know (by the closure of
knowledge under known consequence from two known premises) that her grounds
g are not misleading with respect to the truth of p. In other words, S must be in
a position to know that, though it is possible for g to be true while p is false, that
possibility does not obtain. But how could S be in a position to know this whenever
she knows that p while knowing that her grounds for believing that p are fallible?
In the absence of a satisfying answer to this question, we may conclude that the
original supposition—namely, that S has fallible knowledge that p—cannot be true,
and so fallibilism about knowledge is false.
Even if fallibilism is false, this does not show that the skeptical argument from
fallibility succeeds. For there is another possible response to the argument from
fallibility, and that is Albritton’s own response, namely, that many of our beliefs are
formed in a way that is not fallible. It simply is not possible for those beliefs, formed
as they were, to be false, even if other beliefs formed in apparently similar ways
actually were, or are, false. (In order to avoid the generality problem, we should
specify that, when we here speak of “the way” in which a belief was formed, we
mean to be describing those features of its fixation that are relevant to the epistemic
status of the belief.) Epistemological disjunctivists like McDowell adopt this anti-
skeptical strategy when they claim that knowledgeably held beliefs are only those
that register the fact that they represent: a belief may seem to the believer to do this
without actually doing it, but any belief that fails to do this is not knowledgeably
held, however it seems to the subject. So, on this “infallibilist” view, a knowledgeably
held belief is one that is formed in a distinctive way, and the way it is formed is such
that no false belief can be formed in that way.
5b Responses to the Argument from Underdetermination

The most common response to the argument from underdetermination involves
the claim that, even if our perceptual evidence is logically consistent with skeptical
hypotheses, it nonetheless provides evidence against those hypotheses in favor of
our commonsensical beliefs about the external world, and that is because, even if
skeptical hypotheses can be constructed that do explain that perceptual evidence,
our commonsensical beliefs about the external world provide a much better
explanation of that same perceptual evidence. Why is this? Some philosophers have
argued that our commonsensical beliefs about the external world provide a simpler
or more ontologically or ideologically parsimonious explanation. But it seems that
the skeptic can rebut these claims by designing her competing hypothesis to be as
simple and as parsimonious as our commonsensical beliefs about the external world.
Vogel (1990), however, points out an important difference between the former
(which he calls the “computer skeptical hypothesis,” or CSH) and the latter (which
he calls the “real world hypothesis,” or RWH). The RWH explains our perceptual
evidence by appeal to the locations of objects, the CSH explains that same evidence
by appeal to some property analogous to location—call it “pseudo-location.” Much
of the RWH’s explanatory power depends on the necessary truth that no two material
objects can occupy the same location at the same time. If the CSH is to enjoy all the
explanatory success of the RWH, it will need to appeal to some analogous principle,
that is, that no two causes of our perceptual evidence can occupy the same “pseudo-
location” at the same time. But either the CSH explains this analogous principle or
it doesn’t. If it does, then the CSH explanation of our perceptual evidence is more
complicated. If it doesn’t, then the CSH explanation of our perceptual evidence
is not as deep or powerful. Either way, the CSH explanation of our perceptual
evidence is not as good as the RWH explanation (Vogel 1990: 664–665).
While Vogel may have succeeded in showing that the skeptic cannot provide
the best explanation of our perceptual evidence, this rebuts the argument from
underdetermination only when that argument is understood as assuming that our
ordinary beliefs about the external world do provide some explanation of our
perceptual evidence, where the latter is accessible to us in terms that don’t presuppose
the existence of an external world. But that assumption has been challenged by
Neta (2004), who argues that the “explanatory gap” that has been raised, for
example, by Levine (1983) and Chalmers (1996), as a problem for explanations of
conscious phenomenology in terms of physical processes poses a challenge to any
attempt to explain our perceptual evidence, understood in terms that don’t have any
commitments to the existence of an external world, by appeal to any physical events
or processes. A way to put this worry is that, even if the skeptic cannot provide a
good explanation of our perceptual evidence, so conceived, neither can our ordinary
beliefs about the external world. The latter can explain our perceptual evidence
only insofar as such evidence is described in terms that do presuppose the existence
of external things, for example, as perceptual relations to external things. In other
words, the solution to this problem of underdetermination demands that we think
of our perceptual evidence in terms that presuppose the existence of external things.
But in that case this “response” to the skeptical problem of underdetermination is
no response at all.
5c Responses to the Argument from Closure

Though many philosophers have attempted to rebut the argument from closure,
nothing close to consensus has been achieved on the correct reply. One common line
644 SKEPTICISM
of reply, issued by Nozick (1981) and Dretske (1971), among others, is simply to
deny that knowledge is closed under known implication. Nozick and Dretske each
develop an account of knowledge that has as a consequence that knowledge is not
closed under known implication. According to Dretske, S knows that p only if S’s
belief that p is based on conclusive grounds for p, where grounds for p are conclusive
just in case S would not have those grounds if p were not true. According to Nozick,
S knows that p if and only if (a) S believes that p by means of some method M, and
(b) S would not believe that p by means of M if p were not true, and (c) S would
believe that p by means of M if p were true. Nozick and Dretske each argues for
his account by showing that it issues correct verdicts about the various cases that
emerged in the post-Gettier literature on knowledge. But one consequence of both
of their accounts is that knowledge is not closed under known implication. And this
consequence, as Kripke (2011) has pointed out, is not easy to accept. Consider, for
instance, the following example: you see a carpet store’s truck pull up into your
neighbor’s driveway, and the delivery people bring a new red carpet out of the
truck and into your neighbor’s home. If your neighbors had not bought a new red
carpet, they would have bought a blue carpet, and you would have seen it (just as
you are now seeing the red one), and so would not have believed that they bought a
red carpet. On both Dretske’s and Nozick’s account, you know that your neighbors
just bought a new red carpet. But suppose that, if your neighbors had not bought
a carpet, they would have paid delivery people to drive the carpet store’s truck
into their driveway and unload a new red carpet into their home anyway, just for
display purposes, and in order to fool you and their other neighbors into thinking
that they bought a carpet. Then, on both Dretske’s and Nozick’s account, you do
not know that your neighbors just bought a carpet. So, by Dretske’s and Nozick’s
lights, this is a situation in which you know (by looking) that your neighbors have
just bought a new red carpet, but you do not know that they have bought a carpet.
The implausibility of this consequence damages any attempt to rebut the argument
from closure by rejecting the closure principle.
Another way to rebut the argument from closure, without rejecting the closure
principle, is to reject the premise that we cannot know that the skeptic’s hypothesis
is false. On this view, defended by Sosa (1999), Williamson (2000), Wright (2004),
Lehrer (1974), Pritchard (2005), Byrne (2004), and others, for any hypothesis that
the skeptic designs—any hypothesis the falsity of which is obviously implied by
the claims about the external world that we ordinarily take ourselves to know, and
according to which our perceptual experiences are just as they actually are—we
know that the hypothesis is false. Of course, this position raises two questions.
First, how is it that we manage to know that the skeptic’s hypothesis is false?
(For instance, how do I manage to know that I am not a brain in a vat being
electrochemically stimulated to have the very experiences that I’ve been having?)
And second, why are so many philosophers tempted into making (what is, by the
lights of this position) the mistake of thinking that we do not know the skeptic’s
hypothesis to be false?
The first question tends to be answered in one of two ways. Externalists like Sosa
and Williamson take it that our knowledge of the falsity of the skeptic’s hypothesis
does not require us to be able to justify our belief that the skeptic’s hypothesis is false:
all that such knowledge requires is our bearing the relevant modal relation to the
fact that we are not (say) mere brains in vats. In contrast, internalists like Wright and
Pryor (2000) take it that our knowledge of the falsity of the skeptic’s hypothesis—
or, at least, our internalist justification or warrant for believing that the hypothesis
is false—is based on some considerations that are available to us upon reflection.
(Wright and Pryor hold very different views about what those considerations are:
Wright takes them to be non-empirical considerations concerning what sorts of
things we must accept if we are to be rationally successful at all, and Pryor takes
them to be empirical.)
The second question has been answered in many different ways. Sosa, for
example, thinks that the following true principle
S knows that p only if it would not easily happen that S falsely believes that p is
easily confused with the following false principle
S knows that p only if S would not believe that p if p were false.
While the false principle implies that we do not know the falsity of the skeptic’s
hypothesis, the true principle has no such implication. Nonetheless, since we confuse
the two principles, we also are confusedly led to the view that we do not know the
falsity of the skeptic’s hypothesis.
Roush (2010) has a different diagnosis: the skeptic’s hypotheses are typically
ambiguous. On one reading, they are incompatible with the claims that we ordinarily
take ourselves to know, but we can easily know that they are false. On the other
reading, they are not incompatible with the claims that we ordinarily take ourselves
to know. For instance, I ordinarily take myself to know that I have hands. But is the
claim that I have hands inconsistent with my being a brain in a vat? Of course not:
that hypothesis leaves it open that I am a brain in a vat with hands!
But there is yet another diagnosis of why so many philosophers have been tempted
into the view that we do not know the skeptic’s hypotheses to be false, and this other
diagnosis forms the basis of a third way of rebutting the argument from closure.
According to this diagnosis, our attributions of knowledge are held to different
standards in different situations. And while our ordinary positive attributions of
knowledge cannot be correct relative to the strictest standards of knowledge, they
can be correct relative to the more relaxed standards that govern our everyday
talk and thought. This view can be fleshed out in different ways, depending on
whether the variation in standards is taken to be a variation in the truth conditions
of our ordinary knowledge-attributing sentences (as DeRose (1995), Lewis (1996),
Cohen (1988), Schaffer (2005), Neta (2002), and Blome-Tillmann (2014) would
say) or whether it is instead a variation in the acceptability of asserting our ordinary
knowledge-attributing sentences (as Fogelin (1994), Williamson (2000), Stanley
(2005), and Turri (2010) would say). This way of rebutting the argument from
closure accepts the closure principle outright, and then takes our ordinary claims to
knowledge to be true and/or acceptable only relative to standards that are less than
fully strict. Even if some version of this view is correct—and it is controversial that
any version of it is (see Reed (2009), for opposition)—it’s not clear to what extent
646 SKEPTICISM
skepticism about the external world is committed to denying this view. Perhaps
this sort of view should be thought of not so much as a rebuttal of the argument
from closure, but rather as an attempt to show how the skeptical conclusion of that
argument can be understood in a way that doesn’t impugn the positive knowledge
attributions that we ordinarily issue.
5d Responses to the Argument from Dogmatism

By far the most common response to the argument from dogmatism was the one
originally issued by Harman (1973): we can know that p even if we cannot rationally
disregard apparent counterevidence to p, but that is because when we gain apparent
counterevidence to p, we thereby lose our knowledge. According to this popular
view, knowledge that p does rationally permit the knower to disregard apparent
counterevidence to p, but if and when such counterevidence comes into view, then
the subject ceases to know, and is no longer rationally permitted to ignore that
counterevidence. For example, suppose that I know, on the basis of testimony, that
the current American secretary of state is Rex Tillerson. Then, so long as I continue
to know that fact, I am rationally permitted to ignore evidence to the contrary.
For instance, so long as I continue to know that the current American secretary of
state is Rex Tillerson, I am rationally permitted to ignore news reports to the effect
that Tillerson has suddenly resigned. However, when I hear such news reports,
I no longer continue to know that the current American secretary of state is Rex
Tillerson, and so I am no longer rationally permitted to ignore those reports. This
view allows that many of the things that we ordinarily take ourselves to know are
things that we do know, even if we are not rationally permitted to ignore apparent
counterevidence to them.
The problem with this view has been pointed out recently by Maria Lasonen-
Aarnio (2014). Let’s grant, she says, that knowledge that p does rationally permit
one to ignore counterevidence to p. Let’s also grant, as seems immensely plausible,
that it’s possible to know that p is true when the balance of one’s evidence provides
very compelling support for p, even if one has a very slight bit of counterevidence
to p. Now, in such a case, what should the defender of Harman’s view say? Should
she say that, since the subject continues to know that p despite having some
slight counterevidence to p, she is rationally permitted to ignore her very slight
counterevidence to p? That seems implausible: rationality requires us to distribute
our credence in light of all our relevant evidence, and it should not matter whether
that evidence comes in all at once, or the positive evidence all comes in first (to
give us knowledge) and only later does the very slight negative evidence come in.
So should the defender of Harman’s view instead say that knowledge is lost as
soon as the knower acquires even the slightest bit of counterevidence to p? That
also seems implausible: presumably, it’s possible to know things even without being
absolutely certain of them, and it’s also possible to continue to know things of which
we were absolutely certain even once we acquire evidence that rationally requires
us to be only ever so slightly less than absolutely certain of them. So the defender
of Harman’s view cannot explain why someone who knows that p is not rationally
permitted to ignore even very slight counterevidence to p, counterevidence the

possession of which is consistent with their knowing that p.
But if the Harman response to the skeptical argument from dogmatism does not
work, then how can we reply to that argument? Another line of reply is implicit in
the view of McDowell (1982), who grants that knowledge is indefeasible: when you
do know that p, you are rationally permitted to ignore counterevidence to p, and
your knowledge is not defeated by your obtaining such counterevidence. Part of why
this reply is so unorthodox is that it seems, on the face of it, implausible to claim that
our ordinary knowledge can never be defeated by counterevidence. Can I not lose
my knowledge that Rex Tillerson is the current American secretary of state if I come
to hear seemingly credible news reports that he has resigned and been replaced?
Suppose I do come to hear such news reports, and then learn, on the basis both
of those reports and subsequent corroboration, that Tillerson has in fact resigned
and been replaced: what would I say then about my earlier claim to know that Rex
Tillerson is the current American secretary of state? I would say that I only thought
I knew, but I did not really know. Now suppose that I come to hear these same news
reports, and then have not yet corroborated them but am still investigating, and so
am no longer sure what to believe: what would I say then about my earlier claim to
know that John Kerry is the current American secretary of state? Once again, I would
say that I thought I knew, but I did not really know. On McDowell’s view, this later
claim is exactly right. What happens when I gain sufficient counterevidence to p is
not that I cease to know that p, but rather I learn that I do not, and did not, know that
p. If you know that p, then you will not gain evidence that would make it rational
to doubt that p. If you do gain such evidence, then you did not know that p. (We
are assuming here that p is a proposition the truth of which is constant over time.)
Fortunately for our epistemic position, many of the things that we ordinarily take
ourselves to know are never put in doubt by counterevidence.
6 CONCLUSION: HOW THE VERY IDEA OF

THE EXTERNAL WORLD CAN GENERATE
SKEPTICAL WORRIES
What all four of the skeptical arguments above have in common is that they include
premises the plausibility of which depends on the thought that external things are
the way they are independent of how they perceptually appear to us: that’s why
the beliefs that we form on the basis of those appearances can be false, why our
beliefs about external things are underdetermined by our perceptual evidence, why
we cannot know that we’re not receiving those appearances in the way described
by some skeptical hypothesis, and why we cannot rationally disregard apparent
counterevidence to our beliefs about external things.
But notice that, while it follows from Moore’s characterization of external
things that such things can be as they are without being perceived, it does not
follow from that characterization that our perceptual states can be as they are
without external things. It is easy to confuse these two thoughts, but they are not
648 SKEPTICISM
logically equivalent. We can accept the first thought, but reject the second. In this
section, I canvass three famous attempts to do this, and thereby to undermine
skepticism about the external world.
According to the “content externalism” defended by Putnam (1981), the
contents of our mental representations are metaphysically determined by the causal
relations that we bear to features of our environment. It is only because we spend
our lives in causal contact with water that so many of our thoughts are about water.
Our doppelgangers on Twin Earth, who spend their lives in causal contact with a
substance that is perceptually indistinguishable, but distinct, from water, have mental
representations that are about that distinct substance, and not about water. Thus,
on Putnam’s view, there is a metaphysical limit to how wrong our beliefs about the
external world can be, and that limit is imposed by the fact that the contents of those
beliefs depends on facts about the external world. More specifically, as Putnam
put it: “If we can consider whether [the hypothesis that we are brains in vats being
electrochemically stimulated to have the very experiences that we’ve had] is true or
false, then it is not true. . . . Hence it is not true” (Putnam 1981: 8).
Davidson (1984) argues for a form of psychological externalism by appeal to the
premises that (i) what we think is just what the method of radical interpretation,
when ideally executed in application to all of our observable behavior, would claim
that we think, and (ii) the method of radical interpretation delivers the result that
most of the interpreter’s beliefs are true by the lights of the interpreter, and (iii) an
omniscient being could, in principle, execute the method of radical interpretation.
From these three principles it follows that, for any particular thinker, what they
think is just what the omniscient interpreter would claim (on the basis of radical
interpretation) that they think, and most of those beliefs would be true by the
lights of the omniscient interpreter. But what is true by the lights of an omniscient
interpreter is true. Therefore, for any particular thinker, most of her beliefs are true.
This could only be the case if the external world were at least very roughly the way
we believe it to be. Therefore, the external world is at least very roughly the way
we believe it to be.
Both Putnam’s content externalism and Davidson’s psychological externalism
impose a metaphysical limit to how wrong our beliefs about the external world
can be, but this limit is consistent with various narrow skeptical hypotheses (e.g.,
five minutes ago I made the transition from being a normally embodied human
to being a brain in a vat). In contrast, on McDowell’s disjunctivist view, it’s true
not only that our representations depend globally on some features of the external
world; it’s also true that particular representations depend on particular features of
the external world. For instance, when I look at my hands in the clear light of day,
I see my hands, and when I see my hands, I have a perceptual experience that is
different in kind from any perceptual experience that I could have if I did not have
hands. It’s true that I could have a convincing hallucination as of having hands:
in such a case, I would falsely believe myself to be seeing my hands. It might also
be true that when I dream about, or imagine, my hands, I falsely believe myself
to be seeing my hands. Nonetheless, when I do in fact see my hands, I can know,
simply by reflecting on my perceptual state itself, that I’m seeing my hands and
not merely dreaming, imagining, or hallucinating them. On this disjunctivist view,

while it can still be true that external things (like my hands) can exist without being
perceived, it’s not true that I could occupy my current perceptual states (my current
seeing of my hands) without the existence of the perceived external things. If this
disjunctivist view is correct, then we can grant the skeptic that knowledge of the
external world requires infallibility, no underdetermination, closure under known
implication, and indefeasibility, but still deny that the skeptic can gain any mileage
from these requirements.
REFERENCES
Albritton, Rogers. 2011. “On a Sceptical Argument from Possibility.” In E. Sosa and E.
Villanueva (eds.), Philosophical Issues (The Epistemology of Perception) 21: 1–24.
Armstrong, David. 1973. Belief, Truth and Knowledge. Cambridge: Cambridge University
Press.
Austin, J. L. 1961. “Other Minds.” In his Philosophical Papers, 76–116. Oxford: Clarendon
Press.
Ayer, A. J. 1956. The Problem of Knowledge. London: Hammondsworth.
Berkeley, George. 1975. Three Dialogues between Hylas and Philonous, edited by
P. H. Nidditch. London and Melbourne: Everyman’s Library.
Blome-Tillman, Michael. 2014. Knowledge and Presuppositions. Oxford: Oxford University
Press.
University Press.
Byrne, Alex. 2004. “How Hard Are the Skeptical Paradoxes?” Noûs 38: 299–325.
Chalmers, David. 1996. The Conscious Mind. New York: Oxford University Press.
Chisholm, Roderick. 1982. “The Problem of the Criterion.” In his The Foundations of
Knowing, 61–75. Minneapolis, MN: University of Minnesota Press.
Cohen, Stewart. 1988. “How to Be a Fallibilist.” In J. Tomberlin (ed.), Philosophical
Perspectives 2 (Epistemology), 91–123. Atascadero, CA: Ridgeview.
Davidson, Donald. 1984. “Belief and Its Basis in Meaning.” In his Inquiries into Truth and
Interpretation, 141–154. Oxford: Clarendon Press.
Descartes, René. 1993. Meditations on First Philosophy, translated by D. Cress. Indianapolis,
IN: Hackett.
Dretske, Fred. 1971. “Conclusive Reasons,” Australasian Journal of Philosophy 49: 1–22.
Dretske, Fred. 1981. Knowledge and the Flow of Information. Cambridge, MA: MIT Press.
Fogelin, Robert. 1994. Pyrrhonian Reflections on Knowledge and Justification. Oxford and
Ginet, Carl. 1980. “Knowing Less by Knowing More.” In P. French, T. Uehling, and
H. Wettstein (eds.), Midwest Studies in Philosophy V: Epistemology. Minneapolis, MN:
University of Minnesota Press, 151–161.
Goldman, Alan. 1988. Empirical Knowledge. Berkeley, CA: University of California Press.
650 SKEPTICISM
Goldman, Alvin. 1986. Epistemology and Cognition. Cambridge, MA: Harvard University
Press.
Harman, Gilbert. 1973. Thought. Princeton, NJ: Princeton University Press.
Kripke, Saul. 2011. “Nozick on Knowledge.” In his Collected Papers, Volume 1. Oxford:
Lasonen-Aarnio, Maria. 2014. “The Dogmatism Puzzle,” Australasian Journal of Philosophy
92: 417–432.
Lehrer, Keith. 1974. Knowledge. Oxford: Oxford University Press.
Levine, Joseph. 1983. “Materialism and Qualia: The Explanatory Gap,” Pacific Philosophical
Locke, John. 1975. An Essay Concerning Human Understanding, edited by P. H. Nidditch.
Lycan, William. 1988. Judgment and Justification. Cambridge: Cambridge University Press.
McDowell, John. 1982. “Criteria, Defeasibility, and Knowledge,” Proceedings of the British
Academy 68: 455–479.
McDowell, John. 1994. Mind and World. Cambridge, MA: Harvard University Press.
Moore, G. E. 1959a. “Certainty.” In his Philosophical Papers, 223–246. New York: Collier.
Moore, G. E. 1959b. “Proof of an External World.” In his Philosophical Papers, 126–148.
New York: Collier.
Moore, G. E. 1959c. “Four Forms of Scepticism.” In his Philosophical Papers, 193–222.
New York: Collier.
Neta, Ram. 2002. “S Knows that P,” Noûs 36: 663–681.
Neta, Ram. 2003. “Contextualism and the Problem of the External World,” Philosophical
Neta, Ram. 2004. “Skepticism, Abductivism, and the Explanatory Gap.” In E. Sosa and
E. Villanueva (eds.), Philosophical Issues 14, 296–325.
Press.
Plato. 1976. Meno, translated by G. Grube. Indianapolis, IN: Hackett.
Pritchard, Duncan. 2005. Epistemic Luck. Oxford: Oxford University Press.
Putnam, Hilary. 1981. Reason, Truth, and History. Cambridge: Cambridge University
Press.
Reed, Baron. 2009. “A New Argument for Skepticism,” Philosophical Studies 142: 91–104.
Roush, Sherrilyn. 2010. “Closure on Skepticism.” Journal of Philosophy 107: 243–256.
Russell, Bertrand. 1957. “The Relation of Sense-Data to Physics.” In his Mysticism and
Logic, 140–173. Garden City, NY: Anchor Books.
Russell, Bertrand. 1959. The Problems of Philosophy. Oxford: Oxford University Press.
Schaffer, Jonathan. 2005. “Contrastive Knowledge,” Oxford Studies in Epistemology 1:
235–271.
Sextus Empiricus. 1996. Outlines of Pyrrhonism, translated by Benson Mates. Oxford and
Sosa, Ernest. 1999. “How to Defeat Opposition to Moore.” In J. Tomberlin, Philosophical
Perspectives 13: Epistemology, 141–153. Cambridge, MA and Oxford: Blackwell.
Stanley, Jason. 2004. Knowledge and Practical Interests. Oxford: Oxford University Press.
Turri, John. 2010. “Epistemic Invariantism and Speech Act Contextualism,” Philosophical
Review 119: 77–95.
Unger, Peter. 1975. Ignorance. Oxford: Clarendon Press.
Vogel, Jonathan. 1990. “Cartesian Scepticism and Inference to the Best Explanation,”
Williams, Michael. 2001. Problems of Knowledge. Oxford: Oxford University Press.
Wright, Crispin. 2004. “Warrant for Nothing (and Foundations for Free?),” Aristotelian
Society Supplement 78: 167–212.
CHAPTER FORTY-FIVE
Disjunctivism
and Skepticism
DUNCAN PRITCHARD AND CHRIS RANALLI
1 THE NATURE OF PERCEPTUAL EXPERIENCE

I see a particular book on the shelf. How should I characterize the nature of this visual
perception? According to naïve realism, what explains what it is like for me to see
this book is the book itself and its manifest properties—the properties that I can see,
like its color, shape, size, texture, and so on. In particular, naïve realism maintains
that the objects in the world around us—objects like trees, books, tablecloths, and
so on—enter into our experiences of them as constituents. On this view, it is the
external objects and their manifest properties which explain the “phenomenal
character” of our perceptual experiences—what Thomas Nagel (1974) called those
features of experience such that there is something that it is like to undergo them.
Most philosophers reject naïve realism in favor of some form of intentionalism
about the nature of experience.1 According to intentionalism, sense experience is an
intentional state, like belief or thought. Consider a now famous expression of the
view from Christopher Peacocke:
A visual perceptual experience enjoyed by someone sitting at a desk may
represent various writing implements and items of furniture as having particular
spatial relations to one another and to the experiencer, and as themselves having
various qualities. . . . The representational content of a perceptual experience has
to be given by a proposition, or set of propositions, which specifies the way the
experience represents the world to be. (Peacocke 1983: 5)
Intentionalism thus holds that perceptual experiences have representational
content—contents which are about something, or represent the world as being some
way—which permits an assessment of them as successful or unsuccessful, veridical
or unveridical, or true or false.
When framed in the terms of the old debate between direct realism and indirect
realism, both naïve realism and intentionalism are versions of direct realism because
neither view maintains that there are surrogate objects or properties that we
experience which mediate our perceptual experiences of mind-independent external
DISJUNCTIVISM AND SKEPTICISM 653
objects and their properties. For the naïve realist, objects and properties in the world
around us can enter into our experiences as constituents. Our perceptions of those
objects are “direct” in the sense that those objects in part constitute our perceptual
experiences when perceived. However, while the intentionalist does not characterize
perceptual experience as a relation to mind-independent objects and properties, she
doesn’t thereby commit herself to characterizing it as a relation to other kinds of
objects and properties. Instead, perceptual experience is a representational state,
and just as we can believe that snow is white without the belief first being about
something else distinct from snow and its whiteness, so too we can have a perceptual
experience in which we represent that snow is white without the experience first
being about something else distinct from snow and its whiteness.
If we describe the nature of perceptual experience like this, it’s not clear what could
make the mental state “indirect” in any philosophically significant sense. In this way,
while the naïve realist and the sense-datum theorist agree that the nature of sense
experience is relational—experience is a relation to objects and properties—they
disagree on the nature of what it’s a relation to. And this distinction helps sharpen
the difference between naïve realism and intentionalism. For the intentionalist, sense
experience is not a relation to mind-independent external objects and properties, but
not because it is a relation to mind-dependent objects and properties. Instead, sense
experience is not a bona fide relation, in which both relata exist. So, while the naïve
realist and the sense-datum theorist agree about the relational character of sensory
experience, the naïve realist and the intentionalist, contra the sense-datum theorist,
agree that sense experience is “direct,” in the sense that we can perceive external
material objects without first perceiving something else ontologically distinct from
them, like sense-data.
The central reason for rejecting naïve realism is some form of the argument from
hallucination. As a general template, the argument can be framed as a paradox.
Naïve realism appears to best articulate how perceptual experience strikes us as
being on reflection. That is, it just strikes us that during visual perception, say,
those objects and properties around us that we see and attend to are themselves
part of the very visual perception. As P. F. Strawson (1974) once put the point,
it seems like if we try to attend to the objects and properties of our experiences,
we thereby attend to the objects and properties in the world around us. But naïve
realism is apparently shown by the argument from hallucination to be inconsistent
with two other plausible theses: experiential naturalism (EN) and the common kind
assumption (CKA).2
According to experiential naturalism, our sense experiences are subject to what
Mike Martin (2004: 39) calls the natural “causal order”—that is, are subject to
causal stimulation, just like other material objects, properties, or states of affairs. In
particular, a suitable stimulation to the brain can cause an experience in the subject
which—at least in principle—is indistinguishable from a perception.
Moreover, the common kind assumption claims that the kind of mental event that
occurs when I perceive an external material object is a kind of event I can undergo
even in the absence of the objects or properties that I perceived. Naturally, one
might think that this thesis gives credence to the idea that it’s possible to undergo
654 SKEPTICISM
hallucinations which are indistinguishable from veridical perceptions because such

mental events are simply events of the same kind being instantiated in two different
kinds of cases: one in a case where what causes the experience is a mind-independent
object, the other a suitable brain-stimulation.
The argument from hallucination against naïve realism then moves as follows.
(CKA) implies that whatever the nature of the sense experience during perception,
that kind of mental event can occur during hallucination. According to naïve realism,
however, perceptual experience is a kind of mental event which is constitutively a
relation, and in particular, a relation to mind-independent, external objects. From
(CKA) and naïve realism alone, there is no inconsistency. But once we add the
plausible (EN), (CKA) and naïve realism are rendered inconsistent. After all, naïve
realism tells us that perceptual experience is constitutively relational. So by (CKA),
that kind of mental event can occur in cases of hallucination. But then there must exist
both relata of the sensory relation. However, (EN), together with (CKA), tells us that
brain-stimulation alone is sufficient to cause a sensory experience of the kind that is
instantiated during perception. But the conditions necessary to cause a hallucinatory
experience are not sufficient to cause the existence of the relata, namely, the subject
and the object of the subject’s awareness. And, ex hypothesi, hallucinations aren’t
relations to anything. So, from (CKA) and (EN), naïve realism is false.
Disjunctivism about the nature of sense experience can be seen as a tool for
blocking the negative conclusion of this argument—viz., that naïve realism is false
(Martin 2006: 354). According to disjunctivism, the nature of the experiences we
have while veridically perceiving some object differs in kind from the nature of the
experiences we have when undergoing indistinguishable hallucinations. Instead of
thinking of the nature of sense experience as what is metaphysically in common
between cases of veridical perception and cases which fall short of it, we can instead
think of the metaphysical nature of sense experience as divergent across “good”
cases like veridical perception and “bad” cases like hallucination. According to
this line, there are no metaphysical commonalities between the disjoint cases of
perception and hallucination from which a general theory of the nature of sensory
experience can be given.3
With disjunctivism in hand, the proponent of naïve realism can respond to the
argument from hallucination as follows. It is true that brain-stimulation alone can
cause us to have experiences which are indistinguishable from veridical perceptions.
But it does not follow from this that that kind of mental event, the indistinguishable
hallucination, has the same nature as the kind we have when we veridically perceive
external material objects and properties. Instead, we can characterize the visual
experience as of a coffee cup, say, as either a state in which we perceive the coffee
cup or as one in which it seems to us just as if we do, as when we suffer from an
indistinguishable hallucination.4 In this fashion, disjunctivism allows one to maintain
that the explanans of the relevant features of experiences in cases of veridical
perception, such as its phenomenal character, can be external mind-independent
objects and properties, even if it fails to be that in cases which fall short of veridical
perception, such as hallucination and illusion.
However, disjunctivism need not be understood as a view committed to naïve

realism. One can be an intentionalist and accept disjunctivism. For example, one might
want to characterize perceptual experience as a distinctive kind of representational
state that cannot be instantiated in cases which fall short of veridical perception,
such as hallucination.5 Following a view like Bertrand Russell’s (1912; 1917) about
the nature of our grasp of propositions, we might think that we grasp some types
of propositions in virtue of being acquainted with some of the constituents of those
propositions. If we add that external mind-independent objects and properties can
be constituents of those kinds of propositions, and we also add that those kinds
of propositions can be the content of a perceptual experience, this allows one to
formulate a distinctive kind of disjunctivism about the nature of sense experience in
virtue of appealing to the different kinds of representational content an experience
can instantiate in cases of veridical perception, but not in cases of hallucination.6
This argument highlights how, even if naïve realism entails disjunctivism, the
converse need not be true. Disjunctivism is compatible with intentionalism, at least
if we take the central component of intentionalism to be the thesis that sensory
experience has representational content, and it is its representational content which
bears the burden of explaining the nature of sensory experience. Indeed, this helps
us to further sharpen the distinction between naïve realism and intentionalism.
What is central to naïve realism is the thesis that external material objects and
properties, and our sensory relation to them, bear the burden of explaining the
nature of sensory experience. But what is central to intentionalism is that it’s the
representational content of experience which bears that burden, and ipso facto not
that sensory experience is a relation to external material objects and properties,
even if it were. Of course, one can be an intentionalist who believes that the
representational content of veridical perception is a kind of content where we are
acquainted with—and so related to—some of the components of the content, where
some of the components of the content are external material objects and properties.
But note that this kind of intentionalism obscures the distinction between naïve
realism and intentionalism, and it’s not clear that it doesn’t simply collapse the
distinction. So, whether a significant distinction between the two would remain at
that point is still quite controversial and difficult to answer.7
2 SKEPTICISM AND DISJUNCTIVISM

The epistemological problem of the external world is the problem of explaining how
knowledge of the external world is possible given certain obstacles that make it look
impossible. One obstacle to this kind of knowledge can be derived from a certain
intuitive requirement on knowledge. For example, we might think that in order to
know that p we have to know that all of those propositions which we know to be
incompatible with p are false. For example, consider the proposition that I’m a bodiless
brain in a vat, stimulated to have experiences which are indistinguishable from the
corresponding veridical perceptions that I think I’ve had. Call this proposition (BIV).
Now consider the following argument:
656 SKEPTICISM
(P1) If I know that I have hands, then I know that ~(BIV).

(P2) I don’t know that ~(BIV).
Therefore,
(C) I don’t know that I have hands.
The following principle can be proffered in favor of (P1):
Competent Deduction Closure
If S knows that p, and S competently deduces from p that q, thereby forming a
belief that q on this basis while retaining her knowledge that p, then S knows
that q.8
We might think that this principle—“closure” for short—just gives expression to
our intuition that competent deduction from known propositions can be a means
of extending our knowledge to the known consequences of those propositions.
Since we know that having hands implies that we are not bodiless, and so not
bodiless brains in vats, the closure principle implies that, if we do know that we
have hands, then we can know that ~(BIV). But ~(BIV) seems to be just the sort
of proposition we cannot know to be true. After all, how could I know that I’m
not a bodiless brain in a vat if being the victim of that kind of scenario would
produce in me just the sorts of experiences I would have were I an embodied
human being, active in the familiar world around us? This makes it look like I don’t
know that ~(BIV). But this together with (P1) entails that I don’t know that I have
hands, which is counterintuitive.
How disjunctivism fares against this problem depends on how we understand
the skeptical challenge. For example, if we frame the problem of the external world
as arising out of the supposed fact that sense experience underdetermines the
choice between various propositions about the world around us and (known to be
incompatible) competitor propositions, then disjunctivism about the nature of sense
experience might be able to provide a response to this challenge. If we conceive of
the nature of sense experience in general as disjunctive—as dividing into states in
which either we perceive external material objects or states in which we seem to
do so but fail (as with indistinguishable hallucinations)—then why couldn’t some
sense experiences put us in a position to know propositions about the world around
us, and so put us in a position to infer that ~(BIV) from our knowledge that those
propositions are true, even if our capacity to do so is fallible?
Call this the simple disjunctivist response. An immediate problem for the simple
disjunctivist response is that it’s not at all clear that a core component of the problem
of the external world depends on a view about the relation between sense experience
and the external world.9 Moreover, even if it is a core component of the problem, it’s
not clear that mounting a competing thesis is sufficient to remove the challenge. After
all, certain arguments make it look plausible that sense experience underdetermines
the choice between propositions about the world and their corresponding skeptical
competitors. The disjunctivist would have to speak to these arguments.10
Another concern is that the disjunctivist about sense experience just misses the
point that the problem of the external world is an epistemological problem, arising
out of core epistemological intuitions. For example, (BIV) might express a scenario
where the subject is not perceiving the world around her for how it is (since she’s a
bodiless brain in a vat, even though it doesn’t seem to her that she is). But what is
at issue with (BIV) is not the fact that she cannot perceive the world around her for
how it is, but that whether or not (BIV) is true, it’s not clear that we could come to
know that ~(BIV). The reason need not be that the kinds of experiences we have
in either case are experiences of the same nature. Instead, it can be that we cannot
distinguish between cases in which we do get things right from cases in which we
seem to get things right but nevertheless fail.
For example, one might argue in support of the following principle, which we
will call perceptual discrimination:
Perceptual Discrimination (PD)
If S knows that p on the basis of perceptual experience, then S can discriminate
cases in which p from cases in which ~p.
For example, if I know that a goldfinch is in the garden on the basis of visual
experience, then I must be able to discriminate the case where it’s a goldfinch in
the garden from other kinds of birds being in the garden. If not, what makes it the
case that the visual experience supplies me with the knowledge that it’s a goldfinch
rather than some other bird, or something less specific like that a bird is in the
garden? With (PD), there is a quick move available to (P2) (i.e., ~K~(BIV)). After
all, I cannot discriminate cases in which I see hands from cases in which I am a
bodiless brain in a vat who seems to see hands.11
Moreover, we can leave (PD) to one side and just consider the closure principle.
This latter principle seems to reflect a general epistemological intuition, and it
implies that if we know propositions about the world around us, such as that we
have hands, then we can know ~(BIV), at least provided we make the competent
deduction. But we might just take it as primitive that we don’t know that ~(BIV).
It might just reflect a basic epistemological intuition: an intuition as basic as that
not all propositions can be known. If this is right, then the proponent of the simple
disjunctivist response will need to explain why it isn’t primitive, even if initially it
strikes us as being primitive.
How might the proponent of the simple disjunctivist response overcome this
challenge? One thought is that she should speak directly to epistemological concerns.
For example, she could say that our perceptual relation to mind-independent
external objects is sufficient for knowledge-level justification: justification which, if
we have it, would put us in a position to know. After all, what else could be better
reason for believing that I have hands than just my hands themselves being part of
the very sensory episode that I enjoy? But why is disjunctivism any better off than,
say, its intentionalist rival in arguing that perceptions provide us with knowledge-
level justification?
More generally, there is a strong intuition in favor of the idea that there’s nothing
about the “directness” of a sensory perception that makes it any better a candidate
for knowledge-level justification than a matching hallucination. The thought here
finds a clear expression in these remarks by Earl Conee:
658 SKEPTICISM
It remains strongly intuitive that perception does not provide any better reason
for an external world belief than would be provided by a matching hallucination.
There is no a priori limit on the potential verisimilitude of hallucination. Whether
S1 is in perceptual state P1 or the matching hallucination H1, it appears to S1
exactly as though T1 is true. Intuitively, that is as good a basis for reasonable
belief in T1 as is the veridical perception. Thus, since the strength of epistemic
justification is equal to the strength of reasonableness, a perception does not
justify more strongly than does a matching hallucination. (Conee 2007: 18)
Like the previous criticism, this criticism challenges the epistemological benefits of
the disjunctive conception of sensory experience. It says that even if the sensory
experiences implicated in perception and hallucination are metaphysically different
kinds of states, what remains epistemologically problematic is the fact that we
cannot introspectively distinguish them.
Notice that this challenge appeals to the following intuitive principle:
Same Epistemic Status (SES)
If a perceptual state E provides S with an epistemic standing for her belief that
p, then any perceptual state R which is (for S) introspectively indistinguishable
from E provides at least as much of an epistemic standing for S to believe that p.
If we accept (SES), then the epistemological benefits of the disjunctive conception
of sensory experience dissipate.
3 EPISTEMOLOGICAL DISJUNCTIVISM
Let us now consider a distinctively epistemological form of disjunctivism, which
we will refer to as epistemological disjunctivism. Like its metaphysical counterpart,
epistemological disjunctivism is composed of a negative thesis about the fundamental
difference in epistemic significance between successful and unsuccessful cases of
sensory experience and a positive thesis about the nature of the epistemic significance
of the sensory experience in the successful case.
Epistemological disjunctivism’s negative thesis consists of its rejection of (SES),
which we just encountered. In particular, epistemological disjunctivists will contend
that in pairings of introspectively indistinguishable good and bad cases, the subject
concerned has a much stronger epistemic standing for her belief in the good case
than in the bad case.
(SES) can look appealing because, given that good and bad cases are
indistinguishable, it is hard to fathom how the epistemic standing of one’s belief
in the good case could be better than the epistemic standing of one’s belief in
the corresponding bad case.12 Nonetheless, epistemological disjunctivism rejects
this thesis. In particular, it argues that we should not limit the epistemic support
available for our beliefs to the low level of epistemic support which is available in
the bad case. Instead, we should regard the epistemic standing of one’s belief in the
corresponding good case as being far superior.13
In specifying that the epistemic support is different in the two cases, however,
this still leaves open a number of options in terms of what constitutes this difference.
For example, is the difference between the subject’s epistemic standing in the
good case and the bad case one of kind or only of degree? Either way, what is
crucial to epistemological disjunctivism is the idea that there is epistemic support
available to one in the good case that’s unavailable in the bad case. Call this the core
epistemological disjunctivist thesis.
It is an interesting question how metaphysical and epistemological disjunctivism
relate to one another. They are certainly dialectically affiliated in the sense that, if
one endorses metaphysical disjunctivism, then one will be naturally inclined to also
endorse epistemological disjunctivism, and vice versa. But whether they actually
entail one another is a different matter. In particular, it seems at least possible to
hold the one thesis while rejecting the other, even though such a combination of
views may well be philosophically quite uncomfortable.14
It should be clear that epistemological disjunctivism’s prospects for dealing
with the problem of radical skepticism are more promising than metaphysical
disjunctivism. Recall that one core issue facing the latter proposal was that it was
attempting to confront an epistemological problem and yet its own epistemological
ramifications were moot. For epistemological disjunctivism, there is no such
difficulty. In particular, by rejecting (SES) epistemological disjunctivism directly
challenges a claim which seems to be presupposed in the radical skeptical problem.
Radical skeptical scenarios, after all, are extreme examples of bad cases. And yet
part of what is motivating the skeptical thought that we are unable to rule out such
scenarios is that our epistemic standing in the good case is no better than it would
be in a corresponding radical skeptical scenario (on account of the latter being
indistinguishable from the former). Once we have rejected (SES), however, then
what would prompt us to concede this point to the skeptic? Why not instead insist
that so long as one is in the good case, then one’s epistemic support is substantially
greater than it is in the skeptical bad case, and hence that one has an epistemic basis
on which one can reject this error-possibility?
Whether such a line of argument is tenable will to a large extent depend on the
positive thesis that the epistemological disjunctivist offers about the nature of the
epistemic support in the good case. In particular, the problem facing epistemological
disjunctivism on this score is that if the positive account is cast along epistemic
externalist lines then the view becomes pedestrian, while if it is cast along epistemic
internalist lines the view will strike many as absurd.
Consider first an epistemic externalist rendering of epistemological disjunctivism,
whereby the different epistemic standing of the subject’s belief in the pairs of good
and bad cases are not reflectively accessible to the subject.15 Such a view would be
pedestrian because it is entirely normal for epistemic externalists to hold that a
subject’s epistemic standing differs across good and bad cases in this way. After all,
on epistemic externalist views the epistemic standing of one’s beliefs is dependent
on non-reflectively accessible factors, such as the reliability of the belief-forming
process through which one acquired one’s belief. Hence it is unsurprising that in
660 SKEPTICISM
the bad cases, where those externalist factors do not obtain, the subject’s epistemic
standing will tend to be much worse than it is in the corresponding good case.
A further upshot of the foregoing is that an epistemological disjunctivism which
is wedded to epistemic externalism will offer a response to the problem of radical
skepticism which is no more plausible than the standard epistemic externalist responses
which many find unpersuasive.16 In particular, the normal epistemic externalist line
on radical skepticism is to treat our knowing as being in a sense contingent on factors
beyond our ken. That is, while the skeptic insists that knowledge is (for the most part
anyway) impossible, the epistemic externalist anti-skeptic argues that knowledge is
possible, just so long as we are in the right kind of conditions which would sustain
it. Whether we are in such conditions, however, is not held to be something that
we can determine by reflecting on the nature of our epistemic position, since these
conditions concern facts about the world such as whether we are indeed reliably
forming our beliefs (as opposed to being a brain in a vat who merely thinks that she
is). The combination of epistemological disjunctivism and epistemic externalism is no
different on this score. If we are indeed in the good case rather than the bad, then we
are in the market for knowledge, contra the skeptic. But whether we are in the good
case depends on factors outwith our reflective ken.17
Hence, if epistemological disjunctivism is to offer anything distinctive to the
contemporary debate about radical skepticism, then it will need to be cast along
epistemic internalist lines. Such a view would hold that the superior epistemic
standing available to the subject’s belief in the good case is reflectively accessible
to the subject. One can straight away see a difficulty looming for this approach,
which is why it is thought unattractive (indeed, simply unavailable). For given that
pairs of good and bad cases are by definition introspectively indistinguishable to the
agent, then how can it be that in the good case one’s superior epistemic standing is
reflectively accessible?
We will examine this issue by looking at a particular version of epistemological
disjunctivism which is cast along epistemic internalist lines. This is the approach
offered by John McDowell (e.g., 1995) and, following him, Duncan Pritchard (e.g.,
2012a).18 On this view, one’s perceptual knowledge can consist, in paradigmatic
cases at least, in one’s possession of rational support that is both reflectively
accessible and factive. More specifically, in such cases one perceptually knows that
p in virtue of seeing that p, where that one sees that p is one’s reflectively accessible
rational support for believing that p.
On this form of epistemological disjunctivism we thus get a particularly strong
rendering of the epistemic standing of the subject’s belief in the good case, in that it
is an epistemic standing which actually entails the truth of the believed proposition.
Clearly this is an epistemic standing which is unavailable to the subject in the bad
case, since as bad cases are usually described the target belief is false.19
This form of epistemological disjunctivism is distinctive in another respect too,
in that the very idea that there can be factive yet reflectively accessible reasons of
this kind is thought highly controversial. The claim that knowledge consists of the
possession of reflectively accessible rational support marks the view out as a kind of
epistemic internalism.20 But, on standard forms of epistemic internalism, the facts
to which one has such special epistemic access are by their nature not such that
they entail specific claims about the subject’s environment (but rather concern, for
example, the subject’s mental states, narrowly understood). In contrast, whereas
epistemic externalism does allow that knowledge can be in virtue of an epistemic
standing which entails specific facts about the subject’s environment, these facts
(such as facts about the reliability of the process through which the subject acquired
her belief) are not held to be reflectively accessible to the subject. By allowing that
one’s epistemic standing can be both reflectively accessible and also entail specific
facts about one’s environment, this form of epistemological disjunctivism is thus a
highly non-standard version of epistemic internalism.
Is this view tenable? We noted above that there is at least a strong prima facie
reason for thinking not. For how is one to square this form of epistemological
disjunctivism with the claim that pairs of good and bad cases are introspectively
indistinguishable to the subject? On this view, in the good case one’s rational support
for believing that p can be that one sees that p, where it is reflectively accessible to
one that one has this rational support. But if there are facts reflectively accessible to
the subject which are only available in the good case (and the subject is in a position
to know that this is so on purely a priori grounds), then surely the subject can,
through an entirely reflective process, come to know that she is in the good case as
opposed to the bad case. Isn’t that in direct conflict with the stipulation that pairs of
good and bad cases are introspectively indistinguishable?21
There is a further, but related, problem in play here. For consider our subject’s
perceptual knowledge that there is a table before her in virtue of her possession of
the relevant factive reason. Given that this rational support is reflectively accessible,
and given that the subject is also presumably in a position to know on a purely
a priori basis that this rational support is factive, then what is to stop the subject
acquiring a completely non-empirical route to the knowledge that there is a table
before her? But that seems absurd. How can it be possible to come to know specific
facts about one’s environment through purely reflective processes?22
Let us grant for the sake of argument that these problems can be satisfactorily
resolved and consider the distinctive approach epistemological disjunctivism, so
understood, offers us for the problem of radical skepticism.23 On this view it is not
just that one’s belief has a better epistemic standing in the good case than it does
in the (skeptical) bad case. Rather, one’s belief in the good case has an epistemic
standing which is factive (i.e., which entails the truth of the proposition believed)
and which is also reflectively accessible to the subject. It follows that the subject has
excellent—indeed, decisive—rational grounds available to her to prefer her everyday
beliefs to radical skeptical alternatives. Hence, this version of epistemological
disjunctivism can take a broadly “Moorean” line with radical skepticism and insist
that we can know the denials of radical skeptical hypotheses after all. Moreover, this
view can diagnose the enduring attraction of radical skepticism as being due to our
having implicit philosophical commitments to such theses as (SES), theses which the
epistemological disjunctivist argues we should jettison.24
One concern with this way of dealing with the problem of radical skepticism
is that it can strike one as too strong. On this view, for example, it seems
662 SKEPTICISM
somewhat mysterious that the problem of radical skepticism has had such a
hold on the philosophical imagination for so long. Doesn’t the radical skeptic
succeed in highlighting anything important about the human epistemic condition?
Moreover, given the difficulties which plague epistemological disjunctivism on
this construal—problems which we have set to one side for now—this response to
radical skepticism has an essentially contingent status. This inevitably diminishes
somewhat the intellectual comfort that might be offered by this way of responding
to radical skepticism.25,26
NOTES
1. Although support for naïve realism has grown. See Snowdon (1990), Campbell
(2002; 2011), Martin (2004; 2006), Brewer (2007; 2011), and Fish (2009b).
2. This is how Martin (2004; 2006) formulates the argument from hallucination.
3. Note that the disjunctivist is not claiming that no theory of the nature of sensory
experience can be given. Instead, she is claiming that no theory of the nature of
sensory experience can be given which appeals to what is metaphysically common
between perception and hallucination. Instead, she might recommend that the
nature of sensory experience in general should be explained epistemologically: as
our inability to know introspectively a perception from a metaphysically different
kind of state. See Martin (2004) for an expression of this kind of theory.
4. Put another way, we might say that cases of hallucination that are indistinguishable
from veridical perception are states in which we are not only misled about how
the world is, but about how we are experientially: introspectively, we are apt
to judge that we enjoy a veridical perception of some object, when in fact we
suffer a mere appearance of a veridical perception of some object. See McDowell
(2008) for this way of characterizing the relationship between hallucination and
introspective knowledge.
5. See Martin (2002: 395n).
6. Cf. Martin (1997).
7. For recent work on whether the distinction is as deep as some philosophers of
perception seem to suggest, see Brewer (2011) and Robinson (2012); For some useful
recent overviews of the literature on disjunctivism, see Haddock & Macpherson
(2009), Byrne & Logue (2009), Fish (2009a), Soteriou (2009), Brogaard (2010),
and Dorsch (2011).
8. This is essentially the formulation of the closure principle put forward by Williamson
(2000a: 117) and Hawthorne (2005: 29).
9. See Wright (2008) in particular for this view.
10. One might feel that this objection turns on the debate between the disjunctivist
response to the argument from hallucination and illusion, and its rivals. After all,
the disjunctivist will say that what makes it look like sense experience in general
underdetermines the truth-value of propositions about the external world is an
argument from hallucination/illusion, blocking the support for the thesis that
compelling arguments establish that sense experiences underdetermines the truth-
value of those kinds of propositions.
11. See Pritchard (2010; 2012: part two) for further discussion of discrimination-style
principles like (PD) and their role in an account of perceptual knowledge.
12. We are in effect here motivating (SES) by appeal to the so-called “new evil genius”
thesis—roughly, that one has the same extent of epistemic support for one’s beliefs
that one’s envatted counterpart enjoys for her beliefs. This is meant to be a core
epistemological intuition that all epistemological theories must either accommodate
or explain away. Epistemological disjunctivists opt for the latter horn—see, for
example, Neta & Pritchard (2007) and Pritchard (2012: passim). The locus classicus
for discussions of the new evil genius thesis is Lehrer and Cohen (1983). For an
excellent survey of recent work on this topic, see Littlejohn (2009).
13. As McDowell (e.g., 1995) famously tends to put the point: we should reject a
“highest common factor” conception of epistemic standing such that the epistemic
standing enjoyed by one’s beliefs in the good case is no better than that available for
one’s beliefs in the corresponding bad case. See Pritchard (2012: part one) for an
elaboration of McDowell’s argument.
14. Various commentators have argued that epistemological disjunctivism does not
in itself entail metaphysical disjunctivism. See, for example, Snowdon (2005),
Millar (2007; 2008), Byrne and Logue (2008), McDowell (2008: 382n), and
Pritchard (2008). In particular, the claim in this regard is often that epistemological
disjunctivism is compatible with a causal theory of perceptual experience, as
defended, for example, by Grice (1961) and Strawson (1974). For further discussion
of the logical connections between metaphysical and epistemological disjunctivism,
see Haddock & Macpherson (2008), Byrne and Logue (2009), Fish (2009a),
Soteriou (2009: esp. §2.4), and Pritchard (2012: part one).
15. Note that we are here casting the epistemic externalism/internalism distinction along
broadly accessibilist lines. That is, roughly, we are understanding an internalist
epistemic condition such that its obtaining should be reflectively accessible to
the subject, and hence treating an externalist epistemic condition such that it is
not reflectively accessible to the subject. Although this is a fairly standard way
of understanding the epistemic externalism/internalism distinction, it is far from
being uncontroversial. This way of construing the distinction should suffice for
our purposes, however. For a useful recent overview of the epistemic externalism/
internalism distinction, see Vahid (2011).
16. See Stroud (1989; 1994) for a subtle articulation of this kind of concern regarding
standard epistemic externalist responses to radical skepticism.
17. For a variety of epistemological disjunctivism of an epistemic externalist stripe
which is not pedestrian, see Williamson (2000), though note that he does not
himself describe his view as being a variety of epistemological disjunctivism.
What distinguishes this view from standard epistemic externalist proposals in this
regard is that it is wedded to some further distinctive philosophical claims, such
as that knowledge is a mental state. See Pritchard (2011b) for more on this point.
Interestingly, some commentators, such as Brueckner (2010), explicitly run together
the McDowellian-style epistemological disjunctivism put forward by Pritchard (e.g.,
2012) and the kind of position defended by Williamson.
18. See also McDowell (1982; 1986; 1994a,b; 1995; 2002a; 2008; 2011) and Pritchard
(2007; 2008; 2009; 2011a,b).
664 SKEPTICISM
19. Note that we are glossing over some of the nuances regarding the good/bad case
distinction which are relevant to at least Pritchard’s rendering of this form of
epistemological disjunctivism. See Pritchard (2012: part one) for more details. We
are also side-stepping the issue of what the subject’s epistemic standing is in the bad
case on this view. What is salient for our purposes is just that it falls well short of
the factive rational support available to the subject in the good case.
20. Of an accessibilist variety at least (see note 16). See Neta and Pritchard (2007)
and Pritchard (2011a; 2012: part one) for further discussion of how this form of
epistemological disjunctivism should be understood in relation to the epistemic
externalism/internalism distinction.
21. For a more extensive discussion of this difficulty, which he refers to as the
“distinguishability problem,” see Pritchard (2012: part one).
22. For a more extensive discussion of this difficulty, which he refers to as the
“access problem,” see Pritchard (2012: part one). The problem in play here is
structurally akin to the so-called McKinsey problem which is held to afflict the
combination of (some forms of) content externalism and first-person authority.
See McKinsey (1991).
23. Pritchard (2011a; 2012: passim) claims that these problems—along with a third
problem, which he refers to as the “basis problem” can be resolved. Note that the
basis problem turns on an issue which we haven’t explored here—viz., whether
seeing that p is a way of knowing that p. See Ranalli (2014) for a discussion of the
issues in this regard.
24. See Pritchard (2008; 2009; 2012: part three) for a development of this way of
exploiting epistemological disjunctivism to deal with the problem of radical
skepticism. For a critical appraisal of this proposal, see Schönbaumsfeld (2013).
25. In more recent work, Pritchard (2015) has argued that epistemological disjunctivism
of this variety should only be thought of as applying to that variety of radical
skepticism which trades on the claim that the rational support we have for our
beliefs is underdetermined with respect to radical skeptical scenarios.
26. This chapter was written as part of the AHRC-funded “Extended Knowledge” project
which is hosted by the University of Edinburgh’s Eidyn Philosophical Research
Centre, and we are grateful to the AHRC for their support of this research. Thanks
also to Baron Reed for detailed comments on an earlier version of this chapter.
REFERENCES
Brewer, Bill. 2007. “Perception and its Objects,” Philosophical Studies 132: 87–97.
Brewer, Bill. 2011. Perception and its Objects. Oxford: Oxford University Press.
Brogaard, Berit. 2010. “Disjunctivism,” Oxford Bibliographies: Philosophy, DOI:10.1093/
OBO/9780195396577-0033.
Brueckner, Anthony. 2010. “¬K¬SK.” In his Essays on Skepticism, 367–381. Oxford:
Byrne, Alex and Heather Logue. 2008. “Either/Or.” In A. Haddock and F. Macpherson
(eds.), Disjunctivism: Perception, Action, Knowledge, 57–94. Oxford: Oxford University
Press.
Byrne, Alex and Heather Logue. 2009. “Introduction.” In A. Byrne and H. Logue
Disjunctivism: Contemporary Readings, vii–xxix. Cambridge, MA: MIT Press.
Campbell, John. 2002. Reference and Consciousness. Oxford: Oxford University Press.
Campbell, John. 2011. “Relational versus Kantian Responses to Berkeley’s Puzzle.” In J.
Roessler, H. Lerman, and N. Eilan (eds.), Perception, Causation, and Objectivity, 35–50.
Conee, Earl. 2007. “Disjunctivism and Anti-Skepticism,” Philosophical Issues 17: 16–36.
Dorsch, Fabian. 2011. “The Diversity of Disjunctivism,” European Journal of Philosophy
19: 304–314.
Fish, William. 2009a. “Disjunctivism.” In B. Dowden and J. Fieser (eds.), Internet
Encyclopedia of Philosophy, http://www.iep.utm.edu/d/disjunct.htm.
Fish, William. 2009b. Perception, Hallucination, and Illusion. Oxford: Oxford University
Press.
Grice, H. P. 1961. “The Causal Theory of Perception,” Proceedings of the Aristotelian
Society (suppl. vol.) 35: 121–152.
Haddock, Adrian and Fiona Macpherson. 2008. “Introduction: Varieties of Disjunctivism.”
In A. Haddock and F. Macpherson (eds.), Disjunctivism: Perception, Action, Knowledge,
Hawthorne, John. 2005. “The Case for Closure.” In E. Sosa and M. Steup (eds.),
Contemporary Debates in Epistemology, 26–43. Oxford: Blackwell.
Lehrer, Keith, and Stewart Cohen. 1983. “Justification, Truth, and Coherence,” Synthese
55: 191–207.
Littlejohn, Clayton. 2009. “The New Evil Demon Problem.” In B. Dowden and J. Fieser
(eds.), Internet Encyclopedia of Philosophy, www.iep.utm.edu/evil-new/.
Martin, M. G. F. 1997. “The Reality of Appearances.” In M. Sainsbury (ed.), Thought and
Ontology, 77–96. Milan: Franco Angeli.
Martin, M. G. F. 2002. “Transparency of Experience,” Mind & Language 17, 376–425.
Martin, M. G. F. 2004. “The Limits of Self-Awareness,” Philosophical Studies 120: 37–89.
Martin, M. G. F. 2006. “On Being Alienated.” In T. S. Gendler and J. Hawthorne (eds.),
Perceptual Experience, 354–410. Oxford: Oxford University Press.
McDowell, John. 1982. “Criteria, Defeasibility and Knowledge,” Proceedings of the British
Academy 68: 455–479.
McDowell, John. 1986. “Singular Thought and the Extent of Inner Space.” In P. Pettit and
J. McDowell (eds.), Subject, Thought and Context, 137–168. Oxford: Clarendon Press.
McDowell, John. 1994a. “Knowledge by Hearsay.” In B. K. Matilal and A. Chakrabarti
(eds.), Knowing from Words: Western and Indian Philosophical Analysis of Understanding
and Testimony, 195–224. Dordrecht: Kluwer.
McDowell, John. 1994b. Mind and World. Cambridge, MA: Harvard University Press.
McDowell, John. 1995. “Knowledge and the Internal,” Philosophy and Phenomenological
Research 55: 877–893.
McDowell, John. 2002a. “Knowledge and the Internal Revisited,” Philosophy and
McDowell, John. 2002b. “Responses.” In N. H. Smith (ed.), Reading McDowell: On Mind
and World. London: Routledge.
666 SKEPTICISM
McDowell, John. 2008. “The Disjunctive Conception of Experience as Material for a

Transcendental Argument.” In A. Haddock and F. Macpherson (eds.), Disjunctivism:
Perception, Action, Knowledge, 376–389. Oxford: Oxford University Press.
McDowell, John. 2011. Perception as a Capacity for Knowledge. Milwaukee WI: Marquette
University Press.
McKinsey, Michael. 1991. “Anti-Individualism and Privileged Access,” Analysis 51: 9–16.
Millar, Alan. 2007. “What the Disjunctivist is Right About,” Philosophy and
Millar, Alan. 2008. “Perceptual-Recognitional Abilities and Perceptual Knowledge.” In
A. Haddock and F. Macpherson (eds.), Disjunctivism: Perception, Action, Knowledge,
Nagel, Thomas. 1974. “What Is It Like to Be a Bat?,” Philosophical Review 83: 435–450.
Neta, Ram, and Duncan Pritchard. 2007. “McDowell and the New Evil Genius,” Philosophy
Peacocke, Christopher. 1983. Sense and Content: Experience, Thought and Their Relations.
Pritchard, Duncan. 2007. “How to be a Neo-Moorean.” In S. Goldberg (ed.), Internalism
and Externalism in Semantics and Epistemology, 68–99. Oxford: Oxford University
Press.
Pritchard, Duncan. 2008. “McDowellian Neo-Mooreanism.” In A. Haddock and
F. Macpherson, Disjunctivism: Perception, Action, Knowledge, 283–310. Oxford:
Pritchard, Duncan. 2009. “Wright Contra McDowell on Perceptual Knowledge and
Skepticism,” Synthese 171: 467–479.
Pritchard, Duncan. 2010. “Relevant Alternatives, Perceptual Knowledge and
Discrimination,” Noûs 44: 245–268.
Pritchard, Duncan. 2011a. “Epistemological Disjunctivism and the Basis Problem,”
Philosophical Issues 21: 434–455.
Pritchard, Duncan. 2011b. “Evidentialism, Internalism, Disjunctivism.” In T. Dougherty
(ed.), Evidentialism and its Discontents, 362–392. Oxford: Oxford University Press.
Pritchard, Duncan. 2012. Epistemological Disjunctivism. Oxford: Oxford University Press.
Pritchard, Duncan. 2015 Epistemic Angst: Radical Skepticism and the Groundlessness of
Our Believing. Princeton, NJ: Princeton University Press.
Ranalli, Chris. 2014. “Luck, Propositional Perception, and the Entailment Thesis,” Synthese
191: 1223–1247.
Robinson, Howard. 2012. “Relationalism versus Representationalism: How Deep is the
Divide?” Philosophical Quarterly 62, 614–619.
Russell, Bertrand. 1910. “Knowledge by Acquaintance and Knowledge by Description,”
Russell, Bertrand. 1912. The Problems of Philosophy. London: Home University Library.
Schönbaumsfeld, Genia. 2013. “McDowellian Neo-Mooreanism?,” International Journal
for the Study of Skepticism 3: 202–217.
Snowdon, Paul. 1990. “The Objects of Perceptual Experience,” Proceedings of the
Aristotelian Society (suppl. vol.) 64: 121–150.
Snowdon, Paul. 2005. “The Formulation of Disjunctivism: A Response to Fish,” Proceedings

of the Aristotelian Society 105, 129–141.
Soteriou, Mathieu. 2009. “The Disjunctive Theory of Perception.” In E. Zalta (ed.), Stanford
Encyclopedia of Philosophy, http://plato.stanford.edu/entries/perception-disjunctive/.
Strawson, P. F. 1974. “Causation in Perception.” In his Freedom and Resentment, 73–93,
London: Methuen.
Stroud, Barry. 1989. “Understanding Human Knowledge in General.” In M. Clay and
K. Lehrer (eds.), Knowledge and Skepticism, 31–50. Boulder, CO: Westview Press.
Stroud. Barry. 1994. “Scepticism, ‘Externalism,’ and the Goal of Epistemology,” Proceedings
of the Aristotelian Society (suppl. vol.) 68: 291–307.
Vahid, Hamid. 2011. “Externalism/Internalism.” In S. Bernecker and D. H. Pritchard
(eds.), Routledge Companion to Epistemology, 144–155. London: Routledge.
Williamson, Timothy. 2000. “Scepticism and Evidence,” Philosophy and Phenomenological
Research 60: 613–628.
Wright, Crispin. 2008. “Comments on McDowell’s ‘The Disjunctive Conception
of Experience as Material for a Transcendental Argument.’” In A. Haddock and
F. Macpherson (eds.), Disjunctivism: Perception, Action, Knowledge, 390–404. Oxford:
CHAPTER FORTY-SIX
Inductive Skepticism:
A Functional Investigation
RUTH WEINTRAUB
1 INTRODUCTION
I wish to present a new way of responding to skepticism and consider its application
to inductive skepticism. The chapter is structured as follows. I present Hume’s
argument in support of inductive skepticism (Section 2) and draw some distinctions
between types of responses to skepticism (Section 3). Having explained what I
mean by a “functional” response, I illustrate its application to Hume’s inductive
skepticism (Sections 4–5).
2 HUME’S SKEPTICAL ARGUMENT

AGAINST INDUCTION
Hume’s argument against induction is presented in the course of his (complex)
discussion of causality in the Treatise (1739: 1.3.4), and more succinctly and
perspicuously in the Enquiries (1777: 35–36):
(1) The conclusion of an inductive argument isn’t logically entailed by
its premises. “Concerning matter [sic] of fact and existence,” Hume
claims, “there are no demonstrative arguments . . . since it implies no
contradiction that the course of nature may change . . . and the trees will
flourish in December. . . . If we be, therefore, engaged by arguments to
put trust in past experience . . . these arguments must be probable only”
(1777: 35).
(2) Every inductive argument assumes that nature is uniform; “that the future
will be conformable to the past.”
(3) The principle of the uniformity of nature, which must be warranted if
induction is to be justified, cannot be justified a priori.
(4) The principle of the uniformity of nature cannot be justified a posteriori
since such a justification would be circular. It would itself be inductive,
INDUCTIVE SKEPTICISM: A FUNCTIONAL INVESTIGATION 669
inferring nature’s uniformity tout court from its observed uniformity, thus
presupposing—like any inductive argument—that nature is uniform.
3 SOME DISTINCTIONS BETWEEN TYPES OF

RESPONSE TO SKEPTICISM
The first distinction I wish to draw is between combative and acquiescent responses
to a skeptical claim. A combative response attempts to rebut the skeptical
arguments adduced in support of the skeptical claim and show that the claim is,
in fact, false. (These are two distinct aims.) This can be done in (at least) three
ways. Russell (1948: 9) claims that we are justified in taking “for granted that
scientific knowledge, in its broad outlines, is to be accepted,” because the skeptical
doctrine cannot lastingly and steadily be believed, and anyone who preaches it is
being “frivolous and insincere.” This is, perhaps, the least promising of combative
responses. If skepticism is “logically impeccable” (Russell 1948: 9), the insincerity
of its proponents does not show that it is false or that the arguments adduced in
its support aren’t cogent.
The second way of rejecting the skeptical conclusion is to take its implausibility
as a reason for thinking that it is false (Reid 1764; Moore 1953; Chisholm 1982;
Greco 2000). The third is to offer a diagnosis of the flaw in the seemingly cogent
skeptical argument. It can be illustrated by considering the argument (in the form
of a dilemma) adduced by Sextus Empiricus, the ancient Greek skeptic, to “set aside
the method of induction”:
When they propose to establish the universal from the particulars by means of
induction, they will effect this by a review either of all or of some of the particular
instances. But if they review some, the induction will be insecure, since some of
the particulars omitted in the induction may contravene the universal; while if
they are to review all, they will be toiling at the impossible, since the particulars
are infinite and indefinite. [PH II 204; I follow Bury’s (1933) translation]
In a similar (skeptical) vein, Popper suggests that “any conclusion drawn in this
[inductive] way may always turn out to be false: no matter how many instances
of white swans we may have observed, this does not justify the conclusion that all
swans are white” (1972: 27, original italics).
The argument can be rebutted (thereby exemplifying the third version of the
combative strategy) by pointing out that justification is fallible. We can concede
that the conclusion of an inductive argument may be false even if its premises are
true, and that in this induction differs from, indeed falls short of, deduction. But
that doesn’t mean, as the skeptic is (implicitly) assuming, that inductive grounds do
not provide good reasons. The term “reason,” we will point out (Edwards 1949),
doesn’t mean “logically conclusive,” and inductive grounds, despite being logically
inconclusive, may well provide perfectly good reasons, thereby rendering their
conclusions rational. Taking a risk, we will remind the skeptic (and ourselves), may
be eminently reasonable. It is rational, for instance, to take a medicine one thinks
670 SKEPTICISM
is very likely, although uncertain, to cure a horrible disease even if it sometimes has
pretty unpleasant side-effects.
In contrast to a combative response to skepticism, an acquiescent response accepts
the skeptical claim (as true). This might engender distress, as it does in Hume. When
ruminating about his skeptical findings, he says:
The wretched condition, weakness, and disorder of the faculties, I must employ
in my enquiries, encreases my apprehensions. And the impossibility of amending
or correcting these faculties, reduces me almost to despair. (1739: 264)
But this (woeful) reaction isn’t mandatory. Even if we accept the skeptical claim, we
might try to take the sting out of it. Perhaps—initial appearances notwithstanding—
the skeptical claim isn’t as threatening as it seems to be, because it doesn’t deny us
something valuable.
We can neutralize the skeptical claim in two ways. The first is illustrated by
Sextus Empiricus. He is unperturbed by the skeptical conclusion because he is not
interested in the cognitive endeavors that it impugns. He aims to attain peace of
mind (ataraxia), an aim that acceptance of the skeptical conclusion may actually
promote (PH I 25–30).
Even if (contentiously) Sextus’s psychological claim is sound, it is irrelevant
if our perspective is epistemological (rather than practical, for instance). But the
skeptical claim about induction may have no sting even if our interest is cognitive.
The fact that induction is unjustified doesn’t constitute a threat to our rationality
if the practice the claim brands as irrational isn’t ours. This eliminativist response
to inductive skepticism provides us with a distinct way of rejecting Broad’s (1952:
143) memorable claim that induction is “the glory of science and the scandal of
philosophy.” Inductive skepticism may well be “a skeleton in the cupboard of
Inductive Logic” (Broad 1952: 143), but the cupboard isn’t ours!
In this (eliminativist) vein, Popper (1972) claims that science doesn’t invoke
induction. It progresses by formulating bold hypotheses and subjecting them to rigorous
tests. The only mode of reasoning it requires, Popper claims, is deduction. Predictions
(“observation statements”) are deductively derived from hypotheses, and, when
observationally refuted, a theory is deductively falsified: if T├ E, then ~E├ ~T. We
may now cheerfully concede the skeptical conclusion—that induction is irrational. Since
science does not invoke induction, its rational credentials are not thereby impugned.
We could shrug aside, or at least belittle the significance of, the skeptical claim,
even if it concerns our practice, a practice that we deem to be valuable. This might
sound paradoxical, but I have in mind the possibility that the skeptical claim pertains
to an epistemological concept (“justification,” “knowledge”) that is needlessly
stringent. Our beliefs, which never count as “justified” (“known”) when assessed in
the light of our concept of justification (knowledge), can, nonetheless, measure up to
the requirements of some other concept that is well-worth falling under (cognitively
speaking). To see whether the skeptical claim denies us something valuable, we need
to know what the point is of imposing the requirements the concept embodies, what
role it plays in our cognitive lives. Focusing on such considerations is the hallmark
of the functional response to skepticism.
The functional approach to the study of a concept is attractive for three reasons.
First, it will deliver a better understanding of the concept than that provided by
(extensionally adequate) necessary and sufficient conditions for its application.1
Second, it enables us to bypass the (vexing) question as to the real meaning(s) of
the term “justification” (“knowledge”), in the discussion of which so much energy
has been invested.2 Surely it is far more important to decide which (epistemic)
labels are worth fighting for, and whether these battles can be won. These questions
may be answered while remaining neutral about the actual meaning(s) of the term
“justified” (“knows”). Of course, we should—as a result of the investigation—adopt
a useful concept of justification (knowledge), but for that we don’t need to know
whether we are retaining or supplanting an existing one.
The third (and final) advantage of a functional investigation is that it might
(depending on its outcome) enable us to contend with an impeccable skeptical
argument. The skeptical conclusion may be true with respect to a concept that the
functional analysis reveals to be inappropriate, or at least, to have a non-skeptical
alternative that is “good enough.”
The functional perspective changes the rules of engagement with the skeptic.
Correct usage of the relevant term (“justification,” “knowledge”) is no longer
necessary for success, and this can work to the advantage of both sides. The skeptic
is often accused of distorting our epistemic standards (Schmitt 1992: 7–10). He is
likened to someone who claims that there are no physicians in New York and turns
out to mean by the term “physician” someone who can cure any disease in ten
seconds (Edwards 1949). Once this nonstandard usage is discerned, it is argued,
it becomes clear that the skeptic is not disputing with us, that he is “changing the
subject” (BonJour 1985: 37). But from the functional perspective, the skeptic cannot
be silenced in this way. If the concept he is invoking is epistemologically important,
our failure to measure up to it is damning, even if our concept is more permissive
and significantly less vulnerable to skepticism.
The converse is also true. Suppose the skeptic is using an epistemological
concept (knowledge or justification) correctly and exposes its skeptical implications.
Nonetheless, like any tool, a concept may turn out to be ill-suited for the use to
which it is put. And our investigation could reveal that our concept is needlessly
stringent, and that we are doing well according to a new concept that is suitably
stringent.
Since we aim to assess the usefulness of proposed concepts, rather than to analyze
existing ones, we will be indifferent to linguistic intuitions—an important tool
employed in the old-fashioned style of analysis. That we do not count as justified a
lucky guess shows that truth is insufficient for justification (knowledge) as we use
the term(s), but considerations of a different kind should be invoked to show that
counting a lucky guess as “justified” (known) leaves out something epistemologically
crucial.
The suggestion that we should focus on the point of concepts is not new. The
function and significance of knowledge attributions is explored by Craig (1987),
Sartwell (1992), Papineau (1992), Brandom (1998), and Kaplan (2008), and that
of justification by Alston (1988), Lycan (1988: ch. 7), Plantinga (1990), Bergmann
672 SKEPTICISM
(2000), and Weiner (2005), to name but a few. But my aim is significantly different.
First, my interest is not restricted to extant concepts, so the outcome of the
investigation may be a revision of our usage. Second, although I will be considering
a particular (reliabilist) concept of justification, it is by way of illustrating the
functional response and the way it transforms the debate. Other concepts could
turn out to be more appropriate.
If in our analysis we are not bound by our concepts, it might be wondered,
how do we determine whether a proposed concept is adequate? We cannot, ex
hypothesi, appeal to our intuitions about its application in particular cases, and
general platitudes about epistemic justification—for instance, that it is an evaluative
concept that has something to do with truth—do not suffice to pick out a sufficiently
determinate concept.
The answer is twofold. First, we can consider the point of concepts picked out by
proposed analyses of our concept(s) of justification. Although we are not concerned
with the extensional adequacy of an analysis, still, it picks out a concept that is of
the right sort, and worth considering (functionally). Second, we can discern features
of a concept of justification from skeptical arguments. We can then consider the
point of having a concept with these features and the loss attending the adoption of
a concept that lacks them. For instance, if the skeptic invokes a necessary condition
for justification (that he aims to show cannot be satisfied), we will consider the point
of imposing the condition.
With these clarificatory points in mind, I will now consider Hume’s argument
against induction from the functional perspective. From this perspective, we cannot
assume that the argument is concerned with justification (and not, for instance,
knowledge), because we are interested in (epistemological) concepts that may not
even be ours. But Hume’s argument is best construed as targeting the justification (as
determined by our concept) of inductive inferences.3 So to facilitate the exposition,
I will call the concepts I will be considering concepts of justification. These may
well not be our concept(s) of justification, since my interest is in those that violate
conditions that the argument imposes, conditions the skeptic thinks—with at least
some plausibility—are incorporated by our concept(s).
4 A FUNCTIONAL RESPONSE TO HUME’S

SKEPTICAL ARGUMENT AGAINST INDUCTION
I start by slightly modifying the argument, so as to make it more perspicuous and
easier to consider functional responses to it. The reformulated argument will also
enable us to see how the functional strategy can be applied to other skeptical claims
(pertaining, for instance, to deduction, sense-perception, and inference to the
best explanation).
What does it mean to say that nature is uniform? It can’t mean that all
generalizations (not even those that we can formulate in our language) are true.
First, some generalizations, “All ravens are black” and “All ravens are white,” for
instance, are mutually contradictory.4 Second, if it did mean that, there would be
no need to rely on positive instances of a generalization to infer its truth. Accepting
nature’s uniformity would render induction otiose. Finally, if this is how nature’s
uniformity is construed, then a single counterexample to a generalization would
falsify the principle of uniformity. But we don’t take a black swan to refute the
principle of uniformity (Ayer 1972: 22)!
The suggestion that nature is uniform if most generalizations (formulable in our
language) are true fails for a similar reason. If all ravens are black, then “All ravens
are white,” “All ravens are green,” etc., are all false. So every true generalization
brings in its wake several that are false.
The suggestion that nature is uniform if as many generalizations as possible are
true fails, too. It entails that if there is one black raven, all ravens are black. But we
don’t take the observation of a black raven and a white raven (for instance) to show
that nature is not uniform.
If the principle of uniformity is invoked when we reason inductively, it must
mean that sufficiently many generalizations (that we can formulate in our language)
are true. The idea is that there aren’t too many generalizations with both positive
instances and counterexamples. The principle is quite vague, but this is harmless.
The argument purports to show that the invocation of the principle is illicit whatever
the degree of uniformity the principle assumes (so long as it is positive, i.e., nature
isn’t perfectly chaotic, and the principle is not a logical truth).
Note, next, that the principle of uniformity is true if and only if induction is
reliable. Putting these two insights together, we can reformulate the argument
against induction:
If a belief derived inductively is to be justified, the argument’s first (“internalist”)
premise suggests, induction must be justifiably believed to be reliable.5 Without “a
patent of trustworthiness—a soundness argument—for a putative rule,” Howson
(2000: 28–29) argues, “you have no reason for confidence in any reasoning
which employs it.”
How are we to establish that induction is reliable (that nature is uniform, to use
Hume’s terminology)? Since the claim is contingent6 and (very) general, it cannot be
basic: it must, that is, be justified inferentially (Fumerton 1995; Van Cleve 2003).
But this inference, Hume argues, is inductive, based on induction’s track record.
So—by the first premise—it is subject to the same requirement: to be justified,
induction must first be shown to be reliable. And we have a vicious circularity:
induction’s reliability must be established prior to reasoning inductively, but it can
only be established by inductive reasoning. Inductively derived beliefs, the argument
concludes, are unjustified.
For ease of exposition, I will label this “Hume’s argument against induction” (and
sometimes “the argument against induction”), although Hume’s actual argument is
slightly different.7 The interpretive inaccuracy is harmless in our context.
My interest is in inductive skepticism, but the argument against induction can be
modified so as to apply to other belief-acquisition methods, such as inference to the
best explanation, perception, and deduction:
(1) If a belief, p, derived by a method, M, is to be justified, M must be justifiably
believed to be reliable. (Call this the “internalist” premise.)
674 SKEPTICISM
(2) The belief that M is reliable (call it q) is itself acquired via a method, M'
(which may be identical with M.)
(3) In order for q to be justified, its acquisition method, M', must be justifiably
believed to be reliable.
(4) The justification of p involves a chain of uses of methods, each giving rise to
the belief that the previous use of a method is a use of a reliable method.
(5) The chain of methods required for p to be justified is either circular or
infinite or terminating (with a belief that isn’t justified by another).
(6) None of the possibilities canvassed in (5) makes for genuine justification.
(7) Conclusion: p is unjustified.
Hume’s original argument cannot be generalized in this way (so that it applies to
other acquisition methods in addition to induction). Such a generalization requires
that some “super principle” be added to each inference. But it is not clear what the
analogue to the principle of the uniformity of nature is in the case of inference to the
best explanation or perception. What, for instance, does it mean to say that nature
is intelligible (explicable)? That inference to the best explanation is a reliable belief-
acquisition method, by way of contrast, makes perfect sense.
I propose to consider a concept of justification that doesn’t satisfy the internalist
requirement on the justification of inductive inference—reliabilism—and isn’t,
therefore, vulnerable to the modified version of Hume’s argument. Reliabilists
(Braithwaite 1953; Goldman 1979; Van Cleve 1984; Greco 2000) think that an
inference rule has to be reliable if its invocation is to be justified, but it may be
justifiably used by one who doesn’t believe that it is reliable.
The reliabilist concept of justification, call it R-justification, is satisfied by beliefs
that result from a reliable process—one whose products are typically true.8 Thus,
a belief derived inductively is R-justified if and only if induction is (sufficiently)
reliable, and the premises (of the inductive argument) are (sufficiently) reliably
acquired.
I am not interested in reliabilism as such. It merely serves to illustrate in a very
simple fashion how the functional strategy is to be implemented and how the
functional perspective transforms the debate with the skeptic. We could consider
concepts that violate other premises of the argument, and non-reliabilist concepts
that violate the internalist requirement. Indeed, this is what we should do if our aim
is a comprehensive exploration of the functional response.
From the functional perspective, the focus is on the point of imposing the
internalist requirement. We wish to know why internalist justification matters
and how severe is the (epistemic) loss incurred by the adoption of non-internalist
concepts. From this perspective, familiar, and seemingly very forceful, objections
are rendered irrelevant, and new questions come to the fore.
Our prospects in the dispute with the skeptic are quite bright. His argument,
which may have seemed invincible when couched in terms of the internalist concept
of justification, collapses once our attention turns to R-justification. Our new
concept is unabashedly externalist, and renders the first premise false!
What of an alternative skeptical argument to establish the impossibility of

R-justification? The skeptic’s task now is to show that induction isn’t reliable.
And, again, his prospects aren’t bright: induction’s track record is pretty good.9
Moreover, in pointing this out, we will be taking the offensive and showing the
skeptical claim to be false.
It will, of course, be objected that showing induction’s reliability by appealing
to its track record is circular and, therefore, illicit. Isn’t this one of Hume’s
truistic premises?
The objection is natural but mistaken. Hume’s assumption that circular
justification is illicit is eminently plausible given the internalist requirement. If
induction’s reliability has to be presupposed—as a premise—when invoking
induction, then inductively inferring induction’s reliability is viciously circular. The
circular justification of propositions is non-discriminating: every proposition entails
itself. But in adopting the concept of R-justification, we deny the first premise (of
the modified argument). And the question now is whether the invocation of an
inference rule to show that it is reliable is similarly illegitimate because trivial.
The answer is No. Even if we are allowed to invoke an inference rule in its own
justification, success is not guaranteed. The horoscope may predict the unreliability
of its predictions and the superiority of an alternative method of prediction (Lewis
1971). The same is true of the circular justification of induction. Consider the
inductive rule I1:
Most observed A’s have been B’s
____________________________ I1
The next A will be B

The rule I1 can be defended by invoking the following inductive argument, utilizing
I1 itself:
Most applications of I1 have been successful
____________________________ I1
The next use of I1 will be successful

Note that the premise is contingent, and the self-supporting defense can (logically)
fail. Indeed, it does fail in the case of the following inductive rule, I2:
All observed A’s have been B’s
__________________________ I2
The next A will be B

The (circular) justification fails, since the requisite premise is false:
All applications of I2 have been successful
__________________________ I2
The next use of I2 will be successful10

676 SKEPTICISM
We can now answer Alston’s contention that track-record arguments “do not serve
to discriminate between reliable and unreliable doxastic practices,” even if their
conclusion—that a method is reliable—is (reliabilistically) justified (1993: 17, my
italics). “We can just as well say of crystal ball gazing,” he explains, “that if it is
reliable, we can use a track record argument to show that it is reliable” (1993: 17,
original italics). But what is the “discrimination” that reliabilistically justified beliefs
fail to achieve? It can’t be that the track-record argument makes justification too
easy to achieve, that it can be invoked to vindicate doxastic methods that are
“crazy”—crystal ball gazing, for instance. The success of the argument requires that
the method be self-certifying and reliable. Crystal ball gazing fails (we think) at least
the second condition. (It will pass the first only if the crystal ball certifies itself. And
this may fail to obtain.)
We are not yet home and dry. Competing rules can be invoked in their own
defense (Salmon 1957: 46). Consider the anti-inductive rule, AI, which allows us
to infer from “Most observed A’s have been B’s” that the next A will not be a B.
We can construct an argument with true premises utilizing AI to show that AI will
be successful:
Most applications of AI have not been successful
_________________________ AI
The next use of AI will be successful

The answer is that self-certification, “immodesty” (Lewis 1971), isn’t sufficient for
the justification of an inference rule. In citing it, I simply want to show that invoking
an inference rule to show its reliability isn’t trivial. And the fact that another
(“crazy”) inference rule is also self-certifying doesn’t put it on a par with induction.
To be sure, its proponents might also attempt to use it to infer its reliability. But
whereas our belief in induction’s reliability is R-justified, their belief in the reliability
of their rule isn’t! I am assuming that induction (unlike anti-induction, for instance)
is reliable. If it isn’t, then our (circular) way of establishing its reliability won’t work:
its conclusion won’t be R-justified. But the point still stands. A circular vindication
isn’t trivial, and cannot be invoked equally successfully by rivals that are in serious
opposition.11
Here are the two most telling objections that have been leveled against reliabilism.
The first is the extensional inadequacy of the analysis. The demon’s victim shows,
Cohen (1985) claims, that reliabilism is too stringent. And several examples have
been adduced to show that reliabilism is too lenient. The most familiar one is
BonJour’s clairvoyant agent, who thinks she has no clairvoyant powers and seems
intuitively unjustified in her (reliably acquired) beliefs (1985: 43). But my favorite
is Vogel’s Roxanne (2000: 313–314). Her car has a reliable gas gauge, and she uses
it to form beliefs about how much gas there is in the tank of her car. She also forms
beliefs about the gauge’s readings by looking at it. Both beliefs are reliably acquired:
the gauge and her perception are (sufficiently) reliable. By conjoining these two types
of belief, Roxanne forms several beliefs of the form: The gauge reads “F” (“E”) and
F (E). She now reasons inductively from these conjunctions and concludes that the
gauge is reliable. Since induction is reliable, her belief is justified by reliabilist lights
but is intuitively (resoundingly) unjustified.12
Perhaps reliabilists can explain away these putative counterexamples. But from
our functional perspective, we can ignore them even if we find them convincing. We
are not trying to capture our concept of justification, and are allowing for a revision
of our concept(s).13
But there is a difficulty we cannot shrug aside, the revisionary nature of our
proposal notwithstanding: the “generality problem,” described by Heller (1995:
501) as “the most formidable challenge to . . . [the] reliability theory.” Reliabilists
define the (degree of) justification of a belief as the (degree of) reliability of the
process via which it was acquired. Thus, a belief is justified to the extent that it is
engendered by a doxastic process the relative frequency of truths among whose
products is (sufficiently) high.
The problem is that a particular belief is a token of indefinitely many process
types, with different attendant frequencies of truth. My belief that the sun will
rise tomorrow is inductively formed, formed by a woman, formed on Tuesday, etc.
And the reliabilist needs a principled way of picking out a unique description if his
notion of justification is to be well defined.
The revisionary reliabilist can bypass the generality problem, at least when he
is contending with Hume’s skeptical argument against induction. The argument
pertains to the use of induction, which is an acquisition method, and not a belief
that falls under (indefinitely) many descriptions. (The same is true for the reliabilist
who is attempting to contend with skeptical arguments that purport to impugn
other acquisition methods: perception, deduction, inference to the best explanation,
and memory). But the generality problem will resurface when he tries to vindicate
particular beliefs derived by induction—that the sun will rise tomorrow, for instance.
For this, it is not enough to vindicate the inductive inference rule. The belief will
have other descriptions, and the reliabilist will have to eliminate them if he is to
invoke the R-justification of induction. He will also have to assign the belief in the
premise (that the sun has hitherto always risen) an acquisition method (perception
or testimony) so as to show that it is R-justified. Without this, the conclusion of the
inductive inference will only be conditionally R-justified—that is, justified if the
premise is justified.
5 TAKING STOCK
We can now attempt to assess our epistemic situation from a reliabilist perspective,
remembering that the purpose of the (possibly revisionary) exercise is to see how
well we would fare relative to an alternative, non-internalist, justificatory standard.
We are able to fend off Hume’s skeptical challenge. But this—by itself—is not
enough to allay our worries: skepticism is true relative to another standard of
justification, one that does include the internalist requirement. So should we be
concerned?
From the functional perspective, the question is what we lose by relinquishing
the internalist requirement for justification and why R-justification matters (by way
678 SKEPTICISM
of compensation). The internalist requirement may have more than one rationale,
but here is one suggestion. To decide whether a claim is true (and the belief in its
truth is justified), one must test it by reference to a criterion of truth. I will now
explain and defend the suggestion.
The criterial conception of justification is as old as epistemology. According to
Sextus Empiricus, “Those who claim for themselves to judge the truth are bound
to possess a criterion of truth . . . For no matter of dispute is to be trusted without
judging” (AM VII 340; I follow Bury’s (1935) translation).
Descartes, too, searches for a criterion of truth: “Everything which I clearly and
distinctly perceive is of necessity true. . . . I was ignorant of this rule for establishing
the truth” (CSM II: 48–49, my italics). Descartes’s conception of a criterion differs
from Sextus’s in two respects. His criterion is infallible: a belief that satisfies it is
true. And it is universal, applicable to all propositions, whereas Sextus allows for
two propositions being adjudicated by reference to two different criteria.
It is fairly easy to see that the criterial conception of justification brings in its
wake the internalist requirement. In applying a criterion, one must believe that it is
reliable, sufficiently well correlated with that which it serves to identify. Thus, the
physician who uses red spots as a criterion for measles believes that having red spots
is correlated with measles. So if acquisition by induction is to function as a criterion
of truth, the property inductively acquired must be believed to be (sufficiently well)
correlated with truth; induction, that is, must be believed to be reliable. And since
this belief is a premise in an argument supporting the belief (say) that the sun will
rise tomorrow, it must be justified. We have now got the internalist requirement.
(There is an additional requirement the criterial conception engenders, which is not
pertinent to our concern. In properly applying a criterion to decide that X is, indeed,
the case, one must believe that X satisfies the criterion.)
For Descartes’s criterion, clarity and distinctness (however understood), we
get an analogous requirement: to be justified in a belief, the agent must justifiably
believe that clarity and distinctness are reliable indicators of truth. (He must also
believe that the belief is clear and distinct.)14 Appropriate requirements go with
other (putative) truth criteria.
The criterial conception of justification renders it unattainable. Hume’s argument
shows this for acquisition by induction as a criterion, and Sextus’s criterion argument
shows it more generally:
Those who claim for themselves to judge the truth are bound to possess a criterion
of truth. This criterion, then, either is without a judge’s approval or has been
approved. But if it is without approval, whence comes it that it is trustworthy?
For no matter of dispute is to be trusted without judging. And if it has been
approved, that which approves it, in turn, either has been approved or has not
been approved, and so on ad infinitum. (AM VII 340)15
How serious is our predicament? Sosa (1994) thinks that the very fact that a desire
turns out to be necessarily unattainable means that we should give it up. If an
objective turns out “to be unfulfillable,” he asks rhetorically, “Is it reasonable . . . to
insist that somehow the objective is still worthy?” Well, it may well be. Of course,
in such a case, one would be advised “to put [one’s] time to better use” (Sosa 1994:
287), stop treating it as an “objective” in the practical sense. It is, of course, silly, to
look for a truth-criterion. But that is different from deeming it worthless. It seems to
me that there is nothing incoherent (as opposed to strange) about wishing the circle
could be squared, even if we know it is mathematically impossible. What is silly is
to look for ways of doing it. The same is true for the unsatisfiable desire to have a
truth-criterion. And it isn’t even strange.
If we cannot dismiss the criterial conception of justification because it renders
justification unattainable, we must (more messily) consider whether Sextus Empiricus is
right in thinking that without a criterion we are like “people . . . searching for gold in a
dark room containing many treasures. . . . None of them will be convinced that he has
lighted on the gold, even though, in fact, he has lighted upon it” (AM VII 52). Of course,
given the criterial conception of justification, the analogy is apt: we are never “justified.”
But from the functional perspective, this needn’t constitute a serious predicament even
if our concept of justification is criterial. And, conversely, it may be a predicament if our
concept isn’t criterial but should be. To assess our epistemological situation, we should
consider the point of non-criterial concepts of justification, reliabilism, for instance.
Our question, more precisely, is whether we should value a belief’s being R-justified.
A positive answer will complete the vindication of an acquiescent response to Hume’s
argument, enabling us to accept its conclusion with equanimity. Remember, the skeptical
challenge to the R-justification of induction fails on two counts: the argument adduced
in its support isn’t sound and its conclusion is false.
There are two (complementary) responses the reliabilist can give so as to meet this
challenge. The first invokes the fact that R-justification is truth-conducive, so a belief
that is R-justified is likely to be true. This is, of course, a good thing. Note, too, that
not every truth-conducive property of beliefs enables us to meet the (corresponding)
skeptical challenge. Consider the theological concept of justification, G-justified,
which applies to a belief just in case God holds it. This concept is perfectly truth-
conducive (assuming—for the sake of the example—that God exists), because God
is doxastically perfect: he believes all truths and no falsehoods. But our position vis-
à-vis the G-justification skeptic is inauspicious. To be sure, the skeptic cannot show
that God does not believe (say) that the sun will rise tomorrow. But neither can we
show that He does (so that our belief that it will rise is G-justified). The skeptical
argument against the R-justification of induction, by way of contrast, is unsound
and its conclusion is false.
We might be able to find yet another way in which R-justification is valuable
by considering the Meno problem (raised in Plato’s dialogue), which challenges us
to explain why knowledge is more valuable than (mere) true belief. The relevance
of the Meno problem to ours is easy to discern, the differences between the two
notwithstanding. The Meno problem pertains to the value of knowledge, and ours
to that of R-justification. But this only means that we will not be interested in
the full plethora of solutions to the Meno problem, only in those that pertain to
knowledge construed as true belief reliably acquired. And we can also take in our
stride the fact that an answer to the Meno problem might invoke a feature of
knowledge that—unlike R-justification—presupposes truth (while going beyond
680 SKEPTICISM
it). We will only consider solutions to the Meno problem that locate the added
value of knowledge in reliable acquisition.
Zagzebski claims (2003) that reliable acquisition has only instrumental value.
Forming beliefs in a reliable manner is valuable because they are more likely to be
true. So a true belief isn’t more valuable than a true belief reliably acquired, just as
a good cup of coffee made by a reliable coffee machine isn’t more valuable than an
equally good cup of coffee that is produced by an unreliable coffee machine.
Brogaard (2006) demurs. The coffee-machine analogy isn’t apt, she claims,
because it is based on a flawed conception of the good, according to which two
things that are identical in their intrinsic properties have the same intrinsic value.
Although this is true in the case of coffee, she continues, it isn’t true in general,
because the distinction between intrinsic and instrumental value isn’t exhaustive. As
Rabinowicz and Roennow-Rassmussen (1999) point out, value may derive from an
object’s being extrinsically related to something else that we value. For instance, we
value Princess Diana’s dress more than an exact copy because it belonged to Diana.
In a similar fashion, we value one true belief over another because it was produced
by a reliable cognitive trait, which is independently valuable.
I will not attempt to adjudicate the Meno question (with the reliabilist slant we
have adopted on it). I use it so as to provide another illustration of the way the
“functional turn” transforms the debate with the skeptic: the “value question” now
becomes part and parcel of our engagement with the skeptic. It is sometimes claimed
that an analysis of knowledge that cannot explain its distinctive value is thereby
shown to be erroneous (Zagzebski 1999; Sosa 2000). This may be doubted, but even
if it is true, there is—from a functional perspective—a much more important lesson
to be learned from such a failure. It shows that we should be unperturbed by failing
to satisfy the concept the analysis determines, and that skeptical arguments targeting
it can be shrugged aside. An acquiescent strategy is here appropriate.
6 CONCLUSION
I do not have a verdict to offer with respect to the success of the functional response
to Hume’s skeptical argument. I considered only one rationale for a premise of the
argument and only one concept that doesn’t satisfy it (and renders the argument
unsound). But I have pointed out the way the discussion should be conducted so as
to yield a verdict.
ACKNOWLEDGMENT
I am grateful to the editors for their painstaking and perceptive comments. This
research was supported by the ISRAEL Science Foundation (grant No. 366/14).
NOTES
1. Craig (1987) goes even further. He suggests that the concept of knowledge cannot be
given a traditional analysis, because it doesn’t have (informative) conditions that are
(individually) necessary and (jointly) sufficient for its application.
2. I use the plural, because it is claimed (Swinburne 2001; Alston 2005) that there
are (in English) several concepts of justification. Similarly, there may well be more
than one concept of knowledge.
3. We may be misconstruing Hume. Perhaps he isn’t claiming that induction isn’t
justified, only arguing against a rationalist conception of justification (Garrett
1997). But the purpose of my exploration isn’t interpretive. My concern is with an
argument that may be suggested by Hume’s text, even if it isn’t his.
4. I owe this point to Baron Reed.
5. I add yet another way of using the term “internalism.” It is hitherto used in
(at least) five different ways. There is, first, the doctrine that requires for the
justification of a belief (inference rule) a second-order justified belief about the
factors that make first-order justification possible (Van Cleve 2003). This usage
is closest to mine. The difference is that unlike mine, Van Cleve’s doesn’t require
the second-order belief to be justified before (independently of) the first-order
belief in order for it to be legitimately invoked in its justification. According to
the second variety of internalism, epistemic status supervenes on facts about the
subject’s mental states. According to the third, the subject must be able to tell
by reflection that his belief (inference rule) is justified. According to the fourth,
the subject constitutes the conditions of justification: a belief (inference rule) is
justified just in case the subject thinks it is. The last two doctrines Schmitt (1992)
aptly labels “accessibility internalism” and “perspectival internalism,” respectively.
The fifth sense is found in Plantinga (1990) and Sosa (1994). They use the term to
denote the doctrine that a belief can be justified and amount to knowledge only
through the backing of reasons or arguments. On this usage, internalism stands in
opposition to foundationalism.
6. Hume belabors this point, but to us the claim seems truistic.
7. The difference is that in Hume’s original argument, there is premise circularity:
the principle of the uniformity of nature is assumed in the proof of itself. The
modified argument, by way of contrast, purports to discern in our practice epistemic
circularity: a faculty (belief-acquisition method) is used to prove itself reliable. I am
grateful to Baron Reed for pressing me to clarify this.
8. We do not here have to decide whether the preponderance of truths is to be found
in a method’s actual outputs, potential outputs in all possible worlds, or (Alston
2005) potential outputs in (sufficiently) similar worlds.
9. This is not true in general. Hume’s argument doesn’t conclude that induction isn’t
reliable. But the “bad lot” argument against inference to the best explanation (van
Fraassen 1989: 142–150) purports to show that it is an unreliable inference rule.
It takes us to the best explanation among those we have considered. And even if
the best explanation is likely to be true, we have no reason to suppose that the
best explanation is to be found among the theories we have considered. Indeed,
van Fraassen argues (1989: 146), it is very likely that among the explanations we
haven’t considered there are many that are as good as (or even better than) the
explanation that is the best among those we have considered. Stanford (2006)
adduces historical evidence to support the claim that scientists choose from
theories to which there are unconsidered rivals that explain equally well or even
better. For instance, the evidence available to proponents of Aristotelian science
was equally well (or even better) explained by Newtonian science (which they
didn’t consider).
682 SKEPTICISM
10. Black (1958) points out that a circular justification of a rule involves no formal
circularity. But this difference between the circular justification of an inference rule
and that of a proposition doesn’t seem significant.
11. A rival rule to induction that diverges from it only very occasionally can be invoked
to show its own reliability. But the two rules are roughly equally reliable, and the
conclusions of the two (circular) arguments have the same truth-value. So this failure
of self-certification to be unique isn’t damaging. To be sure (as Baron Reed pointed
out), the two rules may diverge with respect to some important claim, and for every
claim, there will be two very (but imperfectly) reliable rules that diverge over it. But
this is beside the present point—that reliabilist self-certification is genuine: it works
only for reliable methods.
12. Conjunctions of reliabilistically acquired beliefs are typically not as likely to be
true as the conjuncts. Two methods may count as reliable by reference to some
threshold, yet the reliability rate of the combined method might fall below it. But
we may suppose that overall, conjunction is sufficiently reliable.
13. Is Quine (1973) a revisionary reliabilist? There are three reasons for resisting
the suggestion. First, it is not clear that he repudiates the old (validatory) project
(answering the old skeptic): “Our liberated epistemologist ends up as an empirical
psychologist, scientifically investigating man’s acquisition of science. . . . Yet it is
no gratuitous change of subject matter, but an enlightened persistence rather in the
original epistemological problem. It is enlightened in recognizing that the skeptical
challenge springs from science itself, and that in coping with it we are free to use
scientific knowledge. The old epistemologist failed to recognize the strength of his
position” (1973: 3). Since the skeptic is invoking science against us, we can answer
him using science. Second, to the extent that Quine’s project (explaining knowledge)
is revisionary, it recommends that epistemology jettison normative questions
altogether, replacing them with empirical ones. Revisionary reliabilism, by way of
contrast, is still normative: it replaces our concept of justification with another.
Finally, Quine’s aim is to explain “how it is that man works up his command of that
science from the limited impingements that are available to his sensory surfaces”
(1973: 3), to explain the acquisition of knowledge. To say that a belief is reliably
acquired isn’t terribly explanatory.
14. According to Van Cleve (1979), Descartes doesn’t think that the agent must
believe a belief is clear and distinct if it is to be justified. It is, rather, the
epistemologist who identifies clarity and distinctness as features that render a belief
(externalistically) justified.
15. If Descartes doesn’t subscribe to the full-blooded criterial conception of justification
(see n. 11), he isn’t vulnerable to Sextus’s argument.
REFERENCES
Alston, William. 1993. The Reliability of Sense Perception. Ithaca: Cornell University Press.
Alston, William. 2005. Beyond “Justification”: Dimensions of Epistemic Evaluation. Ithaca:
Cornell University Press.
Ayer, A. J. 1972. Probability and Evidence. London: Macmillan.
Bergmann, Michael. 2000. “Externalism and Skepticism,” Philosophical Review 109:

159–194.
Black, Max. 1958. “Self-Supporting Inductive Arguments.” In R. Swinburne (ed.), The
Justification of Induction, 127–134. Oxford: Oxford University Press.
University Press.
Braithwaite, R. B. 1953. Scientific Explanation. Cambridge: Cambridge University Press.
Brandom, Robert. 1998. “Insights and Blindspots of Reliabilism,” The Monist 81: 371–392.
Broad, C.D. 1952. “The Philosophy of Francis Bacon.” In C.D. Broad, Ethics and the
History of Philosophy, 117–143. London: Routledge and Kegan Paul.
Brogaard, Berit. 2006. “Can Virtue Reliabilism Explain the Value of Knowledge?”
Canadian Journal of Philosophy 36: 335–354.
Chisholm, Roderick. 1982. Foundations of Knowledge. Minneapolis: University of
Minnesota Press.
Cohen, Stewart. 1985. “Reliability and Justification,” The Monist 68: 149–158.
Craig, Edward. 1987. “The Practical Explication of Knowledge,” Proceedings of the
Aristotelian Society 87: 211–226.
Descartes, René. 1628. Philosophical Works of Descartes, vol. 2, translated by John
Cottingham, Robert Stoothoff, and Dugald Murdoch. Cambridge: Cambridge
University Press, 1985.
Edwards, Paul. 1949. “Russell’s Doubts about Induction,” Mind 68: 141–163.
Fumerton, Richard. 1995. Metaepistemology and Skepticism. Lanham, MD: Rowman &
Littlefield.
Garrett, Don. 1997. Cognition and Commitment in Hume’s Philosophy. Oxford: Oxford
University Press.
Goldman, Alvin. 1979. “What is Justified Belief?” In G. Pappas (ed.), Justification and
Knowledge: New Studies in Epistemology, 1–23. Dordrecht: Reidel.
Greco, John. 2000. Putting Skeptics in Their Place. Cambridge: Cambridge University Press.
Heller, Mark. 1995. “The Simple Solution to the Problem of Generality,” Noûs 29: 501–
515.
Howson, Colin. 2000. Hume’s Problem. Oxford: Clarendon Press.
Hume, David. 1739. A Treatise of Human Nature, 2nd edn., edited by L. A. Selby-Bigge.
Oxford: Clarendon Press, 1975.
Hume, David. 1777. Enquiries Concerning Human Understanding, 3rd edn., edited by
L. A. Selby-Bigge. Oxford: Clarendon Press, 1978.
Kaplan, Mark. 2008. “Austin’s Way with Skepticism.” In J. Greco (ed.), The Oxford
Handbook of Skepticism, 348–371. Oxford: Oxford University Press.
Lewis, David. 1971. “Immodest Inductive Methods,” Philosophy of Science 38: 54–63.
Lycan, William. 1988. Judgement and Justification. Cambridge: Cambridge University
Press.
Moore, G.E. 1953. “Hume’s Theory Examined.” In G.E. Moore, Some Main Problems of
Philosophy, 108–126. London: MacMillan.
Papineau, David. 1992. “Reliabilism, Induction and Scepticism,” The Philosophical
684 SKEPTICISM
Plantinga, Alvin. 1990. “Justification in the 20th Century,” Philosophy and Phenomenological
Research, supplement vol. 1: 45–71.
Popper, Karl. 1972. The Logic of Scientific Discovery, 4th edn. London: Hutchinson.
Quine, W. V. O. 1973. The Roots of Reference. La Salle, Illinois: Open Court.
Rabinowicz, Wlodek and Toni Rønnow-Rasmussen. 1999. “A Distinction in Value:
Intrinsic and for its Own Sake,” Proceedings of the Aristotelian Society 100: 33–49.
Reid, Thomas. 1764. An Inquiry into the Human Mind. In W. Hamilton (ed.), The Works
of Thomas Reid, 5th ed., vol. I, 95–211. Edinburgh: McClachlan and Stewart, 1846.
Russell, Bertrand. 1948. Human Knowledge: Its Scope and Limits. London: George Allen
and Unwin.
Salmon, Wesley. 1957. “Should We Attempt to Justify Induction?” Philosophical Studies
8: 33–48.
Sartwell, Crispin. 1992. “Why Knowledge Is Merely True Belief,” Journal of Philosophy
89: 167–180.
Schmitt, Frederick. 1992. Knowledge and Belief. London: Routledge.
Sextus Empiricus. 1933. Outlines of Pyrrhonism, translated by R. G. Bury, The Loeb
Classical Library, London: W. Heinemann.
Sextus Empiricus. 1935. Against the Logicians, translated by R. G. Bury, The Loeb Classical
Library, London: W. Heinemann.
Sosa, Ernest. 1994. “Philosophical Skepticism,” Proceedings of the Aristotelian Society,
Supp. 68: 263–290.
Stanford, P. Kyle. 2006. Exceeding our Grasp: Science, History and the Problem of
Unconceived Alternatives. Oxford: Oxford University Press.
Swinburne, Richard. 2001. Epistemic Justification. Oxford: Oxford University Press.
The Philosophical Review 88: 55–91.
Van Cleve, James. 1984. “Reliability, Justification and Induction,” Midwest Studies in
Van Cleve, James. 2003. “Is Knowledge Easy—or Impossible? Externalism as the Only
Alternative to Skepticism.” In S. Luper (ed.), The Skeptics, 45–60. Hampshire: Ashgate.
van Fraassen, Bas. 1989. Laws and Symmetry. Oxford: Oxford University Press.
Vogel, Jonathan. 2000. “Reliabilism Leveled,” Journal of Philosophy 97: 602–623.
Weiner, Matthew. 2005. “Why Does Justification Matter?” Pacific Philosophical Quarterly
86: 422–444.
Zagzebski, Linda. 1999. “What Is Knowledge?” In J. Greco and E. Sosa (eds.), Epistemology,
92–116. Oxford: Blackwell.
Zagzebski, Linda. 2003. “The Search for the Source of the Epistemic Good,” Metaphilosophy
34: 12–28.
CHAPTER FORTY-SEVEN
Skepticism about
A Priori Knowledge
OTÁVIO BUENO
1 INTRODUCTION
A priori knowledge is, roughly, knowledge that does not depend on experience
(except perhaps for whatever experience that may be needed to acquire the relevant
concepts). This is a rough formulation since, when analyzing the notion of a priori
knowledge, some care is needed to characterize the independence-from-experience
condition so that what is taken to be known independently of experience (such as
mathematical and logical results) turns out to be known a priori (according to the
proposed analysis of a priori knowledge), and what is taken to be known based on
experience, and thus is dependent on experience (such as particular physical traits of
the world), is not known a priori (according to the relevant account). The challenge
is then to secure that the proposed analysis is extensionally adequate.
Two questions then emerge: (a) Can we have a priori knowledge? (b) Can we know
that—or, at least, determine whether—we have such knowledge? Corresponding
to these two questions, there are two forms of skepticism regarding the a priori:
(aʹ) One can provide arguments to the effect that, despite appearances to the
contrary, one doesn’t have knowledge of the a priori, such as mathematical claims
and claims about the validity of arguments. This is a priori skepticism (see Beebe
2011). Alternatively, (bʹ) one can provide arguments to the effect that conceptual
analyses (characterizations, definitions) of a priori knowledge (see, for instance,
Kitcher 1980, 2000; and, for an account of a priori justification, Casullo 2003)
are ultimately problematic. This is skepticism about a priori knowledge. And if it
is unclear what a priori knowledge ultimately is, it is similarly unclear how we can
determine whether we have any such knowledge.
In this chapter, I will consider both forms of skepticism regarding the a priori,
focusing in particular on skepticism about a priori knowledge. After all, by
challenging the very concept of a priori knowledge, it seems to me that this form
of skepticism is more basic. Given the centrality of a priori knowledge to so many
cognitive endeavors (from mathematics and logic to philosophy), the presence of
686 SKEPTICISM
these forms of skepticism highlights a central issue that needs to be addressed by

those who claim to have the corresponding knowledge.
2 A PRIORI SKEPTICISM
Most people think they know several mathematical claims, including simple
arithmetical results, such as that 71 is the sum of 13 and 58. Most people think they
also know whether a given conclusion follows logically from certain premises, such
as that q follows logically from p and if p then q. Since such mathematical and logical
results are arguably instances of what is known a priori, most people think they have
a priori knowledge (although, unless they have some philosophical training, they are
unlikely to describe that knowledge in these terms).
A priori skepticism challenges such knowledge claims, and any others that are
based on a priori considerations (Beebe 2011). The form of the argument for it is
the same as the familiar brain-in-a-vat or evil-demon arguments, in which, first, a
necessary condition for knowing something about the world is that one knows that
one’s experiences are not being generated in a brain-in-a-vat scenario or by an evil
demon. Support for this premise is provided by the closure principle, according
to which (roughly) one knows what follows logically from things that are already
known and that are established on the basis of that knowledge. Second, it turns out
that one is unable to know that one is not in such skeptical scenarios. After all, one’s
perceptual experiences are insensitive to such scenarios: one would believe that one is
experiencing the world even if one weren’t (as the case would be if one were in such
skeptical situations). The result is that one doesn’t know anything about the world.
The argument for a priori skepticism is entirely analogous. In fact, it is an
application of the familiar evil-demon argument to a priori knowledge claims, such
as knowing the validity of modus ponens. As James Beebe notes:
(P1) If I know that modus ponens is correct, then I know that my belief that
modus ponens is correct is not based on faux intuitive experiences induced
in me by a bumbling evil demon.
(P2) I don’t know that my belief that modus ponens is correct is not based on
faux intuitive experiences induced in me by a bumbling evil demon.
(C) Therefore, I don’t know that modus ponens is correct (Beebe 2011: 590;
the numbering of the premises and the conclusion of this argument has been
changed; of course, nothing hangs on this).
Interestingly, the argument above does not depend on establishing the invalidity
of modus ponens. Those who take this form of inference to be valid would insist
on the impossibility of demonstrating the inference’s invalidity: on their view, any
such attempt would inevitably lead one into a contradiction. Rather, the argument
is meant to challenge one’s knowledge that modus ponens is valid (or correct) on
the grounds that one’s belief in the validity of this inference could be based on a
bumbling evil demon. Whether modus ponens is ultimately valid or not, one cannot
know that it is valid.
SKEPTICISM ABOUT A PRIORI KNOWLEDGE 687
There is also a different route to a priori skepticism (Beebe 2011: 593–594). It

emerges from the consideration of alternative logical and mathematical practices,
in which established mathematical and logical truths may not hold. This route
surfaces from Ludwig Wittgenstein’s work (see Wittgenstein 1956, as understood
by Stroud 1965; for a discussion, see Beebe 2011: 593–594). From the perspective
of those who adopt the traditional view regarding logical and mathematical
truths—according to which these truths are metaphysically necessary—one may
not understand under what conditions mathematical and logical truths would be
false. However, one’s inferential and calculation practices, which shape the relevant
logical and mathematical results, could differ significantly from the traditional
ones. This would generate genuine alternatives to these practices, so that despite
the agreement that modus ponens is valid, one would not agree that from p and if
p then q, it would follow that q; one could also follow the rule “adding 2” in such a
way that after reaching 1,000, the rule continues as 1,004, 1,008, etc.
It is ultimately a contingent fact that we developed the mathematical and logical
concepts we currently have. Had the history of our species been different, we could
have developed correspondingly different concepts. As Wittgenstein points out:
I am not saying: if such-and-such facts of nature were different people would
have different concepts (in the sense of a hypothesis). But: if anyone believes
that certain concepts are absolutely the correct ones, and that having different
ones would mean not realizing something that we realize—then let him imagine
certain very general facts of nature to be different from what we are used to, and
the formation of concepts different from the usual ones will become intelligible
to him. (Wittgenstein 1956: I 141; quoted in Beebe 2011: 594)
It is possible then, in light of the contingency of the way in which mathematical
and logical concepts have been developed, that these concepts could have evolved
differently, thus resulting in different logical and mathematical practices (see
also Stroud 1965: 513, and Beebe 2011: 594). As a result, one doesn’t know the
correctness of one’s current mathematical and logical practices.
It seems to me that a priori skepticism provides a significant challenge to those
who claim to know mathematical and logical results. But the challenge depends
on the version of the challenge that is advanced. The second, contingency-based,
formulation of this skeptical challenge seems to me more effective. The concern
about the first, brain-in-a-vat or evil-genius-based formulation, is that it invites a
particular sort of Moorean response. One could easily complain that we have more
reason to believe in the truth of the relevant mathematical and logical results than
we have to believe in the truth of the second premise, (P2), of the argument for a
priori skepticism, according to which we don’t know that our relevant logical or
mathematical beliefs are not based on experiences induced by an evil demon (or
that emerge from a brain-in-a-vat scenario). Of course, as a principled response to
skepticism, this move blatantly begs the question, since it assumes that we know
precisely those items that the skeptical argument challenges. From this point of view,
the response is clearly ineffective. But one can think of it as expressing a particular
attitude: one that dismisses the entire skeptical argument as being so general, so
688 SKEPTICISM
detached from the particular mathematical and logical issues under consideration,
that it verges on being irrelevant. The response reinforces the point that, due to its
generality, there is very little in the first argument for a priori skepticism that engages
with the content of logical or mathematical beliefs per se. The fact that the argument
is a particular instance of the usual brain-in-a-vat or evil-demon argument clearly
illustrates this point. In the end, the Moorean response is perhaps better thought of
as an expression of the difficulty of taking such an argument very seriously.
In contrast, the contingency-based formulation of a priori skepticism is more
relevant. It engages with the content of the relevant mathematical and logical
concepts. It is the fact that these concepts could have been developed differently
that allows for the possibility that alternative logical and mathematical results could
have been entertained.
However, despite the advantage of this formulation of a priori skepticism, it
can be made stronger. There is no need to leave the argument, as Wittgenstein,
Stroud, and Beebe have left it, as a mere possibility, namely, that alternative
mathematical and logical practices, shaped by different concepts, could have
emerged. These possibilities are, in fact, actual. Challenges have been raised to the
validity of modus ponens (McGee 1985) as well as to the correctness of classical
mathematics, by intuitionists (Dummett 2000), predicativists (Weyl 1918/1994),
and paraconsistentists (Mortensen 1995; for a discussion, see da Costa, Krause,
and Bueno 2007). Even classical arithmetical results have been challenged within
these circles (Priest 1997, 2000). The alternative logical and mathematical practices,
posited as mere possibilities in the formulation of a priori skepticism above, are, in
fact, not only possible, but actual. This means that one cannot claim to know, so easily,
traditional mathematical and logical results, since they have all been challenged.
Of course, the challenge is more focused and targeted at particular logical and
mathematical results. In contrast to the first argument for a priori skepticism, this is
a local rather than a global skeptical challenge. It is far more relevant, specific, and
cannot be so easily dismissed.
As an illustration, consider the challenge one can pose to modus ponens using the
2016 US presidential election as a context for the counterexample (this is adapted
and updated from McGee 1985):
(P1) If a Republican wins the election, then it if is not Donald Trump who wins,
it will be Ted Cruz.
(P2) A Republican will win the election.
(C) If it is not Donald Trump who wins, it will be Ted Cruz.
The premises of the argument are clearly true: Ted Cruz is the closest Republican
contender to Donald Trump, and a Republican will win (and, in fact, did win) the
election. However, the conclusion is clearly false: if it is not Donald Trump who
wins the election, it will be Hillary Clinton, rather than Ted Cruz, who does it. (In
fact, Clinton eventually won the popular vote by about 3 million votes!)
This local, substantive form of a priori skepticism is clearly something to be
reckoned with.
3 EMPIRICISM AND A PRIORI KNOWLEDGE

Before considering skepticism about a priori knowledge, it’s important to dispel a
concern about a priori knowledge that has often been raised. Traditionally, from
Sextus Empiricus through David Hume, skepticism and empiricism have often been
closely connected. And a priori knowledge (and a priori justification) have often been
a source of puzzlement particularly for those who are sympathetic to empiricism.
How is it possible to have such knowledge (and justification) independently of
empirical information? Those who are unmoved by these concerns insist that the
content of the relevant statements, since their truth does not depend on empirical
factors, is non-empirical, too, and therefore considerations to support the statements
in question need not, and in fact should not, invoke empirical reasons, that is,
reasons based on empirical information.
Note that empiricists, just for being empiricists, need not be worried about a
priori knowledge (or justification). Suppose empiricism is understood as involving
the requirement that all information about the world is based on experience. There
are at least two ways in which empiricism is not in tension with the a priori (a priori
knowledge and justification): (a) when experience is understood in a suitably broad
sense or (b) when the offending statements, which are known or justified a priori,
are not about the world.
With regard to (a), consider that among the experiences in question there are
intellectual seemings (or intellectual intuitions; see Chudnoff 2013). In this case, one
could have suitable experiences about objects of a non-empirical sort. Admittedly,
it would require a correspondingly broader empiricism to accommodate such a
broader notion of experience, and not every sort of empiricist will be willing to
embrace this proposal.
With regard to (b), one could argue that (pure) mathematical statements (those
that are only about numbers, functions, sets, categories, and other mathematical
objects and structures) are not about the world, since none of the objects invoked
in such statements are empirical. If the truth of mathematical statements does not
depend on what goes on in the empirical world, and if such statements are not about
the empirical world either, there is nothing in these statements that can contravene
empiricism—understood in the more restricted sense as a conception that is only
concerned with information about the empirical world. This move clearly limits the
scope of empiricism, which is then ultimately restricted to empirical objects alone.
But that makes a priori knowledge (and justification) safe for empiricists.
At this point, one can be concerned about (allegedly) synthetic a priori statements,
that is, statements about the world whose knowledge and/or justification do not
depend on empirical information about it (the world). Understandably, such
statements raise concerns for empiricists. After all, if the statements in question are
about the world, how can knowledge of their truth be independent of what goes
on in it? Consider a statement such as “The standard meter is one meter in length.”
This is clearly a statement about the world, but, the argument goes, it is known
independently of experience, given that the standard meter specifies what one meter
is. In response, it can be pointed out that knowledge of this statement ultimately
690 SKEPTICISM
depends on traits of the world, namely, the particular length of the standard meter
(whatever it turns out to be). So, it is unclear that this statement is a priori in the
end: it is known on the basis of knowledge of the length of the standard meter (an
empirical fact about the world). Not surprisingly, empiricists have traditionally been
skeptical about the existence of synthetic a priori statements in the first place.
It is, of course, an issue whether mathematical statements are synthetic a priori.
Kant famously argued that they are. But empiricists have very little reason to
grant that, and may have concerns about the way Kant conceptualized the notions
of analytic and synthetic. In fact, whether the subject is contained or not in the
predicate crucially depends on how subjects and predicates are understood (for
instance, are they formulated intensionally or extensionally?), and what the proper
notion of containment is (e.g., should it be understood set-theoretically or in some
other way?). However these notions are spelled out, arguably (pure) mathematical
statements are not about the (empirical) world at all, given that they concern abstract
mathematical objects, which are neither spatiotemporally located nor causally active.
That every metric space is a topological space (but not vice versa) is a fact about the
relevant spaces (which are abstract if they exist at all) and not about any particular
(partial) spatial instantiation of such spaces. Any such concrete (partial) instantiation
depends on a suitable physical interpretation of the abstract notions of metric and
topological spaces. Depending on how such interpretations are worked out, different
instantiations end up being provided. Just consider the different metrics (metric
functions) that characterize different metric spaces. The way in which such metrics
are physically interpreted yields different instantiations of the corresponding spaces.
The same metric space is compatible with a variety of distinct physical instances.
These considerations illustrate that (pure) mathematical statements do not
depend on concrete objects for their truth or falsity, and thus neither knowledge
nor justification of mathematical objects depend on knowledge or justification of
empirical facts. The result is that, properly understood, a priori knowledge and a
priori justification need not pose particular difficulties for empiricists—assuming, of
course, that there are any.
We can now address the second kind of skepticism regarding the a priori: it
challenges the very characterization of a priori knowledge. We start by considering
an influential attempt at providing an analysis of this concept.
4 ANALYZING A PRIORI KNOWLEDGE

A priori knowledge, we noted, is taken to be knowledge independent of experience.1
Although the intuitive idea underlying the notion of a priori knowledge seems clear
enough, it turns out to be extremely difficult to characterize properly what a priori
knowledge is.
In what follows, I illustrate this by arguing that a classic, and prima facie plausible,
characterization of a priori knowledge—the one provided by Philip Kitcher (1980,
2000)—fails. Kitcher’s account has been criticized for being too strong (see, e.g.,
Harper 1986). As I argue here, however, the account turns out to be too weak, as
it’s ultimately unable to distinguish empirical and a priori knowledge. In the end,
the proposal seems unstable, since it seems to over-generate (turning clear cases of
empirical knowledge into a priori knowledge) or under-generate (failing to classify
clear cases of a priori knowledge as such). This raises a partial doubt, in light of the
failure of this account, regarding the prospects of characterizing the notion of a
priori knowledge more generally.
Kitcher’s account of a priori knowledge was developed as part of an investigation—
and, ultimately, a defense—of the claim that mathematics is not a priori; thus,
making it safe for empiricists.2 So, one of the challenges the account needs to meet
is not to have stacked the deck in the first place. Kitcher is, of course, aware of the
problem, and he explicitly tries to avoid this difficulty.
The proposed analysis is meant to capture the two crucial features of a priori
knowledge: (i) the fact that this type of knowledge doesn’t depend on experience,
while acknowledging that (ii) some of the concepts involved in a priori knowledge can
only be obtained from experience. To accommodate (ii), one introduces the notion
of experiences that are “sufficiently rich for an agent X to entertain the proposition
p.” That is, X has experiences that yield all the relevant concepts invoked in p. To
accommodate (i), the requirement of independence of experience, one introduces
the idea that the same type of processes that produce, in the actual world, the belief
that p in X will warrant that belief in all (epistemically) possible worlds. In this way,
the justification for an a priori belief does not depend on experience, given that
processes of the same type will warrant that belief in different possible worlds.
The definition is formulated in two steps. First, a priori knowledge is characterized
in terms of a priori warrant. Second, a priori warrant is then formulated in terms
of a process. Here is the account (which revisits and improves on an earlier account
provided in Kitcher 1980):
X knows a priori that p iff X knows that p and X’s knowledge that p was produced
by a process that is an a priori warrant for p.
α is an a priori warrant for X’s belief that p just in case α is a process such that for
any sequence of experiences sufficiently rich for X for p
(a) some process of the same type could produce in X a belief that p;
(b) if a process of the same type were to produce in X a belief that p, then it
would warrant X in believing that p;
(c) if a process of the same type were to produce in X a belief that p, then p.
(Kitcher 2000: 67)
As Kitcher notes, if only condition (a) is met, we obtain weak a priori knowledge.
If conditions (b) and (c) are satisfied as well, we obtain strong a priori knowledge.
Is this definition adequate? The responses have been unanimous in stressing that the
definition is too strong (see, e.g., Parsons 1986; Harper 1986, Summerfield, Casullo,
and Peacocke 1994). Talking about a particular a priori warrant, William Harper
insists that it might “even approach the outrageously high standards set out by Philip
Kitcher in his attempt to explicate a priori knowledge” (Harper 1986: 268 n. 8).3
As it turns out, however, I don’t think that Kitcher’s standards for apriority are
too high. If anything, as will become clear below, they are not high enough.
692 SKEPTICISM
5 SKEPTICISM ABOUT A PRIORI KNOWLEDGE:

SOME DIFFICULTIES
The main problem with this account of a priori knowledge is that, ultimately, it
cannot distinguish a priori and empirical knowledge, and so, it fails to deliver an
extensionally adequate characterization of the a priori. To see why this is the case,
let’s first highlight what might be taken to be, initially, a benefit of the view.
Recall that, according to this account, the notion of a priori warrant is defined in
terms of the notion of process. But the latter is left completely open in this proposal.
No constraints are imposed on what counts as a process. (The term seems to be
taken as a primitive.) It might be argued that the lack of constraints here is not
without reason. It’s sensible to expect that any account of a priori knowledge should
leave it open which processes yield this type of knowledge, whether they are proofs,
conceptual analyses, or perhaps some sort of introspection.4 So, the openness of the
notion of process might be taken as an advantage of the proposal.
Moreover, the notion of process had better be left open indeed. After all, it’s not
clear that a process adequate to define the notion of a priori warrant could be exactly
characterized without somehow presupposing the notion of the a priori: either the
processes in question are extensionally equivalent to a priori warrants or they are
not. If they aren’t, these processes clearly fail to characterize, in an extensionally
correct way, the notion of an a priori warrant. If, however, the processes under
consideration are extensionally equivalent to a priori warrants, then in virtue of
what does the extensional equivalence obtain? To circumscribe correctly the class
of a priori warrants, one needs to have already some sort of grasp of what counts as
an a priori warrant, but this means that, ultimately, the notion of a priori has to be
presupposed. To avoid this sort of circularity, it’s not by chance that the notion of
process is then left completely open.
But this openness comes with its price, and it’s in virtue of this openness that
the account is unable to distinguish a priori and empirical knowledge. Let’s start
with weak a priori knowledge. Consider the process of knowledge acquisition about
quarks provided by particle accelerators, and consider a community of physicists
X who, by using particle accelerators, have now formed the belief p that quarks
have been detected. Of course, the community of physicists, understanding what
quarks are, has had a series of experiences sufficiently rich for p. The trouble is that
particle accelerators yield a process of detection of quarks that immediately satisfies
condition (a) in the definition above. After all, for any sequence of experiences
sufficiently rich for the community of physicists X for the belief that p (i.e., that
quarks have been detected), there is always a process of the same type—such as, a
particle accelerator built in the other side of the planet—that could produce in X a
belief that p.5 In other words, according the account, particle accelerators provide
us with (weak) a priori knowledge of quarks!
Kitcher may grant this point, given that he himself acknowledges that the notion
of weak a priori knowledge is far from adequate. But more than a simple inadequacy,
this suggests that the notion of weak a priori knowledge is importantly problematic:
in the end, it does not seem to be a notion of a priori knowledge. If the latter cannot
be distinguished from empirical knowledge, there’s something deeply unsettling

with the proposed account.
But perhaps this is only a feature of weak a priori knowledge, and strong a priori
knowledge may fare better. Alas, the same problem also plagues the latter notion.
To see why this is so, let’s examine the particle accelerator case a little further.
Is condition (b) in the account above satisfied in this case? This seems to be so.
Let’s consider all epistemically possible worlds—in particular, worlds in which the
community of physicists X has experiences sufficiently rich for the belief in quarks
p, and in which it is known that quarks exist. We should now consider processes of
the same type as those that, in the actual world, produce in X the belief that p. But
what would count as a process of the same type in this case? Certainly, at least other
particle accelerators built in these epistemically accessible worlds. Clearly, any such
particle accelerators also warrant the belief that quarks have been detected, given
that it is a process of the same type as the one in the actual world. After all, if the
processes are of the same type, presumably the same propositions will be true in each
of them. (Compare: if two structures are isomorphic, the same propositions will
be true in each structure.6) Note that this doesn’t require infallibility: a warranted
belief can, of course, be false. But epistemic processes of the same type will validate
the same beliefs; that is, in part, what makes them processes of the same type.
In other words, epistemic processes are individuated, in part, by the beliefs they
validate. Those beliefs that are true in one process will be true in other processes of
the same type; those that are false in one process will be false in other processes of
that type as well. As a result, in each of these worlds, the belief p, being generated
by the same type of process, will be warranted. Condition (b) is met.
In response, it might be argued that cases of empirical knowledge fail to satisfy
condition (b). If it is merely nomologically (as opposed to metaphysically or logically)
necessary that the world has a certain structure, then there will be worlds in which
particle accelerators don’t give us a window into the structure of the world. Particle
accelerators in those worlds will be like witch detectors in the actual world. There
are no witches to be detected, and we are not justified in thinking that the output
of a witch detector tells us anything about the natural world. Ditto for particle
accelerators in worlds where the laws of nature are different from the laws of nature
in the actual world.
This response, however, faces two difficulties. First, if in some worlds particle
accelerators fail to warrant X in believing that quarks exist, that means that, in
these worlds, the accelerators in question end up not being processes of the same
type as those that produce the corresponding beliefs in the actual world. After
all, in the actual world, particle accelerators do warrant the physics community in
believing in the existence of quarks. And if the processes are indeed of the same
type, presumably they will warrant the same propositions. In other words, condition
(b) is met—by contraposition.
It may be argued that the condition to the effect that processes of the same type
warrant the same propositions yields too easy a response to skepticism. If perception
is a process that yields warranted beliefs in the actual world, but does not in an
evil-demon world, then the process relied on in the latter is not of the same kind as
694 SKEPTICISM
the one invoked in the former. After all, true beliefs are generated in one case, and
false ones in the other. Similarly, testimony arguably generates true beliefs, at least
in the typical cases. In contrast, a testimony that conveys false information, and
thus yields false beliefs, would be a different type of epistemic process, given the
distinct outcomes in the two cases (true vs. false testimonies). But this would make
skeptical challenges extremely easy to address, since the challenges can be set aside
as engaging with a different kind of epistemic process rather than with veridical,
truth-conducive procedures.7
In response, arguably one does not perceive in an evil-demon scenario, given
that the beliefs that are yielded are not produced, as in the case of perception,
by the objects they are about. Since this is a constitutive feature of perception,
the experiences one undergoes in an evil-demon scenario are of a different kind,
despite the many similarities they share with perception, particularly regarding
their phenomenology. (The same point goes through in the case of hallucinations.)
Similarly, a false testimony, as opposed to a true one, doesn’t typically yield true
beliefs. Despite the similarities between the testimonies of a liar and of a truth-
teller, there is a significant epistemic difference between them: in one case, truth
is preserved and transferred; in the other, it isn’t. It’s not unreasonable, thus, to
consider them as distinct epistemic processes.
Note, however, that this doesn’t provide a response to skepticism. After all, one
typically doesn’t know when a testimony is true or when it is false, nor does one
know whether the scenario one experiences is of an evil demon or not. And without
these pieces of knowledge in place, the skeptical challenges still stand.
Returning now to the case of worlds involving quarks, a second difficulty
emerges. If in some worlds quarks don’t exist (just as witches don’t exist in the
actual world), then, in these worlds, we had better not have knowledge of them.
After all, how could we have knowledge of an object that doesn’t exist? So, in these
worlds, we couldn’t have a priori knowledge of quarks, given that we simply don’t
have knowledge of them. (Recall that, on this account, for one to know a priori that
p, one needs first to know that p.) This seems to be good news for the account. The
trouble, however, is that in the worlds in which quarks do exist, processes of the
same type as those that produce the belief in quarks in the actual world will also
warrant X in believing in the existence of quarks. After all, it was the nonexistence of
quarks that undermined X’s warrant in p. In other words, in those worlds in which
quarks exist, the warrant is no longer undermined. But, in this case, condition (b)
will be, once again, satisfied.
How about condition (c): is it met? Recall that we are considering all epistemically
possible worlds in which the community of physicists X has experiences sufficiently
rich for the belief in quarks p. We are also considering processes of the same type
as those that produce in X the belief that p. Given these assumptions, for exactly
the same reason discussed above, in each of these worlds the proposition p (that
quarks have been detected) will also come out true. Hence, this satisfies condition
(c). After all, is it possible that p is not true? Well, note that the first requirement
of the account of a priori knowledge is that X knows that p. Typically, this entails
that p is true. So, p’s truth follows from the assumption that X has knowledge that
p. Moreover, given that the processes under consideration are of the same type (as
those that yield the belief that p in X), the same propositions will come out true in
each of these processes (and as noted above, this doesn’t assume infallibility). But p
is one of these propositions, and p is true in the actual world. Hence, condition (c)
is also met. The result is then a counterexample to the proposed account of strong
a priori knowledge. After all, it is typically assumed that knowledge of quarks is
empirical rather than a priori. However, for the reasons provided above, according
to the proposal under consideration, we have strong a priori knowledge of quarks!
In the end, it seems that even the strong version cannot distinguish empirical and a
priori knowledge.
6 SKEPTICISM ABOUT A PRIORI KNOWLEDGE:

EPISTEMIC COMMUNITIES, MODALITY,
AND WARRANT
6a Epistemic Communities on the Fringes
Let me consider some ways in which one could try to resist the counterexample
above. Adapting a point made in Kitcher (2000), one could suggest that, among the
epistemically possible worlds, there is one in which the community of physicists X
is actually on the fringes, and their methods of knowledge acquisition about quarks
are taken by everyone else outside the community X to be mistaken. In this case,
condition (b) of the account won’t be satisfied, given that X won’t be warranted in
believing that p—or, at least, X won’t be taken to be warranted in believing that p.
As a result, as one would expect, the account won’t entail that we have strong a
priori knowledge of the detection of quarks.
In response, note first that it’s not clear that the proposed situation shows that
condition (b) is not met. After all, to establish that (b) is not satisfied, we need to
provide a process of the same type as the one that produces in X the belief that p,
but which fails to warrant X in believing that p. Moving the community of physicists
X to the fringes doesn’t establish that their methods for believing that p no longer
warrant this belief. The fact that the methods are not taken to be warranted by
anyone else outside X doesn’t entail that the methods aren’t warranted. Condition
(b) is a claim about the warrant of the methods (or processes), not a claim about
whether these methods are taken to be warranted.
Second, if considering the community X to be on the fringes counts as an epistemic
possibility that undermines condition (b), exactly the same move can be made in the
case of mathematics. After all, consider a possible world in which a community of
mathematicians Xʹ uses methods that are taken by everyone else as being mistaken,
and so these methods are not taken to warrant Xʹ in believing that pʹ (say, that
inaccessible cardinals exist). This would then establish that Xʹ doesn’t have strong a
priori knowledge of the existence of such mathematical entities.8
As a result, it follows from the account that in order for us not to have strong a priori
knowledge of the existence of quarks (by allowing the possibility of communities of
physicists on the fringes), we can’t have strong a priori knowledge of the existence of
696 SKEPTICISM
certain sets (given the possibility of communities of mathematicians on the fringes).

In other words, for the account not to over-generate—that is, for it not to turn clear
cases of empirical knowledge into a priori knowledge—it ends up under-generating—
that is, it fails to classify clear cases of a priori knowledge as such. Thus, once again, a
priori and empirical knowledge are not properly distinguished by the account.
6b Truth: Contingent and Necessary Matters

But perhaps there is an important aspect of the distinction between a priori and
empirical knowledge that the account above is able to capture. According to the
traditional view about mathematics, mathematical objects are necessary objects:
they exist in all possible worlds. And so, in particular, they exist in all epistemically
possible worlds. Hence, in the case of mathematics, as one would expect, condition
(c) can be met easily, given that the consequent of (c)—being necessarily true—
will make (c) true. But the same doesn’t happen in the case of contingent objects,
such as quarks. After all, quarks only exist in some worlds, and presumably only in
some epistemically accessible worlds. So, when we deal with contingent matters,
condition (c) is not immediately met.
In response, note first that it’s an advantage of the proposed account that it
doesn’t identify a priori knowledge with knowledge of necessary truths (see Kitcher
2000: 69). Since Kripke (1980), it has become clear that apriority and necessity
are different notions. After all, we may know a priori things that are contingent
(such as our own existence), and we may fail to know—and hence fail to know
a priori—necessary things (such as extraordinarily complex mathematical results).
Hence, taking mathematical objects as necessary doesn’t lend any support per se to
the proposed definition of a priori knowledge.
Second, even in the case of mathematics, the fact that condition (c) might be
easily met is irrelevant. After all, as noted, to have a priori knowledge we need first
to have knowledge (see the formulation of Kitcher’s account above). And given that
we may not be able to know a particular mathematical result, for example, due to
its complexity, we will not have a priori knowledge of this result, either. Exactly the
same point applies to empirical knowledge.
6c Warrant: Contingent and Necessary Matters

But let’s follow a little further the traditional account that insists that mathematical
objects exist in all possible worlds. Presumably, in all epistemically accessible worlds,
any process of the same type as those that produce in X a belief that p will also
warrant X in believing that p. After all, no empirical constraints play any role in
the warrant process, and presumably, nothing that could happen in any world
would change the warrant in question. As a result, if we deal with necessary entities,
condition (b) can be met without difficulty. But the same doesn’t happen in the
case of contingent objects, such as quarks, which presumably exist only in some
epistemically accessible worlds. So, perhaps this aspect of the distinction between a
priori and empirical knowledge—the fact that the latter, but not the former, depends
on what goes on in each world—can be secured via this account.
But it’s not clear that the account captures even this much. After all, consider
again the possibility of a world with a community of mathematicians on the fringes,
so that the techniques developed to warrant mathematical beliefs fail to warrant
them. This possibility would undermine the claim that even mathematical results are
known a priori. Hence, mathematical knowledge, just as empirical knowledge, turns
out to be world-dependent.
To overcome this difficulty, one could reject the possibility of epistemic
communities of mathematicians on the fringes, by insisting that they are
epistemically impossible. But if this move is made, then epistemic communities of
physicists in the fringes will also have to be ruled out. After all, it will simply beg
the question to claim that, given that physics is not known a priori, the relevant
epistemic possibilities can be entertained! But, as a result, and as we saw above, if
the possibility of epistemic communities of physicists is not entertained, we will end
up having a priori knowledge of physical objects! Once again, the proposed analysis
is unstable: it either over-generates or under-generates.
7 CONCLUSION
Can the criticism of this account of a priori knowledge be generalized to other
attempts to characterize this type of knowledge? Or is this criticism restricted only
to this particular characterization? It is possible that the point made here has a
broader scope, although I will not try to argue for this claim here. Suffice to note
that, in general, attempts to characterize an intuitive notion in an extensionally
adequate way are typically plagued with difficulties, as challenges to proposed
conceptual analyses of knowledge (Gettier 1963), modality (Shalkowski 1994),
and logical consequence (Etchemendy 1990) clearly show. A priori knowledge is no
exception in this respect.
In conclusion, although the intuition that motivates one to develop an account
of a priori knowledge is clear—a priori knowledge is knowledge independent of
experience—it’s not so apparent that one can turn this intuition into a problem-
free characterization of a priori knowledge. In particular, if we judge from the
difficulties faced by Kitcher’s account, it’s much less obvious that one can articulate
an extensionally adequate account of a priori knowledge that captures the above
intuition. Given this difficulty, the problem of how to characterize a priori knowledge
remains open, and thus, in the end, so does the answer to the question: just what is
a priori knowledge?
In light of the considerations discussed in this book, it seems that both a priori
skepticism (particularly in its local form) and skepticism about a priori knowledge
are alive and well.9
NOTES
1. As also noted, this is not meant to include the concepts required to understand the
propositions that are known a priori. After all, we may know a priori propositions
whose concepts can only be obtained from experience.
698 SKEPTICISM
2. See Kitcher (1983) for his defense of this view about mathematics.
3. Similar complaints can be found in, e.g., Parsons (1986) and Summerfield, Casullo,
and Peacocke (1994).
4. In the end, maybe the issue regarding the “sources” of a priori knowledge shouldn’t
be settled a priori!
5. Note that condition (a) only requires the existence of a particular process that could
produce in X the relevant belief that p, not a process that has actually done so.
6. In other words, isomorphic structures are elementarily equivalent.
7. My thanks to Baron Reed for pressing this point.
8. Kitcher agrees, of course, with this conclusion. But if he made this move as part of the
characterization of a priori knowledge, he would be clearly stacking the deck.
9. My thanks go to Diego Machuca and Baron Reed for all their help and feedback on
earlier versions of this work.
REFERENCES
Beebe, James. 2011. “A Priori Skepticism,” Philosophy and Phenomenological Research 83:
583–602.
Boghossian, Paul and Christopher Peacocke (eds.). 2000: New Essays on the A Priori.
Butts, Robert (ed.). 1986. Kant’s Philosophy of Physical Science. Dordrecht: Kluwer
Academic Publishers.
Casullo, Albert. 2003. A Priori Justification. New York: Oxford University Press.
Chudnoff, Elijah. 2013. Intuition. Oxford: Oxford University Press.
da Costa, Newton, Décio Krause, and Otávio Bueno. 2007. “Paraconsistent Logics and
Paraconsistency.” In D. Jacquette (ed.), Philosophy of Logic, 791–911. Amsterdam:
North-Holland.
Dummett, Michael. 2000. Elements of Intuitionism. 2nd edn. Oxford: Clarendon Press.
Etchemendy, John. 1990. The Concept of Logical Consequence. Cambridge, MA: Harvard
University Press.
Harper, William. 1986. “The A Priori and Materially Necessary.” In Kant’s Philosophy of
Physical Science, 239–272. Dordrecht: Kluwer Academic Publishers.
Jacquette, Dale (ed.). 2007. Philosophy of Logic. Amsterdam: North-Holland.
Kitcher, Philip. 1980. “A Priori Knowledge,” Philosophical Review 89: 3–23.
Kitcher, Philip. 1983. The Nature of Mathematical Knowledge. New York: Oxford
University Press.
Kitcher, Philip. 2000. “A Priori Knowledge Revisited.” In P. Boghossian, and C. Peacocke
(eds.), New Essays on the A Priori, 65–91. Oxford: Clarendon Press.
Kripke, Saul. 1980. Naming and Necessity. Cambridge, MA: Harvard University Press.
McGee, Vann. 1985. “A Counterexample to Modus Ponens,” Journal of Philosophy 82:
462–471.
Mortensen, Chris. 1995. Inconsistent Mathematics. Dordrecht: Kluwer.
Parrini, Paolo (ed.). 1994. Kant and Contemporary Epistemology. Dordrecht: Kluwer
Academic Publishers.
Parsons, Charles. 1986. “Review of The Nature of Mathematical Knowledge,” Philosophical
Review 95: 129–137.
Priest, Graham. 1997. “Inconsistent Models for Arithmetic: I, Finite Models,” Journal of
Philosophical Logic 26: 223–235.
Priest, Graham. 2000. “Inconsistent Models for Arithmetic: II, The General Case,” Journal
of Symbolic Logic 65: 1519–1529.
Shalkowski, Scott. 1994. “The Ontological Ground of the Alethic Modality,” Philosophical
Review 103: 669–688.
Stroud, Barry. 1965. “Wittgenstein and Logical Necessity,” Philosophical Review 74:
104–118.
Summerfield, Donna, Albert Casullo, and Christopher Peacocke. 1994. “The Origins of the
A Priori.” In P. Parrini (ed.), Kant and Contemporary Epistemology, 47–72. Dordrecht:
Kluwer Academic Publishers.
Weyl, Hermann. 1918/1994. The Continuum: A Critical Examination of the Foundations
of Analysis. Translated from the 1918 German edition by Stephen Pollard and Thomas
Bole. New York: Dover.
Wittgenstein, Ludwig. 1956. Remarks on the Foundations of Mathematics, edited by G. H.
von Wright, R. Rhees, and G. E. Anscombe, and translated by Anscombe. Oxford:
CHAPTER FORTY-EIGHT
Skepticism about
Other Minds
ANIL GOMES
1 INTRODUCTION
Skepticism about others’ minds is the claim that we have no knowledge of other
people’s mental lives. This can sometimes strike us as a manifest truth. We occasionally
find ourselves in situations where we feel cut off from what other people think and
feel, and when this happens it can seem like the bringing to the surface of a cognitive
estrangement which is always there. Thus Lily Briscoe, in Virginia Woolf’s To the
Lighthouse, expresses frustration at her inability to enter into “the chambers of
[another’s] mind and heart. . . . How, then, she had asked herself, did one know one
thing or another about people, sealed as they were?” (1998: 44). Such knowledge is
unattainable, she concludes.
Yet such skepticism should also strike us as strange and revisionary. For we do
take ourselves to know many things about what other people think and feel. I know
that Arun hopes for a Labor victory; I know that Priya is angry with her sister;
I know that Ayesha suffers from back pain. This knowledge is part of our everyday
dealings with other people. And it is a central part of many of those relationships
that give our lives significance: part of what makes Ayesha such a good friend is
that she knows when a particular conversation topic is upsetting those around her
and moves to change the topic. Skepticism about others’ minds conflicts with our
ordinary conception of our relation to other people.
My aim in this chapter is to consider what reason we have to endorse skepticism
about others’ minds. I’ll argue that we need to distinguish two distinct ways in
which one might motivate such skepticism and provide some reason for focusing on
the second of these motivations. I’ll set out some of the responses to the skeptical
problems before providing my own sense of which avenues of further exploration
will be most fruitful.
An initial point of clarification: which knowledge claims does skepticism about
others’ minds rule out? I take the knowledge in question to be any claim that
ascribes a mental feature to another person. Since the category of a mental feature
is so broad, this is itself an expansive category: the ascription of beliefs, feelings,
SKEPTICISM ABOUT OTHER MINDS 701
character traits, emotions, actions, and activity all fall under the heading of mental
features, broadly construed. Those who write on this topic tend to concentrate on
our knowledge of what other people believe and feel. I will follow that restriction in
what follows. But we should not assume that the same considerations apply to our
knowledge of each and every member of this diverse class.
2 THE PROBLEM OF ERROR

One common way to raise a skeptical problem about others’ minds is by appeal to
cases of pretense (e.g. McDowell 1998c, drawing on a tradition emanating from
Wittgenstein 1992). Cases of pretense are ones in which someone presents themselves
as having a mental feature which they in fact lack. A judgment based on how things
seem in such cases will result in a false belief. But—the skeptic argues—your evidential
basis for this judgment is the same as that in the case in which you judge someone
to have a mental feature that they do have. Since your evidence does not suffice for
knowledge in the bad case, it cannot suffice for knowledge in the good case.
Note first that there is something odd about using pretense as an exemplar of this
skeptical scenario. Another person can pretend to have a mental feature that she
lacks only if she has other mental features such as a desire to simulate and the belief
that this is the way to do so. For this reason, a case of pretense does not support
the claim that we have no knowledge of others’ minds, for I can reason as follows:
either she is pretending or she is not. If she is not pretending, then she really does
instantiate the mental feature that she appears to instantiate. If she is, then she
must have some other beliefs and desires. In either case, she has a mind. The case
of pretense is ineffective in motivating complete skepticism about others’ minds
for the same reason that Descartes judged the evil demon hypothesis ineffective in
motivating skepticism about his own existence.1
This implication is avoided by choosing an alternative skeptical scenario, such
as one in which those around us are robots or philosophical zombies. I will refer
to such cases as ones of error, understood as ones in which it seems to you as if
someone has a particular mental feature which they in fact lack. Unlike cases of
pretense, certain cases of error are compatible with it not being possible for a subject
to have knowledge of others’ minds.
With this in mind, we can formulate the skeptic’s argument. Where ascriptions
to another of a mental feature are of the form “o is F,” let “the bad case” designate
a case in which it seems to you as if someone has a particular mental feature which
they in fact lack and you judge mistakenly that o is F, and let “the good case”
designate the case in which it seems to you as if someone has a particular mental
feature which they do in fact have and you judge accurately that o is F. Let two cases
be phenomenally indistinguishable when things are phenomenally the same for the
subject in the two cases. Then we can reason as follows:
1. The good case and the bad case are phenomenally indistinguishable.
2. If the good case and the bad case are phenomenally indistinguishable, then
the subject’s evidence is the same in the two cases.
702 SKEPTICISM
3. If the subject’s evidence is the same in the two cases, then the epistemic
status of her belief is the same in the two cases.
4. In the bad case, you don’t know that o is F.
5. In the good case, you don’t know that o is F. (From 1, 2, 3, and 4)
This argument generalizes to take in all mental phenomena that fall under cases
of error. It can thus be used to show that we don’t have any knowledge of other
people’s mental lives.
How should an anti-skeptic reply to this argument? (4) follows from the
assumption that knowledge entails truth. That leaves us with (1), (2), and (3). Some
forms of epistemological externalism reject (3): they hold that factors external to a
subject can affect the epistemic status of her beliefs without changing her evidence.
Evidential externalist responses reject (2): they hold that the evidence available
to a subject can differ even when things seem to her to be the same. Disjunctive
responses reject (1): they hold that how things seem to the subject is not the same in
the two cases for in the good case, “the appearance that is presented to one . . . is a
matter of the fact itself being disclosed to the experiencer” (McDowell 1998c: 387).
The rejection of (1) is compatible with the good and bad cases being subjectively
indiscriminable—in the sense that in both cases one can’t know, through reflection,
that one is not in the good case—but subjective indiscriminability does not entail
phenomenal indistinguishability.
I won’t evaluate these responses here. What should be clear, from the way in
which I’ve set up the problem, is that this style of skeptical problem arises in many
domains. In the case of perceptual knowledge, the bad case is one of illusion or
hallucination. In the case of knowledge of the past, it is one in which the world came
into existence five minutes ago with all my apparent memories included. Skeptics
claim that the indistinguishability of such cases from genuine cases undermines our
claims to knowledge. This form of argument generalizes.
Furthermore, the forms of response mentioned above also apply in each of these
domains. Evidential externalists about perceptual knowledge hold that the evidence
in the good case is not the same as the evidence available in the bad case; disjunctive
responses hold that how things seem to one is different in the two cases. This fact of
wider application is made more salient by the fact that John McDowell’s influential
response to the skeptical problem arises in the context of drawing a link between the
problem of other minds and the problem of perceptual knowledge. Once one has
noted this generalization, it is hard not to sympathize with Fodor’s comment that
“it’s gotten harder to believe that there is a special problem about the knowledge of
other minds (as opposed to other anything elses)” (1994: 292).
What are the implications for skepticism about others’ minds? My provisional
conclusion is that we have seen nothing, thus far, to motivate the thought that we
have a distinctive philosophical problem of other minds—one that turns on aspects
of our mentality and its expression to others. This conclusion is only provisional
because it may be that the plausibility of any particular application of the responses
mooted above will depend on facts specific to the domain in question: perhaps a
disjunctive solution is plausible for perceptual knowledge but not knowledge of
others’ minds, say. But, as a general moral, we can say that what we have here is a
problem about knowledge, one that can be raised in the case of our knowledge of
others’ minds. And while this is a way of motivating skepticism, it doesn’t motivate
a distinctive skepticism about others’ minds.
3 THE PROBLEM OF SOURCES

In the rest of this chapter, I want to motivate an alternative way of raising a skeptical
problem about others’ minds. It is a way of raising the problem that I think is both
more faithful to the history of the problem as discussed in twentieth-century analytic
philosophy and turns on issues distinctive to our mentality.
Let me start with the history. When one looks at discussions of the problem of
other minds at the start of the twentieth century, issues about pretense and deception
are largely absent from the debate. Instead, the starting point is a supposed contrast
between the way we know our own minds and the way in which we know the minds
of other people. Here are two examples, spanning the first half of the twentieth
century. First, G. F. Stout, introducing his discussion of knowledge of others’ minds
by means of an implied contrast with the way in which we know our own mind:
To introspect is to attend to the workings of one’s own mind. . . . No-one can
directly observe what is passing in the mind of another. (Stout 1898: 14, 20)
Second, A. J. Ayer, characterizing the problem of other minds:
We may take as our starting-point the propositions that I can have direct
knowledge of my own experiences and that I cannot have direct knowledge of
anyone else’s. (Ayer 1953: 193)
Each of these writers adverts, in different ways, to a certain contrast between the
way in which we know our own mind and the way in which we know the minds of
others. The exact nature of this contrast is largely unimportant except for dialectical
purposes. What matters, for each of them, is that the contrast shows up a problem in
accounting for the source of our knowledge of another’s mind. The thought is this:
we have a way of knowing about our own mental lives. But this source of knowledge
cannot be used to find out about the mental life of another person. If we do not have
a source of such knowledge, then we do not have knowledge of others’ mental lives.
We can best formulate this line of thought as an antinomy of three claims:
(S1) If you know that p, then there is a means by which you know that p.
(S2) There is no means by which one can know truths about another’s mind.
(S3) We know truths about another’s mind.
These three claims are jointly incompatible. From (S1) and (S3) it follows that
there is a means by which we know about other people’s mental lives; this is in
contradiction with (S2).
Skepticism about others’ minds is a rejection of (S3): we have no knowledge
about other people’s mental lives. Most participants in the debate reject (S2): it is
704 SKEPTICISM
certainly the most obviously contestable, and I will focus on it in what follows. But
it will be useful to say something about (S1), a claim which is rarely articulated but
which I take to be assumed in the debate.
J. L. Austin (1961) notes that whenever someone makes a claim to knowledge, it
is appropriate to ask “How do you know?” Timothy Williamson (2000: 252–253)
agrees, and Quassim Cassam (2009: 112) calls this “something like a conceptual
truth.” The natural way of taking this question is as requesting identification of the
source or means of one’s knowledge. The linguistic fact about the appropriateness
of a certain question reflects a metaphysical fact about knowledge: namely, that if
someone knows that p, then there is a means by which she knows it.
Challenges to (S1) arise in cases in which the “How do you know?” question
is claimed to be inappropriate, thus identifying potential counterexamples to the
thesis. Consider the assertion “I am in pain.” Certain readers of Wittgenstein have
taken the question “How do you know?” to be inappropriate in response to this
assertion. One response is to conclude that it is false to say that a subject knows she
is pain. But an alternative, canvassed by Stuart Hampshire, is to hold that a subject
can know that she is in pain without there being a means by which she knows it
(1969: 282–283). Donald Davidson makes the same claim for other cases of self-
knowledge (1991: 212).
I will put these challenges to one side. They arise specifically in the context of
certain cases of self-knowledge, and the considerations adduced do not motivate
the failure of (S1) for other types of knowledge. Moreover, (S1) has a great deal of
plausibility. If one thinks of states of knowledge as part of the natural world, then
there must be a causal story one can tell about how subjects get into them. Those
who are sympathetic to some form of naturalized epistemology will hold that telling
such a story identifies the means by which a subject knows. Those who want to
reject skepticism about others’ minds should deny (S2).
4 PERCEPTION AND INFERENCE

Why would one think that there is no means by which we know truths about
another’s mind? One reason would be if there were some general condition
on knowledge which couldn’t be met by truths about another’s mind. Say, for
example, that knowledge requires a causal interaction between a subject’s belief
and that which is known. This condition has been thought to raise a problem
in explaining our knowledge of domains from which we are causally isolated,
for example, mathematical truths (Benacerraf 1973: 671–673) or moral values
(Mackie 1977: 38–39). If others’ minds were causally isolated from the physical
world, then there would be no means by which we could know about them, and
(S2) would hold.
This suggests that (S2) follows from certain conceptions of the mental, such as
those on which the mental features are attributes of an immaterial substance causally
isolated from the physical world. It may also follow from those conceptions of
the mental that deny the existence of immaterial substance but hold that particular
mental features such as the qualitative character of consciousness are causally
isolated from the physical world. I take it that these conceptions of the mental are
neither popular nor attractive, so put this motivation of (S2) to one side.
The discussion in Ayer and Stout offers us an alternative way of motivating (S2).
Consider the standard way in which we know about our own mental lives. Call
this method introspection. It is necessarily first-personal, that is to say, it cannot be
used to find out about the mental features of another person. So the source of our
knowledge about own mental life is not one that can provide us with knowledge
of another’s mind. I take this to be the reason why some of those writing on this
issue focus on the asymmetry between self-knowledge and our knowledge of others’
minds: it shows up the fact that one prevalent source of knowledge of mental
features cannot be used to gain knowledge of another’s mental life.
How, then, do we know about others’ minds? Other people and their mental
features are aspects of the world around us. And when we think about how we gain
knowledge about the external world, there seem to be two broad methods of which
we make use: perception and inference. Perceptual knowledge is knowledge based on
the basic deliverances of the senses, as when I come to know that the chairs are blue
by looking at them. Inferential knowledge is knowledge that takes us beyond what is
given in experience, as when I come to know that there was a party at my neighbor’s
house on the grounds that it is the best explanation of the contents of her recycling
bin. Perhaps there are interesting philosophical questions about how exactly we are
to distinguish these sources of knowledge, but they look to be distinct.
Is our knowledge of others’ minds perceptual or inferential? This is where the
case for (S2) has been thought to get its bite. For one recurring theme in the literature
is that neither of these sources of knowledge can enable us to gain knowledge about
another person’s mental life. Against perception it is said that we cannot perceive the
mind of another person: at best we can perceive her behavior. So we cannot be said
to have perceptual knowledge of her mental life. But against inference it is claimed
that this experience of behavior isn’t secure enough to ground an inference to the
presence of others’ minds. If neither perception nor inference can serve as a source
of knowledge of other people’s mental lives, then there is a problem in accounting
for the source of such knowledge.
What reasons are given for ruling out these sources of knowledge? Take perception
first. In comparing the perceptual and inferential responses, Price comments that
“it must strike every unprejudiced person that [the perceptual theory] is much the
odder of the two. . . . It is extremely paradoxical” (Price 1931: 54). This sense of
absurdity is often shared: what would it be to perceive another’s mental life? Colin
McGinn offers further considerations in support: “While propositional perceptual
reports sometimes seem natural in specifying my cognitive relation to the mental
states of another, direct object perceptual reports (‘I saw the pain in his foot’) seem
definitely wrong” (McGinn 1984: 123 n.2). If this linguistic phenomenon tells us
something about the objects of perception, knowledge of others’ minds cannot be a
form of perceptual knowledge.
What about the claim that inference cannot serve as a source of such knowledge?
Various arguments have been put forward in support of this claim. It has been
claimed that we do not appear to engage in inferential reasoning; that infants and
706 SKEPTICISM
animals can know things about others’ minds without possessing the cognitive
powers necessary to engage in inference; that we do not know enough about other
people’s behavior to provide a sound base for inference (Price 1931: 55–56). More
sophisticated objections hold that one cannot engage in inference in cases where it is
logically impossible to verify the conclusion, or that the link between behavior and
mind is known only from one’s own case and that this is not a secure enough basis to
secure inferential justification (Ryle 1949: 15–16; Malcolm 1958: 969). If one finds
these reasons persuasive, knowledge of others’ minds is not a form of inferential
knowledge. It is the supposed inadequacies of perceptual and inferential models of
knowledge that motivate skepticism about others’ minds.
Standard responses to this problem opt to deny one or other of these claims:
either they claim that perception can be a source of knowledge of others’ minds,
or—as is more common—they claim that the experience of behavior can ground
inferential knowledge about another’s mind. Either option amounts to a denial
of (S2): there is a means by which we know truths about another’s mind. Let us
consider each in turn.
First, perception. Although it has been the less popular of the two options, the
claim that our knowledge of others’ minds is a form of perceptual knowledge has
been defended throughout the twentieth century (Duddington 1918; Dretske 1973;
McDowell 1998c; Cassam 2007a: ch.5; McNeill 2011). It was a prominent part of
the phenomenological tradition, and some recent defenders of a perceptual account
have drawn on the work of Scheler, Husserl, and Merleau-Ponty in presenting their
proposals (Gallagher and Zahavi 2008; Krueger and Overgaard 2012).
Much of this work is primarily defensive. Consider the following objection to
perceptual accounts: other people’s mental lives can’t be perceived; therefore, our
source of knowledge of such lives can’t be perception. There are two responses to
this claim in the literature. Dretske and Cassam accept the premise of the objection
while rejecting its conclusion: they hold that one can see that another person is
angry and thus come to know that another person is angry without seeing her anger.
Others hold that our perception of others extends beyond their behavior to take in
their mentality: other people’s mental lives are perceptible.
In defending the first option, Dretske offers the following analogy: one can
see that a piece of metal is hot without seeing its heat. Similarly, one can see that
another person is angry without seeing her anger. These claims rely on the account
of epistemic seeing—seeing that something is the case—set out in Dretske (1969). If
one accepts Dretske’s account of epistemic seeing, it is possible to see that someone
is angry and thereby come to know that she is angry without seeing her anger.2
The second form of response holds that certain mental features can be seen.
This is sometimes supported, contra McGinn, by appeal to the facility of certain
perceptual locutions in reporting the source of our knowledge of others’ minds:
“I heard the anger in her voice”; “I saw the pain in her eyes” (Green 2007; Stout
2010). The claim has been developed in different ways: some hold that mental
states are expressed in behavior and thus perceptible (Green 2007); others hold
that behavior is partly constitutive of mentality, such that in seeing others’ behavior
one sees their mental features (Krueger and Overgaard 2012). These claims turn
on wider issues in the philosophy of mind about the relation between mental states
and behavior.
The above comments are defensive: they aim to defend the perceptual model
against a natural and prominent objection. Are there any positive reasons for
adopting a perceptual account of our knowledge of others’ minds? One can find
two lines of thought in the literature. The first is that it seems to us as if our access
to other people’s mental lives is perceptual and that we should take this appearance
seriously. This is particularly stressed by those who draw on the phenomenological
tradition (e.g., Krueger and Overgaard 2012: 248), but it might also be thought
to be part of John McDowell’s influential presentation of a perceptual account.
As McDowell puts it, “We should not jib at, or interpret away, the commonsense
thought that . . . one can literally perceive, in another person’s facial expression or
his behavior, that he is in pain, and not just infer that he is in pain from what one
perceives” (McDowell 1998b: 305). This explains why the bulk of this literature
is primarily defensive: if our starting point is the commonsense thought that our
access to others’ minds is perceptual, then all we need do is defend that starting
point against objections.
The second motivation relates to what is sometimes called “the conceptual
problem of other minds.” This chapter is concerned with the epistemological
problem of other minds, the problem of accounting for our ability to know about
others’ mental lives. The conceptual problem of other minds concerns the problem
of explaining our ability to think about others’ mental lives. This latter problem is
often traced back to a set of remarks in Wittgenstein’s Philosophical Investigations
(1953: §243–315). Details of this supposed problem needn’t concern us here, but
some philosophers who address the conceptual problem take it that explaining our
ability to think about others’ mental lives requires us to take other people’s mental
lives to be perceptible.3 And this, they hold, supports the claim that our knowledge
of others’ mental lives is perceptual.4
Let me turn now to inference. There have been two historically popular ways of
cashing out the thought that our knowledge of others’ minds is based on inference.
According to analogical accounts, my beliefs about another’s mental life are warranted
in virtue of my recognizing that my thoughts and feelings cause certain patterns of
behavior, perceiving other people exhibiting those same patterns of behavior, and
concluding—by analogy—that when other people exhibit that behavior, they also
instantiate those thoughts and feelings (Mill 1979; Russell 1948). According to best
explanation accounts, my beliefs about another’s mental life are warranted in virtue
of the hypothesis that the person instantiates those mental features being the best
explanation of her observable behavior (Pargetter 1984).
J.S. Mill is often claimed to have provided the paradigm statement of an analogical
account. He writes:
I conclude that other human beings have feelings like me, because, first, they have
bodies like me, which I know, in my own case, to be the antecedent condition of
feelings; and because, secondly, they exhibit the acts, and other outward signs,
which in my own case I know by experience to be caused by feelings. (1979: 190)
708 SKEPTICISM
This argument is typically presented as a case of enumerative induction. I observe

that, in my own case, behavior type B is caused by feeling type F; I conclude that for
all humans, behavior B is caused by feeling F; I observe another person exhibiting
behavior B and conclude that she instantiates feeling F. But so presented, it is
vulnerable to the objection that observation of one positive instance is not sufficient
to warrant belief in a universal statement. According to this objection, the argument
from analogy no more justifies my believing that other people have thoughts and
feelings than my having a mole under my left arm justifies my believing that everyone
has a mole under their left arm (Malcolm 1958: 969; Putnam 1975: 342–343).
There are two ways to respond to this objection. First, one might hold that
enumerative induction from a single case is sometimes permissible. Hyslop and
Jackson (1972: 175), for example, hold that enumerative induction from a single
case is permissible when one already knows that the relation between the two
features in question is causal. Second, one might deny that the argument from
analogy really proceeds from consideration of a single case. Ayer (1956: 219–222),
for instance, holds that the analogical inference rests on a multitude of correlations
between instances of mental properties and instances of behavior, albeit instances all
confined to a single person.
Whatever the merits of these responses, the single-case objection is a poor
objection to Mill, whose discussion is much more sophisticated and interesting than
typically presented.5 Central to Mill’s discussion—which bears reading beyond the
sample of text quoted above—is the thought that we are dealing with the class
human beings. Human beings are members of a natural kind, and Mill himself is a
member of that kind. Since the laws of nature that govern our mental features apply
to all members of the natural kind uniformly, Mill takes himself to be licensed in
moving from the identification of those features of himself which hold in virtue of
him being a member of that kind to the reasonable assumption that other members
of the kind will instantiate the same features. This is comparable, Mill thinks, to
the chemist who makes pronouncements about the properties of a new chemical
element after having conducted one proper experiment (Mill 1973, quoted in
Thomas 2001: 519).
Mill’s positive proposal is closer to the best explanation accounts according to
which our belief in others’ mental states is justified in the same way as our belief in
scientific unobservables: on the grounds that the postulation of such entities best
explains the observed data. In the case of others’ minds, the observed data is the
movement of their bodies, and the best explanation is the ascription to others of
mental states. This is perhaps the dominant account of our knowledge of others’
minds in the psychological literature, but it raises a number of questions. First,
does inference to the best explanation warrant the beliefs of those who do not have
explicit beliefs about the explanatory role of mental states—a group which I take
to include most of us before we took up philosophy or psychology? Second, can it
warrant the beliefs of children, say, who lack the conceptual resources to hold beliefs
about the explanatory connections between mental states and observable behavior?
And finally, is the ascription of mental states genuinely the best explanation of other
people’s behavior, or are there other neuroscientific explanations that crowd it out?
Two final comments on the perceptual and inferential approaches. First, it should
be common ground that some of our knowledge of others’ minds is inferential.
We sometimes puzzle over another’s strange behavior, before concluding that it is
explained by some fact about her mental life. This knowledge is based on inference.
Or consider the way in which historians come to know what those in the past
believed. Perhaps cultural historians of the British Empire know whether Dadabhai
Naoroji—the first British Asian MP—regretted leaving mathematics for politics. But
if they do so, it is because they have inferred it from historical evidence. The debate,
then, should not be phrased as whether our knowledge of others’ minds is either
perceptual or inferential, but given that some of it is inferential, whether it is all so.
Second, in presenting the opposition between these accounts, I have taken it for
granted that there is a distinction between perceptual and inferential knowledge.
How are we to characterize this distinction? Defenders of a perceptual model
sometimes seem to assume that a piece of knowledge is inferential if it is the result
of a process of inferential reasoning (Cassam 2007a: 160; cf. Price 1931: 55). But
a piece of knowledge can be supported by inferential relations even if it is not the
result of inferential reasoning (Brandom 2002: 95–98; Pryor 2004: 190). My belief
that you’re at home, for example, may come to me unbidden upon seeing your car
in the driveway; in this sense, it is not the result of a process of inferential reasoning.
But its epistemic justification may still lie in my warranted belief in an inferential
relation between the proposition that your car is in the driveway and the proposition
that you are at home. We need a fuller characterization of the differences between
perceptual and inferential knowledge if we are to confidently evaluate the reasons
for thinking that one or other is the source of our knowledge of others’ minds.
5 NEXT STEPS
I finish with some more general comments on skepticism about others’ minds. As
I’ve presented the issue, the aim is to identify the source of our knowledge of others’
mental lives. The recent debate confines itself to discussion of perception and
inference as possible sources of knowledge. This marks a change from the discussion
conducted at the start of the previous century where other sources of knowledge,
such as telepathy, were given serious consideration. Let me conclude by noting two
ways in which broadening our horizons may help this debate.
First, testimony. When one thinks about the way in which we come to know about
others’ mental lives, it is not perception or inference that stand out. Rather, other
people tell us things—and among the things they tell us are facts about their mental
life. Could testimony be a source of knowledge of others’ mental lives? One might
think not for a simple reason: knowledge gained through testimony presupposes
that one knows that the other person has a mind, so it can’t be used to explain how
one gains such knowledge (Stroud 2000b: 101; Cassam 2007a: 157). But this form
of reasoning has a parallel in discussions about perceptual knowledge: knowledge
gained through perception presupposes that one knows that one is perceiving (Stroud
2000c: 131–133). And the presupposition claim about perception has been the subject
of much discussion: it looks to lead to skepticism about perceptual knowledge, and
710 SKEPTICISM
many have argued that it should be rejected (McDowell 1998d, 1998e). It is worth
considering whether the lessons of this debate can be applied to the case of gaining
knowledge of others’ minds through testimony and, if so, whether testimony could
serve as a basic source of knowledge of others’ minds (Gomes 2015).
Second, variety. I’ve assumed throughout that there is one way—or better, that
there is some fundamental way—by which we come to know about others’ minds.
But perhaps this is false: perhaps there is no general way by which we come to know
about the minds of others. Let us draw a distinction between two different ways
of grouping known facts: on the basis that such knowledge has been gained from
the same source, or on the basis that it all concerns the same domain. Call these
two ways of classifying knowledge source-knowledge and domain-knowledge. Some
examples of source-knowledge include the facts I know on the basis of perception;
the facts that you know on the basis of reading Wikipedia; the facts that Priya knows
through divine revelation. Some examples of domain-knowledge include everything
I know about cars; that which you know about Sri Lanka; the things you know
about the Catholic Church.
The difference between these two ways of classifying known facts is not always
made clear, as when philosophers move interchangeably between discussing
perceptual knowledge and knowledge of the external world. But we should be
careful to mark the distinction, for it need not be the case that some knowledge
picked out via a certain domain has all arisen from the same source. And, indeed,
although there are cases where all the things we know about a particular domain
were gained from the same source—“Everything I learned about love I learned from
romance novels”—this is the exception rather than the norm. Given that knowledge
of others’ thoughts and feelings is a type of knowledge picked out via its domain, we
have no grounds for thinking that it is all gained via the same source.
Does such variety have implications for the skeptical challenge? If the aim is to
identify the means (perhaps plural) by which we know about others’ minds, then we
need some way of individuating sources of knowledge. This is particularly important
in a context in which psychological and neuroscientific investigation is providing
us with a wealth of information on the mechanisms by which we ascribe mental
states to one another (Nichols and Stich 2003). There have been discussions of how
to individuate sources of knowing in the epistemology literature (Nozick 1981:
179–185; Audi 2002; Cassam 2007b), but presentation of the problem of other
minds often proceeds as if we have an unproblematic understanding of the means of
knowing which are antecedently available for consideration. Addressing skepticism
about others’ minds may require renewing interest in this epistemological question.6
NOTES
1. I owe this point to Paul Snowdon.
2. For criticisms and discussion of Cassam and Dretske on this point, see Gomes (2010)
and McNeill (2011).
3. See Avramides (2001) and Gomes (2011) for introductions to the conceptual problem.
4. See Malcolm (1958) for an influential presentation of this thought.

5. See Thomas (2001) for an excellent discussion of Mill’s text.
6. Many thanks to Nick Jones, Rory Madden, Matt Parrott, Ian Phillips, and Lee Walters
for helpful comments.
REFERENCES
Audi, Robert. 2002. “The Sources of Knowledge.” In P. Moser (ed.), The Oxford Handbook
of Epistemology, 71–92. Oxford: Oxford University Press.
Austin, J. L. 1961 (1946). “Other Minds.” In his Philosophical Papers, 76–116. Oxford:
Avramides, Anita. 2001. Other Minds. London: Routledge.
Ayer, A. J. 1953. “One’s Knowledge of Other Minds.” In his Philosophical Essays, 191–
214. London: Macmillan.
Ayer, A. J. 1956. The Problem of Knowledge. London: Macmillan
Benacerraf, Paul. 1973. “Mathematical Truth,” Journal of Philosophy 70: 661–679.
Brandom, Robert. 2002. “Non-inferential Knowledge, Perceptual Experience, and
Secondary Qualities: Placing McDowell’s Empiricism.” In N. Smith (ed.), Reading
McDowell: On Mind and World, 92–105. London: Routledge.
Cassam, Quassim. 2007a. The Possibility of Knowledge. Oxford: Oxford University Press.
Cassam, Quassim. 2007b. “Ways of Knowing,” Proceedings of the Aristotelian Society 107:
339–358.
Cassam, Quassim. 2009. “What is Knowledge?” Royal Institute of Philosophy Supplement
84: 101–120.
Davidson, Donald. 1991. “What is Present to the Mind?” Philosophical Issues 1: 197–213.
Dretske, Fred. 1973. “Perception and Other Minds,” Noûs 7: 34–44.
Dretske, Fred. 1969. Seeing And Knowing. Chicago: University of Chicago Press.
Duddington, N. A. 1918. “Our Knowledge of Other Minds,” Proceedings of the Aristotelian
Society 19: 147–178.
Fodor, Jerry. 1994. “Jerry A. Fodor.” In S. Guttenplan (ed.), A Companion to the Philosophy
of Mind, 292–300. Oxford: Blackwell.
Gallagher, Shaun and Dan Zahavi. 2008. The Phenomenological Mind: An Introduction to
Philosophy of Mind and Cognitive Science. London: Routledge.
Gomes, Anil. 2009. “Other Minds and Perceived Identity,” Dialectica 63: 219–230.
Gomes, Anil. 2011. “Is There a Problem of Other Minds?” Proceedings of the Aristotelian
Society 111: 353–373.
Gomes, Anil. 2015. “Testimony and Other Minds,” Erkenntnis. 80: 173–183.
Green, Mitchell. 2010. “Perceiving Emotions,” Proceedings of the Aristotelian Society,
Supplementary Volume 84: 45–61.
Hampshire, Stuart. 1979 (1969). “Some Difficulties in Knowing.” In T. Honderich and M.
Burnyeat (eds.), Philosophy as It Is. Harmondsworth: Penguin.
Hyslop, Alec and Frank Jackson. 1972. “The Analogical Inference to Other Minds,”
American Philosophical Quarterly 9: 168–176.
712 SKEPTICISM
Krueger, Joel and Søren Overgaard. 2012. “Seeing Subjectivity: Defending a Perceptual
Account of other Minds,” ProtoSociology: Consciousness and Subjectivity 47: 239–262.
Mackie, J.L. 1977. Ethics: Inventing Right and Wrong. London: Penguin.
Malcolm, Norman. 1958. “Knowledge of Other Minds,” Journal of Philosophy 55: 969–
978.
McDowell, John. 1998a. Meaning, Knowledge and Reality. Cambridge, MA: Harvard
University Press.
McDowell, John. 1998b (1978). “On ‘The Reality of the Past.’ ” In his Meaning, Knowledge
and Reality, 295–313. Cambridge, MA: Harvard University Press.
McDowell, John. 1998c (1982). “Criteria, Defeasibility and Knowledge.” In his Meaning,
Knowledge and Reality, 369–394. Cambridge, MA: Harvard University Press.
McDowell, John. 1998d (1986). “Singular Thought and the Extent of ‘Inner Space’.” In his
Meaning, Knowledge and Reality, 228–259. Cambridge, MA: Harvard University Press.
McDowell, John. 1998e (1995). “Knowledge and the Internal.” In his Meaning, Knowledge
and Reality, 395–413. Cambridge, MA: Harvard University Press.
McGinn, Colin. 1984. “What is the Problem of Other Minds?” Proceedings of the
Aristotelian Society, Supplementary Volume 58: 119–137.
McNeill, Will. 2012. “On Seeing That Someone is Angry,” European Journal of Philosophy
20: 575–597.
Mill, J. S. 1973 (1843). A System of Logic Ratiocinative and Inductive, Collected Works of
John Stuart Mill, Volume VII, ed. J. M. Robson. Toronto: University of Toronto Press.
Mill, J.S. 1979 (1865). An Examination of Sir William Hamilton's Philosophy, Collected
Works of John Stuart Mill, Volume IX, ed. J. M. Robson and A. Ryan. Toronto:
University of Toronto Press.
Nichols, Shaun, and Stephen Stich. 2003. Mindreading: An Integrated Account of Pretence,
Self- Awareness and Understanding Other Minds. Oxford: Oxford University Press.
Press.
Pargetter, Robert. 1984. “The Scientific Inference to other Minds,” Australasian Journal of
Philosophy, 62: 158–163.
Price, H.H. 1931. “Our Knowledge of Other Minds,” Proceedings of the Aristotelian
Society 32: 53–78
Pryor, James. 2004. “There Is Immediate Justification.” In M. Steup and E. Sosa (eds.),
Contemporary Debates in Epistemology, 181–202. Oxford: Blackwell.
Putnam, Hilary. 1975. “Other Minds.” In his Mind, Language and Reality: Philosophical
Papers, vol. 2, 342–361. Cambridge: Cambridge University Press.
Russell, Bertrand. 1948. Human Knowledge: Its Scope and Limits. London: Routledge.
Ryle, Gilbert. 1949. The Concept of Mind. London: Hutchinson.
Stout, G. F. 1898. Manual of Psychology. London: University Correspondence College
Press.
Stout, Rowland. 2010. “Seeing the Anger in Someone’s Face,” Proceedings of the Aristotelian
Society, Supplementary Volume 84: 29–43.
Stroud, Barry. 2000a. Understanding Human Knowledge. Oxford: Oxford University Press.
Stroud, Barry. 2000b (1989). “Understanding Human Knowledge in General.” In his
Understanding Human Knowledge, 101–121. Oxford: Oxford University Press.
Stroud, Barry. 2000c (1996). “Epistemological Reflection on Knowledge of the External

World.” In his Understanding Human Knowledge, 123–138. Oxford: Oxford University
Press.
Thomas, Janice. 2001. “Mill’s Argument for other Minds,” British Journal for the History
of Philosophy 9: 507–523
Wittgenstein, Ludwig. 1953. Philosophical Investigations. Oxford: Basil Blackwell.
Wittgenstein, Ludwig. 1992. Last Writings on the Philosophy of Psychology. Vol. 2:
The Inner and the Outer. Oxford: Blackwell.
Woolf, Virginia. 1998 (1927). To The Lighthouse. Oxford: Oxford University Press.
CHAPTER FORTY-NINE
Moral Skepticism
RICHARD JOYCE
1 WHAT IS MORAL SKEPTICISM?

The moral skeptic denies (or at least refuses to affirm) that anyone has moral
knowledge. Moral skepticism could, of course, fall out of a more ubiquitous
skepticism according to which we lack knowledge in general, but as an interestingly
distinct position it is the view that there is something problematic about moral
knowledge in particular. If one assumes that for S to know that p is (at least) for S to
(i) believe that p (ii) truly and (iii) with justification, then there are several different
ways of being a moral skeptic:
(1) Noncognitivism: the denial that moral judgments are beliefs.
(2) Error theory: the acceptance that moral judgments are beliefs while denying
that these beliefs are ever true.
(3) Justification skepticism: the acceptance that moral judgments are beliefs
while denying that we are ever justified in holding these beliefs.
If one takes moral realism to be the thesis that we sometimes succeed in making
objectively true moral judgments, then the first two forms of moral skepticism
count as moral anti-realist views.1 However, moral skepticism and moral realism are
neither contraries nor contradictories. First, consider the constructivist view that
we sometimes succeed in making moral judgments that are true but lack any robust
claim to objectivity. Such a view leaves room for moral knowledge and yet remains
(by the above criteria) a form of antirealism. Second, consider the justification
skepticism just described in 3. One might lack justification for holding moral beliefs
that are in fact objectively true. Such a view satisfies the criteria both for moral
realism and for moral skepticism.
To clarify moral skepticism further, let us consider an ancient distinction. The
Academic skeptic assents that we lack knowledge and therefore suspends judgment
on many matters, whereas the Pyrrhonian skeptic withholds assent even to the claim
that we lack knowledge.2 One might think that this distinction would “scale down” to
the moral domain, perhaps matching some version of modern moral skepticism, but
it does so only clumsily. Let us define “moral knowledge” as knowledge of first-order
MORAL SKEPTICISM 715
moral claims, such as “Shoplifting is wrong,” “One ought not break promises without
good reason,” and so forth. The natural thought would be that the Academic moral
skeptic assents that we lack moral knowledge (and therefore suspends judgment on
moral matters), whereas the Pyrrhonian moral skeptic withholds assent even to the
claim that we lack moral knowledge. What is noteworthy is that the proposition that
we have/lack moral knowledge is not itself a first-order moral claim (but is rather an
epistemic claim about moral claims), and therefore such a Pyrrhonian moral skeptic
must doubt something more than just first-order moral claims. In other words, one
cannot be Pyrrhonian about just first-order moral claims. The classical Pyrrhonian
skeptic poked fun at his Academic opponent for withholding assent to many claims
while making a seemingly unaccountable exception for the universal epistemic
claim that there is no knowledge. But the Pyrrhonian moral skeptic is robbed of this
stratagem, for the Academic’s suspension of judgment for moral claims is perfectly
compatible with the wholehearted assertion of the higher-order claim that there
exists no moral knowledge.
While a modern skeptic may side with the Academic in affirming that we have
no moral knowledge, she does not necessarily therefore suspend judgment on all
first-order moral matters. The moral error theorist, for example, may base her
confident denial of moral knowledge on her equally confident denial that any moral
judgments are true. Faced with a moral proposition like “Shoplifting is morally
wrong,” the error theorist does not suspend judgment, but affirms its falsity. (She
also, of course, affirms the falsity of “Shoplifting is morally right” and “Shoplifting
is morally permissible”—so her affirmation must not be taken to imply any kind of
practical tolerance of shoplifting.)3
Nevertheless, something with the flavor of the Academic/Pyrrhonian distinction
can exist at the level of moral skepticism, for one can either be confident that
there is no moral knowledge or be quite undecided on the matter. One may, for
example, be undecided whether moral judgments are beliefs or rather some kind
of affective/conative mental state, in which case one will be undecided whether
noncognitivism is true, and hence potentially undecided whether there is such
a thing as moral knowledge. This is not, however, a terribly interesting kind of
position to hold, since it may reflect a simple contingent and temporary ignorance
on the subject. (I, for example, know nothing of advanced set theory, and
therefore maintain a healthy but mundane skepticism on any controversial claim
from that domain.) A more interesting kind of skepticism would hold that there
is something undecidable about these matters.4 While this latter view is coherent
(and possibly plausible), modern metaethicists who hold a position implying the
nonexistence of moral knowledge have tended to be more assured in advocating
their favored arguments, and therefore the three skeptical positions listed above
are usually put forward with some confidence, meaning moral skepticism as it is
discussed in the philosophical literature generally has a more Academic color. (In
other words, Pyrrhonian skeptics would probably accuse modern moral skeptics
of dogmatism.) In any case, let us discuss the three metaethical versions of modern
moral skepticism in turn.
716 SKEPTICISM
2 NONCOGNITIVISM
Noncognitivism has its roots in the nineteenth century, and arguably goes back to
Hume,5 but it first crystallized in the 1930s in works by Rudolf Carnap (1935), Ayer
(1936), and Stevenson (1937).6 The noncognitivist claims that moral utterances are
not really assertions, despite appearances to the contrary. Noncognitivism can be
put forward as a semantic thesis—that the meaning of moral sentences cannot be
properly rendered in the indicative mood—or put forward as a thesis concerning
the pragmatics of moral language—that moral sentences are not used with assertoric
force. According to the former understanding, moral sentences, such as “Shoplifting
is wrong,” need to be transcribed into some other non-indicative format, such as
“Don’t shoplift!” or “Boo to shoplifting!” What is metaethically important about
these renderings is that (i) they are no longer truth-evaluable, (ii) reference to moral
properties (e.g., wrongness) has disappeared, and (iii) they are no longer items of
belief. These same features are claimed of moral discourse according to a pragmatic
understanding of noncognitivism. Imagine that one were to utter the sentence
“Shoplifting is wrong” as a line in a play. The meaning of the sentence would remain
the same as ever—it would be a mistake to transcribe the sentence into some other
format—yet the sentence would not be put forward as a truth (i.e., it would not be
asserted), the speaker would not be committed to the existence of wrongness (any
more than “Once upon a time there lived a dragon” commits the storyteller to the
existence of dragons), and it would not be believed, either by utterer or by audience.
The pragmatic noncognitivist claims all this for ordinary moral utterances. For our
current purposes, the key noncognitive claim is that moral judgments are not items
of belief, implying that there can be no such thing as moral propositional knowledge.
Noncognitivism faces many challenges. If moral sentences are not truth-evaluable,
then it is difficult to make sense of their appearance in standard logically complex
contexts, such as negations or conditionals (Geach 1965). We frequently form views
about the world on the basis of our moral judgments, yet if moral judgments are
essentially conative (as opposed to doxastic), then our doing so would appear to be
nothing more than a form of wishful thinking (Dorr 2002). If there are no moral
beliefs, then it appears that we should reject as false or incoherent claims like “John
believes that Hitler was evil.” If moral judgments are nothing more than expressions
of the speaker’s desires, then what practical authority could such judgments purport
to have over others? Indeed, why would we engage in serious moral debate at all, if
all we are doing is trying to express conflicting feelings?7
The noncognitivism of the mid-twentieth century tended to be unapologetic
about these consequences, seemingly willing to bite the bullet that many aspects
of everyday moral discourse are flawed and its speakers confused. Contemporary
noncognitivism is more conciliatory, attempting to win the right for many of these
seemingly cognitivist features for moral discourse. Simon Blackburn’s quasi-realist
program, in particular, aims to provide legitimacy for talk of moral truth, moral
belief, moral properties, moral assertions, and moral knowledge, all within a
noncognitivist framework (Blackburn 1993, 1998). Regarding moral knowledge,
for example, Blackburn claims that we know things when our confidence will
not be overturned. One’s disapproval of shoplifting, say, may not be as assured as

one’s disapproval of punching babies—after all, one may be willing to accept the
possibility that shoplifting may, by some unforeseen but reliable causal chain, result
in overall economic benefit, in which case one may be willing to countenance the
possibility of ceasing to disapprove of shoplifting. One is considerably less likely to
countenance this possibility for punching babies. Thus, Blackburn maintains, one
can claim to know that punching babies is wrong in a way that one does not claim to
know that shoplifting is wrong—though both associated moral judgments are mere
expressions of attitude rather than belief.
The potential instability of this quasi-realist program has been pointed out
several times (Dreier 2004; Lewis 2005), inasmuch that if quasi-realism is successful
it becomes difficult to differentiate the position from the cognitivist perspective
from which it is supposedly distinct. Perhaps the problem here arises from shifting
and imprecise characterizations of the term “noncognitivism.” If, for example, one
defines “noncognitivism” as the view that moral judgments are not truth-evaluable,
then obviously earning the right to the literal use of the full-blown truth predicate
for moral judgments is precluded. It would seem Blackburn himself has a somewhat
different understanding of the anti-realist stance for which he aims to earn the
trappings of realist talk, and on at least one occasion he has expressed a lack of
enthusiasm for “noncognitivism” as a label for his position; he tends to prefer the
title “projectivism.”8 It is clear, though, that if his quasi-realist program earns for
morality a claim to genuine knowledge, then the resulting position is no longer a
skeptical one (a result Blackburn will welcome).
3 ERROR THEORY
Error-theoretic moral skepticism is the thesis that while our moral discourse does
function assertorically (contra the noncognitivist’s view), the world lacks the
properties and relations that are necessary to render these assertions true. A good
analogy is atheism: the atheist denies that when someone utters, say, “God loves us”
the speaker is merely expressing his/her feelings; rather, religious speakers really are
making assertions about the world. (Thus the atheist is a cognitivist about theistic
discourse.) However (the atheist thinks), the world is not furnished with the items
necessary to make “God loves us” true; religious discourse suffers from a massive
and systematic error. Similarly, the moral error theorist thinks that, in order for
our moral judgments to be true, the world would need to provide properties like
goodness and badness, rightness and wrongness, evil and virtue—which it simply
does not. The error theorist doesn’t merely think that the world lacks objective
moral properties, but that it lacks subjective moral properties. (To press the analogy,
the atheist doesn’t merely hold that the gods fail to exist objectively, but that they
fail to exist subjectively, too.)
The error theory was coined and clearly articulated by John Mackie in 1977,
though one can find awareness of the viewpoint running back to the ancients.
Even Brutus is reported to have died with the despairing words “Wretched virtue;
718 SKEPTICISM
although you were nothing but a name, I practiced you as something real!”9 Several
centuries earlier, Anaxarchus espoused a moral view that has been interpreted as
an error theory (Warren 2002: 81).10 Generally speaking, although this form of
skepticism has had few staunch defenders, the fact that so much of moral philosophy
can be interpreted as striving to refute the view suggests a perennial awareness of
its possibility.
In order to approach the question of what considerations might lead one to
endorse a moral error theory, it is useful to reflect on what considerations might
lead one to endorse atheism. While a person might be persuaded by a careful
philosophical argument, such as the argument from evil, more likely would be a
general conviction that appeal to divine entities is explanatorily redundant, along
with some vague evocation of a principle of parsimony. This might be coupled
with a set of images about why religious belief emerged for sociological reasons, a
component of which might be the notion that humans are prone to gullibility when
it comes to theistic ideas. The atheist need not claim to have any demonstrative
falsification of theism (though, again, he might try to develop such an argument)—
need not claim to know that there are no gods—but nevertheless thinks that, on
balance, the evidence is in his favor sufficiently to support disbelief being accorded
high credence.
The moral error theorist takes a similar view of moral discourse. Perhaps he
reflects on the thousands of years of trying to make sense of moral facts and judges
the enterprise to have been a spectacular failure. Nothing in the world seems to
require the existence of moral facts (he thinks), which, coupled with a more-or-
less stringent and more-or-less precise principle of parsimony, counsels in favor of
disbelief. Perhaps this pessimistic conclusion is bolstered by a set of images about why
moral thinking might have evolved (at evolutionary, sociological, and psychological
levels), a component of which is the notion that humans are particularly prone to
this kind of gullibility.
The moral error theorist (like the atheist) faces challenges that may force
more careful defense. It goes without saying that many will find it an appallingly
counterintuitive and pernicious position. On the basis of its counterintuitiveness,
opponents might employ an epistemological principle according to which any
sufficiently implausible conclusion should be rejected in preference to overturning
any less implausible premises. Thus, any argument to the conclusion “There is
nothing morally wrong with punching babies” can be rejected out of hand, it will be
claimed, perhaps even without one’s being able to pinpoint its fault, since we know
that accepting the conclusion will require a greater violation of our intuitions than
rejecting any philosophical premise that might appear to entail it (see Enoch 2011).
The error theorist, though, is unlikely to accept this epistemological methodology,
and in any case the principle arguably fails to count against a counterintuitive
skeptical thesis if the advocate of the thesis also has at her disposal a plausible
hypothesis about why the skeptical thesis would seem counterintuitive even though
true (see Joyce 2014).
Other opponents will endeavor to provide a characterization of moral
properties about which the error theorist will find it difficult to locate grounds
for complaint. For the religion analog this is implausible. To the atheist someone
may say “But God is just love, and you believe in love, don’t you?”—to which
it is reasonable to reply adamantly that God is not love (not, at least, in a literal
sense involving the “is” of identity). That the concept God essentially involves
characteristics that the concept love does not is something one can pronounce as
assuredly as that the concept dog involves characteristics that the concept glove
does not. The same sort of reply can be maintained by the moral error theorist
but cannot be pressed with quite the analogous assurance. A hedonic utilitarian,
for example, might say “But moral goodness just is happiness, and you believe in
happiness, don’t you?”—to which the error theorist’s reply “No, moral goodness
is not happiness” is a response that will require rather more deliberation and
argument. In likelihood, what the error theorist will claim is that the concept
moral goodness essentially involves reference to a certain kind of practical
authority that the concept happiness does not (though it is open to the error
theorist to highlight some other difference). Mackie, for example, refers to the
work of rationalist Samuel Clarke, who in the early eighteenth century argued for
(in Mackie’s words) “necessary relations of fitness between situations and actions,
so that a situation would have a demand for such-and-such an action somehow
built into it” (1977: 40). Mackie cashes out this “intrinsic action-guiding” quality
by saying that it would provide “reasons . . . for doing or for not doing something
[that] are independent of that agent’s desires or purposes” (Mackie 1982: 115).
Such reasons, he thinks, can exist only in virtue of normative institutions, in the
same way as one’s reason to move a chess piece in a certain manner exists only
in virtue of some human-decreed system of rules. But moral rules, according to
Mackie, have their reason-giving quality objectively; we do not treat them as
norms of our invention, for to do so would rob them of their practical authority,
which is, arguably, their whole point.
The important elements of this argument are, first, that this “intrinsic-action-
guiding” quality is (according to Mackie) something that moral goodness has
essentially (such that any property lacking the quality simply wouldn’t upon reflection
be recognizable as moral goodness), and, second, that it is a quality that happiness
does not instantiate; therefore moral goodness cannot be happiness. Mackie goes
further, of course, and claims that this quality is instantiated by no actual property;
therefore the property of moral goodness is uninstantiated; therefore no claim of
the form “X is morally good” is true.11 (Whether this is taken to hold contingently
or necessarily is an interesting distinction that I won’t investigate here, except to say
that either will suffice for an error theory.)
The thorniest element of the argument is the conceptual claim that moral
goodness has this intrinsic-action-guiding quality essentially. It is open to the
error theorist’s opponent to acknowledge that many people have believed moral
properties to have this kind of objective authority, while still maintaining that it is
a dispensable component, conceptually speaking. Analogously, it was at one time
almost universally believed that gorillas are violent brutes, but the discovery that the
large primates living in Africa are in fact quite gentle and sociable creatures was not
taken to be the discovery that gorillas do not exist. “Violent brutishness” was never
720 SKEPTICISM
an essential part of the concept gorilla. By contrast, the realization that nobody is or
ever has been in cahoots with the devil was taken to imply that witches do not exist,
since having supernatural powers is an indispensable trait of witches, conceptually
speaking. So are moral concepts more like the concept gorilla or more like the
concept witch? Not only has it proven hard to say, it has proven hard to decide upon
a framework with which to address the question.
4 JUSTIFICATION SKEPTICISM
Let us now turn to the third form of modern moral skepticism, the thesis that moral
judgments are not justified. It is important first of all to establish that we mean
epistemic justification. Suppose a person, S, holds the belief that p on the basis of
good evidence, but would actually be better off, in terms of welfare (understood
however you like), believing that not-p. Then we might say that S’s believing that
p is instrumentally unjustified but, nevertheless, epistemically justified. Or suppose
that S believes that p on the basis of some terribly flawed process of reasoning, but
that having this belief is of great benefit to S. Then we might say that S’s believing
that p is epistemically unjustified but, nevertheless, instrumentally justified. Of
course, precisely what it takes for a belief to be epistemically justified is a matter of
great controversy. (Even my references to S’s being sensitive to the evidence will not
please everyone.)
Speaking generally, one person may be justified in believing that p while another
person is not justified in believing that p. Someone may lack justification for her
belief but at a later date gain justification, or have justification but at a later date
lose it. Nobody, I suppose, will object to the claim that some people sometimes lack
justification for some of their moral beliefs (assuming that moral judgments are
beliefs). Probably few will deny the somewhat bolder claim that everyone sometimes
lacks justification for some of their moral beliefs. Maybe we’ll be willing to accept
that there are some people who lack justification for all their moral beliefs. But the
interesting skeptical thesis is stronger than any of these: that we all lack justification
for all of our moral beliefs. And one can go stronger still, by maintaining that this
epistemically disheartening situation is permanent, that there is nothing any of us
can do to acquire justification for our moral beliefs. Call the penultimate thesis
“justification skepticism” and the final thesis “strong justification skepticism.” Both
views appear to prescribe a suspension of judgment upon learning that one’s moral
beliefs lack justification. In what follows, I will confine attention to two arguments
that might lead one to endorse justification skepticism, keeping an eye on whether
the arguments also support strong justification skepticism.
Let us begin by considering a person ordinarily lacking justification for one of
her everyday moral convictions. What might lead us to endorse this description of
her epistemic state? Perhaps we think that she hasn’t really considered this moral
matter in any depth but is simply voicing an opinion that she was brought up
with or that her peers all hold. Perhaps we think that she is biased or emotionally
invested in the matter in a way that distorts her judgment. (Witness the eagerness
manifested in people to punish someone when a deeply upsetting crime has been
committed.) Perhaps we are aware that other parties, who are no worse off in terms
of access to information or the capacity to deliberate, have come to a completely
different moral conclusion. In everyday contexts, any combination of these
findings might lead us to the conclusion that an individual’s moral conviction lacks
justification. The proponent of justification skepticism about morality maintains
that, upon reflection, one or more of these features that we ordinarily think of as
removing justification actually applies across the board, to everyone’s moral beliefs
at all times.
Focus, to begin with, on the case when a person’s moral judgment is merely a
matter of unreflectively voicing a view with which she was raised and by which she
is surrounded. Some will claim that this is not sufficient for a lack of justification,
since the moral view in question might nevertheless be an accurate one, picked up
by the epistemic community via reliable truth-tracking mechanisms and transmitted
to the person via reliable causal means; the fact that the individual herself cannot
articulate evidence for her moral view is neither here nor there. Our attention, then,
should move from the individual to her epistemic community: how likely is it that
the community is (or in the past was) reliably tracking moral facts?
At this point questions about the genealogy of moral thinking come to the fore—
not just at the individual or sociological level but potentially from an evolutionary
and biological perspective as well. In recent years, there has been much interest in
hypotheses of moral nativism: the thesis that human moral thinking has evolved as
a discrete psychological adaptation (see Krebs 2005; Joyce 2006; Mikhail 2011;
Boehm 2012). What is noteworthy about these hypotheses is that they explain
why moral thinking evolved in terms of its enhancing social cohesion among our
ancestors—something that is no less plausible for an error theorist than a moral
realist. In other words, moral nativism provides an explanation of why humans
engage in moral judgment in a manner that does not imply or presuppose that the
faculty in question was or is truth-tracking. (And, importantly, it is not a far-fetched-
but-unfalsifiable skeptical hypothesis, like Descartes’s demon, but a perfectly
respectable and plausible empirical theory.)
One might be tempted by an error-theoretic argument at this point: that because
moral facts are not needed to explain anything, not even our making moral
judgments, then one should, via a principle of parsimony, eliminate them from our
ontology. But this is not the argument with which we are concerned here. Rather, the
argument is that because our moral judgments can be explained without reference
to moral facts—because they flow from a non-truth-tracking mechanism—we have
reason to suspend judgment as to their truth. Whether the availability of moral
nativist hypotheses undermines the epistemic status of moral judgments in this
fashion is a matter of current debate (see Street 2006; White 2010; Brosnan 2011;
Joyce 2013).
A second argument for justification skepticism focuses on moral disagreement—
an argumentative schema familiar to classical skeptics. Agrippa’s first Mode
(diaphōnia) states that we must suspend judgment when there is undecided conflict
over a proposition, while several of Aenesidemus’s Ten Modes highlight conflicting
judgments among parties as the basis of Pyrrhonism. Modern epistemology
722 SKEPTICISM
continues to struggle with the question: what impact should disagreement have
on the epistemic status of our beliefs? According to what is called “the Equal
Weight View,” one’s confidence that p should fade upon encountering a dissenting
epistemic peer if one’s prior conditional credence that p is low, where “prior”
means prior to thinking things through and knowing what one’s peers think about
whether p, and “conditional” means conditional on what one has learned about
the circumstances of disagreement (see Elga 2007; also Feldman and Warfield
2010; Machuca 2013). In the case of moral matters, disagreement is ubiquitous.
Norm variation among cultures on topics like infanticide, cannibalism, and incest
are the most obvious and vivid examples to reach for; but one might instead
simply advert to the small number of dedicated moral error theorists, who (when
speaking carefully and sincerely) dissent to every moral claim to which anyone
else wishes to assent. After all, the empirical prevalence of disagreement may not
be an important factor; if it is merely possible that a rational agent with the same
information as you, with the same facilities of discernment and deliberation as
you, could come to believe that not-p, then this may be sufficient to undermine
the epistemic status of your belief that p. If only actual peer disagreement were to
count, then, as Lammenranta (2011: 211) notes, one could render one’s beliefs
justified by killing all actual dissenters (see also Kelly 2005; for criticism, see
Tersman 2013).12
I will end by briefly drawing attention to three features of these arguments for
justification skepticism.
First, they seem particular to the moral realm; they don’t fall out of a more
general epistemic skepticism. Even if human moral thinking is the output of a
mechanism that is non-truth-tracking, evolutionarily speaking, this need not be true
of other evolved psychological mechanisms. Many of the various faculties involved
in perception, for example, have surely evolved to track aspects of the world. (There
would, for instance, be no adaptive gain in seeing distant things as close, or vice
versa.) Also, moral disagreement seems to be different from many other forms of
disagreement. We will (arguably) countenance the possibility of two rational agents
with the same information and the same powers of deliberation disagreeing over a
moral matter in a way that we will not accept the possibility of their disagreeing over
an empirical matter.
The second thing to note about justification skepticism is that in counseling
suspension of judgment on moral matters it differs both from noncognitivist
skepticism—which allows one to carry on confidently making moral judgments—
and from error-theoretic skepticism—which seems to require one to deny moral
propositions. In this respect justification skepticism seems more akin to classical
versions of skepticism.
The final observation is that neither of the arguments outlined—from genealogy
and from disagreement—appears to contain the resources to establish strong
justification skepticism. If they succeed at all, they succeed in removing any benefit
of the doubt that we might otherwise have been inclined to accord our moral
judgments, but there doesn’t appear to be anything in these arguments to exclude
the possibility of justification being instated. Ancient skeptical arguments also

suffered from this limitation. Perhaps the skeptical Modes succeeded in showing
that our current beliefs lack justification, but the conclusion that this justification
is unattainable was more a matter of the rhetoric of the skeptics’ language than an
established thesis. Philo, for example, in On Drunkenness talks of disagreement
being “chronic and hopeless” (in the words of Annas 1998: 196), which does
suggest that the skeptical plight is envisaged to be a permanent human fixture.13 The
crucial phrase in Sextus Empiricus’s Outlines of Pyrrhonism I 165 (and elsewhere)
is translated variously as “unresolvable impasse” (Mates 1996) and “undecidable
dissension” (Annas and Barnes 2000). But the problem is that neither Philo,
Sextus, nor any other ancient philosopher possessed a solid argument to show
that disagreement must always have these qualities; they are just pessimistic. The
ancients, of course, lacked an awareness of the impressive menu of epistemological
theories now available—foundationalism, coherentism, reliabilism, and so forth—
all of which hold out hope of demonstrating that many of our moral beliefs are
justified.
Perhaps what these arguments for justification skepticism achieve is the overthrow
of the force of epistemic conservatism regarding moral beliefs: the idea that some
degree of warrant for moral beliefs comes from the mere fact that we already hold
them. The arguments for justification skepticism would shift the burden of proof
onto the shoulders of the non-skeptic, challenging her to articulate a sound and
defensible account of moral justification. The skeptic may then proceed to reject
these accounts as they are offered, thus showing that the initial skeptical challenge
remains unanswered. Sextus shows awareness of this strategy, since not only does
he use the Modes to throw down the skeptical gauntlet, so to speak, but he also
then backs this up with arguments aimed at exposing the failings of every available
positive theory.14 (An updated version of Sextus’s Against the Ethicists might include
“Against the Foundationalists,” “Against the Coherentists,” etc.) Such a strategy is
by no means doomed to failure, but its advocate can never declare victory, since
one never knows what vindicative epistemological theory the next generation might
devise. The justification skeptic may have to rest content with the open-ended
conclusion that until such a theory is enunciated and persuasively defended, a
suspension of judgment on moral matters is called for.15
NOTES
1. If discourse about X consists of assertions that are sometimes true, then minimal
realism is true of X. If, in addition, these assertions are true in virtue of the
obtaining of objective facts, then robust realism is true of X. Some people doubt
the distinction (Sayre-McCord 1986; Rosen 1994). Here I am privileging robust
realism.
2. The distinction as presented here is widely accepted though possibly not entirely
accurate. For more subtle discussion of the relation between Academic and Pyrrhonian
skepticism, see Part I of this book.
724 SKEPTICISM
3. There are complications here concerning what counts as a “moral proposition.” If S

claims to know that it is not the case that shoplifting is wrong, is this not a claim to
moral knowledge, and therefore is it not incompatible with S’s claim that there is no
such thing as moral knowledge? I do not think so. The error theorist can maintain that
moral propositions are all and only those the assertion of which commits one to the
existence of moral properties. “Shoplifting is morally wrong” is such a proposition,
but “It is not the case that shoplifting is wrong” is not. (“Wilma is a witch” commits
the speaker to the existence of witches; “It is not the case that Wilma is a witch” does
not.) The error theorist can therefore confidently claim to know that it is not the case
that shoplifting is wrong without contradicting the claim that there is no such thing
as moral knowledge. See Pigden (2010) for useful discussion of the nuances of how
the error theory should be described.
4. For views along these lines, see Sinnott-Armstrong (2006); Loeb (2008);
Joyce (2012).
5. I myself doubt that Hume should count in any straightforward way as a noncognitivist.
See Joyce (2010) for discussion.
6. See Satris (1987) for discussion of the nineteenth-century continental roots
of noncognitivism.
7. See Smith (2001) for discussion of some less conspicuous problems for noncognitivism.
8. Blackburn (1996) writes: “I for many years strenuously opposed the label ‘non-
cognitivist,’ and . . . finding the label appropriate is a prime symptom of failing
to stand by the advertised minimalism” (the minimalism referred to being a
central plank of the quasi-realist program). For discussion of the relation between
noncognitivism and projectivism, see Joyce (2009).
9. Reported by both Plutarch and Dio Cassius, though apparently Brutus (if he said
it at all) was paraphrasing something poetically attributed to Heracles. See Nauck
(1889: 910).
10. Though Julia Annas (1986) sensibly cautions against reading modern moral skeptical
concerns into ancient skeptical thinking.
11. Here I have limited attention to moral goodness, but of course the dialectic is
supposed to generalize to any moral property. One might complain that Mackie’s
references to the “intrinsic-action-guiding” quality of moral properties is more
suitable to a discussion of obligation than goodness. This, I think, is a fairly
superficial complaint, and the example could easily be altered to accommodate. See
Joyce (2001: 175–177) for some pertinent discussion.
12. We must again distinguish this epistemological argument from a superficially similar
argument for a moral error theory. Mackie claims that the best explanation of the
phenomenon of widespread intractable moral disagreement is that there are no
moral facts (1977: 36–38). The argument currently under discussion, by contrast,
does not rely on inference to the best explanation, and does not conclude that there
are no moral facts, merely that our beliefs about them lack justification.
13. The Philo mentioned here is Philo of Alexandria, not the Academic skeptic
Philo of Larissa. Still, in the passage cited Philo of Alexandria trenchantly builds
a case against dogmatism, employing arguments that clearly draw on the Greek
skeptical tradition.
14. On Sextus’s skepticism, see the chapter on him in Part I of this book. Cicero’s de
Finibus also contains arguments designed to refute each major moral theory in turn.
15. Many thanks to the editors of this book for a lot of sensible advice.
REFERENCES
Annas, Julia. 1986. “Doing without Objective Values: Ancient and Modern Strategies.” In
M. Schofield and G. Striker (eds.), The Norms of Nature: Studies in Hellenistic Ethics,
Annas, Julia and Jonathan Barnes. 2000. Sextus Empiricus: Outlines of Scepticism.
Ayer, Alfred J. 1936. Language, Truth and Logic. London: Victor Gollancz.
Blackburn, Simon. 1993. Essays in Quasi-Realism. Oxford: Oxford University Press.
Blackburn, Simon. 1996. “Commentary on Ronald Dworkins’ ‘Objectivity and Truth:
You’d Better Believe it,’ ” Brown Electronic Article Review Service in Moral and Political
Philosophy. Accessed August 04, 2014.
Blackburn, Simon. 1998. Ruling Passions. Oxford: Oxford University Press.
Boehm, Christopher. 2012. Moral Origins: The Evolution of Virtue, Altruism, and Shame.
New York: Basic Books.
Brosnan, Kevin. 2011. “Do the Evolutionary Origins of Our Moral Beliefs Undermine
Moral Knowledge?” Biology and Philosophy 26: 51–64.
Carnap, Rudolf. 1935. Philosophy and Logical Syntax. London: Kegan Paul, Trench,
Trubner & Co.
Dorr, Cian. 2002. “Non-cognitivism and Wishful Thinking,” Noûs 36: 97–103.
Dreier, Jamie. 2004. “Meta-ethics and the Problem of Creeping Minimalism,” Philosophical
Elga, Adam. 2007. “Reflection and Disagreement,” Noûs 41: 478–502.
Enoch, David. 2011. Taking Morality Seriously. Oxford: Oxford University Press.
Feldman, Richard and Ted Warfield (eds.). 2010. Disagreement. Oxford: Oxford University
Press.
Geach, Peter. 1965. “Assertion,” Philosophical Review 74: 449–465.
Joyce, Richard. 2001. The Myth of Morality. Cambridge: Cambridge University Press.
Joyce, Richard. 2006. The Evolution of Morality. Cambridge, MA: MIT Press.
Joyce, Richard. 2009. “Is Moral Projectivism Empirically Tractable?” Ethical Theory and
Moral Practice 12: 53–75.
Joyce, Richard. 2010. “Expressivism, Motivation Internalism, and Hume.” In C. Pigden
(ed.), Hume on Motivation and Virtue, 30–56. Basingstoke: Palgrave MacMillan.
Joyce, Richard. 2012. “Metaethical Pluralism: How Both Moral Naturalism and Moral
Skepticism may be Permissible Positions.” In S. Nuccetelli and G. Seay (eds.), Ethical
Naturalism: Current Debates, 89–109. Cambridge: Cambridge University Press.
Joyce, Richard. 2013. “Irrealism and the Genealogy of Morals,” Ratio 26: 351–372.
Joyce, Richard. 2014. “Taking Moral Skepticism Seriously,” Philosophical Studies 168:
843–851.
726 SKEPTICISM
Kelly, Thomas. 2005. “The Epistemic Significance of Disagreement,” Oxford Studies in

Epistemology 1: 167–196.
Krebs, Denis. 2005. “The Evolution of Morality.” In D. Buss (ed.), The Handbook of
Evolutionary Psychology, 747–771. Hoboken, NJ: John Wiley & Sons.
Lammenranta, Markus. 2011. “Skepticism and Disagreement.” In D. Machuca (ed.),
Pyrrhonism in Ancient, Modern, and Contemporary Philosophy, 203–216. Dordrecht:
Springer.
Lewis, David. 2005. “Quasi-Realism Is Fictionalism.” In M. Kalderon (ed.), Fictionalism in
Metaphysics, 314–321. Oxford: Clarendon.
Loeb, Don. 2008. “Moral Incoherentism: How to Pull a Metaphysical Rabbit out of a
Semantic Hat.” In W. Sinnott-Armstrong (ed.), Moral Psychology, Volume 2: The
Cognitive Science of Morality: Intuition and Diversity, 355–385. Cambridge, MA: MIT
Press.
Mackie, John L. 1977. Ethics: Inventing Right and Wrong. London: Penguin.
Mackie, John L. 1982. The Miracle of Theism. Oxford: Oxford University Press.
Mates, Benson. 1996. The Skeptic Way: Sextus Empiricus’s Outlines of Pyrrhonism. Oxford:
Mikhail, John. 2011. Elements of Moral Cognition: Rawls’ Linguistic Analogy and the
Cognitive Science of Moral and Legal Judgment. Cambridge: Cambridge University
Press.
Nauck, August. 1889. Tragicorum Graecorum Fragmenta. Leipzig: Teubner.
Pigden, Charles. 2010. “Nihilism, Nietzsche, and the Doppelganger Problem.” In R. Joyce
and S. Kirchin (eds.), A World Without Values: Essays on John Mackie’s Moral Error
Theory, 17–34. Dordrecht: Springer.
Rosen, Gideon. 1994. “Objectivity and Modern Idealism: What is the Question?” In
M. Michael and J. O’Leary-Hawthorne (eds.), Philosophy in Mind, 277–319. Dordrecht:
Kluwer.
Satris, Stephen. 1987. Ethical Emotivism. Dordrecht: Martinus Nijhoff.
Sayre-McCord, Geoffrey. 1986. “The Many Moral Realisms,” Southern Journal of
Philosophy, suppl. vol. 24: 1–22.
Sinnott-Armstrong, Walter. 2006. Moral Skepticisms. Oxford: Oxford University Press.
Smith, Michael. 2001. “Some not-much-Discussed Problems for Non-cognitivism in
Ethics,” Ratio 14: 93–115.
Stevenson, Charles L. 1937. “The Emotive Meaning of Ethical Terms,” Mind 46: 14–31.
Street, Sharon. 2006. “A Darwinian Dilemma for Realist Theories of Value,” Philosophical
Studies 127: 109–166.
Tersman, Folke. 2013. “Moral Disagreement: Actual vs. Possible.” In D. Machuca (ed.),
Disagreement and Skepticism, 90–108. New York: Routledge.
Warren, James. 2002. Epicurus and Democritean Ethics: An Archaeology of Ataraxia.
White, Roger. 2010. “You Just Believe That Because. . .,” Philosophical Perspectives
24: 573–615.
CHAPTER FIFTY
Religious Skepticism
J. L. SCHELLENBERG
1 WHAT IS RELIGIOUS SKEPTICISM?

Religious skepticism is often understood fairly loosely today. Both in academic
and nonacademic contexts, pretty much anyone who is at all negative about
religion, or who questions it in any way, may be seen as having an attitude that fits
under this label. In the present discussion I adopt a narrower and, I hope, more
illuminating understanding of the term. On this understanding, the term “religious
skepticism” allows us to pick out certain specific attitudes in relation to religion—
obviously associated with skepticism in the history of philosophy and deserving of
discriminating attention—that are associated with being in doubt.
What it is to be in doubt is best understood in relation to belief: ordinary
usage and philosophical discussion alike suggest that someone is in doubt about a
proposition when, and only when, having considered the proposition, she believes
neither it nor its denial.1 (Note that being in doubt about a proposition is not the
same as doubting that a proposition is true, which would usually be taken to entail
or bring one close to belief of its denial.) Thus Suneetra, who on reflection does
not believe that the soul is reincarnated but does not disbelieve this either, is to
be regarded as being in doubt about whether the soul is reincarnated. But is this
the only sort of skepticism to be associated with doubt? I think not. For example,
consider Sam, who finds himself with the belief that there is a God but thinks doubt
would be more appropriate because of the force of certain arguments his desires
strongly resist. Sam may take it upon himself to actively question theistic belief
over a period of time, hoping thereby eventually to do justice to the arguments,
psychologically speaking. And it would be odd to deny that Sam is being skeptical
about the claim that God exists as he does so.
We can deal with this fact by distinguishing between active skepticism and passive
skepticism, saying that Sam exemplifies the former but seeks to reach the latter by,
so to speak, helping the relevant arguments induce a state of doubt in him. (He is
forced to do so because the state of being in doubt, like belief, is fundamentally
involuntary rather than voluntary.)2 This distinction between passive and active
skepticism also neatly permits us to accommodate the commonly used phrase
“suspending judgment,” which is often linked to skepticism but sounds like it refers
728 SKEPTICISM
to something one does—something active. One may voluntarily suspend judgment

about a proposition by refusing to assert that the proposition is true or that it is
false, and resisting inner affirmations on either side too, and this precisely is what
one does when one exemplifies active skepticism.3
So we have passive skepticism, which is involuntarily being in doubt, and active
skepticism, which means voluntarily identifying with doubt in the way I have
described. These ways of being skeptical might of course arise in relation to many
different subjects of inquiry. So we have to ask: what makes either form of skepticism
religious? There are various possibilities here, the most obvious of which are (as my
examples suggest) tied to religious propositions. But this point only invites a further
question: what makes a proposition religious?
Unfortunately, there is no standard answer to this question about the nature of
religion; the matter is controversial. But, pretty clearly, restricting religiousness to
behavior tied up with views about God or gods would be overly narrow: though
they may be in the minority, there evidently are in the world today nontheistic and
nonpersonalistic forms of religion (they can be found, for example, in Hinduism,
Buddhism, and Taoism). Having said that, it is understandable if philosophers
find most interesting those forms of religion that have clear and distinctive non-
naturalistic metaphysical commitments of the sort that belief in a personal god
exemplifies. (Metaphysical naturalism is the view that concrete reality is a single
unified system of natural law.) To deal with this issue here I will assume for the
purpose of our discussion, which is primarily philosophical, that religion is bound
up with the notion that metaphysical naturalism is false. Religious realities, if such
there should be, would in other words be transcendent realities—transcendent
of the natural world. It seems clear that religious people also greatly value the
transcendent toward which they think of themselves as oriented. These metaphysical
and axiological points I think we may safely assume, in a philosophical context. But
to ensure wide relevance for what I say, I will not assume any more than what
they contain.
With that as background, we may continue elucidating religious skepticism by
reference to religious propositions, now taken as claiming or entailing that there is a
religious reality, thus understood. One might say that being skeptical about even one
religious proposition is enough to instantiate religious skepticism. And so on this
option, Suneetra, even if the proposition I mentioned before is the only religious
proposition about which she is in doubt, exhibits religious skepticism. But to make
our discussion of religious skepticism philosophically interesting, we need to construe
the latter as a general stance on religion, not a particular one of the sort exhibited
by Suneetra. Even if Suneetra is in doubt about the reincarnationist proposition, she
may believe many other religious propositions, such as the propositions propounded
by the Catholic Church, to which she has just converted. And it seems implausible
to say she exhibits religious skepticism of any philosophically interesting sort in that
case, for she is then a theist and a Christian! Notice that we are helped here by the
fact that someone who, in a philosophical context, exhibits religious skepticism we
should be able to call a religious skeptic. It would certainly be odd to apply this label
to Suneetra if she is well on her way toward being a full-fledged Catholic.
RELIGIOUS SKEPTICISM 729
So when would it not be odd? To see when, we need to make a shift from
the particular to the general. Suneetra is skeptical about whether one specific
religious proposition is true. The religious skeptic, properly so-called, is skeptical
about whether any religious proposition is true. This is what is required to have
an alternative both to the religious believer, who thinks she has found at least one
religious proposition that merits a more positive response than skepticism (there
are of course innumerable examples of this stance, including many in contemporary
philosophy of religion), and to a certain general sort of religious disbeliever,
who thinks that religion merits, across the board, a more negative response than
skepticism affords, believing all religious propositions to be false (metaphysical
naturalists, whose view is presently in the ascendancy in philosophy more generally,
afford examples here). The religious believer holds that there is truth in religion, the
religious disbeliever holds that there isn’t, and the religious skeptic is unsure.
One must be careful about the details here. Being skeptical about whether any
religious proposition is true is not the same as being skeptical about each religious
proposition; rather many of them—take, for example, the propositions of ancient
Greek religion—are clearly false, and contemporary religious skeptics usually see
this. Nor is it the same as being skeptical about the conjunction of all religious
propositions, which brings them together, linking each to another using the logical
connective “and.” It’s not hard to see that that proposition, properly understood,
will not inspire doubt but rather disbelief, since, given religious disagreement, it
entails innumerable contradictions and must be false. Instead, what we have is
something that, if made precise and clarified in the skeptic’s mind, would amount
to skepticism about the disjunction of religious propositions (also an enormous
proposition including all religious propositions, but this time one that links each to
another with the logical connective “or”). The religious skeptic typically sees a huge
number of religious possibilities, some false, others unknown and so unassessable,
and still others assessable but allowing for no verdict one way or the other. She is
unsure as to whether any of these possibilities is realized: perhaps (at least) one is,
perhaps not.
Various states of mind seem to represent a sufficient condition for being religiously
skeptical in this sense. Consider the proposition “Some religious proposition is true.”
It might not seem itself to be a religious proposition (we would not inappropriately
call it meta-religious). But being skeptical about that proposition should surely
count as a sufficient condition of being a religious skeptic in the desired general
sense, since that proposition is logically equivalent to the disjunctive proposition
mentioned above. For the same reason, so should being skeptical about “There is
a religious reality” (where a religious reality is one whose being realized would be
sufficient to make some religious proposition true). Another sufficient condition
for being skeptical about religion in the desired general sense can be detected in
skepticism about a particular religious proposition which the skeptic regards as
being such as must be true if any religious proposition is true. For many in the West
today, who suffer from shrunken imaginations about what’s possible, that would
be theism, that is, “God exists” (where by “God” is meant a broadly person-like
ultimate divinity). And here we also bump into the term “agnosticism.” Purely at the
730 SKEPTICISM
level of attitude-type, there is evidently a close relationship between “skepticism,”

as we are using the word, and “agnosticism.” But the latter term also has for us,
these days, a strong religious connotation, associated with theism. The agnostic most
often addressed in Western discussions of religion is skeptical about whether there
is a God. And so we may think it is Sam, rather than Suneetra, who exemplifies
religious skepticism, properly construed. And Sam might be a religious skeptic if he
thinks theism (or propositions entailing theism) to represent the only “live” option,
religiously speaking. Many religious skeptics of recent years appear to be agnostics
about theism who make the latter assumption.
Let’s call the general sort of religious skepticism just delineated categorical
religious skepticism. Quite a number of individuals in the history of philosophy
give the appearance of being categorical religious skeptics. In ancient Greece, we
find Sextus Empiricus arguing that “it is inapprehensible whether there are gods”
(PH III 11, trans. in Annas et al. 2000). Moving forward in time we find similar
sentiments in thinkers such as Michel de Montaigne and Pierre Bayle. Ironically,
some of these individuals seem more in tune with relevant cultural facts and
conceptual possibilities than many of us in the twenty-first century, noticing that a
divine reality might have lineaments very different from those of any god or indeed
any reality with which we are acquainted—this theme may be found, for example, in
Montaigne’s “Apologie de Raimond Sebond,” the longest chapter of his famous Les
Essais (Montaigne 1981). Even Bertrand Russell, in the twentieth century, was able
to distinguish between the traditional theistic idea of an all-good and all-powerful
personal deity and another sort of “God” not so defined or restricted, declaring
himself a disbeliever about the former but only an agnostic or skeptic about the
latter: “That sort of God is, I think, not one that can actually be disproved, as I
think the omnipotent and benevolent creator can” (1996: 258). One of my aims in
this chapter is to push us back toward this sensitive and sensible perspective, which
enables the acknowledging of a wide range of religious possibilities.
Something that will help us get there is the recognition that a further way of
being a religious skeptic, related to but distinct from categorical religious skepticism,
remains to be identified. It is perhaps a less obvious sort of religious skepticism,
one that does not get its religious character from being straightforwardly tied to
religious propositions. I call it religious capacity skepticism. This sort of skepticism
involves a skeptical attitude in response to the proposition that we have (sometimes
that we have now, sometimes that we ever will have) the capacities required to
discern religious truth, if such there should be.
Though certainly to be found among a few contemporary religious skeptics,
particularly in the work of Schellenberg (2007), such skepticism is also present in
classical writers—for example, in the work of David Hume, where it can be seen
as one aspect or application of his famous “mitigated skepticism,” which urges us
to limit our thinking to the matters that “are best adapted to the narrow limits
of human understanding” (1975: 162). Hume, like Russell after him, realizes that
arguments against specific religious claims (perhaps including traditional theism)
may be successful in justifying disbelief of certain such claims even if we cannot rule
out that some religious claim—perhaps one of whose content we cannot even form
a clear conception—is true and so must, in general terms, remain religious skeptics.
As can be seen here, capacity skepticism is closely related to categorical skepticism.
The former sort of skepticism can provide one with justification for the latter. After
all, if there may be religious propositions that we are incapable of formulating or
incapable of assessing (perhaps because the relevant evidence is at least for now
beyond our grasp), how are we in a position to rule out that one of them is true?
Or that some such proposition is true and incompatible with religious propositions
known to us, which we are inclined to believe? Religious capacity skepticism will
therefore lead, for those in its grip who understand the state at all, to categorical
religious skepticism.
So we have categorical religious skepticism, which is active and/or passive
skepticism about the proposition that there is truth in religion (or some logically
equivalent proposition), and religious capacity skepticism, which means being
actively and/or passively skeptical about the proposition that human beings possess
the capacities required to discern truth in religion, if such there should be. And
although the former can exist without the latter, they may also go together, with
the former grounded in the latter. A final point we should make note of here is that
where religious skepticism occurs or is discussed in philosophy, we may expect it
to be associated not only with such attitudes but with an epistemological claim: the
claim that religious skepticism is intellectually preferable to both religious belief
and a general religious disbelief of the sort I described earlier in this section. This
claim need not be interpreted in such a way as to imply or suggest that religious
believers and disbelievers have something wrong with them personally, some
defect of intellectual character, perhaps, or the failure to fulfill intellectual duties
of a sort that philosophers often have in mind when they speak of someone being
unreasonable or unjustified in her belief. Rather, what we should think of here is
a proposal as to what is objectively the case—what reason actually best supports
in the way of a response to religious propositions. The religious skeptic, naturally
enough, proposes that the best response is going to be religiously skeptical. I shall
now outline what I regard as the strongest case for this proposal—a case I regard as
being successful.4
2 A SCIENTIFIC CASE FOR RELIGIOUS SKEPTICISM

The case begins with points that have long been emphasized by religious skeptics,
having to do with human limitations. But its true force emerges most clearly when
we add to these certain further points, allowed and indeed encouraged by modern
science, concerning human immaturity.
We have seen already that Hume emphasizes the first sort of point. So we can
take him as an example of someone who would push that part of my case. Human
beings, on this view, have cognitive powers that are severely circumscribed, and are
prone to cognitive errors. Though this is not where they are most fruitfully brought
into play, here too scientific developments can help to buttress the case. Cognitive
psychology has in recent years, in part through the rich ingenuity of its experiments,
revealed many surprising biases in the workings of human minds (Kahneman 2011).
732 SKEPTICISM
More generally, what we have learned about the vicissitudes through which humans
have emerged in evolutionary history, the particular environments to which they
have contingently become adapted, has led some philosophers—notably Chomsky
(1988), whose view is developed by McGinn (1993)—to surmise that the human
brain is quite good at some cognitive tasks, such as those of science itself, and not
so good or perhaps even really bad at others, such as the general and fundamental
problems addressed by philosophy. Now the intellectual questions referenced by
McGinn are of the same sort—general and fundamental—that religious belief and
disbelief purport to answer. And it would not be surprising if religion were an area
of human life that particularly invited cognitive bias. So humans could, by virtue of
the contingent and limited cognitive structures they have to work with, be prevented
from learning whether there are any realities beyond the natural world studied by
science or what such realities—should there be any—would be like. It might of
course be hard to be sure, in any rationally approvable manner, whether we have
been thus prevented. But our limitations may nonetheless be noticeable enough to
support being in doubt about whether this is not the case. What we have here is an
argument for capacity skepticism that undergirds categorical religious skepticism.
Now to such reasoning the religious believer may be inclined to reply in terms of
revelation, perhaps a revelation from God. Precisely because of our human inability
to find God on our own, so it may be said, God reveals Himself to us through
scriptures or human authorities or special experiences or miracles. Or perhaps God
is revealed in the very tendencies of our minds, as exposed by science: the cognitive
science of religion (CSR) has recently documented how “natural” it is for humans
to think of the divine in personal, agential terms, and to believe in God or gods
(Guthrie 1993; Barrett 2004). The religious disbeliever, for her part, may scoff at
doubt, arguing that no singular ability is needed to tell that there obviously is no
God and gods, or insisting that science is so much more impressive, intellectually,
than religion as to justify consigning the latter to the dustbin of history.
But even in their common use of male personal pronouns—such as “Himself”—
for a transcendent spiritual reality, the influence on believers of human limitations
may be observed! And the religious skeptic will wonder whether what CSR has
revealed may not be precisely the sort of limitation she is talking about: being
constitutionally such as to be specially attracted to what may be a very small subset
of religious possibilities. It is in any case question-begging to speak of revelation in
response to an argument purveying doubt as to the accessibility to us of a divine
reality: how can we have any confidence that we would understand the information
that needs to be delivered to us?
The disbeliever’s argument is no better. The various god ideas may only scratch
the surface when it comes to transcendent possibilities. Just the likeness to us of God
and gods should already make any serious religious adventurer ready to move on
in the quest for what, if anything, lies beyond ourselves. (Here too the CSR results
might be illuminating, and in an unexpected way, helping to explain why critics
of religion such as Daniel Dennett (2006: 9) seem just as onesidedly focused on
personalist religious ideas as believers are.) By the same token, the disbeliever subtly
begs the question too. For if our limitations are as serious as the religious skeptic
suggests, mightn’t there very well be a religious reality even if human religion hasn’t
found it? The very unimpressiveness of human religion (supposing the disbeliever
to be right about that) should lead the disbeliever to wonder whether its ideas are,
as she assumes, representative of all there is in the domain of religious possibilities.
What both religious believers and religious disbelievers typically overlook is, as
I have suggested, what Montaigne and Hume and Russell saw: that there might be
innumerable possibilities when it comes to alternative ways of construing reality
beyond nature. Human limitations may lead us to miss one or another or various
sets of such possibilities, or large swaths of evidence pertaining thereto. The relevant
information might be such as we are capable of recognizing, but nonetheless
overlooked, neglected, or inaccessible (by which I mean that the particular
directions of our thinking conspire to obscure it from us). It might also be such as we
are incapable of recognizing: undiscovered or undiscoverable. The importance of
these five ways of missing relevant evidence is itself something we tend to overlook
(Schellenberg 2007).
So much for the Limitation Mode of religious skepticism. Now it’s time to see
how adding the Immaturity Mode allows us to take the case for religious skepticism
to a whole new level. The Limitation Mode is about how we are constituted as finite
beings and the constraints this imposes on investigation into the ultimate and infinite
or at any rate the transcendent. It is not about our stage of development as a species,
and indeed it might inspire doubt at any stage of development (even if some of its
concerns could to some degree be mitigated at higher levels of finite development).
The Immaturity Mode, on the other hand, is all about development—or rather a
(relative) lack thereof. The Limitation Mode argues that human beings, given their
finite constitution, may lack the capacities required for religious or irreligious insight.
The Immaturity Mode argues similarly but with a twist: even if finite beings could
have the relevant capacities, we may not yet have them (Schellenberg 2007; 2013).
“Immature” can be understood in two senses, both quite familiar to us. The first
is illustrated by Theresa, who is called an immature chess player because she has a
tendency to become angry and overturn the chess board. Todd too might be called
immature at chess, but that’s because he just started playing chess six months ago.
The first sort of immaturity might be called normative, the second descriptive.
Now what’s interesting is that the human species, in respect of thinking about
religious matters, is presently immature in both senses. In matters of religious
investigation we have surely done less well than might have been hoped with
the time already allotted to us, but it is also true that—at least when scientific
timescales are brought to bear—we must be regarded as having just got started. To
see the former point, we need only consider the fact that credulous belief, dogmatic
preaching and teaching, and fighting and killing in the name of God or gods are rife
in religion as we’ve known it so far. This normative immaturity is a familiar fact, so
I will not tarry in its precincts. The more interesting and also far less familiar fact is
the descriptive fact about our place in time—one that we will count as representing
a very early stage of development both for human beings and for intelligent life on
our planet when we have reached a proper understanding of scientific time and
internalized it.
734 SKEPTICISM
Clearly we have not yet internalized it. Indeed, this is part of our immaturity. In
our culture we are constantly bombarded with scientific results, but our awareness
and internalization of deep scientific time is still incomplete. It is true that we have
begun to adapt to the fact of a deep past. Daily we are reminded that we come at the
end of a long evolutionary process. But there is little to stimulate the thought that
we come at the very beginning of one too, and that we humans may have inscribed
in the book of intelligent life only its first few lines.
To be thus stimulated, we need to meditate on such uncontroversial scientific
results as the following. Our planet has supported life for about 3.8 billion years.
But only about 5 million years ago did the human lineage diverge from the ape
lineage. You’d have to travel almost all the way through that period, stopping just
100,000 or 50,000 years short of the present, to finally meet beings a lot like us,
who are clearly capable of practicing some form of religion.5 What’s more, you’d
have to make it almost all the way through that 50,000-year span of time to arrive
at the forms of religion, including Christianity, that preoccupy us today.
Now contrast all this, the human religious story of about the last 50,000 years,
still touched by the violence that is in our genes and conspicuously heavy at our
end, with the fact that, according to science, although the sun will eventually scorch
our planet, Earth may remain habitable for as long as another billion years more
(Schroder and Smith 2008). What this means is that the human story, so far, appears
in about the first twenty-thousandth of the portion of deep time that may feature
intelligence on Earth! Long story short: religious experience and inquiry might only
be getting started. When we refer to the twenty-first century, we should think of this
as indicating how very brief and flickering has been the existence of religion so far,
rather than (as is more usual) taking it as grounding claims of ancient venerability.
So what can the religious skeptic do with such facts and with the new sensibility
sponsored by an ability to convert from human to scientific time? Well, there are
consequences for both religious belief and religious disbelief. To religious believers
the skeptic says this: perhaps religious experiences of the future will, because of
much greater future intellectual and spiritual capacities, reveal a transcendent reality
with defining characteristics quite different from those of any such reality we have
yet conceived. To the disbeliever she may of course say the same, but here she may
also put her point directly in opposition to what religious disbelievers commonly
hold concerning their argumentative prowess: for all we know, what we have ruled
out so far in the way of religious ideas, however impressively we may have done
so, is only what beings at a very early stage of thought about such things have been
able to come up with. So let the human species develop some of the more advanced
capacities that may in the future come with a larger brain, or with the overcoming
of violent tendencies, and think then about what the evidence shows. Believing now
is premature!
Perhaps it will be thought that this reasoning is rather too permissive. Doesn’t
it have the unacceptable consequence that advocates of many views we should
confidently dismiss—say, views favoring alchemy—can now propose the possibility
of improved future results and rightly ask us to retreat from disbelief to skepticism?
But at least two conditions necessary for justified dismissal are realized in relation
to such views: the available evidence is against them, and there has been a serious
attempt to ensure that no relevant evidence has been overlooked or neglected.
And even if these same conditions are satisfied for this or that specific religious
claim—say, the theistic claim that there is a person-like God—arguably neither is
satisfied for the very general claim that the religious skeptic is reluctant to dismiss
as false: the claim that in one way or another there is a religious reality beyond
nature. That this is so one might indeed regard as following from the arguments
I have made. Therefore, even if there are other and nonreligious views that should
rightly take comfort from those arguments, they do not include views we ought to
confidently dismiss.
Another objection to my argument says that fallibilism accommodates it well
enough. Fallibilism is a position in epistemology according to which those who
believe a proposition true or who believe it false should (at any rate for many
propositions) admit that they could be wrong, and could be shown to be wrong in
the future, but may believe meanwhile in response to what the available evidence
supports. This view, it seems to me, is overly sanguine, at least in the present context.
Fallibilism makes a “one size fits all” sort of concession to human limitations, refusing
expressions of certainty (“I’m not certain I’m right” is what “I could be wrong”
comes to here). But it lacks the sensitive discriminating quality required to do justice
to the specific sorts of limitations and especially to the immaturities we have found
to be at work in the religious case, which call us to take more seriously what the
unavailable evidence might show. Especially in light of deep time, and recognizing
our very early stage of development, we have to say that on the profound and
general and relatively unexamined issues concerning which, on the basis of available
evidence, religious believers and disbelievers claim insight, the evidence we don’t see
might be vast. And there is no justification for the belief that it points to religion or
for the belief that it points away.
Of course, the state of being in doubt that I have called passive religious skepticism
may not immediately be induced by reflection on such considerations as I have
marshalled here. I think the religious skeptic will quite sensibly prescribe extended
contemplation of our immaturity, both normative and descriptive, for all of us. For
those who see that doubt is appropriate but are prevented by the involuntariness of
belief, this will of course amount to a prescription of active skepticism.
But the religious skeptic has another move up her sleeve—the most powerful
one, which I have left for last. The key here is to see how, although either the
Limitation Mode or the Immaturity Mode might, on its own, be seen as justifying
religious skepticism, these two can also be combined in a certain way to produce an
even more impressive result.
This is how to produce it. Strategically back away from the claim that the
Limitation Mode on its own justifies religious skepticism. Instead, use the points
made there to support just the following unexceptionable claim: that the very least
we can conclude from our limitations is that an extended process of high-quality
religious inquiry instantiated in our midst would be required to justify religious belief
or disbelief for human beings. (The principle here is the same as the one preventing
us from expecting correct answers on a test from a child who has little aptitude for
736 SKEPTICISM
a subject and moreover hasn’t studied.) The clinching move will perhaps be obvious.
For surely the very least we can conclude from our immaturity is that we have not
yet engaged in such inquiry. The human species is both profoundly limited and
profoundly immature. Thus, even more obviously than before, religious belief and
religious disbelief are shown to be intellectually unjustified for beings such as we are.
3 CONCLUSION
What we should conclude, then, is that against the backdrop of a more subtle and
adequate understanding of just what religious skepticism is, and with vital help from
science, the reasoning implicitly or incipiently found in Montaigne, Hume, Russell,
and like-minded others can be brought into the light of day and the religious
skeptic’s proposal wins powerful intellectual support. In our culture, science is often
taken as supporting metaphysical naturalism. But deep time is a wake-up call. The
thinking about human immaturity it inspires and reinforces shows that what science
actually supports is doubt about whether naturalism is true and also about whether
any religious proposition is true. What science supports is religious skepticism.6
NOTES
1. What this means is that skepticism here is not simply or specifically knowledge-
skepticism, which could exist even in someone who happily believes a proposition p
and thinks she is justified in doing so (someone in this condition might still wonder
whether she knows that p). Of course, someone skeptical in my sense will not claim
to know either of the propositions involved in her skepticism if she recognizes that
knowledge entails belief.
2. It should be noted that the “passivity” involved here reflects not the absence of
such actions as are involved in active skepticism (in which case passive and active
skepticism would be mutually exclusive) but rather the involuntariness of doubt just
mentioned.
3. Of course, as my descriptions already suggest, one could be both actively and passively
skeptical (even after he reaches doubt, Sam may engage in active skepticism to keep
himself in that subjective condition). These two attitudes are not mutually exclusive.
But it is also important to note that they don’t have to go together, and in particular to
recognize that active skepticism may occur without passive.
4. The case I here present in outline form is a somewhat revised version of the case
detailed in Schellenberg (2007).
5. For an authoritative survey of recent scientific thinking about these early periods, see
Wade (2006). The behavioral changes I’ve alluded to are visible in the archaeological
record of about 50,000 years ago, and recent genetic evidence suggests that they may
have been spurred by the advent of fully articulate language—something that could
be required for anything more than incipient religiousness.
6. Many thanks to the editors of this book for their helpful comments on a draft of this
chapter.
REFERENCES
Annas, Julia and Jonathan Barnes (ed. and trans.). 2000. Sextus Empiricus: Outlines of
Scepticism. Cambridge: Cambridge University Press.
Barrett, Justin. 2004. Why Would Anyone Believe in God? Lanham, MD: AltaMira Press.
Chomsky, Noam. 1988. Language and Problems of Knowledge. Cambridge, MA: MIT
Press.
Dennett, Daniel. 2006. Breaking the Spell: Religion as a Natural Phenomenon. New York:
Viking.
Guthrie, Stewart. 1993. Faces in the Clouds: A New Theory of Religion. Oxford: Oxford
University Press.
Hume, David. 1975. Enquiry concerning Human Understanding, in Enquiries concerning
Human Understanding and concerning the Principles of Morals, edited by L. A. Selby-
Bigge, 3rd ed. revised by P. H. Nidditch. Originally published in 1748. Oxford:
Clarendon Press.
Kahneman, Daniel. 2011. Thinking, Fast and Slow. New York: Farrar, Straus, and Girioux.
McGinn, Colin. 1993. Problems in Philosophy: The Limits of Inquiry. Oxford: Blackwell.
Montaigne, Michel de. 1981. Les Essais. Originally published in 1580. New York: Georg
Olms, Hildesheim.
Russell, Bertrand. 1996. “The Existence and Nature of God.” In J. G. Slater (ed.),
The Collected Papers of Bertrand Russell, Vol. 10, 253–270. New York: Routledge.
Schellenberg, J. L. 2007. The Wisdom to Doubt: A Justification of Religious Skepticism.
Ithaca: Cornell University Press.
Schellenberg, J. L. 2013. Evolutionary Religion. Oxford: Oxford University Press.
Schroder, K.-P. and Robert Connon Smith. 2008. “Distant Future of the Sun and Earth
Revisited.” Monthly Notices of the Royal Astronomical Society 386: 155–163.
Wade, Nicholas. 2006. Before the Dawn: Recovering the Lost History of Our Ancestors.
New York: Penguin Books.
NAME INDEX
The indexes have been compiled by D. Machuca.
Aenesidemus 3, 5–6, 8, 25–6, 67–78, 114, Blackburn, Simon 716–17

116, 120, 126, 166 Boethius 188–9, 263
Agrippa 6, 71, 141, 431 Bonjour, Laurence 524, 540–1, 544, 571,
Alexander the Great 24, 145 594, 606 n.21, 638, 671, 676
al-Ghāzalī, Abū Ḥāmid Muḥammad xiv–xv, Brueckner, Anthony xv, 528 n.3, 636,
169, 196–207 663 n.18
Alston, William 395, 401, 597, 604, 676 Brunschwig, Jacques 29, 118
Anaxagoras 38 Buridan, John 166–8, 209–10, 213,
Antiochus of Ascalon 6, 67, 70, 82–3, 215–18, 226–7, 229
85–6, 89, 90 n.10, 91 n.22, 122 n.2, Burnyeat, Myles 22 n.18, 61 n.17, 132–7,
324, 326, 331–2 249, 266, 381–3
Antipater 117–18
Arcesilaus of Pitane 3–4, 6, 8, 36–46, Carnap, Rudolf 454, 716
51, 61 n.18, 67, 69, 81, 83, 85–7, 90 Carneades of Cyrene 3–4, 6, 8, 40, 43, 46,
nn.8–9, 114–17, 119, 268 n.5, 320, 51–60, 81–4, 109, 115–18, 178, 190,
324–6, 330, 332–3, 355, 357, 444 254, 284, 325–6, 330, 332–3
Aristippus 16, 21 n.8 Carroll, Lewis 543–4
Aristo of Chios 37 Celsus 7
Aristocles of Messene 16, 19, 24–8, 32–3, Charmadas 83, 88, 90 n.10, 91 n.19
34 n.6, 69, 72–3 Charron, Pierre xiv–xv, 170–1, 247–57,
Aristotle 18, 25, 31, 45, 86, 130–1, 186– 287–8, 290, 296, 320–1, 509–13
90, 192–3, 196, 198–9, 207, 219 n.11, Chisholm, Roderick 395, 550–4, 570–
226, 237, 242, 263–4, 267, 296, 512 1, 593
Armstrong, David 530 n.21, 637 Chrysippus 38–9, 43, 47 n.13, 48 n.14,
Augustine of Hippo 81, 84, 87, 89, 166, 49 n.33, 117–18, 130–1, 358
170, 175–83, 186, 210–11, 218 n.5, Cicero, Marcus Tullius xiii–xv, 3, 6,
235, 268 n.4, 275, 296, 336, 361 8, 36–7, 39–41, 53, 55–6, 58–9,
Aulus Gellius 8, 121 81–7, 89, 93–9, 115, 119, 121,
Austin, J. L. 503 n.21, 704 166–7, 169–70, 175–7, 186, 188,
Avicenna 198–9, 203, 207 190, 196, 235, 240, 255, 296,
Ayer, A. J. 673, 703, 705, 708, 716 320–1, 325
Clarke, Thompson 470–2
Bayle, Pierre 235, 260, 273, 275–6, Clifford, W. K. 95–7, 569–72
279 nn.12, 15, 284, 286, 292 n.3, 295, Clitomachus 6, 8, 43–4, 46, 57, 62 n.24,
355–66, 443, 507–9, 511, 513, 730 63 n.32, 81, 87, 93–5, 99 n.6, 100 n.13,
Berkeley, George 318 n.6, 339 n.8, 116–17, 284
369–78, 402, 454, 637 Cohen, Stewart 542, 607 n.30, 623,
Bernard of Arezzo 167, 210–11, 227 625, 676
740 NAME INDEX
Colotes 17–18, 22 n.17, 37, 45, 47 Harman, Gilbert 638, 646–7

n.4, 119 Hawthorne, John 629–30
Cudworth, Ralph 380, 385–6 Hegel, G. W. F. 260, 407, 430–9, 550,
559 n.1
Davidson, Donald 648, 704 Henry of Ghent 166–7, 209, 218, 219 n.9
Democritus 22 n.21, 38, 114, 267, 290, 297 Heraclitus 37–8, 70–1, 119, 297, 444
Dennett, Daniel 732 Hervet, Gentian 169, 235–6, 508
DeRose, Keith xiv, 602, 623, 627, 631 Hippocrates 110, 297
n.8, 632 n.14 Huet, Pierre-Daniel 320–8, 338, 507, 509
Descartes, René xiv–xv, 18, 154, 165–7, Hume, David xiv, 107, 133, 169, 253,
169, 175–6, 181, 222–3, 228, 243, 253, 279 nn.12, 16, 17, 360, 380–91,
266, 273–4, 276–8, 295, 299–300, 302, 397–400, 402, 411, 420, 438, 443–5,
306–18, 321–4, 326–9, 331–8, 342–4, 507, 509, 513, 565–6, 668–70, 672–80,
346–7, 349–52, 358, 386, 398–9, 402, 689, 716, 730–1, 733, 736
419, 425, 445, 461, 507, 509–13,
523–4, 614, 634, 637, 678, 701, 721 Israel, Jonathan 364–5
Diogenes of Babylon 118
Diogenes Laertius 6, 8, 16, 24, 37–8, 46, Jayarāśi 3, 145, 152–3, 155
68, 72, 103, 145, 234, 263, 323, 331, John Duns Scotus 166, 224–5, 268 n.15
443, 445 John of Salisbury 167, 170, 186–93,
Dretske, Fred 601–3, 628, 644, 706 219 n.9
Empedocles 38, 297 Kant, Immanuel 136, 149, 357, 402, 404,
Epictetus 8 407–8, 410, 413, 419–27, 436, 439,
Epicurus 119, 445, 448 455, 467–8, 484–5, 513, 690
Estienne, Henri 169, 235–7, 240, 243 Kitcher, Philip 690–2, 695–7
n.4, 290 Klein, Peter 538, 558, 561 n.24, 573–4
Eusebius of Caesarea 16, 24, 28, 69, 120 Kripke, Saul 644, 696
Favorinus of Arles 3, 7–8, 120–2 La Mothe Le Vayer, François de 283–91,

Fine, Gail 22 n.18, 22 n.24, 129, 130, 321, 510–12
133, 141, 142 n.8, 143 n.17 Lewis, David 578 n.1, 621, 627, 632 n.14
Fodor, Jerry 702 Lorenzi, Giovanni 166
Fogelin, Robert 565–78, 645
Foucher, Simon 320–1, 323–4, 328–38, Mackie, J. L. 704, 717, 719,
507, 509 724 nn.11–12
Frede, Michael 128–33, 135–6, 141, Malebranche, Nicolas 322–3, 328–9, 333,
380–3, 565 335, 338, 355, 358, 507
Fumerton, Richard 530 n.24, 542, 607 McDowell, John 617 n.5, 640, 642,
n.30 647–8, 660, 663 n.14, 702, 707
Menodotus of Nicomedia 3, 7, 102–12
Galen 7–8, 102, 104–12, 121, 262, 267 Mill, John Stuart 707–8
Gassendi, Pierre 136, 295–303, 321, 328, Montaigne, Michel de xiv, 136, 170–1,
510, 512 232–43, 247–51, 253–6, 284–91, 296,
Gettier, Edmund 159 n.29, 561 n.25, 255, 442–4, 509–10, 512, 550, 565, 730,
612–15 733, 736
Goldman, Alvin 547–8, 605 n.4, 606 Moore, G. E. xiv, 454, 467–7, 482,
n.19, 638 484–6, 489–95, 497–8, 635, 639, 647
Gregory of Rimini, 166, 228 Morison, Benjamin 138–9, 141
NAME INDEX 741
Muhammad 203–4, 206 Sanchez, Francisco 170–1, 260–7, 512

Seneca 8, 255, 296
Nāgārjuna 3, 145, 147–51 Sextus Empiricus xv, 3–8, 14–20,
Nagel, Thomas 438, 652 32–3, 36, 40–6, 53–5, 58,
Niccolò da Reggio 166 68–78, 81, 83, 86–7, 103–4, 107,
Nicholas of Autrecourt 166–7, 209– 111–12, 114, 125–41, 158 n.18,
18, 227 166, 169–71, 196, 218, 234–8,
Nozick, Robert 606 n.24, 628, 641, 644 242, 254, 261, 266, 302, 320,
Numenius 84 323, 325, 355–6, 359, 380–3,
385–6, 388, 390, 432, 442–3,
Ockham, William 166, 225, 227–8, 263 482, 507–9, 550, 572, 669–70,
678–9, 723, 730
Paganini, Gianni 286, 291, 358, 512 Sinnott-Armstrong, Walter 565, 567–9
Parmenides 37, 119, 130 Socrates 6, 16, 36–8, 51, 85–6, 88, 102,
Pascal, Blaise 235, 283, 287, 290, 321, 114, 119, 121, 241, 254, 297, 324–5,
444, 447, 510–11 331–3, 409–12, 444, 448, 640
Peirce, C. S. 616, 618 n.18 Sosa, Ernest 530 nn.21, 23, 549 n.4, 576,
Peter of Ailly 166, 225–30 628–9, 644–5, 678–9
Philo of Alexandria 9 n.8, 72, 115–16, Spinoza, Baruch 342–52, 355, 358,
724 n.13 436, 507
Philo of Larissa 6, 8, 67, 70, 81–9, 115, Śrīharşa 145, 153–5
118, 332 Stroud, Barry 222, 438, 470–1,
Photius 68–70, 72, 75, 166 663 n.17, 688
Pico della Mirandola, Gianfrancesco 235,
296, 510 Timon of Phlius 3–5, 24–33, 37, 67, 73,
Pico della Mirandola, Giovanni 235, 331 78 n.13
Plato 6–7, 36–8, 51, 62 n.28, 68, 70, 82,
85–8, 102, 118–19, 121, 131, 196, 315, Unger, Peter 636
330–2, 386, 436, 444, 679
Plutarch 7, 16–18, 39, 46, 85, 89, Van Cleve, James 405 n.5, 557, 559 n.1,
114–20, 135 561 n.19, 681 n.5, 682 n.14
Polemo 82, 85–6, 88, 120 Vogel, Jonathan 528 n.3, 542, 638,
Popkin, Richard 169–70, 261, 278 n.2, 643, 676
279 n.16, 286–8, 320–1, 333, 338, 350,
364, 506–14 Williams, Bernard 266, 274, 318 n.2
Popper, Karl 669–70 Williams, Michael 391 n.11, 438–9,
Porphyry 186 561 n.21, 570
Protagoras 38, 48 n.27, 444 Williamson, Timothy 530 n.19, 596,
Pryor, James 473–4, 478 n.12, 645 606 n.27, 644, 663 n.18, 704
Putnam, Hilary 648, 708 Wittgenstein, Ludwig 391 n.7, 469,
Pyrrho of Elis 3–6, 8, 21 n.13, 24–33, 481–501, 565, 687–8, 701
36–7, 67–70, 73–4, 76–8, 110–11, 145, Wright, Crispin 472–3, 503 n.25, 645
286, 355–6, 384, 396, 444–5, 448
Xenocrates 82, 86, 118, 332
Quine, W. V. O. 454, 682 n.13
Zeller, Eduard 27, 34 n.8, 443
Russell, Bertrand 454–63, 474, 482, 495, Zeno of Citium 16, 36–7, 39–41, 43–5,
634–5, 637–8, 655, 669, 730, 733, 736 52, 59, 83, 86, 118
SUBJECT INDEX
a priori: intuition 595–6; knowledge 176, 196, 578; moral 716–17; norm of 536,
546, 685–97; reasoning 596; skepticism 630; warranted 578
about the 309–11; warrant 217–18 atheism 196, 257, 279 n.12, 285, 287,
Academic skepticism: Bayle on 357; Fa- 357, 359–60, 717–18
vorinus on 120–2; Foucher on 328–38;
history of ancient 6–7; Huet on 324–8, Buddhism 145–52, 154, 728
338; John of Salisbury on 186–8; neo-
253–4; and morality 714–15; Plutarch Cartesian circle 278 n.3, 528 n.1, 550,
on 118–19, 135; and Pyrrhonism 6, 8, 557, 559 n.1
44, 261; Sanchez and 261–2; in Western certainty xiv, 52, 54, 70, 86, 121,
medieval thought 166–7 168–70, 177, 181, 188, 190, 200,
Academy: New 58, 63 n.33, 81, 85–7, 93, 203, 209, 274, 277, 295, 297–8,
114–18, 175–6, 182, 324–5, 392 n.23; 310, 315, 323–4, 333, 342–9, 351,
Old 6, 82, 85–8, 324–5 399, 409, 411–12, 414, 433–4, 460,
ad hominem argument 40, 45, 171, 209 470, 481, 487–8, 490–4, 496–501,
agnosticism 409, 492, 729–30 510, 512–13, 535, 637, 735
Agrippa’s trilemma 9 n.9, 299–302, Christianity: apologia for 235; Augus-
431–2, 439, 569–72, 636. See also tine’s conversion to 175; Cartesianism
modes, the Five as dangerous to 323; Nietzsche on 445,
aim (τέλος) 5, 8, 14, 62 n.22, 130, 252–3, 447; obedience to 286; paradoxes in 364;
291, 670 principles of 361; skepticism as compat-
aporia 39, 42, 46, 70, 115, 169, 196, 239, ible with 276, 287, 511; truth of 180;
243, 264, 431–2, 443 validity of 286
appearances (φαινόμενα): assent to 127–8, circularity 138, 151, 434, 530 n.23, 539,
133–4, 297; conflicting 70–1, 74, 249; 556–8, 569–70, 572, 604, 673, 692
as criterion of action 73, 127; to follow closure: argument from 632 n.15, 638–9,
53, 73, 112, 240, 405 643–6; principle of 473, 478 n.4, 556,
apprehension (κατάληψις) 16–17, 22 n.24, 600–3, 627–8, 656–7, 686
39, 41–2, 52–3, 57, 59, 70, 86 cogito 181–2, 266, 332, 343, 398, 416,
assent: action without 45–6, 119; to 445, 512–3
a cognitive impression 41; Hume on common sense xiv, 39, 118, 285–7,
380–91; Sextus Empiricus on 128–37; 373–4, 395, 402, 461, 463, 469, 471,
skeptical vs. dogmatic 127–8; suspension 484–5
of 57, 94, 127, 290; varieties of 57–9; contextualism 431, 470–1, 527, 555, 557,
withholding of 38, 41–2, 52, 68, 296–7, 569–70, 575–6, 578, 583, 620–31
397, 714–15 contrastivism 567
assertion 5–6, 31–3, 38, 40, 42, 46, 75,
132, 153, 169, 241, 359, 362, 493, disagreement: argument from 586, 589;
704, 715; bare 138; confusion between deep 501, 527; intractable 501; mode
knowledge and 578; dogmatic 111, from 138, 577; moral 721–2, 724 n.12;
SUBJECT INDEX 743
peer 722; religious 273, 729; skeptical 536–41, 547–8, 555–6, 570–1, 573–5,
method of 242 592–3, 595–6, 598
doubt: Cartesian 489, 494, 509, 511–13;
about the evidence of the senses 84, 201; God: as deceiver 165–6, 168, 209, 215–6,
general 187, 210–11; hyperbolic 266, 222–30, 316–17, 343, 419; knowledge
315–8, 513; levels of 568; method of of 177–8, 199, 207, 345, 349, 351–2,
324–5, 513; methodological 181, 411, 408; omnipotence of 210–11, 229, 315,
445; ordinary 488; philosophy of 119, 336–7, 510; veracity of 169, 344
238–41; reasonable 487, 500
hallucination 150, 177, 419, 524, 648,
Empiricism, medical 3, 7–8, 78 n.13, 694, 702; argument from 653–8. See
102–12, 126, 171, 266–7 also illusion
emptiness (śūnyatā) 148–9, 151 happiness 26, 29–30; and pleasure 16; and
Epicureanism 117, 119–20, 131, 328 wisdom 51, 53; as moral goodness 719;
epilogismos 105, 107–11 as the goal of life 43–4, 54; as tranquility
epistemic peer 501, 585, 722 33, 78, 88, 244 n.14
epistemic superior 585, 587 Hinduism 146, 156 n.3, 728
equipollence (ἰσοσθένεια) 38–9, 42, 46, hinge: commitments 501, 504 n.25;
74–5, 77, 133–5, 139–41, 169, 196, propositions 496, 504 n.25
236, 249, 251, 283, 285–7, 291, 432
error theory: epistemic contextual- ignorance: learned 196, 203–4, 207, 257;
ism’s 626–7, 629; moral 714–15, Socrates’s profession of 38, 51, 85, 241,
717–22 254, 324, 409–11
evidence: agreed-upon 72, 501; criterion illusion 120, 154, 167–9, 214–15, 226,
of 328, 333, 335–6; inductive 602; ob- 318 n.2, 456; argument from 373, 526,
jective 239; perceptual 212, 474, 642–3, 528 n.4. See also hallucination
647; testimonial 585–6 impression (φαντασία): cognitive 40–1, 52,
evidentness (evidentia) 168, 172 n.12, 83–4, 86, 89, 97, 154; persuasive 51–60;
211–14, 216–17, 226, 298 probable 54, 61 nn.17, 19, 84
evidentialism 592–3, 595–6, 598, 603–4 inaction/inactivity (ἀπραξία) 44, 46, 52–3,
evil demon 165, 222, 266, 274, 278 62 n.27, 73, 94, 128
externalism 215, 592–604, 648, 659–60, inapprehensibility (ἀκαταληψία) 7, 15,
663 n.16, 664 nn.21, 23, 702 19–20, 37, 40–1, 46, 47 n.10, 52–3, 75,
experience: perceptual 158 n.21, 318, 83, 103
419, 491, 524–6, 592–3, 595–6, 599, (infinite) regress 150, 152, 155, 236, 239,
601, 637, 644, 648, 652–5, 657, 686; 423, 554–5, 558, 599; argument from
subjective 17–8, 21 n.15, 337, 419–20, 165, 214–15, 401, 404, 535–48; Pyrrho-
422–3, 536; waking 308–9, 464 n.5, 526 nian mode from 138–9, 299, 301, 385,
508, 523, 569, 572–6
faith: Christian 175, 283, 338, 510–11; internalism 592–600, 660–1, 681 n.5
as a condition for understanding 178;
as a means to acquire knowledge 170, justification: a posteriori 45; a priori
447; and reason 179–80, 323–4, 357; 689–91; circular 675; deontological
Reformers’ rule of 509–10; skepticism as 597; doxastic 589; epistemic 170, 216,
preparation for 287 431, 433, 569, 597, 599, 604 n.1,
fideism 119, 178–9, 276, 283, 286–8, 672, 709, 720; fallible 669; inferential
507, 510–11 vs. non-inferential 537–48, 605 n.4;
foundationalism 150, 167, 209–12, internalist 596, 645, 674–7; moral
214–15, 227, 399, 460, 462, 524–5, 723; prima facie 525–6, 530 n.17;
744 SUBJECT INDEX
propositional 589; rational 44, 46, 250, opinions: conflict of 5, 485; dogmatic
267, 365; reliabilist 674–7, 679–80 132, 290; mere 59, 68, 94; neither tell
the truth nor lie 25; to hold no 25, 40,
knowledge: certain 86, 166–7, 169, 176, 53, 59, 381; weak 171, 252, 298
179, 193, 198–9, 203–5, 207, 227,
261–2, 274, 320–1, 334, 351, 494, 510, paradox: Agrippa’s problem as a 569–71,
512; empirical 105, 639–41, 691–3, 577–8; in Christianity 358, 364; the
696–7; intellectual 215; perceptual 215, liar 241; Meno’s 550; practice of 283,
369, 372, 376, 404, 494, 526, 660–1, 285–7
702, 705–6, 709–10; the problem of pathοs (πάθος) 7, 10 n.21, 15–20, 21 n.12,
easy 542, 556–8, 607 n.30; sources 22 nn.18–19, 249
of 147, 150, 152, 201, 705, 709–10; perception 17–18, 109–10, 150–2, 202,
testimonial 153 207, 210–17, 237–8, 265, 306–10,
313–15, 371–4, 398–400, 403–4, 455–7,
language-game 277, 483, 486–91, 462, 593, 604, 634–5, 652–5, 657–8,
493, 500 693–4, 704–6, 709–10; clear and distinct
libertinage 283, 285–9, 291 299–300, 327–8; of the intellect 310;
internal 210, 420; relativity of 215, 400
materialism 152, 176, 242 (see also sensation)
memory: and knowledge of the past 423; perplexity 169, 196, 203, 241, 291
medical Empiricists on 104, 107, pramāṇa (knowledge source) 146–7,
109–11; reliability of 425, 604 149–51
metaphysics: Cartesian 321, 329, 333, presentation. See impression
338, 512–13; of modality 223; nihilist probability 51, 54, 57, 61 n.17, 82, 219
68; as nonsensical 482 n.11, 254, 295, 299, 547, 617 n.3
Methodism, medical 7, 104 Pyrrhonism: and Academic skepticism 8;
modes: against causal explanations 72, of Aenesidemus 67–78; and Arcesilaus’s
137; the Five 5–6, 9 nn.9, 11, stance 37, 40, 42–4, 46; Bayle on 356–8,
137–9, 141, 327, 566–7, 571 (see 362, 364–5; and Carneades’s stance
also Agrippa’s trilemma); the Ten 53, 58; and Charron 171, 249–55; and
5–6, 71–4, 77, 137–9, 237, 242, classical Indian thought 3, 145, 158
276, 287, 296–9, 301, 303, 721; the n.18; and Cyrenaicism 14–16; Gassendi
Two 9 n.11, 137, 158 on 296, 299; and Favorinus’s interest
moral anti-realism 714, 717 in 120–2; Huet on 324–6, 328; history
moral noncognitivism 714–17, 724 n.8 of ancient 4–6; influence on contempo-
rary philosophy 4; and Hegel 431–2;
natural light 326–8, 361–3, 365 and Hume 380–91; and La Mothe Le
nature of things: assertions about the 5; Vayer 283, 286, 290–1; and medi-
cannot be known 15, 26, 295, 297; non- cal Empiricism 7–8, 103–4, 111–12;
evident 77; knowledge of 19, 83, 115, of Montaigne 170–1, 232, 234–42; neo-
202; suspension of judgment about 43, 565–78; and moral disagreement 721,
158 n.18 723; and morality 714–5; and Nietzsche
nihilism: extreme 442, 449; Nāgārjuna’s 442–7, 449; Plutarch’s attitude towards
rejection of 157 n.8; skeptical 149 120; Popkin on the history of 507–9,
no more (οὐ μᾶλλον) 25, 31–3, 74, 511–13; of Pyrrho and Timon 24–33;
240, 382 Renaissance rediscovery of 169–70; of
nominalism 264, 266 Sanchez 171, 260–3, 265, 267; of Sextus
non-assertion (ἀφασία) 35 n.13, 32, 44, Empiricus 125–41; varieties of 5; in
46, 120 Western medieval thought 165–6
SUBJECT INDEX 745
rationalism: Bayle on 357–8, 360–5; in 431–9; modern 186, 278, 731–2; natural
Cartesianism 285; critical 513; excessive 167–8, 189–90, 216, 230
178, 180; medical 7, 102–6, 108–11 scientia 191, 213, 263–4, 266, 298–9
realism: direct 405 n.5, 463; metaphysical self 175, 182, 238, 241–3, 249, 254,
151; minimal 723 n.1; moral 714, 721; 415, 420
naïve 371, 652–5; refutation of 155; self-evident 211, 214, 339 n.4, 401–2,
robust 723 n.1 425, 460, 489, 491
reason: and Academic skepticism 62 self-refutation 26, 73, 298, 412, 555
n.24; Augustine on 178–80; Bayle 356– sensation 5, 25, 27–31, 115, 119–20, 127,
8, 361–5; Charron on 249–50, 252, 214, 249, 311–13, 334–5, 375–6, 386,
255–7; contrast between experience 454, 456, 462
and 102; in the debate between medi- sense-data 277, 454–8, 460–2, 474, 491,
cal Empiricists and rationalists 106–8; 494, 634–5, 638, 653
epistemic value of 170–1; Foucher sense-perception/sensory perception.
on 332–3, 335, 337–8; Hamman on See perception
408–9, 411–15; Huet on 323–5, 328; sign: commemorative 107; the idea of God
Hume on 387–8, 392 n.19; La Mothe as a 352; indicative 302–3; inferential 152
Le Vayer 283–5, 289–91; medical Em- skepticism: about the external world 15,
piricists’ stance on 103, 105, 109–11; 19–20, 165, 168, 177, 211, 222, 226,
medical rationalists’ stance on 102; 266–7, 334, 419–20, 428 n.1, 438,
Reid on 397–9, 403–4; unreliability 467, 489, 491, 494, 597–8, 634–49;
of 68 Academic (see Academic skepticism);
reasons: epistemic 552–9; practical/prag- Cartesian 266–7, 300, 473, 485, 487,
matic 389, 552 497, 565; experimental 446–9; mitigated
reductio ad absurdum 41, 167, 210–11, 53, 63 n.32, 94, 98–9, 280 n.26, 295,
358, 566–7 442, 510, 730; constructive 295; Pyrrho-
Reformation, the 171, 255, 273, 275, 508 nian (see Pyrrhonism); radical 6, 53, 94,
refutation: self- 73, 298, 304 n.11; of 167, 211, 390, 448, 454, 462, 496, 500,
skepticism 168, 175, 177, 266, 277, 583–4, 659–62
335, 510–13; Socrates’s practice of 38; solipsism 20, 181–2, 211, 227, 411
of Stoicism 115, 117 sorites 49 n.33, 52
relativism 243 n.6, 499–500; cultural 498; Stoicism 16, 71, 115, 117–18, 234,
phenomenal 75 242, 250
reliabilism 476, 542, 555–6, 560 n.9, 583, subjectivity 241–3, 248, 250, 433–4
592–6, 598, 601–4, 674, 676 suspension of judgment (ἐποχή): as the
religion: cognitive science of 732; centerpiece of Pyrrhonism 15; and
Pyrrhonism as the best preparation for equipollence 38–9; on moral claims 715,
235; relation between morality and 720, 722–3; and ‘no more’ 32; about
275, 359; relation between reason and the real nature of things 43; on religious
358, 413; relation between wisdom and claims 511; as the skeptic’s aim 8, 42;
religion 256; skepticism as not opposed and tranquility 68, 76, 135, 240, 266,
to 323; true 239, 256, 275, 362 290, 296; universal 6, 41–4, 94, 103,
114, 167, 523; violation of 5; as a volun-
sage: Academic 42–3, 55–6, 59–60; tary decision 171, 249
Charron’s skeptical 249–52, 254, 256;
Stoic 41–2, 45, 52–3, 55, 57, 130–1 Taoism 728
scholasticism 198–207, 239, 276 testimony 139–40, 147, 151–4, 179, 401,
science: constructive 390; demonstrative 457, 460–1, 536, 542–4, 556, 588, 640,
187–8, 190–2; Hegel’s philosophical of 646, 677, 694, 709–10
746 SUBJECT INDEX
theology 6, 126, 202, 223, 226, 255; 512; search for 39, 48 n.18, 51, 88, 97,
Christian 361; fideistic 116; natural 511; 171, 252, 263, 290, 325, 329–30, 387
negative 256–7; post-Scotist 215; Stoic
51, 62 n.29 uncertainty 252, 283–4, 387, 447, 616
tranquility (ἀταραξία) 5, 25, 33, 53, 68–9, underdetermination: argument from 636–8;
73, 78, 132–5, 171, 235–6, 240, 253, of evidence 528 n.3; responses to the
266–7, 290–1, 296, 326, 565, 670 argument from 642–3
truth: apparent 58; criterion of 16, 40, undisturbedness. See tranquility
44–5, 51, 54, 83–5, 181, 296, 300, 302,
321, 324, 326–9, 332, 335–6, 431, Vedānta 145–6, 163–5
508, 550–6, 678–9; innate 512; inquiry via media 295, 302–3, 510
into 77, 97, 299; logical 482, 610, 687; wisdom 51, 53, 56, 59, 70, 96–7, 117,
mathematical 337, 343–6, 481, 610, 242, 255–7, 325–6, 410–12; Academic
687, 704; metaphysical 315, 361, 438; sage’s 60; common 251; divine 255,
moral 350, 359, 716; necessary 336, 288; human 248–9, 255; and opinion
363, 427, 490, 643, 696; objective 498; 53, 62 n.31, 63 n.32; skeptical 248, 288,
religious 198, 201, 350, 730; revealed 320–1, 509, 513; Stoic view of 52

Diego e Machuca Skepticism From Antiquity To The Present 1

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Diego e Machuca Skepticism From Antiquity To The Present 1

Hochgeladen von

Copyright:

Verfügbare Formate

SKEPTICISM

Also available from Bloomsbury

The Bloomsbury Companion to Analytic Philosophy,

The Bloomsbury Companion to Epistemology,

A Critical Introduction to Skepticism, Allan Hazlett

Edited by Diego E. Machuca and Baron Reed

50 Bedford Square 1385 Broadway

First published 2018

© Diego Machuca and Baron Reed, 2018

No responsibility for loss caused to any individual or organization acting on or refraining

British Library Cataloguing-in-Publication Data

ISBN: HB: 978-1-4725-0771-6

Library of Congress Cataloging-in-Publication Data

Cover design: Clare Turner

Typeset by Deanta Global Publishing Services, Chennai, India

General Introduction xiii

Part I: Ancient Skepticism 1

Introduction Diego E. Machuca 3

1 The Cyrenaics and Skepticism 14

2 Pyrrho and Timon 24

8 Menodotus and Medical Empiricism 102

9 Middle Platonism and Skepticism: Plutarch and Favorinus 114

10 Sextus Empiricus 125

11 Skepticism in Classical Indian Philosophy 145

Part II: Medieval and Renaissance Skepticism 163

Introduction Diego E. Machuca 165

12 Augustine and Skepticism 175

13 John of Salisbury’s Skepticism 186

14 Skepticism in Classical Islam: The Case of Ghazali 196

15 Nicholas of Autrecourt and John Buridan on Skepticism 209

16 Divine Deception 222

17 Michel de Montaigne 232

18 Pierre Charron 247

19 Francisco Sanchez: A Renaissance Pyrrhonist against

Part III: Modern Skepticism 271

Introduction Baron Reed 273

20 François de La Mothe Le Vayer 283

21 Gassendi on Skepticism 295

22 Descartes and the Force of Skepticism 306

23 Varieties of Modern Academic Skepticism:

24 Spinoza on Skepticism 342

25 Pierre Bayle 355

26 Berkeley and Skepticism 369

27 A Pyrrhonian Interpretation of Hume on Assent 380

28 Thomas Reid’s Engagement with Skeptics and Skepticism 395

29 Johann Georg Hamann 407

30 Kant and External World Skepticism 419

31 Hegel: Philosophy as a Kind of Skepticism 430

32 Friedrich Nietzsche 442

33 Russell’s Logical Construction of the External World 454

34 Moore and Mooreanism 467

35 Wittgenstein and Skepticism: Illusory Doubts 481

36 Richard Popkin on the History of Skepticism 506

Part IV: Contemporary Skepticism 521

Introduction Baron Reed 523

37 Regress Arguments and Skepticism 535

38 The Problem of the Criterion 550

40 Disagreement and Skepticism 581

41 Skepticism and the Internalism-Externalism Debate 592

42 Skepticism and Fallibilism 609

43 Skepticism and Contextualism 620

44 External World Skepticism 634

45 Disjunctivism and Skepticism 652

(3) Unless x appears invariably F (F), x cannot be F (F) “by nature.”