Reason and Normativity: Theories of Action and Morality

ALZNAUER / TORRALBA (EDS.
) | Theories of Action and Morality 12 REASON AND NORMATIVITY

Edited by Ana Marta González and Alejandro G. Vigo
ALZNAUER / TORRALBA (EDS.) | Theories of Action and Morality

T he essays in this volume address the question of whether we can under-
stand human action without reference to moral norms or values. Although
the authors approach this question in different and sometimes even incompatible
ways, they are united in thinking that it is undesirable or even incoherent to treat
human agency as if it were conceptually independent of value questions.
The editors have attempted to invite contributions that would be interesting to both
philosophers and social theorists. The conjunction of philosophic and sociological
perspectives might help to overcome some of the mutual misunderstandings that
have been fostered by a lack of dialogue between the philosophic and sociological
action theory. The volume includes essays by Terry Pinkard, Sebastian Rödl, Dieter
Schönecker, Ana Marta González, John Levi Martin, Alejandro N. García Martínez,
Sophie Djigo, Teresa Enríquez, and Evgenia Mylonaki.
D ie in diesem Band versammelten Essays erörtern die Frage nach der Mög-
lichkeit des Verstehens menschlichen Handelns ohne den Rückbezug auf
moralische Werte und Normen. Obwohl die Autoren sich dieser Frage auf ganz
unterschiedliche, manchmal divergierende, Weisen nähern, verbindet sie alle die
Annahme, es sei nicht wünschenswert oder sogar inkohärent, das menschliche
Handeln grundsätzlich unabhängig von moralischen Werten zu betrachten.
Die Herausgeber haben sich um eine für Philosophen und Gesellschaftswissen-
Theories of Action and Morality
schaftler gleichermaßen attraktive Beitragssammlung bemüht. Die Verknüpfung
philosophischer und soziologischer Perspektiven könnte zur Klärung gegenseitiger Perspectives from Philosophy
Missverständnisse beitragen, die aufgrund eines mangelhaften Dialogs zwischen
der philosophischen und soziologischen Handlungstheorie erwachsen sind. In die- and Social Theory
sem Band enthalten sind Essays von Terry Pinkard, Sebastian Rödl, Dieter Schöne-
cker, Ana Marta González, John Levi Martin, Alejandro N. García Martínez, Sophie Edited by Mark Alznauer and
Djigo, Teresa Enríquez und Evgenia Mylonaki.
José M. Torralba
ISBN 978-3-487-15387-2 OLMS
cyan = Sonderfarbe 2-fbg (tiefe + Pantone 7473)

(E-Book)
Reason and Normativity

Razón y Normatividad
Vernunft und Normativität
A Series on Practical Reason, Morality, and Natural Law
Escritos sobre razón práctica, moralidad y ley natural
Schriftenreihe zu praktischer Vernunft, Moralität und Naturrecht
edited by / editado por / herausgegeben von

Ana Marta González / Alejandro G. Vigo
Vol. / Band 12
Georg Olms Verlag

Hildesheim · Zürich · New York
2016
(E-Book)
Theories of Action and Morality

Perspectives from Philosophy and Social Theory
Edited by
Mark Alznauer and José M. Torralba
Georg Olms Verlag

Hildesheim · Zürich · New York
2016
(E-Book)
This work and all articles and pictures involved are protected by copyright.
Application outside the strict limits of copyright law without consent
having been obtained from the publishing firm is inadmissible.
These regulations are meant especially for copies,translations and micropublishings
as well as for storing and editing in electronic systems.
Das Werk ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen
des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig.
Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und
die Einspeicherung und Verarbeitung in elektronischen Systemen.
Die Deutsche Nationalbibliothek lists this publication in the

Deutsche Nationalbibliografie; detailed bibliographic data are available
in the Internet at http://dnb.d-nb.de.
Die Deutsche Nationalbibliothek verzeichnet diese Publikation

in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten
sind im Internet über http://dnb.d-nb.de abrufbar.
© Georg Olms Verlag AG, Hildesheim 2016

www.olms.de
E-Book
Umschlaggestaltung: Inga Günther, Hildesheim

Alle Rechte vorbehalten
ISBN 978-3-487-42168-1
CONTENTS
Introduction.................................................................................................... 7
Terry Pinkard
Contrasting Concepts of Agency and the Space of Reasons............................ 15
Sebastian Rödl
Acting as the Internal End of Acting............................................................ 37
Dieter Schönecker
Why there is no Fact of Reason in the Groundwork. Three Arguments...... 55
Ana Marta González

The Recovery of Action in Social Theory: Acting out of Sentiment,
Acting out of Character, Acting out of Interest, Acting out of Will.............. 79
John Levi Martin

Simmel and Rickert on Aesthetics and Historical Explanation...................... 113
Alejandro N. García Martínez

The Realist Sociological Approach to Action.................................................. 149
Sophie Djigo
Leverage and Truth........................................................................................ 175
Teresa Enríquez
Mixed Actions in the Work of Harry Frankfurt.......................................... 209
Evgenia Mylonaki
Practical Knowledge and Perception................................................................. 241
Name Index.................................................................................................. 269

INTRODUCTION
All of the essays in this volume address, in one way or another, the
question of whether we can understand human action without reference
to moral norms or values. Although the authors of these essays ap-
proach this question in different and sometimes even incompatible ways,
they are united in thinking that it is undesirable or even incoherent to
treat human agency as if it were conceptually independent of value ques-
tions.
In this respect, these essays are part of a more general contemporary
reaction against one of the most dominant tendencies in philosophic
and social thought of the last century, which is the tendency to decouple
the concept of action from moral considerations. There are, of course,
very powerful reasons why action has come to be treated in this way.
From the philosophic point of view, it seems natural to assume that ac-
tion is just purpose-driven behaviour and that these purposes need not
be narrowly moral. From the sociological point of view, it has seemed
essential to devise ways of thinking about action that both cover all ac-
tivity, even activity not guided by moral norms, and that are able to
entirely bracket all questions of the validity of moral norms. These
value-neutral approaches to human agency have certainly proved their
value in opening up new avenues of philosophic and sociological explo-
ration. But in the following, it is argued that these achievements have
often blinded us to crucial aspects of agency, obscuring aspects that were
previously better recognized.
And so in rejecting the tendency to separate the theory of action
from moral questions, many these authors often find themselves re-
turning to more classical approaches: to Aristotle’s concept of praxis, to
Augustine’s notion of the will, or to Kant on reflective judgment or
Hegel on the sociality of action. The volume as a whole not only marks
a rejection of more recent history, it is also an attempt at ressourcement –at
renewing the great tradition of philosophic reflection on this topic, and
making it speak to our concerns. According to almost all of the major
8 Mark Alznauer / José M. Torralba
figures in this tradition, the central or paradigmatic form of action is

moral action, and all moral action needs to be understood as necessarily
situated in a social world, whether that be Aristotle’s polis or Hegel’s
modern state. These are two of the insights that have potentially been
lost in the abstractions of Twentieth Century theories of action (whether
of Weberian or Davidsonian descent)– both play a significant role in the
essays that follow.
In addition to addressing the theme of the volume, the editors have
attempted to invite contributions that would be interesting to both phi-
losophers and social theorists. It is certainly the case that some of these
essays will have less crossover appeal than others, but we hope that the
conjunction of philosophic and sociological perspectives that this vol-
ume includes can help overcome some of the mutual misunderstandings
that have been fostered by a lack of dialogue between the philosophic
and sociological action theory.
***
The first essay in the volume, Terry Pinkard’s “Contrasting Concepts of

Agency and the Space of Reasons”, concerns two incompatible ways of
understanding Hegel’s claims about the sociality of agency –an individu-
alist conception and an intersubjectivist one. According to the individual-
ist conception of agency, becoming an agent is like becoming a language
user; although you might depend on others to acquire a language, once
you have developed this ability, you do not rely on others for the exercise
of it, you have it on your own. According to the intersubjectivist concep-
tion, agency is more like a social status than an acquired ability: you cease
to be an agent when you cease to be recognized as one. Pinkard argues
that Hegel finds an element of truth in both of these models, since our
capacities and our being recognized as having such capacities, exist in a
kind of feedback loop which allows agency to develop in new ways. This
hybrid model of agency allows us to understand exactly how our ability
to act is tied to forms of social interaction without forcing us to embrace
an overly radical form of social constructivism.
Pinkard goes on to argue that these two forms of agency correspond
to two different conceptions of what the American philosopher Wilfrid
Introduction 9
Sellars called the “space of reasons”. According to the static conception,

the reasons we have to act or not act have authority independently of
whether this authority is recognized by us. (This, of course, is the con-
ception of the space of reasons we see in Aristotle’s moral realism.)
According to the dynamic conception, by contrast, the space of reasons
is socially constructed and so dynamic. On this model, our reasons to act
change according to our circumstances –they only hold for a place and
for a time. Pinkard argues that although Hegel’s early work features the
former conception of the space of reasons, he came to think that his
own account of agency is only compatible with the second model.
Pinkard encourages us to endorse the position that emerges from this: a
hybrid conception of agency and a dynamic conception of the space of
reasons.
In “Acting as the Internal End of Acting”, Sebastian Rödl attempts
to vindicate a more Aristotelian or Kantian picture of action (a striking
feature of Rödl’s position is that it attempts to synthesize these two
seeming opposed approaches). Rödl’s target in this essay is the intuitive
idea that moral agency (the ability to act for some moral end) involves
controversial capacities, ones that go beyond mere intentional agency
(the ability to act for some purpose). He challenges this instrumental
conception of agency by showing we could not have the capacity to
direct our action toward any end at all if we did not also have the
capacity to take our action itself as our end (or to act “according to a
maxim” as Kant would put it).
Rödl’s account is grounded in a rigorous consideration of the logical
concept of an end. He argues that for a change or movement to be end-
directed or purposive, the end must be the reason for the change or mo-
tion. We see one form this can take in biological phenomena: for ex-
ample, we say “Helicase unwinds DNA so that DNA can replicate itself ”.
But this analysis raises a further question about the end or purpose itself:
that is, we might ask why does DNA replicate itself ? Rödl claims we can-
not answer this question by either by pointing to some efficient cause or
to another externally purposive movement –either of these responses
would simply re-raise the original question. What we need for a genuine
explanation of purposive movement is the notion of an internal end: we
need to say that the DNA replicates itself not to achieve some further
end but because such replication is part of a teleological system which is
the unity of such purposive moments (in this case, the internal end is
life).
Rödl then goes on to show that this distinction between activity
directed to an external end and activity directed to an internal end has a
strict parallel in the case of human agency. What makes human agency
distinct from the kind of purposiveness we see in animal behavior is that
the human agent knows or understands the end to which she is directing
her actions. In some cases, the end is external to the activity: e.g., we
cross the road to get to the other side. Borrowing from Aristotle, Rödl
calls this a poesis. In other cases, the end is internal to the activity: e.g., I
am doing a kind thing because it is the kind thing to do. This, according
to Rödl, is what Aristotle meant by praxis. Just as before, Rödl argues
that the former (poesis) is only possible by virtue of the latter (praxis). So
we cannot have agency at all if we do not have the capacity or power to
act according to internal ends, ends of the sort implied by the moral
evaluation of actions.
In Dieter Schönecker’s “Why there is no Fact of Reason in the
Groundwork. Three Arguments”, we are offered a detailed account of
feature of Kant’s theory of action that is of central concern to the
themes of this volume. This is Kant’s claim that we only know we are
truly free agents, agents who can act against sensible inclinations, be-
cause of the so-called “fact of reason”: our immediate and undeniable
consciousness of being bound to the categorical imperative. Schönecker
both shows the function that the “fact of reason” plays in Kant’s Critique
of Practical Reason, and offers three arguments as to why it would be a
mistake to think that the thesis of the fact of reason is already present in
Kant’s earlier Groundwork. His analysis reveals that Kant’s notion of mor-
al consciousness undergoes a fundamental shift between these works: in
the former, it is understood as in need of justification, whereas in the lat-
ter it is understood as immediately justified.
The next essay, Ana Marta González’s “The Recovery of Action in
Social Theory”, is also concerned to reject theories of action that fail to
recognize, as she puts it, that every action is moral by default. But she
approaches this issue through an engagement with contemporary socio-
logical literature. Her departure point is a recent observation made by
Colin Campbell that contemporary sociological theories of action tend
to lack a sufficiently robust conception of human subjectivity. The heart
Introduction 11
of the problem, according to González, is an inadequate understanding

of the role of motive in individual actions: following Weber, contempo-
rary theorists focus on the intellectual or social dimensions of action (its
intention or intersubjective meaning) at the expense of the psychological
and physical elements that make each action unique.
In the more positive part of her essay, González shows that certain
classical philosophic theories of agency –those we find in Aristotle,
Hume, and Kant– can help remedy these blind spots in the sociological
approach to action since they include a rich store of reflections on the
role that inclinations, passions, and habits play in moral action. She
concludes by contrasting these philosophic approaches to agency –which
presuppose that moral action is the paradigmatic form of action– with
Max Weber’s famous typology of action in Wirtschaft und Gesellschaft,
arguing that the differences among the action types he identifies is better
understood in terms of psychological categories rather than in terms of
their intended meanings. This allows us to see how we can have our
Weber without taking on board the biases that his account has generated.
In “Simmel and Rickert on Aesthetics and Historical Explanation”,
John Levi Martin challenges one of the central features of post-Weberi-
an sociological methodology: its attempt to approach and analyse real
human values with conceptual tools that are understood as mere analyt-
ical constructs or ideal types. Martin argues, however, that this self-con-
ception of sociology makes it difficult or even impossible to understand
why the results of such inquiry would need to be taken into account by
ethical actors, for if reality can be conceptualized in an indefinite number
of equally valid ways, it would appear that there is no need for any given
agent to accede to the perspective afforded by one of these at the ex-
pense of others.
Martin argues that at the base of this flawed conception of socio-
logical methodology is a highly problematic neo-Kantian approach to
concept formation, one given its most powerful and influential defense
by Heinrich Rickert. Martin shows how certain neo-Kantians abandoned
Kant’s (more promising) theory of reflective judgment in favour of a
more exclusive focus on determinative judgement, which understands
judgment as a matter of the subsumption of an intuition under an al-
ready existing concept. They reconceived determinative judgment, how-
ever, in a voluntaristic fashion. Although the concepts used in such judg-
ments were not understood as completely arbitrary, they were under-

stood to be rooted in selective abstraction, a process that necessarily
picks out certain aspects of reality while screening or suppressing others.
This model of concept-formation made it difficult to grasp individu-
als in their individuality. Although solving this problem was one of cen-
tral aspirations of Rickert’s work, Martin shows that Rickert was never
able to formulate an adequate response. When Max Weber adopted the
Rickert’s basic approach, he did so without recognizing what Rickert’s
failure should have revealed to him: which is that a solipsistic model of
concept-formation is incompatible with the intersubjective validity of
the cultural sciences. Martin ends his essay with the provocative sugges-
tion that Simmel’s reflections on the intrinsically second-personal nature
of love offer us the best way forward.
In “The Realist Sociological Approach to Action”, Alejandro N.
García Martínez analyses two major contemporary social theorists who,
he thinks, are most able to do justice to the role that moral values play in
the social world: Margaret Archer and Pierpaolo Donati. García
Martínez shows how Archer’s morphogenetic approach and Donati’s re-
lational theory tackle the problem of structure and agency; the problem
of determining how much freedom we have as individuals in the fact of
the social structures and impersonal forces that appear to constrain our
agency. He shows that in responding to this problem, both endorse a
kind of sociological realism: a belief that social reality is partly indepen-
dent of our beliefs about it. He then goes on to argue that both theories
can be enriched by incorporating the insights of the other, and that this
offers us a kind of social theory which avoids reductionism about hu-
man value without sacrificing distinctively sociological insights into how
the social world can constrain human agents.
The final three essays in the book address certain well-known prob-
lems in the philosophy of action: the debate about internalism and exter-
nalism, the problem of mixed actions, and the nature of practical knowl-
edge. In “Leverage and Truth”, Sophie Djigo asks how is it possible to
have a reason to do something and yet fail to do it. She suggests that dis-
putes about this problem often turn on different conceptions of what it
is to have a reason. Some have argued that having a reason just is having
a motivation to do something. If we fail to act on a reason, some other
motivation must be interfering. Others have insisted that having a reason
Introduction 13
is a matter of understanding certain normative claims and does not re-

quire that an agent have any particular motivation to act on these claims.
Djigo shows the power and limits of these approaches through a careful
analysis of Augustine’s account of the “half-wounded will” in the Confes-
sions. She offers several reasons why we need to maintain the possibility
of reasons that we have no motivation to act upon, and explains the pos-
sibility that we might be indifferent to such reasons either as a kind of
forgetfulness or as refusal to give leverage to my commitments.
In the next essay, Teresa Enríquez uses Aristotelian notion of a
‘mixed action’ as a way to trace the development of Harry Frankfurt’s
theory of action. A mixed action, according to Aristotle, is one that is
not entirely voluntary or entirely non-voluntary. Although Frankfurt
himself never explicitly uses this term, Enríquez shows that his changing
thoughts on the nature of action can be illuminated by considering how
he accounts for behavioral phenomena that fall into this broad but pro-
blematic class. This approach allows her to tease apart several different
senses in which a Frankfurtian action could fail to be unambiguously vo-
litional. They range from more garden-variety cases where one’s desires
conflict with each other, to the more interesting cases where the unity of
our will is threatened not from our desires but from a deep ambivalence
about what we truly will.
This volume concludes with Evgenia Mylonaki’s essay, “Practical
Knowledge and Perception”, an essay that addresses a potential tension
between Elizabeth Anscombe’s conception of intentional action and Iris
Murdoch’s account of moral perception. According to Anscombe, we
know what we are doing non-observationally: by bringing it about. But
according to Murdoch, knowing what we should do is a matter of recep-
tivity: it is a passive perception of the circumstances we are in. It is
tempting to think that if practical knowledge is a knowledge that brings
about what it understands, as Anscombe argues, then Murdoch must be
wrong to think that moral perception is a form of practical knowledge.
On such a view, moral perception must be a species of reflective knowl-
edge: perhaps a combination of ordinary receptive knowledge along with
knowledge of what the rules of morality require of us. Mylonaki argues
against this two-factor view of moral perception, showing that moral
perception is a fully distinct from ordinary receptive knowledge. She
then offers an alternative account of moral perception that shows it to
be, despite appearances to the contrary, a species of practical knowledge

as Anscombe understood it.
***
The majority of the essays included in this volume were delivered at the
conference “Theories of Action and Morality” held at Universidad de
Navarra (Pamplona, Spain) in September 2012. For the sponsorship of
both the conference and the publication of this volume we would like to
acknowledge the Institute for Culture and Society, Universidad de Navar-
ra (research project “Natural Law and Practical Rationality”) and the
Government of Spain (research projects “Moral Philosophy and Social
Sciences”, ref. FFI2009-09265); “Action, Emotion and Identity”, ref.
FFI2012-38737-C03-01). Two of the papers in the volume were com-
missioned to researchers of these projects.
The editors would like to thank Daniel Doyle, Patricio Fernández,
Luis Placencia, Will Small, David Zapero Maier, and Stephan Zimmer-
mann who were discussants at the conference; and also María Carolina
Maomed for her assistance in the editorial work on the manuscript.
Finally, we extend our sincere gratitude to Alejandro Vigo and Ana Mar-
ta González for their advice and support in the organiza-tion of the con-
ference and the publication of this volume.
MARK ALZNAUER
Northwestern University
Evanston, IL, United States
JOSÉ M. TORRALBA
Universidad de Navarra
Pamplona, Spain
CONTRASTING CONCEPTS OF AGENCY
AND THE SPACE OF REASONS
TERRY PINKARD
“To put law over man is a problem in politics

which I compare to that of squaring the circle in
geometry. Solve this problem well, and the govern-
ment based on this solution will be good and with-
out abuses. But until then, be sure that where you
believe you are making the laws rule, it will be the
men who are ruling”.1
Agency has come to occupy a special and central place in contemporary

philosophical discussions. The term (“agent”) itself is a peculiar English
word in its philosophical usage. It seems to have started its life meaning
that which causes something else. It then came to mean somebody car-
rying out the work of another (as his “agent”). After a short while, it
then came to mean an entity moving itself to thoughts and actions. As a
philosophical term, it also seems to have few counterparts in non-phi-
losophical uses in other European languages. Nonetheless, its general
meaning in Anglophone inspired philosophical discourse seems to have
boiled down to something like this. An agent is a creature that acts ac-
cording to its intentions, where its intentions are governed by the reasons
the agent has for acting. After that, just about everything else one can say
about agency is philosophically contentious. This general conception also
suggests a particular project: If we get clear about what it is to have an
intention and what it is to act in terms of a reason, we should be rela-
tively clear about everything that is baseline significant about agency it-
self. Not unsurprisingly, there is by now an unsurveyable amount of
1 Rousseau, J. J., Considerations on the Government of Poland and on Its Planned Reformation,
in Rousseau, J. J., The Plan for Perpetual Peace, On the Government of Poland and other Writings
on History and Politics, translated by C. Kelly and J. Bush, edited by C. Kelly, Lebanon, NH,
University Press of New England, 2005, pp. 170-171.
16 Terry Pinkard
literature seeking to clarify what an intention is and what kinds of rea-

sons there are.
Two competing conceptions of agency come out of this, and both of
them reflect what on the surface seem to be conflicting accounts of
what it is to be responsive to reasons and to act in light of them (or to
have reasons be efficacious in one’s actions). On the one hand, it is obvi-
ous that much of what we count as responsiveness to reasons is some-
thing we learn, and we start out in life not in full possession of the ca-
pacities for such responsiveness. However, on one conception, we be-
come agents (or become more fully agents) when we learn to do this,
and once we have learned to do this, we are agents all on our own. On
this conception, a primate’s acquiring this ability to be responsive to rea-
sons is analogous to that of an agent acquiring a language. A person typ-
ically learns, say, English by being immersed in a community of English
speakers. What counts as speaking English depends on the possible rec-
ognition of other English speakers. However, once has acquired the
competence to speak English, one can be a speaker even if all the other
speakers around one vanish (just as Marie Smith was the last speaker of
Eyak).2 Having become one, one could in principle remain a rational
agent even if all others have vanished from the scene.3 For the sake of
marking this distinction, let us call it an “individualist” conception.
On the other conception, we never lose our dependence on the
shared norms in which we live. Our own subjectivity precipitates, as it
were, out of intersubjectivity. Agents respond to social norms, and agen-
cy itself, as the capacity to respond to social norms, is itself another so-
cial norm. Or, if the point is put in a more naturalist way of looking at
things, agents are creatures who learn to respond to social norms in the
sense of facts about social expectations. For the sake of demarcation, we
can call this an “intersubjectivist” conception.
Where things move away from truism to contention occurs at least
when it comes to specifying what it is to be responsive to reasons as
Obituary of Marie Smith in The Economist, Feb 7th 2008.

2
This is a version of an argument McDowell makes against Pippin in “Towards a

3
Reading of Hegel on Action in the ‘Reason’ Chapter of the Phenomenology”, in

McDowell, J. H., Having the World in View: Essays on Kant, Hegel, and Sellars, Cambridge,
Mass., Harvard University Press, 2009, pp. 66-184.
Contrasting Concepts of Agency and the Space of Reasons 17
such. One way of taking the intersubjectivist position is to see agency as

essentially responding to social facts. Human beings, so it is said, are
outfitted by evolution to respond to matters such as social expectations
as norms, as guides for action. Responsiveness to reasons is just a partic-
ular and very complicated kind of responsiveness to environmental cues.
In the tradition of thought that comes down from Kant to Hegel and
even to Heidegger, responsiveness to reasons is not fully assimilable to
responses to social cues.4 There is a kind of self-consciousness that has
to be at work in responsiveness to reasons. One must be able to discern
at any given point that what one took to be a good reason for belief, for
feeling, or for action was in fact not a reason at all.5 (John McDowell’s
phrase “responding to reasons as reasons” captures that idea nicely and
succinctly). Responding to a reason in that sense is not just responding
to an environmental cue, it is taking something to justify something else,
and thus ultimately taking it as fodder for inference (even if no inference
is in fact drawn). Self-consciousness, in this sense, is not an awareness of
oneself as an object or even as a different kind of object from those in
the natural world. Self-consciousness in this sense consists in locating
oneself in a social space of norms that license moves to other norms
and give warrant to certain actions.6 In doing so, there also has to be a
recognition on the part of the agent that the space of reasons always
outstrips all the particular considerations that might be on offer as puta-
4 See Pippin, R. B., Hegel on Self-Consciousness: Desire and Death in Hegel’s Phenomenology of
Spirit, Princeton, Princeton University Press, 2010; Pinkard, T. P., Hegel’s Naturalism: Mind,
Nature, and the Final Ends of Life, Oxford, Oxford University Press, 2012, p. 240.
5 Another statement from McDowell: “The image of self-legislation had better be
compatible with an image according to which something in the makeup of rational sub-
jects enables them to recognize an authority that the norms of reason possess anyway.
The authority of the norms is not created by being recognized. The point of the self-leg-
islating image is just that subjecting oneself to their authority is not handing over control
of the relevant areas of one’s life to a foreign power. The power is in oneself, in whatever
it is about one that enables one to recognize that the norms are authoritative. But that
their authority deserves to be recognized does not depend on one’s recognizing it. Seen
in this light, the autonomy idea is a version of the basic commitment of rationalism”
(McDowell, J. H., “Self-Determining Subjectivity and External Constraint”, in McDowell,
J. H., Having the World in View: Essays on Kant, Hegel, and Sellars, p. 105).
6 See Pinkard, T. P., Hegel’s Phenomenology: the Sociality of Reason, Cambridge, Cambridge
University Press, 1994, p. vii, p. 451.

18 Terry Pinkard
tive reasons. It is, after all, a familiar fact that what has been taken as a
reason by highly intelligent sensitive people for centuries can turn out
not to be a reason at all. Racism and sexism are only two of the most ob-
vious examples. Ultimately, an agent has to be able to step back, so it
seems, and decide for themselves whether such and such warrants such a
belief or action.7
This might suggest the authority of reasons is relative to the social
space in which they move, and thus that all authority, even of reason,
itself only exists as socially recognized authority.8 However, if all author-
ity is authority only by being recognized as such, an infinite regress
quickly springs up. If A is authoritative vis-à-vis B, A cannot acquire its
authority from B. A’s authority must be bestowed on it from C, ad
infinitum. The regress, so it seems, would be stopped only by there being
something that is authoritative but whose authority is not dependent on
anything else –that is, on its being unconditionally authoritative. (Some-
thing like either the space of reasons, God’s commands, or the early ide-
alists conception of intellectual intuition would be the major contestants
for what would be unconditionally authoritative in that sense). Thus, so it
seems, in the case of agency, agency must get its authority from some-
thing or someone that does not require recognition from the agent for its
authority. The authority of the space of reasons has to already be in
place for agents to recognize another as acting under its authority.
In his discussion of self-consciousness in the 1807 Phenomenology,
Hegel contrasts two conceptions of what it is to be acting in the space
of reasons. On the one hand, there is the conception of a self-conscious
7 See McDowell, J. H., “Autonomy and Its Burdens”, in Dottori, R. (ed.), Autonomy of
Reason? Autonomie der Vernunft?: Proceedings of the V Meeting Italian-American Philosophy,

Münster, Lit Verlag, 2009, pp. 191-199: “To count as responsive to reasons as the reasons
they are, one must be able to step back from the fact that a certain circumstance, for in-
stance perceived danger, inclines one in a certain direction, for instance towards fleeing,
and raise the question whether one should be so inclined”.
8 In one of his statements of his view, Robert Pippin seems to inclining that way:
“The radical Hegelian claim, which need not be an issue here, is that all having such au-
thority amounts to is being acknowledged –under the right conditions and in the right
way– to have such authority” (Pippin, R. B., Hegel on Self-Consciousness, pp. 76-77). A lot
turns on his phrase, “in the right way”. The premise is not self-evidently true. Various re-
ligious conceptions and certain forms of Platonism could hold that there is an eternal au-
thority that does not depend on recognition from anything or anybody.
agent who appeals to reasons to satisfy his desires. This entails not
merely experiencing the desire as a lack (say, of nutrition) but as some-
thing that may or may not figure as important in an agent’s economy of
desire –that is, may or may not serve as the basis of a commitment to
something.9
Desires can be impeded in that the object of desire can resist. Resis-
tance, however, is not the same as challenging the desire-as-a-reason.
Whatever reasons there may be not catching a fish and eating it, the fish
itself cannot raise them. As far as appealing to the space of reasons goes,
the individual agent is vis-à-vis the fish self-sufficient. He need not ap-
peal to anything outside of the space of reasons he inhabits and which
the fish does not. Another self-conscious agent, however, can offer that
kind of challenge. He can, for example, withhold recognition, and in
such a case, if all authority requires recognition, then the self-sufficient
authority of such an agent is undermined or denied. When another self-
conscious agent –that is, an agent who is capable not merely of re-
sponding to reasons but to making avowals of where he stands in the
space of reasons– shows up, the situation changes from “Am I getting
what I want?” to that of “Who is setting the rules for this situation?”.
If there is no unmediated access to the space of reasons, and all such
appeal to such a space is always made from within a determinate space
of reasons –what both Wittgenstein and Hegel called a form of life–
then the possibility of struggles over recognition is always present. In
such a struggle, the infinite regress of recognition –that A recognizes B,
but to have the authority to recognize B, A must be recognized by C,
who has such authority only by being recognized by D, etc. (Although it
might look as if this could be solved by making the relation of recog-
nition transitive and therefore reciprocal –such that A recognizes B, B
recognizes C, and C recognizes A, where transitivity would be supposed
to supply full and mutual recognition10 –that will not work unless one of
9 This is Robert Brandom’s distinction. See Brandom, R., “The Structure of Desire
and Recognition: Self-Consciousness and Self-Constitution”, in Grøn, A. – Raffnsøe-

Møller, M. – Sørensen, A. (eds.), Dialectics, Self-Consciousness, and Recognition, Aarhus, NSU
Press, 2009, pp. 140-171. See also Pippin, R. B., Hegel on Self-Consciousness.
10 This is Brandom’s solution in Brandom, R., “The Structure of Desire and
Recognition: Self-Consciousness and Self-Constitution”. If no individual has any author-

ity to begin with, expanding the circle will not grant more authority, even though it might
20 Terry Pinkard
the participants already has the authority to start the chain. I cannot be-
come king of the world by recognizing one of my friends who recog-
nizes another as having the authority to recognize me as king of the
world).
Without any further mediation in the form of practice-based or insti-
tutionally shared commitments, the regress will only be practically
stopped by one of claimant’s willingness to put recognition of him as
“the” authority in such basic matters ahead of everything else, even life
itself, and the other claimant, fearing for his life, accepting the authority
of the first one. (Note that he accepts the other’s authority, not merely
the other’s power to compel him, even though the authoritative figure
will be able to remain in authority only insofar as he continues to wield
power. Otherwise, the conflict and struggle will continually begin again,
as in fact it usually does). In that case, the philosophical puzzle of the in-
finite regress established by claiming authority is stopped not rationally
but irrationally, violently. The person who has seized authority –let us call
him “the master”– has also established de facto authority.11 He in effect
says: I issue entitlements, but others need not entitle me to do so. I am
self-entitled, entitled by nature, entitled by god, or perhaps as in neo-
Hegelian Marxism, entitled by history. You must recognize me, but I do
not recognize you.
Two agents confronting each other who are operating only within the
first conception need know nothing about power. The only force they
recognize is the unforced force of the better argument. Had they but
world enough and time, they could wait an infinite time to decide which
argument had the better force. The space of reasons is, after all, infinite
in the sense that, metaphorically speaking, it recognizes no authority be-
yond itself and thus no authoritative limits outside of itself.
bestow the appearance of authority on one or more of them. This at first might seem
counter-intuitive: If we form a club (say, the readers of this footnote club), and we elect
one of us as the chair of the club and even give her extended powers (such as the power
to expel members for refractory behavior), it might seem as if that is all there is to the
story. That assumes that any of us has the authority to start a club in the first place.
11 He may or may not also claim that he and he only is better situated or better suited
to understand just what the space of reasons requires in these particular circumstances.
That is the role played by historical context and ideology.
Agents are not infinite in anything like that sense. They never have
world enough and time. Finite agents need a finite solution to an infinite
problem, and they thus must deal in the world of power. In practice,
what counts as a reason will be to a large extent a function of the
boundaries of the form of life in which it is given, and that will depend
on exercises of power. Agents stand in both conceptions. Reason seems
to outstrip all finite, contingent exercises of itself, and what counts as a
reason will always appear as a function of the forms of life of the finite,
contingent people who rely on them. There is “reason” as a metaphoric-
al substantive only insofar as there are beings who give and ask for rea-
sons. However, this metaphorical substantive of “reason” also outruns
all attempts to pin it down to only those reasons given at any particular
time by particular people. To continue with the substantivized metaphor,
“Reason” seems to fragment itself into contingent historically bounded
practices of giving reasons and into an overarching conception of an un-
bounded reason that is immune to the failures its historically conditioned
instances have suffered.
On the whole, this looks like an antinomy. On the one hand, there is
the obvious fact that people must make choices about what counts as a
reason under conditions that do not allow for an infinite probing of the
space of reasons. What counts as a reason is always relative to the social
and historical position of the agent. On the other hand, any such appeal
to limited reasons must confront the nature of authority which tends to
generate an infinite regress that has to end in the self-establishing nature
of rational authority itself (that is, in the impossibility of getting outside
the space of reasons to justify that space itself).
There are two ways to take this, and Hegel at one point in his devel-
opment took both of them.12 One way is to see that the space of rea-
sons is eternal and unchanging and that we come to elucidate its struc-
ture only gradually and always only partially. The space of reasons is
eternal and unchanging –it is an ideal to which we infinitely approximate.
12 The shift from the first view to the second view takes place roughly around 1804-
1805 as he moves away from his Schellingian conceptions into the mature ones that find
expression in the 1807 Phenomenology.
22 Terry Pinkard
The eternal space of reasons always outstrips all finite attempts at giving
articulation to it, but we can get better or worse at various times at iden-
tifying what it requires.13 We ourselves may develop over time to become
more rational as we come to see that considerations that we had thought
counted for much in fact were not good reasons at all (again, the exam-
ples of sexism and racism come to mind as the more obvious contempo-
rary examples). On what has been disparagingly called the “Whig inter-
pretation of history”, the movement of history represents increasing
progress in discovering what this eternal reason requires.
We may develop, but Reason stays the same.
Or we could take it that the space of reasons itself has a develop-
ment. That is, if what counts as a reason depends on the structures of
recognition, failed recognition, and withheld recognition at work in
specific social and political shapes, then it as these shapes develop, so
would the structure of reason. The inherent normativity of a claim to
reason –that it outstrips all its specific embodiments– means that it is al-
ways capable of such development.14 As it develops, some types of con-
siderations that were reasons fall out, and the explanation for why they
were taken wrongly to be reasons falls on the side of social, historical
and perhaps even psychological considerations.
13 There is also the alternative of denying that the space of reasons can in any way be
as “unbounded” as it claims and in fact must rely on something more or less beyond the
human space of reasons, say, in a deity who issues commands and to whom obedience is
required (a “master of all masters”, as it were). This is a common enough claim in vari-
ous religious approaches, and it goes beyond what is at stake in rationalist “stoical” con-
ceptions. It holds that what is at stake is an eternal deity giving structure to things, of
which the human space of reasons is one inherently limited aspect. Johannine Chris-
tianity, for example, identifies God with the Logos (“Reason”) and claims that the Logos
became flesh in Jesus. The supra-human authority of the Logos took on human form
and thus mediated between Reason (the Logos) and the finite human world. How this
mediation is to be done is, however, a mystery.
14 Reason, as a competence for making inferences and providing grounds for judg-
ments, always carries within it the possible demand for a further ground. However
bounded it always is, it seeks to make that boundedness intelligible. How is intelligible
that the space of reasons, as always embodied, always outstrips its particular embodi-
ments is another issue, calling for something like a ground-level account of the logic, as it
were, of such rational normativity itself.
Reason develops, and as it develops, so do we.

Hegel held the first of these views until about 1805 or 1806, and then
he abruptly changed into holding the second of these views as he was
writing the Phenomenology in Jena.15 This change in his views came about
by way of his confrontation with Schelling’s views during the period of
their joint collaboration in Jena.
What especially drew Hegel’s attention was what at the time drew
everybody’s attention: Schelling’s creation of what Schelling took to be a
new form of philosophy itself as Naturphilosophie. The Naturphilosophie
was not a “philosophy of nature”, but supposedly a way of thinking
through nature itself to a conclusion about subjectivity. What drove
Schelling to this occurred in three steps. In the first step, for Schelling,
like all the idealists in the wake of Kant, the third antinomy between nat-
ural causation and freedom loomed large, since it seemed that Kant had
drawn exactly the right conclusion from it.
If we take nature as we must take it, namely, as a deterministic causal
system, then there is no room for freedom in such a nature unless agents
can be shown to exercise a special form of causality that is distinct from
all other natural forms of causality. That is, to be free, humans must be
able to exercise a causality over their own actions that is not itself the
result of some other causal series. Like the other idealists, Schelling also
found Kant’s own transcendental solution to the problem to be un-
satisfactory, namely, the idea that we must practically commit ourselves
to a conception of ourselves as free without being able to theoretically
comprehend how such freedom is possible, and that transcendental
idealism, with its commitment to the realm of unknowable things in
themselves, provides just the right framework for holding such a view.
If one rejects Kant’s transcendental idealist thesis that, first, we can
know nothing about things in themselves, and, second, this underwrites
the permissibility of committing ourselves to the idea that we do indeed
15 The reasons for this shift are explored in Pinkard, T. P., “From Schelling’s
Naturalism to Hegel’s Naturalism”, in Haag, J. – Wild, M. (Hrsg.), Übergänge, diskursiv oder

intuitiv?: Essays zu Eckart Försters Die 25 Jahre der Philosophie, Frankfurt a. M., Klostermann,
2013. I by and large accept, with some major caveats, Förster’s argument about the cru-
cial role played by Hegel’s new acquaintance with Goethe’s writings on metamorphosis in
plants.
24 Terry Pinkard
have this special causality –nobody can show that we can, and nobody
can show that we cannot– then one is stuck with the antinomy itself.16
What was crucial for Schelling’s second step was his general commit-
ment to the idea that what had to change was not our conception of
subjectivity but our idea of nature. The concept of nature had to be ex-
pansive enough for there to be room within it for rational activity. This
concept of nature, moreover, had to be compatible with what the natural
sciences claimed about nature. This coalesced with Schelling’s insight
that both Kant and Fichte had not exhausted all the possibilities for ide-
alism. On Schelling’s view, both Kant and Fichte had argued that the dis-
tinction between subjects and objects was not itself a matter of their
being two different kinds of things or “stuff ”. It was a matter of sub-
jects themselves introducing the distinction by taking up a rational stance
to the world of objects. That is, they took the distinction between sub-
jects and objects not to be an objective distinction but instead a sub-
jective one instituted by the subjects themselves. However, Schelling took
a remark by Kant in his third Critique to be indicative of a third possi-
bility. Kant remarked that in the experience of natural beauty we are
oriented by the “indeterminate concept of the supersensible substrate of
appearances” that is itself “neither nature nor freedom and yet is linked
with the basis of freedom, the supersensible” and this opened the door
to the third alternative: The opposition between subject and object is
neither objective nor subjective but something neutral between them.17
Schelling’s third step was the development of the Naturphilosophie
itself. The natural sciences of the time, so Schelling argued, already
showed us the way. Newtonian physics had given a precise account of
what was at work in matter in motion. However, the discoveries about
electricity and magnetism in the later eighteenth century had shown, or
so Schelling could convincingly argue at the time, that there were forces
that could not be explained simply in terms of matter in motion. More-
over, there were elements of chemistry –as the science that studies how
new substances are created out of the combinations of other sub-
16 The in’s and out’s of Kant’s transcendental idealist way out are of course compli-
cated, and we can leave them aside here.
17 See Kant, I., Critique of Judgment, translated by W. S. Pluhar, Indianapolis, Hackett,
1987, § 40, p. 159f.

stances– that could not be explained simply by electricity and magnetism

itself, and likewise the structure and behavior of organisms could not be
explained by chemistry. Mechanics, physics, chemistry and biology had
made astonishing progress in their own domains, but they could not
offer an account of how it was that those domains hung together as a
whole. Schelling’s rather bold hypothesis was that there were something
like metaphysical forces at work in nature in additional to the forces dis-
covered and tested by empirical science, and that these forces could be
discerned only by a form of “intellectual intuition”. Nature developed
into domains that embodied the unity of competing metaphysical forces
until they reached an “indifference point” at which they generated the
new domains. Even though physics can empirically study how heat, light,
electricity and magnetism function and interact, mechanics itself (matter
in motion, billiard balls colliding with each other) could not explain how,
say, magnetism arises. Schelling called these metaphysical forces Potenzen,
the higher powers that nature creates out of its lower powers. For
Schelling, the a priori intellectual intuition that delineates the various
Potenzen is like the a priori intuition that Kant invoked for arithmetic and
geometry. Just as one must construct triangles and tangents to provide a
proof in geometry (where one just “sees” it), the Naturphilosoph con-
structs proofs of the Potenzen by drawing on the natural sciences to show
where the gaps in the domains of nature must be filled in. Ultimately, a
Naturphilosophie gives us a conception of what Kant had called the un-
conditioned and what Schelling more exuberantly called the “absolute”
as nature propelling itself through various indifference points until “life”
tips over into “subjectivity”, and nature gives a rational account of itself
in philosophy. For Schelling, “nature” itself is the “unconditioned”, but
only nature as thinking of itself in that way, and nature thinking of itself
in that way is the object of an “intellectual intuition”, not a discursive
thought.
In his early work in Jena, Hegel took over the Schellingian concep-
tions of the Potenzen and put them to work for his own purposes. After
Schelling’s departure from Jena in 1803 and the strained personal rela-
tions between them, Hegel dropped the language of the Potenzen and be-
gan to work out his own ideas about a logic of such accounts instead of
a Naturphilosophie in the Schellingian sense. He laid out his objections to
Schelling rather obliquely, but it was clear that he had Schelling in mind
26 Terry Pinkard
when he characterized that conception of the absolute as the night in

which all cows are black.18 However, Hegel kept the idea that we had to a
conception of nature that had room for rational activity within itself,
even though Hegel’s naturalism developed differently and distinctly from
Schelling’s naturalism.
If not nature as working itself through its “Potenzen”, what then is
Hegel’s absolute? The most obvious answer is: Geist, spirit, mindedness.
But what is Geist? We can get a clue to this from the lectures Hegel was
giving as he was commencing with the writing of the Phenomenology and
as he was working the text out. These are the lectures on nature and
spirit from 1805-06. When in the lectures Hegel introduces the concept
of spirit (as proceeding out of “the organic”), he uses a metaphor of
“the night” emerging into daylight (a metaphor to which he later often
returned). If the first activity of spirit consists in its representing a world
external to its own activities, then the first activity of spirit is its aware-
ness in intuition of an objective world distinct from itself, a distinction it
draws itself, not a distinction that is itself an observable part of nature.
This kind of intuitive awareness is partially continuous with that of ani-
mal awareness of itself and its environment.19 However, viewed at this
level of abstraction, the agent is a “subject-object”, a body perceived
from the “inside” of a subjective quasi-animal awareness that projects
outward its intention to act in the world, and in distinguishing its
intuitions-as-representations from the object represented within one
consciousness, the agent goes beyond such animal normativity. (To in-
18 In his lectures on Naturphilosophie, Hegel also expressed his dismay at this reliance
on intellectual intuition: “Wenn es dem Bewußtsein nicht saurer gemacht würde, die
Wahrheit zu erkennen, sondern man sich nur auf den Dreifuß zu setzen und Orakel zu
sprechen brauchte, so wäre freilich die Arbeit des Denkens gespart” (Hegel, G. W. F.,
Enzyklopädie der philosophischen Wissenschaften, II, Frankfurt a. M., Suhrkamp, 1969, p. 18,
§246 Zusatz). In his later lectures on the history of philosophy, Hegel had this to say
about Schelling: “Oft braucht Schelling die Form Spinozas, stellt Axiome auf. Man will,
wenn man philosophiert, daß es so ist, bewiesen haben. Wird aber mit der intellektuellen
Anschauung angefangen, so ist das Assertion, Orakel, das man sich gefallen lassen soll,
weil die Forderung gemacht ist, daß man intellektuell anschaue” (Hegel, G. W. F.,
Vorlesungen über die Geschichte der Philosophie, III, Frankfurt a. M., Suhrkamp, 1969, p. 435).
19 “...er ist, wie das Tier, die Zeit, die für sich ist, und ebenso Freiheit der Zeit”
(Hegel, G. W. F., Jenaer Realphilosophie. Vorlesungsmanuskripte zur Philosophie der Natur und des
Geistes von 1805-1806, edited by J. Hoffmeister, Hamburg, Meiner, 1967, p. 179).
dicate this, Hegel even calls it an act of the imagination representing

things, making reference to Kant’s idea of a faculty that has a foot in
both sensibility and the understanding).20 Hegel calls such an intuitive
representation now regarded as a representation –that is, as being pos-
sibly true or false– an “image”, das Bild. To distinguish this kind of
awareness of things as an embodied “subject-object” from a more
classical picture of the mind as always aware of itself, Hegel uses the
metaphor of the “night” for the way in which our representational ac-
tivities are not to be construed as internal mental events about which we
are directly aware but as elements of an embodied activity of repre-
senting the world to ourselves. They are, to use a shorthand that itself
needs further elaboration, competencies and not internal mental states.
As he puts the same point much later in his Encyclopedia, the fluency
involved in such competencies “consists in having the particular knowl-
edge or kind of activities immediately to mind in any case that occurs,
even, we may say, immediate in our very limbs, in an activity directed out-
wards”.21
Moreover, when Hegel discusses in 1806 how spirit in its activity goes
beyond its intuitive representations of itself and the world, he says it
does so by virtue of “naming” things. To “name” something in this 1806
Hegelian sense is to enter into the space of reasons. To make this point,
Hegel notes that “this is the first power of creation exercised by spirit. Adam
gave a name to all things. This is the sovereign right (Majestätsrecht) and
the first taking-possession of all nature –or the creation of nature out of
Spirit [itself]”.22 The very idea that spirit’s Majestätsrecht consists in
20 “...so ist er vorstellende Einbildungskraft überhaupt” (Hegel, G. W. F., Jenaer Real-
philosophie, pp. 179-180).

21 Hegel, G. W. F, Enzyklopädie der philosophischen Wissenschaften, I, §66; Hegel, G. W. F.,
The Encyclopaedia Logic, with the Zusätze: Part I of the Encyclopaedia of Philosophical Sciences with
the Zusätze, translated by T. F. Geraets, W. A. Suchting and H. S. Harris, Indianapolis,
Hackett, 1991, p. 115. In the passage cited, Hegel goes on to add, “In all these cases,
immediacy of knowledge not only does not exclude mediation, but the two are so bound
together that immediate knowledge is even the product and result of mediated knowl-
edge”.
22 “Dies ist die erste Schöpferkraft, die der Geist ausübt; Adam gab allen Dingen
einen Namen, dies ist das Majestätsrecht und erste Besitzergreifung der ganzen Nature,
oder das Schaffen derselben aus dem Geiste” (Hegel, G. W. F., Jenaer Realphilosophie, p. viii,
p. 290, p. 183). That the talk of “Kraft” is not there in the “Self-Consciousness” section
28 Terry Pinkard
“naming” things, and that this is what is meant by speaking of “the crea-
tion of nature out of spirit” is all the more evidence of Hegel’s robust
rejection of Schelling’s metaphysical monism, that is, of Schelling’s ac-
count of nature and spirit emerging according to the metaphysical laws
governing a self-conscious Spinozistic substance. That it was “Adam”
who “named” all things and not something else (for example, God) is
also evidence that Hegel had in mind something like human agency
thinking about its own conditions which is to serve as the “absolute”,
not Schelling’s “nature”. To put this point in more deflationary terms,
“spirit” first “names” things and then reflects on its own activity of
naming and what this means for how it is to think of itself and the
world. As Hegel puts it in the 1805-06 lectures, “the world, nature, is no
longer a realm of images that have been internally sublated and have no
being. It is rather a realm of names”, and the activity, or movement, of
spirit consists in “the insubstantial (stofflose) relating of names to
names”.23
In his mature system (especially in the Logic), Hegel thought that the
conception of agency first outlined in the 1805-06 lectures could be
more clearly laid out in terms of the kinds of accounts at work in such a
conception of agency.
First, there is the inescapable argument that the authority of reason
has to terminate in something like reason’s being the last step in an argu-
ment about who or what has authority. There is no “positive”, simply
“given” beginning to thought that is not subject to reason’s own “un-
bounded” take on itself. The “doctrine of being” in the Science of Logic
makes this argument. Hegel thinks he has shown there that all the forms
of argument in which such infinite regresses occur do not warrant any
kind of appeal to “intellectual intuition” or “givenness”. What emerges
out of the logic at work in the “doctrine of being” is that the space of
reasons must be authoritative without having to take its authority from
of the Phenomenology also suggests that the lectures were sketched before the writing of
the book.
23 “Die Welt, die Natur ist nicht mehr rein Reich von Bildern, innerlich aufgehobene,
die kein Sein haben, sondern ein Reich der Namen” (Hegel, G. W. F., Jenaer Realphilosophie,
p. 184). “Das Festhalten nun einer solchen Beziehung des Namens und der Namen ist
DIE STOFFLOSE BESCHÄFTIGUNG und Bewegung des Geistes mit sich” (Hegel, G. W. F.,
Jenaer Realphilosophie, p. 187).
anything other than itself. Or, to put it in Hegel’s preferred terminology,

there is a valid conception of the infinite that does not require it to be
another “positivity”. The infinite regress that arises in elementary
thought about the qualitative differences among things does not require
such a given, and the idea of the infinite as it appeared in the calculus of
the time also does not require any such givenness. The “doctrine of
being” concludes with an affirmation that the space of reasons is un-
bounded.
Second, for reasons to be reasons, they must however show them-
selves to be real, to be at work in the world. The space of reasons “must
appear”, to put it in more Hegelian terms. More concretely put: The in-
tentions informed by the reasons of which the agent is conscious and
which direct his behavior must make an appearance in the phenomenal
world to be real at all. This comes about in what Hegel calls the “doc-
trine of essence”. To be real, agency must express itself in the world.
Agents do this by causing themselves to act and by expressing their rea-
sons for action in the shape the actions take. The space of reasons, to be
real, must always take a determinate, bounded shape in the phenomenal
world.24
Third, for agency to be real and express itself, it must move in the
space of reasons, and the space of reasons itself must have a way of in-
telligibly articulating itself. In particular, for the agent to be rational, he
must be able to translate what is generally required of him into a deter-
minate, historically bounded action that is expressive of the shape reason
has taken in a form of life. The unbounded space of reasons has to limit
itself in action for it to be real at all.25 (Hegel argues this in the “doctrine
of the concept”, the third and final “book” of the Logic).
24 The kind of infinite regress encountered in the dialectic of “Being” is the more
usual type. In “Essence”, it is that of the back and forth from inferring from appearances
to what explains them and inferring from explanatory structures to how they must ap-
pear. What is behind appearances, for example, may look to be what is independent in
the explanatory relationship, but it continually turns out that whatever determinateness it
has, it has only in relation to appearance and thus is not independent of its appearance.
25 Hegel, G. W. F., Hegel’s Science of Logic, translated by A. V. Miller, London – New
York, Humanity Books, 1969, p. 746: “The finitude of rationality has, as remarked, this
side, that the end enters into relationship with the presupposition, that is, with the exter-
nality of the object. In the immediate relation to the object, it would itself enter into the
sphere of mechanism or chemism and thereby be subject to contingency and the loss of
30 Terry Pinkard
This kind of laying out of various ways of accounting for the relation
between reasons and action argues that what seems like an irreconcilable
opposition is actually intelligible. The space of reasons must express it-
self in particular concrete ways in the world to be real at all (or at least to
be wirklich, “actual” in Hegel’s own special terminology). This view itself
has suggested to Hegel’s readers that he has in mind a kind of neo-
Platonic Christian view of the space of reasons as the divine logos that
makes itself flesh and provides the propulsive movement to human his-
tory. However, there is nothing in this kind of account that entails that
there is a divine logos at work in that sense. In the most trivial sense, the
space of reasons was always possible in the world as it took shape from
the Big Bang, but that does not mean except for the most extreme modal
realists that there already was “a” logos waiting for its fulfillment, any
more than the fact that there are English speakers means that English
had to have existed in the eons before the earth took shape, waiting for
its fulfillment. To argue for the development of the space of reasons as
occurring within a natural context of primates learning to give and ask
for reasons does not commit one to a Johannine metaphysics of “rea-
son”.26
That, however, returns us to the question Hegel then had to ask:
How could reason develop? After all, once masters have installed them-
selves as authorities, they have powerful motivations not to throw the
considerations that keep them in power into question. As the kind of
desiring, finite, not completely rational people that they (and everybody
else) are, they will more likely than not find little to recommend a change
in the space of reasons. They will want to get what they want, and to
have good reason to do just that. That in itself provides them with a rea-
son for having other, non-authority-bearing people to work for them.
They command, others obey. They want, and others get them what they
want. From time to time, surely, some of them (the masters) also get in
its determination as the concept that exists in and for itself”. See Hegel, G. W. F.,
Wissenschaft der Logik, II, Frankfurt a. M., Suhrkamp, 1969, pp. 452-453.
26 The most subtle elaboration of this interpretation of Hegel as asserting a kind of
view of divine rationality is Rolf-Peter Horstmann’s conception of Hegel’s “monism of

reason”. See Horstmann, R.-P., Die Grenzen der Vernunft: Eine Untersuchung zu Zielen und
Motiven des deutschen Idealismus, Frankfurt a. M., Klostermann, 1991.
each other’s way, and violence brings about a new set of that group
getting what it wants by getting rid of or demoting one set. That in itself
does not fundamentally change the social positions themselves even if it
does from time to time fundamentally change the people occupying
those positions. There are still those who want, and others to get them
what they want, even if the speaking roles from time to time shift actors.
Those who have to get them what they want are those who work on
things, and working on things requires them to bend their own wills not
merely to the wills of the masters who call for the work but also to the
natural materials on which they work. In learning how to fashion natural
material into works that are functional and pleasing (or even beautiful),
the worker turned artisan is compelled by the material into seeing that
there are standards at work that go beyond those contained in the ex-
isting patterns of recognition. In working on natural things and shaping
them, the artisan establishes a sense of doing things well that is re-
sponsive to the demands that the natural world places on us. Nature may
seem from the theoretical point of view to be divisible in a thousand
arbitrary ways, but the worker has to learn that natural materials place
specific rational demands on him that transcend his or anybody else’s
powers. In working on the material world, the artisan experiences a unity
between his ability to confer new statuses on matter that must none-
theless bend themselves to the laws governing matter if they are to suc-
ceed. It is one thing to create something that fills a social status, such as
“a chalice”. It is another thing to understand that one can make a chalice
only if one pays attention to the way clay can be fired, wood can be cut,
or glass can be made and blown. The masters may find that their various
ways of giving accounts to each other collide and come into conflict, but
those ways of giving accounts are malleable. What sets the development
of reason into motion is the idea of there being standards that transcend
not merely those of the masters who have others work for them, but
transcend even the vocabularies in which the masters converse.27 The
history of ideas is also a history of material culture.
27 Richard Rorty’s conception of “changing vocabularies” in history is a contem-
porary pragmatist version of the idealist conception of history as driven by changes in

ideas. The alternative conception (barely even) sketched here does not deny that changes
in ideas are crucial to the development of “reason” (to stay with the metaphorical sub-
32 Terry Pinkard
The idea that reason itself develops (and we along with it) rests on a
conception of reasons as the standards that emerge first from primates
adjusting their behavior to achieve what it is they want and adjusting
their behavior in light of how it coordinates with or contrasts with oth-
ers. With human language in play, a new standard not merely of suc-
cessful action but also of truth as “getting it right” in light of standards
that outstrip our current accounts of them comes to play a crucial role.
Human primates become agents as they recognize each other as agents
by learning to move in the space of reasons. They acquire new norma-
tive statuses as they develop themselves within the developing space of
reasons, and these in turn provide agents with new competences in the
space of reasons.
The two conceptions of agency are those of agents as self-sufficient
reasoners and as non-self-sufficient agents within a structure of recog-
nition that fixes their statuses. I have argued that both conceptions are
necessary, and that although there are two conceptions of the space of
reasons, the conception of reason developing itself is more consistent
with holding both conceptions together and not letting the tension be-
tween them end up undermining the space of reasons itself.
This is not the standard picture at work in philosophical discussion.
The more standard picture is, as we might put it, “stoic”: The space of
reasons is eternal and unchanging. We aspire to it, but we finite creatures
–although we may from time to time get better at it– never fully achieve
it. At most, we learn to discern some of its outlines, and we acknowledge
our own fallibility in doing so.28 The counterpart to this standard picture
stantive) but merely proposes that such changes are mediated by material culture. See
Rorty, R., Contingency, Irony, and Solidarity, Cambridge, Cambridge University Press, 1989, p.
xvi, p. 201.
28 McDowell, J. H., “Autonomy and Community. Some remarks on the second move-
ment of Brandom’s sonata” (draft): “Rationality made its appearance in the world when
our ancestors came to have a capacity to recognize the authority of reasons. The author-
ity of reasons had to be there to be recognized, and hence possibly misrecognized, as
soon as anyone had a capacity to recognize it. (No sooner, by all means; that is the appli-
cation to this case of Brandom’s point about the pre-human world.) In its first incarna-
tion, even more than at later stages in its development (including ours), the capacity
could yield knowledge of the layout of only some regions of the space of reasons, with
other regions only distortedly in view or not in view at all. Like any human capacity to
know, this capacity is partial and fallible. And knowing it to be partial and fallible is part
despairs of the space of reason at all and instead is more “skeptical”:

There is only power and the sites at which power is inscribed so as to
disguise itself as reason when in fact it is something else. These are, from
the Hegelian perspective, two contrasting concepts of agency, both of
which are within their own contexts necessary. The Hegelian claim is that
an idea of the space of reasons as itself developing, rather than being
static and eternal, is necessary to hold both of these contrasting concep-
tions together.
So Hegel contentiously said, “philosophy also is its own time grasped
in thoughts” (so ist auch die Philosophie ihre Zeit in Gedanken erfaßt), which
means, if the above is correct, that it is grasping what is required of it by
a space of reasons that develops itself only in the actual practices of hu-
mans giving and asking for reasons. All of this is a prelude to a recon-
sideration of Hegel’s philosophy of history in light of the rejection of
idealist theories of history from Marx until the present.
Georgetown University
Washington D. C., United States
of what it is to have it at all. Even at the dawn of rationality, those who had the capacity
to recognize the authority of reasons already knew that what is a reason for what was
something they and their fellows could be ignorant or wrong about, not a subject matter
that simply reflected, because it was instituted by, the subjective stances that presented
themselves as views of it”.
34 Terry Pinkard
References
Brandom, R., “The Structure of Desire and Recognition: Self-

Consciousness and Self-Constitution”, in Grøn, A. – Raffnsøe-
Møller, M. – Sørensen, A. (eds.), Dialectics, Self-Consciousness, and
Recognition, Aarhus, NSU Press, 2009, pp. 140-171.
Hegel, G. W. F., Jenaer Realphilosophie. Vorlesungsmanuskripte zur Philosophie
der Natur und des Geistes von 1805-1806, edited by J. Hoffmeister,
Hamburg, Meiner, 1967.
——, Enzyklopädie der philosophischen Wissenschaften, II, Frankfurt a. M.,
Suhrkamp, 1969
——, Vorlesungen über die Geschichte der Philosophie, III, Frankfurt a. M.,
Suhrkamp, 1969.
——, Wissenschaft der Logik, II, Frankfurt a. M., Suhrkamp, 1969.
——, Hegel’s Science of Logic, translated by A. V. Miller, London – New
York, Humanity Books, 1969
——, The Encyclopaedia Logic, with the Zusätze: Part I of the Encyclopaedia of
Philosophical Sciences with the Zusätze, translated by T. F. Geraets, W. A.
Suchting and H. S. Harris, Indianapolis, Hackett, 1991.
Horstmann, R.-P., Die Grenzen der Vernunft: Eine Untersuchung zu Zielen und
Motiven des deutschen Idealismus, Frankfurt a. M., Klostermann, 1991.
Kant, I., Critique of Judgment, translated by W. S. Pluhar, Indianapolis,
Hackett, 1987.
McDowell, J. H., “Self-Determining Subjectivity and External
Constraint”, in McDowell, J. H., Having the World in View: Essays on
Kant, Hegel, and Sellars, Cambridge, Mass., Harvard University Press,
2009, pp. 90-107.
——, “Towards a Reading of Hegel on Action in the ‘Reason’ Chapter
of the Phenomenology”, in McDowell, J. H., Having the World in View:
Essays on Kant, Hegel, and Sellars, Cambridge, Mass., Harvard University
Press, 2009, pp. 166-184.
——, “Autonomy and Its Burdens”, in Dottori, R. (ed.), Autonomy of

Reason? Autonomie der Vernunft?: Proceedings of the V Meeting Italian-
American Philosophy, Münster, Lit Verlag, 2009, pp. 191-199.
——, “Autonomy and Community. Some remarks on the second
movement of Brandom’s sonata” (draft).
Pinkard, T. P., Hegel’s Phenomenology: the Sociality of Reason, Cambridge,
Cambridge University Press, 1994, p. vii, p. 451.
——, Hegel’s Naturalism: Mind, Nature, and the Final Ends of Life, Oxford,
Oxford University Press, 2012, p. 240.
——, “From Schelling’s Naturalism to Hegel’s Naturalism”, in Haag, J. –
Wild, M. (Hrsg.), Übergänge, diskursiv oder intuitiv?: Essays zu Eckart
Försters Die 25 Jahre der Philosophie, Frankfurt a. M., Klostermann, 2013.
Pippin, R. B., Hegel on Self-Consciousness: Desire and Death in Hegel’s
Phenomenology of Spirit, Princeton, Princeton University Press, 2010.
Rorty, R., Contingency, Irony, and Solidarity, Cambridge, Cambridge
University Press, 1989.
Rousseau, J. J., Considerations on the Government of Poland and on Its Planned
Reformation, in Rousseau, J. J., The Plan for Perpetual Peace, On the
Government of Poland and other Writings on History and Politics, translated
by C. Kelly and J. Bush, edited by C. Kelly, Lebanon, NH, University
Press of New England, 2005.
ACTING AS THE INTERNAL END
OF ACTING
SEBASTIAN RÖDL
In contemporary literature, the concept of an end is often used to distin-

guish between action directed toward an end and action determined by a
maxim.1 The latter then appears to be aimless or even contrary to our
ends. And demanding to act contrary to ends seems paradoxical. So we
get the “deontological paradox”.
The concept of an end is used often. It is rarely examined.2 We will
see that what today is above all called “an end” is an external end, but
that external ends are only possible through internal ones and that an
end is therefore, first of all, an internal end. We will further see that act-
ing, as far as it is determined by a maxim, is an internal end. If acting is
so fundamentally end-directed that it is itself an end, then there cannot
seem to be a “deontological paradox”.
I proceed in three steps. I explain the concept of an end, then that of
an internal end and finally of maxim-guided action as an internal end.
1. End
The concept of an end determines what falls under it according to its

logical form. In one essential use it refers to a kind of movement, of
kinesis. The concept of movement is itself a logical concept; it refers to a
1 Michael Ridge explains: “We are assuming all reasons are teleological and hence al-
ways involve an end”, and, “Suppose one embraces the tempting view that all reasons for
acting must be teleological in form, which is to say that any principle underwriting a reason
for acting must individuate actions in terms of the states of affairs that they promote”.
The end of an action is therefore a “state of affairs” that is being promoted through this
action. See Ridge, M., “Reasons for Action: Agent-Neutral vs. Agent-Relative”, The
Stanford Encyclopedia of Philosophy (Winter 2011 Edition), ed. by E. N. Zalta,
<URL=http://plato.stanford.edu/archives/ win2011/entries/reasons-agent/>.
2 The Stanford Encyclopedia of Philosophy does not have an entry for “end”, nor for
“telos”.
38 Sebastian Rödl
form of predication, in which the progressive and the perfective aspect

are contrasted to each other. We represent this form by a schema: x is
doing A, x has done A.3 Some examples:
An avalanche is sliding down the slope. An avalanche has slid down

the slope (and has buried the valley village beneath itself).
The ice on the lake is melting in the sun. The ice has melted.
A cell is dividing. It has divided.
Frederik is reciting a poem. He has recited it.
A progressive predication represents something as moving towards

an end, the perfective as having reached this end.4 Both exclude each
other: as long as something has not yet reached the end it is moving to-
wards, it is moving, and as long as it is moving, it has not yet reached the
end it is moving towards. As the cell is dividing, it has not yet divided,
and as it has divided, it is not dividing any more. As soon as a movement
has reached its end, it is no more. Then something other than it is, name-
ly, its result. A movement is therefore finite in the sense that it is nothing
but its bringing itself to an end. A kinesis is, as Aryeh Kosman expressed
it, “on a suicide mission”.
X is doing A is not a simple, but a structured reality. When x is doing
A, it has done something; not A –it is still doing that–, but something
3 We can use “doing” as a schema because it itself exhibits the contrast of aspect. In
this, it differs from “being”.

4 We use a predication of the form in question to represent this form: It is moving
there, It has moved there. That is always the case when a logical form is articulated. As
Wittgenstein remarks, “It is so and so” can represent the general form of a proposition
only because it is itself a proposition. And “is” or “esti” can express the formal character
of the finite verb, or predication as such –as it does in Aristotle– only because it is itself
a finite verb. That is why the kind of explanation just given –that the progressive predi-
cation represents something as moving towards an end, the perfective, that it has reached
it– is a hint. It points towards the general form shared by the examples. It points towards
it and does not represent it through itself, for it only gives a further example, one that
does not draw much attention to its content and so directs us to the form. It is important
to notice the limited utility of such hints and thus the limited meaning of the words by
which we make them. In examining, for instance, the logical category of causality, we
should orient ourselves to propositions like “The sun is heating the stone” –that is Kant’s
example–, not to propositions like “This causes that”.
Acting as the Internal End of Acting 39
else, let us say, A’. But that it is doing A is not exhausted in that. Some-
thing is still missing, with which x, having done it, will have done A. As
the ice is melting at the moment, a layer has melted. But the ice has not
yet melted when that has happened; it has only melted when it has com-
pletely dissolved. As the avalanche is sliding down the slope, it has loos-
ened up and has slid a little. But it has not yet slid down, but only when it
has hit the valley. That something is happening, that is something that al-
ways goes beyond what has already happened. Kosman says a kinesis is
“on a suicide mission”. We have discussed the suicide. That a kinesis is on
a “mission” is the image for the fact that it goes beyond anything that
ever has happened and is given.
As x is doing A, it has done A’. That does not only mean that it is
doing A and that it has done A’. It means: that it is doing A consists
among other things in the fact that it has done A’. That it is doing A is
real and realized in the fact that it has done A’. When referring to this re-
lation, we use the sign “in”: x is doing A in that it has done A’. The nexus
that “in” refers to belongs to kinesis as its internal structure.
There are indefinitely many propositions x has done A’, that are con-
nected to any given proposition x is doing A in the manner in question.
The relation is thus one of something general to something particular.
For that is the general concept of the general: it is the unity of indefi-
nitely many. The form of predication of kinesis thus gives rise to a kind
of generality.
In the simplest case, A and A’ are not distinguished in their concep-
tual, but rather only in their spatial determination. A trunk drifting
downstream has drifted up to here and is drifting on. A log that is burn-
ing, has burned this much and is burning on. The ice on the lake has
melted until here and is melting on. Furthermore, in the simplest case
the conceptual determination does not define the spatial one, but the
latter depends on external conditions. The trunk drifts down to the dam,
there it cannot pass any further. The log burns until nothing more is left
of it. That is the simplest case. However, the logical relation we refer to
with “in” is itself a genus of species. And therefore kinesis is a genus of
species: movements are distinguished as movements, namely according
to their respective structure as movements. We encounter a special kind
of movement in life processes. (We perceive that they are life processes.
We will later say why they are). As DNA replicates itself by way of mito-
40 Sebastian Rödl
sis,5 a series of things happen, at the same time and one after the other,
like:
DNA replicates itself as the double helix unwinds by means of Heli-

case.
Here A and A’ differ in their conceptual determination. And when

spatial determinations enter the scene, they are defined by conceptual de-
terminations. The double helix is unwound; the base pairs are separated
one after the other. This process is completed not when there happens
to be no base pair left, but when the chromosome is completely repli-
cated. Before then the bit of the cord already replicated is not unwound.6
The DNA replicates itself in that the double helix unwinds by means
of Helicase. The “in” signifies: it is not only the case that DNA repli-
cates itself and that Helicase unwinds the double cord. Rather, that DNA
replicates itself means, among other things, –it is real and realized in
this– that Helicase unwinds the cord. But, furthermore, and that distin-
guishes this kind of kinesis, we understand why that happens by linking it
to that which it realizes. As we look on what is happening, we ask: Why
is Helicase unwinding the double cord? Answer: the DNA is replicating
itself, and Helicase is separating the cords, so that on each cord a copy
of the respective other can be synthesized. DNA is replicating itself in
that Helicase is unwinding the cords in such a way that the Helicase is
unwinding the cords because the DNA is replicating itself.
The concept of an end is a logical concept. To explicate it, we must
describe the logical form to which it refers. This is how we explicated the
concept of kinesis: by describing the logical form that defines it. In the
case of kinesis, this form is a unity of concepts in judgment, a kind of
5DNA replicates itself. Here is the “itself ” of life: the cell divides itself, the flower
opens itself, the muscle contracts itself, etc. We will soon turn to describing the logical
character that this “itself ” signifies.
6 Therefore, the workings of Helicase are formally different from that of fire: fire
burns as long as there is something there that burns. But Helicase does not separate
DNA double cords as long as there are some there. Rather, it does it until the chromo-
some is replicated. Aristotle noticed this (and used it to explain why the soul is not fire):
in contrast to fire, life processes are determined by a concept. See Aristotle, De Anima, B
4, 416a15-18.
predication. In the case of the concept of an end, it is a unity of judg-

ments. The unity in question is the one we have described: x is doing A
in that it is doing A’, and in such a way that x is doing A’ because it is
doing A. This logical form defines the concept of an end: it refers to
that which appears to the left of “in” and to the right of “because”. The
correlative concept of means refers to that which is on the other side,
respectively. DNA replicates itself –that is the end; that Helicase un-
winds the double cord is the means to this end. And so we have already
said: Helicase unwinds DNA so that on each cord a copy of the respec-
tive other can be synthesized.
Our “in” refers to a species of the relation of the general to the par-
ticular. Here we have a species of this species. The logical relation here is
at the same time explanatory: the particular does not only fall under the
general, but the latter explains the former, and that characterizes the
manner in which it falls under the general. The end, defined generally, is
therefore the general, in so far as it explains that which falls under it and
in which it is real. That is Kant’s definition of an end: “An end is the ob-
ject of a concept in so far as this concept is regarded as the cause of the
object (the real ground of its possibility); and the causality of a concept
in respect of its object is finality (forma finalis)”.7 And: “the concept of
an object, so far as it contains at the same time the ground of the actu-
ality of this object, is called its end”.8
7 Kant, I., Critique of Judgment, trans. by J. C. Meredith, ed. by N. Walker, Oxford,
Oxford University Press, 2007, §10, 5:220, p. 51.

8 Kant, I., Critique of Judgment, Introduction, IV, 5:180, p. 16. In order to be com-
pletely general, Kant describes the highest genus of the species in question: he wants to
speak in such a way that not only an action concept can be a concept of an end, but also
the concept of an artifact, and so that the definition includes natural ends as well as ratio-
nal ends. Thus he uses “object” and “concept” so generally that they refer to nothing but
the relation of the general to that which falls under it. In the cited passages, Kant calls
the concept sometimes the end and sometimes the object. That is because he views the
end sometimes as efficient and at other times as accomplished. Because Helicase is un-
winding the double cord, we can ask, What is the end?, and answer, DNA replication.
Here the end is the general, the concept. If everything goes well, the process reaches its
end and DNA is replicated, we understand precisely this, that DNA has replicated itself,
as the end of that which led up to it. That DNA has replicated, however, is an object, is a
case of the concept of DNA replication. Kant defines purposiveness correlatively to the
concept of the end as causality of a concept with regard to its object: what is explained
through the general, under which it falls, that is, through its concept, is thereby a means.
42 Sebastian Rödl
We have developed the concept of an end by describing a logical

form of everyday and scientific judgments and so the form of the reality
that is known in such judgments. We assume that these judgments artic-
ulate knowledge, or more precisely, that they are capable of articulating
knowledge. It is unimportant to us if DNA replicates itself by means of
Helicase’s unwinding of the cords. Maybe it proceeds differently. How-
ever, a revision of this theory would propound judgments of the same
form. Therefore, if we disregard the possibility of a general skepticism,
we recognize, through that manner in which we have developed the con-
cept of an end, the reality of causality of ends, its reality in what the
judgments in question represent.
There is an opinion in the literature that causality according to ends
cannot be real. However, every biological treatise is full of teleological
talk from the first to the last sentence, and nobody would want to claim
to know a priori that biology as such and as a whole is untrue. Thus this
opinion becomes a motive for the attempt to reduce the causality of
ends to another one, to that of efficient causes, the teleological to me-
chanical causality. We can say two things about this.
First, it is thought that a causality of ends cannot be real because it is
thought to be a causality of something that is yet to come in the future,
and so is not yet there, and causes something that precedes it in time.
That is nonsense indeed. And no sane mind has ever thought that some-
thing that is not is the cause of something that is. The causality of ends
is the causality of the general with respect to that which falls under it, it
is the causality of the concept with respect to its object. Causality ac-
cording to ends does not explain the present by the future and thus the
real by non-real, but rather the particular by the general and thus the real
by real.
Secondly, it is nonsense to reduce the causality of ends to the cau-
sality of efficient causes. For, teleological causality is not on the same
level as mechanical causality. The former is above the latter; the mechan-
ical relation is contained in the former as subordinated. This is because
the suitability of the means is a nexus of efficient causation. We say that
the Helicase unwinds DNA. That is how we describe the mechanical
causality of Helicase: it breaks up base pairs. At the same time we say
that it does this because DNA replicates itself in this way. We can ex-
press this as follows: the concept, that is the end, determines the efficient
cause to causality. (And that is how Kant expresses it.) Helicase is the ef-
ficient cause; it is efficacious in separating base pairs, that is its causality.
It is determined to this by the end; the end determines it to be effica-
cious in this manner: Helicase separates the base pairs because DNA
replicates itself. The same holds for the causes of all effects by which
DNA replicates itself; it holds for this because it holds for all, and it
holds for all, because it holds for each. For, the end is the general as ef-
ficacious and it is therefore efficacious as general. The causality of ends
is therefore this: that a unity of mechanical effects, a unity determined by
the concept in question (in our example the concept of DNA replica-
tion) determines a totality of efficient causes to causality. However, an
efficient cause is not a unity of efficient causes. Wanting to reduce the
causality of ends to the causality of efficient causes is just as reasonable
as wanting to reduce the general to the particular.
2. Internal End
We have determined the concept of an end in general. But there is a dif-

ference between an internal and an external end. The former is primarily
an end.
A kinesis is, as Kosman says, “on a suicide mission”; it is just that –
putting an end to itself. Once it has reached its end, it itself is no more,
but rather its result. As a movement is there in such a way as to make
space for its result, it is not an end in itself, but only through its result. It
is an end as a means to its result, which is primarily the end. As the result
of a kinesis is different from the kinesis, it is an external end. It is external
to that of which it is an end, namely the movement that is the means to
it. The reason for the externality of the end of a kinesis resides in this,
that the progressive and the perfective aspect contrast with each other.
That is why the result of the movement is external to the movement, the
end is external to its means. If there is an internal end, its predication
will manifest a different form. Aristotle distinguishes kinesis and energeia
and defines the latter by a predication in which the progressive aspect
includes the perfective aspect.9 That will be the predication of an inter-
9 See Aristotle, Metaphysics, Theta 6, 1048b18-34.

44 Sebastian Rödl
nal end. For, that the progressive aspect includes the perfective means
that the means includes the end. Therefore the means is internally pur-
posive and the end is an internal end. We represent this form of pred-
ication by a schema and divide “to do” in “kinein” and “energein”. In
English we could distinguish doing from activity. Words are irrelevant.
Their meaning lies in the logical form that they signify. Besides x kinei A
there is therefore also x energei H, besides x is doing A there is also x is ac-
tive in the manner H. We develop the concept of energeia, the concept of an
internal end, first formally; then we give examples.
In the case of an energeia, the progressive aspect includes the perfec-
tive aspect. The predication therefore has the character of a progressive
proposition; it manifests the logical structure that defines the progressive
aspect.
X energei H in that x kinei A.
As in the case of kinesis, this nexus, “in”, is not an external relation,

but the internal structure of the energeia, which is founded in its form
of predication. And as above, the “in” signifies not only that x energei H
and also x kinei A. But, that x energei H resides in this –x energei H is real
and realized in– that x kinei A. Furthermore, the progressive aspect in-
cludes the perfective aspect. X is active H does not realize itself by striv-
ing towards an end that is different from itself and which excludes it
from reality. Rather, x is active H realizes itself in such a manner that it al-
ways remains itself and maintains itself in itself. An activity does not give
out explaining movements in which it realizes itself, and when it ends,
then per accidens. It does not contain its end in the way in which a move-
ment does so.
The “in” in x energei H in x kinei A signifies to a teleological structure;
the logical relation is equally explanatory. Therefore an energeia is an end.
While the end of a kinesis is a result different from the movement, an
energeia is an end in itself. An activity is an end in itself, one that does not
wait for something that is other than itself, but that has always already
been reached as long as the activity carries on. An activity is an internal
end.
Now, the internal end is not on the same level as the external end.
Just as the causality of ends contains and underlies the causality of effi-
cient causes, so the internal end contains and underlies external ends. We
see this when we ask what can explain a purposive movement. This is to
ask how the result of a movement can be an end. For, what explains a
purposive movement is that by virtue of which its result is an end.
A purposive movement does not have a mechanical cause. It would
have to be the cause of its purposiveness; it would have to be the reason
why a unity of efficient causes explains their causality, and so explain
why the result of a movement is an end. A mechanical cause can only be
subordinated to a teleological movement as a means. We can ask, Why
are the double cords unwinding? and answer, Because the Helicase is
separating the base pairs from each other. If nothing has gone wrong,
that is happening because DNA is replicating itself. When we continue
asking, Why is the DNA replicating itself, we would cite efficient causes,
that trigger this process. But they do not stand on their own. If nothing
has gone wrong, they are subordinated to an overarching kinesis as
means. We would explain the causality of the efficient causes in question
by saying that the cell is dividing, which demands that its core divides,
which demands that the DNA replicates itself, which is why that is hap-
pening which is triggering this process, whatever that is.10
We have already encountered a manner of explaining a purposive
movement, for this mode of explanation is contained in the movement
as its internal structure:
X is doing A’ because it is doing A.
A purposive movement is explained by another, overarching move-

ment. As the latter is an end, it explains how the movement subordinated
to it and its result can be an end, namely as a means to the former super-
ordinate movement. The Helicase is unwinding DNA; that is a purposive
kinesis. Its result –the cords are separated– is a means to a superordinate
kinesis: DNA is replicating itself. This kinesis and its result are again
10 This is not the place to explicate this further, but we see that the idea that animal
perception is an input and its movement an output –and thus perception an efficient
cause of the movement– makes no sense. For, the movement in question is purposive,
wherefore no efficient cause can explain it. Perception is, rather, the formal character of
a species of teleological structure, a species of an end and or of purposiveness.
46 Sebastian Rödl
means to a superordinate one: the cell is dividing. And so on. However,

this cannot be the only form of explanation. Otherwise the explanation
would never come to rest and thus the series of explanations would not
begin. If there is nothing in which the question after the reason of a
purposive movement can come to rest, then nothing is a step on the way
to an answer to this question. Therefore, if –per impossibile– a purposive
movement could only be explained by another purposive movement,
then we would not understand any explanation of that sort, and the con-
cept of an external end would be empty.
A purposive movement cannot be explained by an efficient cause, but
by an end. It can be explained by an external end, but that is not the only
way. A purposive movement is ultimately explained by an internal end, by
an activity. As an internal end is an end, it explains the teleological struc-
ture; it explains why the result of the kinesis in question is an end. As it is
an internal end, it does not again raise the question which it answers. The
logical form that underlies the purposive kinesis and makes it possible is
therefore the form that the concept of the internal end signifies and that
Aristotle calls “energeia”. The concept of the external end contains the
concept of the internal end as its ground, the concept of the purposive
movement contains that the concept of activity.
We can further determine the concept of an internal end by ex-
plaining how it prevents the regress of external ends. It is appealing to
think that a regress is brought to an end by a magical member of the
series that steps up at the beginning or the end of the series and in which
the progression comes to a halt in a magical manner. That is never the
case. A regress arises when the form of the nexus of the members of a
series is their logical form, the form that therefore demands that the
series progresses. Thus the purposive kinesis contains in virtue of its logi-
cal form a nexus of external ends: x does A in that it does B in that it
does C in that it does D, etc. In virtue of this very form every kinesis on
which a series like this can hang, points to something other than itself as
its ground. Whenever there is a regress, a member of the series is in its
logical nature incapable of stopping it. Instead, a regress is avoided as the
unity of the series underlies the totality of its members and their connec-
tions.
We have seen how the causality of ends contains the causality of effi-
cient causes and underlies it: an end is a unity of effects of efficient
causes, and this unity is not an aggregate, but a principle: it explains that
which is collected under it. It is precisely in this manner that the internal
end contains the external end: it is a unity of external ends, which is not
an aggregate, but a principle: the ground of the external ends collected
under it.
An activity is an end, that is, something general that is efficacious and
thus is efficacious as general. Therefore it explains a purposive move-
ment by explaining a totality of such movements. Being a totality of
purposive movements, the unity of this totality has a determinate form.
So far, we have represented the teleological nexus of movements pro-
gressively: DNA is replicating itself in that Helicase is separating the
cords. The same can be said generally: DNA replicates in such a way that
Helicase separates the cords. The activity, that is the underlying unity of
the series of teleological movements, resides in this generality. Now, the
general teleological nexus also seems to progress indefinitely; in this case
it would not be a unity. However, as it is general, this series is not a
temporal sequence. And so it can be a unity by returning to itself. The
question, What is first, the hen or the egg? has no answer because it asks
generally. This hen comes before this egg, she laid it, and this egg comes
before that hen, she hatched from it. But the hen and the egg do not have
a temporal location. That is why the hen can be for the sake of the egg
and the egg for the sake of the hen, and so the hen for the hen and the
egg for the egg. In the series that returns to itself each of its members is
a means as well as an end.11 The logical form of energeia, in which the
progressive aspect includes the perfective aspect, is therefore more pre-
cisely this logical form: a teleological system in which each member is
means and end of every other member.
An activity is the unity of a totality of purposive movements. As this
totality has a unity, all its movements are ultimately the same, namely, the
one activity that is real and realized in them. This activity is always the
same in and through indefinitely many movements; in them it remains it-
self and in them it maintains itself as itself. What is always the same is
the teleological system we described. An energeia concept contains as its
form the representation of such a system.
11 The concept of a series is here to be understood abstractly. It can be a net and a
circle of circles.
48 Sebastian Rödl
Aristotle distinguishes kinds of energeia, which form a scale: dzen,

aisthanestai, dialegein, noein; to live, perceive, judge, think. By developing the
concept of energeia from that of the purposive kinesis as the concept that
underlies the latter one and makes it possible, we have determined the
first of these kinds: dzen. A specific energeia concept of this kind refers to
a determinate life, a way of life, a life-form. For x energei H we can thus
write x lives H, where “H” refers to a determinate life, for instance x lives
the ivy life, the chickadee life, the donkey life, or x chickadee-lives, ivy-lives, etc. We
have noticed above that the teleologically structured movements that we
encounter first are life processes. Now we understand why this is so. Pur-
posive movements are external ends; they are only possible through an
internal end; and the first kind of internal end is life.12
The determination of the content of an energeia concept is empirical
knowledge of a way of life. While the knowledge in question may be
fragmentary, the representation of a totality of movements that are all
means and the end of all is its form. Let us remember our example.
Helicase is unwinding the cords. It is determined to do this by the gen-
eral that is real and realized in it: DNA is replicating itself. If we follow
the series of teleological nexus we will find that it leads to something
that makes sure (or contributes to it) that those things happen that must
happen so that Helicase is where it needs to be and does what it needs to
do to divide the base pairs. The series of ends to which the workings of
Helicase is subordinated as a means returns to just these workings of
Helicase. It is only in this way that it is not an accident that those things
happen that must happen so that Helicase does what it needs to do. And
only when this is not an accident are its workings a purposive movement.
Therefore, whatever is a means and an end in a series of purposive
movements, it is subordinated to a totality in which everything is a means
as well as an end. And this totality is its own end, an end in itself.
We can explain a purposive movement by relating it to a further
movement or by relating it to the activity that underlies it. In the former
case the movement appears as externally purposive. Its end is not itself,
12 Therefore it is not illuminating to define life by the teleological structure of life
processes. For this structure is first only that of external purposiveness. However, the in-
ternal purposiveness that underlies it is nothing other than life, so that such a definition
at best defines life as life.
but its result. In the latter it appears as internally purposive. Its end is
itself, namely the life that is real and realized in it. The second consider-
ation is the deeper one; it shows the deeper ground. Internal purposive-
ness is the truth of external purposiveness.
3. Good Action
Kinesis is a species of the logical relation of the general to the particular:

x is doing A in that x has done A’. The purposive kinesis is a species of this
species: the logical relation is as such explanatory. Rational action is a
species of this species.
Rational action is action according to a conceptual representation. A
conceptual representation according to which its subject acts is an act of
will, wanting. There is the opinion that wanting is an efficient cause of
action performed according to this wanting. There are thousand ways of
seeing that this is wrong. We see it here by noticing that a purposive
movement as such does not have an efficient cause, and that rational act-
ion is purposive. The wanting, which cannot be an efficient cause, in-
stead determines the kind of teleological structure that is rational action.
Rational action manifests the internal structure of movement: Hans
is hanging up wallpaper in the room by lathering the wallpaper length
with glue. The structure is teleological: Hans is lathering the wallpaper
length because he is hanging up wallpaper. Furthermore, the logical-
causal relation resides in this that he who is acting understands it.13 Hans,
who is lathering the wallpaper because he is hanging up wallpaper in the
room, understands that he is lathering it because he is hanging up wall-
paper. That he understands this is nothing added to that which he under-
stands. It is not as if he is lathering the wallpaper because he is hanging
up wallpaper, as it were anyway, and is then lucky to notice this, too. No,
rather, it is only by understanding that he is lathering the wallpaper be-
cause he is hanging it up, that this is the case at all. The structure of ra-
tional action is thus not only logical (kinesis, the genus), not only logical-
13 As I say in my book, it is formally represented. By this I mean that its being rep-
resented is a character of its form. See Rödl, S., Self-Consciousness, Cambridge, MA,
Harvard University Press, 2007, Chapter 2.
50 Sebastian Rödl
causal (a species of this genus: purposive kinesis), but logical-causal-un-

derstood (a species of this species: rational action).14
That is rational action as species of purposive movement. We can
equally introduce a rational species of activity. Rational activity is an ac-
tivity whose relation to the movements in which it is real resides in its
being understood. The internal structure of the rational energeia is not
only logical-causal, but logical-causal-understood.
Aristotle’s words for the rational kinesis and energeia are poiesis and
praxis, respectively. The energeia verb appropriate to rational action is
therefore not the general dzen, but specifically prattein or eu prattein. The
“eu” makes explicit that prattein is an internal end. For “good” refers to
an end insofar as it is conceptually represented.15 Praxis is energeia that is
not only an end in itself as energeia, but, as a rational energeia, it is an end
in itself in such a way that it is realized through the representation of it-
self as an end, and specifically, as an internal end. A subject of a rational
energeia has therefore as such the concept of an end in itself, more pre-
cisely the concept of an activity that is an end in itself, and has it through
this very activity.
The first energeia concept is x lives H, where H refers to a manner of
living. We determine such a concept by recognizing: x lives in such a way
14 The person lathering the wallpaper because he is hanging it up, understands that he
is lathering the wallpaper because he is hanging it up. And because he understands the
relation as teleological, he represents hanging up wallpaper as an end and lathering as an
apt means to this end. The former is a wanting, the latter a judgment. In his action, in his
logically-causally-understood structured movement, his wanting and his judgment are
therefore brought together. The action is a unity of both, his wanting and his judgment,
which is itself understood. Generally speaking, a conclusion is nothing but a representing
of two conceptual representations together, two logoi, and in such a way that this repre-
senting together is sufficient for a third representation. Therefore we can say: an action is
a conclusion, whose premises are a wanting (of an end) and a judgment (on the means).
It is fitting to call such reasoning practical reasoning.
15 That is why Kant calls a life form a natural end, but not a natural good, for the in-
ternal end that is an arational life-form is not an end by being understood. –That also ex-
plains why rational action is performed sub specie boni. This is not a psychological prop-
osition, which can be refuted through empirical observation of the depravity of human
motives. (This wrong understanding of the kind of proposition we are dealing with un-
derlies the recent criticism of it.) It describes the logical form which the concept of ratio-
nal action signifies. It explains that this concept describes a doing –a movement or an ac-
tivity– whose teleological structure resides in being understood.
that it... and fill in the blank with explanations that indicate how the kind
of life in question proceeds. The concept of a rational energeia or praxis is
x acts E, whereas “E” refers to a way of acting. We determine the con-
cept in x acts so that she...
Kant calls an energeia concept that is represented as an end, and is
therefore a praxis concept, a maxim. A maxim can be represented in the
first as well as in the third person: I act so that I..., He acts so that he... But as
it is internal to the action determined by it, the first person represent-
tation is essential to it. In a maxim a subject determines its concept of eu
prattein, of acting well. In so far as an action is related to its underlying
praxis, and that is, to its maxim, it is considered as an end in itself. For a
praxis is an internal end. In contrast, if an action is related to its conse-
quences, it is represented as kinesis and more specifically as poiesis: as
something that consists in putting an end to itself and making space for
its result. When considered in this light, it is not an end in itself, but a
means to this result.
A poiesis as externally purposive is not good in itself, but in virtue of
its result. Therefore a poiesis is not unconditionally good, but under the
condition that its result is good. If the ethical quality of action lies in
this, that it is unconditionally good, then an action does not have this
quality as poiesis, not in virtue of its consequences. A praxis, as internally
teleological, can be good in itself. It has the right logical form for that.
The ethical quality of action is attributed to it as praxis, through its max-
im.
There is the opinion that poiesis and praxis are different in the sense
that either praxis is going on or poiesis. Now, there are acts that are
praxis without being poiesis, like public legislation. But nothing stands in
the way of something’s that is poiesis also being praxis. In the fundamental
case, namely when it has a sufficient ground, a poiesis is also a praxis. For
poiesis and praxis are kinesis and energeia, respectively, rational kinesis and
rational energeia. When we determine what is going on through a kinesis
concept, we consider it as kinesis: DNA is replicating itself. The ultimate
ground of such a kinesis is, however, the energeia that is real in it. The
chickadee chickadee-lives in that its DNA replicates itself. Now we un-
derstand what is going on, not only through a kinesis concept, but
through an energeia concept, chickadee-life, and consider it as energeia. The
same relation holds between poiesis and praxis. When I pump up a tire be-
52 Sebastian Rödl
cause you are not strong enough, my action is on the one hand poiesis: it
aims at a result, the pumped-up tire. On the other hand, it is praxis: by
pumping up the tire, I act according to a maxim to act in such a way that...
The difference is especially visible when my action fails as poiesis. If I am
too clumsy with the pump, or even unintentionally break it, I have
pumped badly; considered under this, a poiesis, concept and as poiesis, my
action is bad.16 Regardless of that, it can be the case that I am acting well
and have acted well; the ethical quality of my action can be irreproach-
able, if its maxim is. Therefore, the fact that consequences are irrelevant
for the ethical quality of action does not mean that he who acts well is
not interested in the consequences of his action. For the action in which
praxis is real is poiesis. I act well by pumping up your tire, which includes
that I want to pump up your tire and thus take an interest in its being
pumped up in the end.
We can characterize consequentialism as considering action only as
poiesis, as externally purposive, as a means to an end. But as poiesis, an
action is not unconditionally good. Therefore consequentialism does not
have the concept of the ethical quality of action, which lies in the fact
that it is unconditionally good. But it also does not have the concept of
externally purposive action. For, the external end presupposes the inter-
nal end and is only possible through the latter. We understand the exter-
nally purposive action only as such if we understand how its result can
be an end. We can explain that if we relate it to an overarching poiesis.
But that cannot be the only way. A result can only be an end by being
part of an activity that is an end in itself.
One gets entangled in a “deontological paradox” if one assumes that
a maxim describes an action as serving an external end. For it is always
possible to think of conditions under which acting according to the max-
im gets in the way of the supposed external end. For example, if one
thinks the maxim “Act so as not to kill innocent persons” characterizes
an action as good that has as its goal that as little innocent people as pos-
sible get killed, then one will be able to come up with situations in which
this end can be reached by killing an innocent person. This does not
show that it is paradoxical to determine acting well through maxims;
16 I owe the example to Anselm W. Müller.

rather, it shows that consequentialism does not possess the concept of

acting well. Speaking of consequentialist ethical systems is a contradictio in
adjecto.17
The concept of a good result, insofar as it is meant to be indepen-
dent and not subordinated to the concept of good activity, is empty. All
concepts of an alleged knowledge of such goodness are equally empty,
for example, the idea of an intuition of values. We see this when we con-
sider what it means that the characterization of a result as good is inde-
pendent from the use of “good” in “acting well”. In order to hold onto
this independence, let us use the letter F instead of “good”. Consequen-
tialism then explains that an action is unconditionally good when it pro-
duces something that is F. It is the final end of human beings to produce
Fs, to maximize Fs. It is obvious that someone who founds his life upon
this maxim is suffering from a pathological obsession. This is so because
of the maxim’s form. It is not the case that the maxim does not manifest
an obsession when the content is labeled with the word “good”. Rather,
it is the other way around: the good then appears to be not an internal
end, but an object of a sick obsession. It is a further escalation of the
confusion if the inclination is felt to escape this obsession by asking
whether it not be demanding too much always to be producing Fs, and
whether it may not be admissible also to do something else for once,
maybe because otherwise one would fall apart. This is the so called
“overdemandingness objection” against consequentialism, to the formu-
lation as well as the rejection of which much ingenuity is applied. The
problem of consequentialism is not, however, that it demands too much.
Rather, the problem is that it represents the logical form of acting as that
of an external end, and thus represents the practical existence of man as
a means to an end. That is not false, but nonsense.18
Universität Leipzig
Leipzig, Germany
17 As is indicated by the title of Anselm W. Müller’s immortal essay “Radical Subjec-

tivity: Morality vs. Utilitarianism”, Ratio 19/2 (1977), pp. 115-32.
18 A German version of the essay has appeared in Rödl, S. – Tegtmeyer, H. (eds.),
Sinnkritisches Philosophieren, De Gruyter, Berlin, 2012.

54 Sebastian Rödl
References
Aristotle, Über die Seele, ed. by H. Seidl, Hamburg, Meiner, 1995.

——, Metaphysik, ed. by H. Seidl, Hamburg, Meiner, 1995.
Kant, I., Critique of Judgment, trans. by J. C. Meredith, ed. by N. Walker,
Oxford, Oxford University Press, 2007.
Müller, A. W., “Radical Subjectivity: Morality vs. Utilitarianism”, Ratio
19/2 (1977), pp. 115-32.
Ridge, M., “Reasons for Action: Agent-Neutral vs. Agent-Relative”, The
Stanford Encyclopedia of Philosophy (Winter 2011 Edition), ed. by E. N. Zalta,
<URL=http://plato.stanford.edu/archives/ win2011/entries/reasons-
agent/>.
Rödl, S., Self-Consciousness, Cambridge, MA, Harvard University Press,
2007.
Rödl, S. – Tegtmeyer, H. (eds.), Sinnkritisches Philosophieren, Berlin, De
Gruyter, 2012.
WHY THERE IS NO FACT OF REASON IN
THE GROUNDWORK. THREE ARGUMENTS
DIETER SCHÖNECKER
Endless disagreements as to whether Kant defends this or that particular

claim in this or that particular text have accompanied the work of Kant
interpretation from the very beginning. Would one have to be regarded
as disreputable and ill-disposed to think that these disagreements simply
spring from the nature of the texts themselves? I think so. For while we
cannot deny that there are different opinions about the question, for
example, of whether Kant is already claiming the existence of a “fact of
reason”1 in the Groundwork of the Metaphysics of Morals (GMS) before he
explicitly makes this claim in the Critique of Practical Reason (KpV), it
would be a genetic fallacy to conclude from this that the question is un-
answerable and simply arises from the unfathomable character of the
texts.
Here I shall argue, on the contrary, that the question can indeed be
answered in a clear and unambiguous fashion: No, this thesis –there is a
fact of pure practical reason– is not presented in GMS. Of course, I am not
so naive as to hope that the following contribution will put an end to this
disagreement. But I would like to formulate my answer to this question
in such a succinct and precise manner that those who propose a different
answer can at least clearly try and indicate where and how my answer is
mistaken. And that is why I speak expressly of “arguments” in this con-
nection. Arguments in the strict sense of the word certainly have the un-
comfortable disadvantage that they sometimes prove not to be sound;
but if they are identified and recognized as such, we can at least see pre-
cisely why they are not sound. The claim to provide a demonstration here
is thus to be understood in this sense: I provide clear arguments to show
exactly why there is no thesis of the fact of reason in GMS, and I pres-
1 Critique of Practical Reason (KpV): 31, 24. Kant is cited from the Academy Edition,
with page reference and line number.

56 Dieter Schönecker
ent them in such a way that those who take a different view on this ques-
tion can then explain exactly why they do so.2
I should like to begin by outlining Kant’s thesis of the fact of reason
as briefly and clearly as possible (section 1); and I shall then present three
arguments against the claim that Kant already defends the thesis of the
fact of reason in GMS; I describe them for short as the phantom argument,
the subjection argument, and the confirmation argument (section 2).
1. Kant’s thesis of the fact of reason: a very brief exposition
What then is ‘the fact of pure practical reason’? It is obvious that this
question cannot be answered in a truly comprehensive sense without
considerable effort. On another occasion I have presented the outline
for a new interpretation which I can summarize here as succinctly as
possible.3 This interpretation involves three fundamental thoughts:
1. The theory of the fact of reason explains our insight into the validity or bind-
ing character of the moral law; it is a theory of justification and as such replaces the
deduction which is offered in GMS. Thus, so Kant argues in KpV, we enjoy
neither an immediate (certain) consciousness that we are not determined
by sensuous incentives in some particular action (negative freedom), nor
an immediate consciousness or experience that we are actually deter-
mined in some particular action by the moral law (positive freedom). We
know about this freedom solely through the moral law; the “consciousness
2 The writings of Dieter Henrich in particular (see references at the end of the chap-
ter) have emphatically and influentially endorsed this different view, namely that in GMS
Kant is already basically defending the thesis of the fact of reason which is expressly
presented in KpV; most recently, Heiko Puls (“Freiheit als Unabhängigkeit von bloß
subjektiv bestimmten Ursachen. Kants Auflösung des Zirkelverdachts im dritten Ab-
schnitt der Grundlegung zur Metaphysik der Sitten”, Zeitschrift für philosophische Forschung, 65
(2011), pp. 534-562), for example, has also endorsed this interpretation. The literature on
the subject is vast and cannot be discussed in detail here.
3 See Schönecker, D., “Das gefühlte Faktum der Vernunft. Skizze einer Interpretation
und Verteidigung”, Deutsche Zeitschrift für Philosophie, Heft 1/2013, pp. 91-107; a briefer
English version can be found in Schönecker, D., “Kant’s Moral Intuitionism. The Fact of
Reason and Moral Predispositions”, Kant Studies Online, Feb. 2013, pp. 1-38.
Why there is no Fact of Reason in the Groundwork 57
of the moral law”4 is “the ratio cognoscendi of freedom”.5 The theory of the
fact of reason is thus a theory as to how we are justified in our convic-
tion regarding the “reality”6 of the moral law. Only when we know that
morality is real, do we also know that freedom is real.
Kant’s decisive thought here is that although there is no deduction of
the categorical imperative, this imperative “is nevertheless firmly estab-
lished of itself [steht... für sich selbst fest]”,7 and that this character of being
‘firmly established of itself ’ must be understood as a non-inferential justifi-
cation of the categorical imperative. For the categorical imperative stands
‘firmly established of itself ’ insofar as it “is given, as it were, as a fact of
pure reason of which we are a priori conscious and which is apodictically
certain”.8 It is in this sense that Kant says:
“It was necessary first to establish and justify [rechtfertigen] the purity of
its origin [i.e. of the categorical imperative] even in the judgment of this
common reason before science would take it in hand in order to make
use of it, so to speak, as a fact that precedes all subtle reasoning about its
possibility and all the consequences that may be drawn from it. [...] But
for this reason the justification [Rechtfertigung] of moral principles as princi-
ples of a pure reason could also be carried out very well and with suffi-
cient certainty by a mere appeal to the judgment of common human un-
derstanding ...” (KpV: 91, my emphasis).
Here Kant speaks twice of the ‘justification’ of the categorical imper-

ative: we must ‘justify’ the ‘supreme practical principle’, and indeed do so
‘as a fact’; and he then refers again to this process of ‘justifying’ of the
moral law when he speaks of ‘the justification of moral principles’. Thus it
is the consciousness of the categorical imperative as a fact of reason
which justifies this imperative in its absolutely valid or binding character.
4 KpV: 121, my emphasis.

5 KpV: 4, footnote.
6 KpV: 47, 15; 48, 6.
7 KpV: 47, 19.
8 KpV: 47.
2. In the consciousness of the moral law the categorical imperative is immediately

given, and thereby recognized. A kommentarische interpretation9 of the pas-
sage10 in which Kant first develops the thesis of the fact of reason in
these specific terms permits us to present Kant’s theses as follows:
(F1) The consciousness of the categorical imperative is a fact of rea-

son.
(F2) The categorical imperative is a fact of pure practical reason.
(F3) The categorical imperative cannot be reasoned out from any an-
tecedent data of reason.
(F4) The categorical imperative forces itself upon us as a synthetic a
priori proposition.
(F5) The categorical imperative is given.
(F6) The categorical imperative is in no way an empirical fact.
(F7) The categorical imperative is the sole fact of pure practical rea-
son.
I should specifically like to emphasize that this talk of ‘givenness’ is

the central aspect which allows Kant his thesis that the categorical im-
perative ‘stands firmly established of itself ’; and that ‘the fact of reason’
signifies nothing other than the said givenness of the categorical imper-
ative. That the consciousness of this imperative is a ‘fact of reason’ thus
means, according to this exemplary passage, that the categorical imper-
ative is ‘given’ to us insofar as it ‘forces itself upon us of itself ’11 without
further mediation through any consciousness of negative freedom and
without further conceptual analysis or logical derivation; and that is also
why it stands ‘firmly established of itself’.
9 For this notion of ‘kommentarische interpretation’, see Schönecker, D., “Textver-
gessenheit in der Philosophiehistorie”, in Schönecker, D. – Zwenger, Th. (Hrsg.), Kant ver-

stehen / Understanding Kant. Über die Interpretation philosophischer Texte, Darmstadt, Wissen-
schaftliche Buchgesellschaft, 22004, pp. 159-181 and Damschen, G. – Schönecker, D.,
Selbst philosophieren. Ein Methodenbuch, Berlin – Boston, De Gruyter, 22013, pp. 203-272.
10 KpV: 31.
11 See also Kant’s Religion within the Boundaries of Mere Reason: “Were this law not given
to us from within, no amount of subtle reasoning on our part would produce it” (Rel.:
26, footnote).
3. The unconditionally binding character or validity of the categorical imperative

is given in the feeling of respect: the ‘fact of reason’ takes the place of any at-
tempted deductive grounding of the categorical imperative. The fact of
reason enjoys a moral self-evidence which is encountered in anyone’s judg-
ment –just as it is in the judgment of the individual in Kant’s famous
‘gallows example’12– and thus serves as ‘the justification of moral princi-
ples’. Kant says hardly anything about how we are supposed to think
about this ‘givenness’ of the categorical imperative, namely about the
way it ‘forces itself upon us’. Only through the broader context of the
argument, and through a further passage at KpV: 91f., does it become
clear that the categorical imperative in its absolute validity is given to us
in the feeling of respect. It is respect which reveals or makes this imper-
ative known to us, and specifically as a law which we must “obey”.13 And
this in turn means: consciousness of the categorical imperative is medi-
ated by respect; and since respect is a feeling, Kant’s thesis of the fact of
reason amounts to this: we recognize the validity or binding character of
the categorical imperative through an immediately given feeling. Thus the
thesis does not merely imply that respect drives us to do what is morally
right, and in that sense is an incentive. Respect is a feeling through which
we know or recognize something; for what we know or recognize here is
that the categorical imperative possesses absolute validity.
Now when Kant speaks about respect he constantly emphasizes the
non-empirical origin of this feeling. Respect is described as the “effect”
[Wirkung] of reason.14 As an ‘effect’, therefore, respect does not precede
reason; in another passage Kant even says that the feeling of respect is
“the effect of consciousness of the moral law”,15 as if there were a ‘con-
sciousness of the moral law’ without the feeling of respect. How is this
compatible with the thesis of the fact of reason that it is respect which
first allows us to know and recognize the categorical imperative in the
first place?16 – It is certainly true in one sense that there can be a ‘con-
12 See KpV: 30.

13 KpV: 92, 17.
14 See GMS: 401, 27; KpV: 79, 4; Metaphysics of Morals: 399, 16.
15 KpV: 75.
16 I should like to thank Christoph Demmerling for raising this question and for the
ensuing discussion.
sciousness of the moral law’ without the feeling of respect. As I am

thinking now about the categorical imperative, and thus am also con-
scious of the latter, I do not experience a feeling of respect. If, as I write,
I think about the categorical imperative without having a feeling of re-
spect, I can nonetheless only rightly claim to be thinking about the
categorical imperative and to recognize its content if I can affirm that I
have had that experience (namely the experience of the imperative char-
acter of the moral law through the feeling of respect); if I had not had
this experience, then I would be in the same position as a blind person
talking about colours.17 The ‘consciousness of the moral law’, in the
sense in which it is relevant to the thesis of the fact of reason, refers nei-
ther to the abstract consciousness of a sceptic, nor to a merely reflective
consciousness in which the feeling of respect is not currently or actually
manifest. Rather, we are talking about the consciousness of someone
who is confronted with an urgent and concrete situation of action and
asks himself how he ought to act, and who thereby experiences the sense
of You Ought18 as he draws up maxims (for himself) with regard to his
conduct. But in what sense precisely is this experience an ‘effect’ of the
moral law or of reason, or even an ‘effect of the consciousness of the mor-
al law’? Let us directly consider the categorical imperative: “So act that
the maxim of your will could always hold at the same time as a principle
in a giving of universal law”.19 It is immediately obvious that this imper-
ative has two components: it includes a demand (‘So act’) and a content,
albeit a formal one (what is thus demanded here, in short, is universali-
zation). It is precisely this connection that makes the moral law into a
synthetic proposition; as an analytic proposition, on the other hand, the
17 See Husserl, E., “Kritik der Kantischen Ethik”, in Husserl, E., Vorlesungen über
Ethik und Wertlehre 1908-1914, Dordrecht – Boston – London, Kluwer, 1988, p. 404: “If
we imagine a being that is, as it were, entirely blind to feeling, just as we know there are
beings which are blind to colour, then the whole moral dimension loses its content, and
moral concepts simply become words without meaning”.
18 It may be mentioned in passing that Kant is clearly a weak internalist where moti-
ves are concerned: the recognition of the moral Ought implies a motivation (I certainly
recognize what is binding here insofar as I feel a respect which possesses motivating
force), but this moral motive may be weaker than other motives, so that the action that is
morally required does not actually ensue. Kant thus recognizes such a thing as weakness
of will.
19 KpV: 30.
moral law is simply descriptive: a perfectly rational being acts in such a

way that the maxims of its will can always hold at the same time as the
principle of a universal legislation. But that it is actually so, that I, as an
imperfectly rational being, ought so to act, is something –this is Kant’s
thesis– that I cannot know or recognize without the feeling of respect; I
know and recognize this Ought in and through the feeling of respect.
But this Ought is an integral component of the categorical imperative
(its normative content); thus I do not know or recognize this imperative
–namely the moral law as imperative for imperfectly rational beings–
without respect. Without respect we do not recognize the imperative
component of the categorical imperative; respect is therefore a necessary
condition for the recognition of this imperative. Yet the other necessary
component is reason. For the other component of the categorical imper-
ative (its content) is indeed nothing that I could grasp through respect or
any other kind of feeling; we do not feel, we think (by means of language
and concepts) that ‘the maxim of one’s will must always be able to hold
at the same time as the principle of a giving of universal law’. But not
only must we possess reason in order to think the categorical imperative,
the content of the latter is itself a product of reason, and insofar as it is
a product of reason then respect is an ‘effect’ of reason. We should not,
however, misunderstand this to mean that anyone who possesses a moral
consciousness can first have a consciousness devoid of respect with
regard to the categorical imperative, which itself would only subsequent-
ly be connected with the feeling of respect; that cannot be the case, for
this imperative is precisely a categorical imperative, and this compelling
character –this sense that one actually ought to choose one’s maxims in
universal terms– is simply not experienced (and thus not known and
recognized) without the feeling of respect. Thus respect is an ‘effect’ of
reason insofar as it is indissolubly and specifically connected with that
rational content rather than with any other sort of object. Here we must
not think of ‘reason’ as an organ which does this or that (as we might in
the case of a sensory organ such as the eye). Rather, we find certain
mental activities and processes within ourselves, some of which are ex-
perienced as spontaneous and some as receptive in character; we expe-
rience ourselves as capable of these activities and processes and describe
these as capacities (sources or faculties); and in this way we come to
speak of ‘reason’ or ‘sensibility’. We find ourselves reacting to situations
of action in terms of this content, namely (the procedure of) universali-

zation –hence Kant’s repeated allusion to ‘common human reason’, not
just in the context of the theory of the fact of reason, but also in the
context of the universalization which is the content of the categorical
imperative. This procedure is quite different from a case of colour per-
ception or a case of feelings; we characterize it as an activity of reason; it
is a rational procedure, and it presents itself as something which is de-
manded of us.
It is not the case that we first recognize the categorical imperative and
then experience respect; on the contrary, respect is “inseparably connected
with the representation of the moral law in every finite rational being”,20
for the normative content is ‘inseparably’ connected with the categorical
imperative. Thus we recognize the normative content of this imperative
in and through respect, or as Kant had already put it in GMS: “What I
cognize immediately as a law for me I cognize with respect”.21 We should
note precisely what is said here: I know or cognize the law with respect,
and it is with respect that I know or cognize it. And that is why Kant
writes in the Metaphysics of Morals, in the ‘Doctrine of Virtue’, that man
“must have respect for the law within himself in order even to think of
any duty whatsoever”.22
However it precisely turns out with Kant’s thesis of the fact of rea-
son, the basic thought behind this thesis is that we know and recognize
the validity of the moral law “immediately”23 and in a way that is “unde-
niable”.24 Whether this immediacy lies in the way that we know and rec-
ognize the moral law through a feeling (of respect), or in some other
way, is irrelevant for our purposes here. What matters is that the validity
or binding character of the moral law is not derived from anything; on the
contrary, we have a ‘consciousness’ of the moral law which is necessary
and sufficient in order to “know” [wissen]25 about the validity or binding
character of the moral law, so that the idea of freedom is ‘revealed’
20
KpV: 80; my emphasis.
21 GMS: 401.
22 Metaphysics of Morals: 403, my emphasis.
23 KpV: 29, 34; my emphasis.
24 KpV: 32, 2; my emphasis.

25 KpV: 4, 10.
precisely through this moral law. In a ‘deduction’, on the other hand, we

are offered a demonstration that the moral law is valid or binding, and this
demonstration does not consist simply in referring to the consciousness
of the moral law; in the context of such an approach the moral law does
not stand ‘firmly established of itself ’, for something else is still required
(further propositions, an explicit argument). It is also important that the
law, as Kant puts it, is ‘undeniable’. The validity or binding character of
the categorical imperative is incontestably (self-evidently) ‘given’ in the
consciousness of this law.26
2. The phantom argument, the subjection argument, and the confirmation argument
There is no doubt what so ever that Kant constantly refers, in GMS as

well as in KpV, to the moral consciousness of (all) human beings, and
specifically also of those who are not philosophically trained or
educated. Thus it is no accident that the initial “Transition from com-
mon rational moral to philosophic moral cognition”27 is accomplished in
Chapter One of GMS, whereby Kant expressly emphasizes that with the
moral law (the categorical imperative) we have found a law which com-
mon human reason “admittedly does not think so abstractly in a univer-
sal form, but which it actually has always before its eyes and uses as the
norm for its appraisals”;28 and Chapter Two likewise begins with the
assertion that “we have drawn our concept of duty from the common
use of our practical reason”.29 In much the same vein Kant observes in a
footnote of KpV that a “reviewer [of GMS] who wished to say some-
thing censuring this work hit the mark better than he himself may have
intended when he said that no new principle of morality is set forth in it
but only a new formula. But who would even want to introduce a new
principle of all morality and, as it were, first invent it? Just as if, before
him, the world had been ignorant of what duty is or in thoroughgoing
26 See KpV: 32, 33.

27 GMS: 392.
28 GMS: 403.
29 GMS: 406.
error about it”.30 Thus both GMS and KpV refer to something like a
(common) “consciousness of a law”31 or a (common) “consciousness of
this fundamental law”;32 in this regard, therefore, there is indeed no dif-
ference between the two works. Yet it would be false to adduce this as
evidence for the claim that Kant already effectively defends the thesis of
the fact of reason in GMS; we must first raise and answer the question
as to how Kant refers to moral consciousness in GMS. It turns out that
the way in which he refers to such a ‘consciousness of the moral law’ –
and this is the decisive point– is quite different in the two works. Thus while
in GMS Kant understands the consciousness of the moral law in such a
way that the moral law cannot possibly stand ‘firmly established of itself ’
in and through that consciousness alone,33 this is precisely what is
claimed for the consciousness of the moral law as Kant understands it in
KpV: it is through this consciousness that we ‘immediately’ and ‘undeni-
ably’ know about the validity or binding character of the moral law in
such a way that the latter indeed ‘stands firmly established of itself ’. In
this consciousness the moral law is ‘given’ in its absolute validity or bind-
ing character, and that is why Kant calls the consciousness of the funda-
mental moral law ‘a fact of reason’. If we can show that precisely this
claim –that we know the validity or binding character of the moral law
immediately and undeniably– does not hold for the moral law or the
consciousness of that law in GMS, then we shall also have shown that
there is no thesis of the fact of reason implied or presented in that text.
Now there is no doubt that there are passages in GMS III which
seem at first sight to suggest that Kant does indeed already defend the
thesis of the fact of reason in the Groundwork.34 Anyone who claims that
30KpV: 8.
31GMS: 449, 7.
32 KpV: 31, 25.
33 For the question as to why Kant chooses in GMS to begin with ‘common moral
knowledge’ in the first place, see Schönecker, D., Kant: Grundlegung III. Die Deduktion des
kategorischen Imperativs, Freiburg – München, Karl Alber-Verlag, 1999; “The Transition
from Common Rational to Philosophical Rational Moral Knowledge in the Groundwork”,
in Ameriks, K., – Höffe, O. (eds.), Kant’s Moral and Legal Philosophy, Cambridge, Cambridge
University Press, 2009, pp. 93-122.
34 Here we are thinking (above all) of the following eight passages: (1) “Now, a hu-
man being really finds in himself a capacity by which he distinguishes himself from all
Kant does nothing of the kind here must show that all these passages
can be interpreted otherwise; I have undertaken to do this elsewhere, or
more precisely, I have attempted to show that they must be interpreted
otherwise.35 And, then again, those who claim that Kant does already de-
fend the thesis of the fact of reason in the Groundwork (and especially in
other things, even from himself insofar as he is affected by objects, and that is reason […]
but reason, on the contrary, shows in what we call ideas a spontaneity so pure that it
thereby goes beyond anything that sensibility can ever afford it” (452, 7; my emphasis
here, and in the following citations). (2) “This better person, however, he believes him-
self to be when he transfers himself to the standpoint of a member of the world of un-
derstanding, as the idea of freedom, that is, of independence from determining causes of the world of
sense, constrains him involuntarily to do; and from this standpoint he is conscious of a good
will” (454, 37). (3) “All human beings think of themselves as having free will. From this come all
judgments upon actions as being such that they ought to have been done even though
they were not done. Yet this freedom is no concept of experience [...] Hence freedom is
only an idea of reason, the objective reality of which is in itself doubtful” (455, 11). (4)
“But the rightful claim to freedom of will made even by common human reason is based
on the consciousness and the granted presupposition of the independence of reason
from merely subjectively determining causes, all of which together constitute what be-
longs only to feeling and hence come under the general name of sensibility” (457, 4). (5)
“So it is that the human being claims for himself a will” (457, 25). (6) “The concept of a
world of understanding is thus only a standpoint that reason sees itself constrained to
take outside appearances in order to think of itself as practical, as would not be possible
if the influences of sensibility were determining for the human being but is nevertheless
necessary insofar as he is not to be denied consciousness of himself as an intelligence and
consequently as a rational cause active by means of reason, that is, operating freely. This thought
admittedly brings with it the idea of another order and another lawgiving than that of the
mechanism of nature, which has to do with the sensible world; and it makes necessary
the concept of an intelligible world (i.e., the whole of rational beings as things in them-
selves)” (458, 19). (7) The idea of freedom “holds only as a necessary presupposition of
reason in a being that believes itself to be conscious of a will, that is, of a faculty distinct
from a mere faculty of desire (namely, a faculty of determining itself to action as an in-
telligence and hence in accordance with laws of reason independently of natural in-
stincts)” (459, 9). (8) “Moreover, to presuppose this will is (as speculative philosophy can
show) not only quite possible (without falling into contradiction with the principle of
natural necessity in the connection of appearances in the world of sense); it is also
practically necessary –that is, necessary in idea, without any further condition– for a
rational being who is conscious of his causality through reason and so of a will (which is
distinct from desires) to put it under all his voluntary actions as their condition” (461,
17). (Kant’s emphases have been in part omitted, while the formulations which seem to
suggest the thesis of the fact of reason have been highlighted).
35 See Schönecker, D., Kant: Grundlegung III. Die Deduktion des kategorischen Imperativs.
GMS III) are duty bound to counter or refute the arguments of those
who claim to prove the opposite –namely that Kant certainly does not
develop or defend the thesis of the fact of reason in the Groundwork. I
would now like to introduce three internally connected arguments in or-
der to make my case: the phantom argument, the subjection argument, and the
confirmation argument. There is also a further argument which I believe is a
very strong one and which has also played a significant role in the recep-
tion history of Kant’s moral philosophy. I only mention the argument in
passing here since it is very complex, and has already provoked various
counter-arguments which would have to be addressed in their own right,
something that cannot be undertaken here.
I am talking about the deduction argument, which can be formulated as
follows:
1. If Kant defends the thesis of the fact of reason in GMS, then

there is no deduction of the categorical imperative in GMS.
2. But there is a deduction of the categorical imperative in GMS.
Therefore, Kant does not defend the thesis of the fact of reason in
GMS.
It seems clear to me that Kant employs the concept of a “deduction”

as a counter-concept to the concept of “fact”; in GMS the categorical im-
perative is grounded specifically through a deductive argument, and is
not grounded ‘immediately’ in terms of the consciousness of the moral
law. There is certainly much that could be said in this connection; but, as
I have said, I shall not pursue this line of argument any further here. But
if the other arguments which I present prove to be convincing, this is a
clear indication that the deduction argument is also convincing (and like-
wise the reverse).
a) The phantom argument

The phantom argument appeals to the essential content of the thesis of
the fact of reason: the validity or binding character of the categorical im-
perative is ‘immediately’ and, above all, ‘undeniably’ given. Since this is so
(or since Kant claims that it is so), we cannot find a single passage in
KpV where Kant would so much as question this validity; Kant already
refers to the “fact” concerned in the Preface.36 And it would indeed be a

glaring contradiction in one and the same book to set out the thesis of
the fact of reason, i.e. to claim on the one hand that the validity or bind-
ing character of the moral law is immediately and undeniably given in the
(common) consciousness of this law, while on the other hand placing the
consciousness of this law and thereby the law itself in doubt after all.
But this is precisely what Kant does in GMS: here he places the con-
sciousness of this law and thereby the law itself in doubt. Of course, to
‘put in question’ or ‘to place in doubt’ the consciousness of the moral
law and its validity does not mean that Kant himself, as an individual and
as a philosopher, personally harboured any actual, genuine, or persistent
doubt with regard to the validity or binding character of the categorical
imperative; he has no such doubt (and never did have throughout his
life). To put the consciousness of the moral law and its validity in ques-
tion means to regard such doubt as possible or real, and to take it as such
a serious philosophical challenge that it is worth presenting someone
who does have doubt in this regard with an argument for overcoming
such doubt.
But since it would be a glaring contradiction if Kant in one and the
same book, let us now say in the Groundwork, were to set out the thesis
of the fact of reason, while at the same time placing in doubt the moral law
(the categorical imperative), it must simply be shown that this latter is
exactly what Kant does: in GMS he places the validity or binding char-
acter of the moral law in doubt –so that he could not set out the thesis
of the fact of reason at the same time without courting that glaring con-
tradiction. The first of our arguments takes its name from what may be
the most prominent passage to present such a possible and overtly ex-
pressed doubt regarding the validity of the categorical imperative. For at
the end of GMS II Kant observes that only someone who “holds moral-
ity to be something and not a chimerical idea without any truth”37 has to
“admit”38 his preceding analyses and thereby the categorical imperative.
But precisely this (that morality is ‘something’ and is no ‘chimerical idea
36 KpV: 6, 12.
37 GMS: 445, 5.
38 GMS: 445, 7.
without any truth’) has not yet been shown; for at the end of GMS II it is
not yet clear that the categorical imperative as a synthetic a priori propo-
sition “is no phantom” [Hirngespinst].39 This passage at the end of GMS II
forms a parallel to the earlier one in GMS II where Kant makes the tran-
sition to the metaphysics of morals. There too Kant emphasizes that “if
duty”40 is a concept with any reality and significance, then it must be un-
derstood as a categorical imperative; but whether this is so has not yet
been shown: “But we have not yet advanced so far as to prove a priori that
there really is such an imperative, that there is a practical law, which com-
mands absolutely of itself and without any incentives, and that the ob-
servance of this law is duty”.41
It is not just that Kant expressly says here that he wishes to ‘prove’ the
reality of the categorical imperative later (in GMS III). No such proof42
–or “establishment” [Festsetzung] as Kant already puts it in the Preface43–
would be necessary if the categorical imperative were a ‘fact’; quite obvi-
ously it has not yet been shown that there really ‘is’ a categorical imper-
ative.44 And what sense could this doubt possibly have if Kant were to
claim at the same time that the categorical imperative is a ‘fact’ and as
such ‘undeniable’?
We can thus formulate the phantom argument as follows:
39 GMS: 445, 8; my emphasis. Kant had already spoken of the idea of morality as a
“phantom” at the very beginning of GMS II (407, 17).
40 GMS: 425, 1; my emphasis.
42 In GMS Kant often speaks of ‘proof ’ etc. as well as of ‘deduction’; see 392, 4;
392, 13; 403, 27; 412, 2-8; 425, 8; 425, 15; 427, 17; 431, 33; 440, 20-28; 445, 1; 447, 30-
448, 4; 449, 27.
44 In GMS (as also in KpV) Kant also speaks in this connection of obligation [Verbind-
lichkeit] (see 389, 12; 389, 16; 391, 11; 432, 31; 439, 31; 439, 33; 448, 34); of the reality
[Realität] (see 425, 14; 449, 26) of the categorical imperative; of its actuality [Wirklichkeit]
(420, 1; see also 406, 15); of its validity [Geltung] (see especially 389, 12; 389, 14; 403, 7;
408, 18; 412, 3; 424, 35; 425, 18; 442, 8; 447, 32; 448, 6; 448, 32; 449, 29; 460, 25; 461, 1;
461, 3); of its correctness [Richtigkeit] (392, 13); of its objective necessity [objektive Notwendigkeit]
(see especially 442, 9; 449, 26; 449, 30); he talks of showing that it really is or transpires
[wirklich stattfinde], of human beings as subject to it [unterworfen] (449, 12). All of these
concepts and expressions can be subsumed under the later formula of the “validity of this
imperative” (461, 12).
1. If Kant defends the thesis of the fact of reason in GMS, then at

no point of the argument in GMS does Kant entertain the possibility
that the moral law is a phantom.
2. Kant does entertain at one point of the argument in GMS II that
the moral law is a phantom.
GMS.
It is obviously impossible to contest the second premise here. On the

other hand, there might initially appear to be a plausible objection to the
first premise. Thus we might argue as follows: while it is quite true that
in GMS II Kant still considers it possible that the moral law is a phan-
tom, and to that extent it is also true that up to this point in the text (i.e.
up to the end of GMS II) the thesis of the fact of reason could not be
set out, this does not exclude the possibility that such a thesis is set out
in GMS III. –Our response to this argument is twofold. In the first
place, we shall soon see that even in GMS III Kant still expressly puts in
question whether there actually ‘is’ such a thing as the categorical impera-
tive. In the second place, it would precisely contradict the fundamental
notion behind the theory of the fact of reason if on the one hand we
were to assert the undeniability of the categorical imperative (and partic-
ularly with reference to our common reason,45 and at however late a
point in the book) while on the other hand denying this undeniability
itself by putting this imperative in question; if the categorical imperative
or the consciousness of its validity is a ‘fact’, then this very conscious-
ness cannot be denied.46 (In the case of a proof the situation is different:
as long as the proof has not yet been provided it is unclear whether the
conclusion is true; and even if it is provided, it may still be treated for
presentational reasons as if it had not carried out. To deny the fact of
reason, even temporarily, would be as if someone actually had a percep-
tion of colour, and had no other reason for the conviction that he has a
45 I shall return to this point later.

46 One can deny the fact hypothetically in the sense that one can imagine what would
be the case if there were no such fact; but that is something different from really
doubting the fact. I would like to thank Elke E. Schmidt for her critical observations on
this point.
perception of colour –and would temporarily deny this was so, but with-
out indicating or even being able to indicate another reason for this con-
viction).
b) The subjection argument

We have already pointed out that in GMS III Kant expressly and repeat-
edly asks the question whether there really ‘is’ a categorical imperative.
He does so above all in the third subsection of GMS III. At this point he
summarizes the argument so far and says:
“But there also flowed from the presupposition of these ideas the
consciousness of a law for acting: that subjective principles of actions,
that is, maxims, must always be so adopted that they can also hold as
objective, that is, hold universally as principles, and so serve for our own
giving of universal laws” (GMS: 449, 7).
It is not immediately clear what Kant means by ‘ideas’ here: are they
the “ideas of morality” mentioned in the heading of the third subsec-
tion?47 But then the claim would be that the ‘consciousness’ of the moral
law follows from the ‘ideas of morality’. Or does he mean the “idea of
freedom” which is discussed in the second subsection,48 and which is
taken up and recapitulated in the first paragraph of the third subsection?
But that idea is referred to in the singular. Or does he perhaps mean the
ideas of morality in the sense of common moral knowledge and its mor-
al concepts?49 Whatever we decide in this regard, if we read this passage
without paying attention to what Kant says in the next paragraph one
might well think that Kant is indeed claiming a fact here, that is, precisely
a ‘consciousness of a law for acting’, and thus a consciousness of the
categorical imperative. But whereas in KpV such a “consciousness”50 is
interpreted as an ‘immediate’ and ‘undeniable’ ‘fact of pure practical rea-
son’, in GMS Kant directly takes up the passage we have just cited, and
47 GMS: 448, 23.

48 GMS: 448, 25.
49 I would like to thank Elke E. Schmidt for this suggestion.
50 KpV: 31, 24.
in spite of the ‘consciousness’ which is acknowledged here takes up the

persisting doubt as to whether this imperative actually possesses validity,
so that human beings must indeed subject themselves to it. Thus Kant
continues (and expresses this question of validity) as follows: “But why,
then, ought I to subject myself to this principle and do so simply as a ra-
tional being, thus also subjecting to it all other beings endowed with rea-
son?”.51
And what follows in this and the next two paragraphs is a questioning
with regard to the validity of the categorical imperative which could
hardly be formulated in a more pointed fashion; the question whether
there ‘is’ a categorical imperative, or whether it is not rather a ‘phantom’,
still awaits an answer. Repeating once again the thought from GMS II
(428 and 445) –the conceptual analysis and more precise determination
of the moral law may be satisfactory, but nothing has yet been gained
with regard to its reality– Kant now writes:
“We should still have gained something considerable by at least deter-

mining the genuine principle more accurately than had previously been
done, but we should have got no further with respect to its validity and the
practical necessity of subjecting oneself to it; for, if someone asked us why
the universal validity of our maxim as a law must be the limiting condi-
tion of our actions [...] we could give him no satisfactory answer” (GMS: 449,
27; my emphasis); and with regard to the claim that we must “hold our-
selves subject to certain laws” Kant says that “we cannot yet see how this is
possible, and hence on what grounds the moral law is binding”.52
These are all questions53 which no one who regards the conscious-
ness of the validity of the categorical imperative as ‘undeniable’ would
ask or even discuss. The subjection argument (which takes its name from
that question regarding validity from GMS: “But why, then, ought I to
subject myself to this principle?”54) can be formulated as follows:
51 GMS: 449, 11.

53 Kant expressly alludes to someone who “asked us” such questions (GMS: 449, 31;
my emphasis) and for whom there is still “no satisfactory answer” (GMS: 450, 2).
54 GMS: 449.
1. If Kant defends in GMS the thesis of the fact of reason, then in

GMS Kant does not ask the question as to why one ought to subject
oneself to the moral law.
2. In GMS Kant does ask the question as to why one ought to subject
oneself to the moral law.
GMS.
It is obviously just as impossible to contest the second premise in this

argument as it was to contest the second premise in the phantom argu-
ment. And indeed the question regarding the validity of the categorical
imperative which is presented in various formulations in the third sub-
section is only a variation of the phantom question posed in GMS. Thus
the subjection argument basically expresses the same thought as the
phantom argument: Kant places the categorical imperative in question in
a manner that is incompatible with the thesis of the fact of reason. Once
again, it might initially look as if there is a plausible objection to this
reading. Thus one might argue that it is certainly true that Kant asks the
question concerning validity, but that the answer to this question is pre-
cisely the thesis of the fact of reason. Our response to this is straight-
forward: if the ‘fact of pure reason’ as a ‘consciousness’ of the validity
of the categorical imperative is present in GMS, then the question of its
validity does not even arise in the first place; but the question of validity
is raised in GMS; therefore there is also no thesis of the fact of reason in
GMS. And indeed the question of validity in GMS is also answered with
a deduction; the deduction is the answer to the question as to how a
categorical imperative is possible, and this question is answered in the
fourth subsection of GMS III.
c) The confirmation argument

We have already briefly pointed out that in KpV Kant often55 refers to
our common rational moral knowledge and the moral experience asso-
55 See KpV: 27; 32; 35; 44; 91; 105; 155.

ciated with it. Here he connects the thesis of the fact of reason with this
common rational moral knowledge. And this is of course no accident.
For the ‘fact’ at issue –namely the consciousness of the absolute validity
of the categorical imperative– is not some special insight or knowledge
on the part of the trained or learned philosopher, but rather a funda-
mental experience that belongs to human existence itself. Just after Kant
has introduced the thesis of the fact of reason in the Remark to §7, he
says in a passage we have already cited: “The fact mentioned above is un-
deniable. One need only analyze the judgment that people pass on the
lawfulness of their actions...”.56 Later he explicitly thematizes this point
when he writes, in another passage already cited: “It was necessary first
to establish and justify the purity of its origin [that of the categorical im-
perative] even in the judgment of this common reason [...] But for this reason
the justification of moral principles as principles of a pure reason could
also be carried out very well and with sufficient clarity by a mere appeal to
the judgment of common human understanding”.57 Thus the ‘fact’ is shown by
the ‘judgment’ of common reason; it reveals or manifests itself in this
judgment. In GMS III the situation is quite different. After Kant has fur-
nished a deduction as an answer to the question as to how a categorical
imperative is possible (in the fourth subsection), he writes: “The practical
use of common human reason confirms the correctness of this deduc-
tion”.58 That is to say: in contrast to KpV, the proof of the absolute va-
lidity of the categorical imperative does not consist in a ‘fact’, which
shows itself in this practical use of common human reason; rather, this
practical use merely ‘confirms’ the deduction. Although, according to
Kant, even the most hardened scoundrel is also “conscious of a good
will”,59 this consciousness is not as such a sufficient reason for regarding
the validity of the categorical imperative as ‘undeniable’. In GMS Kant
does not of course challenge or reject the idea that we human beings
have a consciousness of the moral law –how and why would he ever do
such a thing? But in contrast to the thesis of the fact of reason in KpV,
Kant does not yet credit this consciousness with the epistemological
56 KpV: 32, 2.
57 KpV: 91f.
58 GMS: 454, 20.
59 GMS: 455, 4.
function of ‘immediately’ and ‘undeniably’ vouchsafing knowledge of

the absolute validity of the categorical imperative.
The confirmation argument can be formulated as follows:
1. If Kant defends the thesis of the fact of reason in GMS, then the
practical use of common human reason will not merely confirm the
correctness of the deduction of the categorical imperative.
2. The practical use of common human reason merely confirms the
correctness of the deduction of the categorical imperative.
GMS.
A possible objection here would be to point to KpV: there Kant says

that moral “experience”,60 as exemplified and described by reference to
the famous gallows example, also “confirms” something.61 But if it is
possible that moral experience in a particular context ‘confirms’ some-
thing, although in this context we are (indirectly) directed to the fact of
reason, this shows that the passage regarding ‘confirmation’62 is not that
decisive. I admit that the confirmation argument on its own would per-
haps be relatively weak; but at the same time I think that it acquires
strength in the light of the other two arguments (and especially in the
light of the aforementioned deduction argument that I have not exam-
ined in detail here). As far as the parallel passage in KpV is concerned, I
would simply say this: what is ‘confirmed’ there is “this order of con-
cepts”.63 It is true that the experience described in the gallows example
also reproduces what the fact of reason as the “consciousness of that
moral law” consists in.64 But what is ‘confirmed’ here is not the fact of
reason itself but the ‘order of concepts’ of morality and freedom with
regard to that order of knowledge regarding the concepts of freedom
60KpV: 30, 21.

61KpV: 30, 21.
62 See GMS: 454.
63
GMS: 454.
64 KpV: 30, 3. I have undertaken to elucidate the gallows example in detail in
Schönecker, D., “Kant’s Moral Intuitionism. The Fact of Reason and Moral Predis-
positions”, Kant Studies Online, Feb. 2013, pp. 1-38.
and law and their relationship to one another which Kant is so con-
cerned to describe: the moral law is the ‘ratio cognoscendi’ of freedom,
rather than the other way around. Hence Kant recapitulates his claim in
the example of the subject of the prince: “He judges, therefore, that he
can do something because he is conscious that he ought to do it and cog-
nizes freedom within him, which, without the moral law, would have
remained unknown to him”.65
Anyone who defends the fact of reason approach to GMS must be
able to show that Kant both defends the thesis of the fact of reason in
this text and at the same time leaves open the possibility that the moral law
is a phantom; that Kant at the same time still has room for the question as
to why one ought to subject oneself to the moral law; and that Kant at
the same time can write that the practical use of common human reason
merely ‘confirms’ the correctness of the deduction of the categorical im-
perative. Of course it is true that in GMS Kant does not regard the moral
law as a phantom, and of course he does have an answer to the subjection
question. But that is not the point. The point is that there is no longer
any room for such doubts in KpV, and that is so precisely because the
consciousness of the moral law and thereby the validity of the moral law
is a ‘fact’. If Kant had insisted upon such a ‘fact’ in GMS, or, to put it
another way, if he had set out the thesis of the fact of reason in GMS,
there would no longer be any room for such doubts, even if these
doubts were ultimately allayed; but there is still room in GMS for such
doubts; therefore Kant does not set out the thesis of the fact of reason
in GMS. May we not thus conclude: q.e.d?66
Universität Siegen
Siegen, Germany
65 KpV: 30; my emphasis.

66 A shorter version of this paper was first published in German: “Warum es in der
Grundlegung keine Faktum-These gibt. Drei Argumente”, in Puls, H. (Hrsg.), Kants Recht-
fertigung des Sittengesetzes in Grundlegung III. Deduktion oder Faktum?, Berlin, de Gruyter, 2014,
pp. 1-14. Many thanks to Nicholas Walker for providing the English translation of this
text.
References
Kant’s works are cited from the standard Academy Edition, with the
corresponding page numbers and line numbers (e.g. 31, 24). The English
translations used in the text (which all contain the Academy pagination)
are: Groundwork of the Metaphysics of Morals, Critique of Practical Reason, and
The Metaphysics of Morals, translated by M. Gregor, in: Immanuel Kant,
Practical Philosophy, The Cambridge Edition of the Works of Immanuel
Kant, Cambridge, Cambridge University Press, 1996. Also cited is: Im-
manuel Kant, Religion within the Boundaries of Mere Reason, translated by A.
W. and G. di Giovanni, Cambridge, Cambridge University Press, 1998.
Allison, H. E., Kant’s Theory of Freedom, Cambridge, Cambridge University

Press, 1990.
Damschen, G. – Schönecker, D., Selbst philosophieren. Ein Methodenbuch,
Berlin – Boston, De Gruyter, 22013.
Henrich, D., “Das Prinzip der Kantischen Ethik”, Philosophische Rundschau
2 (1954/55), pp. 20-38.
——, “Über die Einheit der Subjektivität”, Philosophische Rundschau 3
(1955), pp. 28-69. English translation: “On the Unity of Subjectivity”,
in Henrich, D., The Unity of Reason. Essays on Kant’s Philosophy, edited
by R. L. Velkley, Cambrigde, MA, Harvard University Press, 1994.
——, “Der Begriff der sittlichen Einsicht und Kants Lehre vom Faktum
der Vernunft”, in Prauss, G. (Hrsg.), Kant. Zur Deutung seiner Theorie von
Erkennen und Handeln, Köln, Gütersloh, 1973, pp. 223-253.
——, “Die Deduktion des Sittengesetzes. Über die Gründe der Dunkel-
heit des letzten Abschnittes von Kants ,Grundlegung zur Metaphysik
der Sitten’”, in Schwan, A. (Hrsg.), Denken im Schatten des Nihilismus.
Festschrift für Wilhelm Weischedel zum 70. Geburtstag, Darmstadt, Wissen-
schaftliche Buchgesellschaft, 1975, pp. 55-112.
Husserl, E., “Kritik der Kantischen Ethik”, in Husserl, E., Vorlesungen
über Ethik und Wertlehre 1908-1914, Dordrecht – Boston – London,
Kluwer, 1988, pp. 402-418 (Husserliana XXXVIII).
Puls, H., “Freiheit als Unabhängigkeit von bloß subjektiv bestimmten

Ursachen. Kants Auflösung des Zirkelverdachts im dritten Abschnitt
der Grundlegung zur Metaphysik der Sitten”, Zeitschrift für philosophische
Forschung, 65 (2011), pp. 534-562.
Schönecker, D., Kant: Grundlegung III. Die Deduktion des kategorischen
Imperativs, Freiburg – München, Karl Alber-Verlag, 1999.
——, “Textvergessenheit in der Philosophiehistorie” in Schönecker, D. –
Zwenger, Th. (Hrsg.), Kant verstehen / Understanding Kant. Über die Inter-
pretation philosophischer Texte, Darmstadt, Wissenschaftliche Buchgesell-
schaft, 22004, pp. 159-181.
——, “The Transition from Common Rational to Philosophical Rational
Moral Knowledge in the Groundwork”, in Ameriks, K., – Höffe, O.
(eds.), Kant’s Moral and Legal Philosophy, Cambridge, Cambridge Univer-
sity Press, 2009, pp. 93-122. [A translation of “Gemeine sittliche und
philosophische Vernunfterkenntnis. Zum ersten Übergang in Kants
Grundlegung”, Kant-Studien, 88 (1997), pp. 311-333.]
——, “Kant über Menschenliebe als moralische Gemütsanlage”, Archiv
für Geschichte der Philosophie, Heft 2/2010, pp. 133-175 (unter Mitarbeit
von Alexander Cotter, Magdalena Eckes und Sebastian Maly).
——, “Kant’s Moral Intuitionism. The Fact of Reason and Moral Pre-
dispositions”, Kant Studies Online, Feb. 2013, pp. 1-38.
——, “Das gefühlte Faktum der Vernunft. Skizze einer Interpretation
und Verteidigung”, Deutsche Zeitschrift für Philosophie, Heft 1/2013, pp.
91-107.
THE RECOVERY OF ACTION IN SOCIAL THEORY:
ACTING OUT OF SENTIMENT, ACTING OUT OF
CHARACTER, ACTING OUT OF INTEREST,
ACTING OUT OF WILL
ANA MARTA GONZÁLEZ
In his 1996 book, The Myth of Social Action, Colin Campbell wrote that
contemporary social theory is lacking
“a theory of action which is not in the first instance a theory of so-

cial action (that is to say, a theory of communicative action or interac-
tion), and yet which both fully embodies the principle of voluntarism
and recognises the reality and significance of intra-subjective processes –
without, however, being so completely intellectualist as to regard its sub-
ject-matter as composed entirely of decision-making processes (or in-
deed of ‘meaning-making’ processes)” (Campbell, C., The Myth of Social
Action, Cambridge, Cambridge University Press, 1996, p. 156).1
By stressing the need for a theory of action that goes beyond con-
temporary theories of social action, Campbell was actually pointing a bit
further to the need for recovering a stronger notion of subjectivity that
1 He goes as far as saying that presently “the practising sociologist has a choice be-
tween two broad alternatives, neither of which actually focuses on action. On the one
hand, there are those theorists which ignore action in order to concentrate upon the pro-
cesses of calculation and judgement underlying the decisions which individuals take
which lead them to embark on one course of action rather than another. (…) On the
other hand, the principal alternative on offer is represented by theories which strangely
have a similar preoccupation with intellectual processes. However, in this case the focus
is less on decision-making than on meaning and sense-making and hence on those bodies
of knowledge, cognitive resources and processes (particularly as manifest in language and
talk), through which the actor’s experiences of the world are rendered intelligible or
‘meaningful’. (…) However, here too there is a similar lack of concern with the actual
subjective meaning informing the conduct of individual actors whilst, once again, action
as such is not the real focus of study. Indeed, it is noticeable that neither alternative
shows much interest in the emotive or conative dimensions of conduct or in the aeti-
ology of action”. (Campbell, C., The Myth of Social Action, pp. 16-17).
80 Ana Marta González
is not diluted in social interaction.2 At the roots of this situation, ac-

cording to Campbell, is a basic misinterpretation of Weber’s account of
social action as a subcategory of action, and the enthroning of social ac-
tion as the prevalent category instead. Campbell identifies the roots of
this misinterpretation in the influential theories of Schutz and C. Wright
Mills,3 and in certain developments in the philosophy of action derived
from a linguistic turn that obscured the relevance of subjective motives
distinct from end-reasons or intentions.
It is on this latter aspect that I will focus in this chapter.
I think Campbell may be right in criticizing the loss of individual ac-
tion in social action. Further, I think that unless we develop a clear ac-
count of the distinction between action and social action, the adjective
“social” does not make much sense of real agency. I think he is also right
in criticizing the over-intellectualism of much of modern social theory,
although this has been reconsidered and revised in the last few years.
Maybe he is also right in his criticism of the reception of the contem-
porary philosophy of action in social theory.
However, I think that philosophical accounts of human action are
richer than what his vision suggests. Indeed, while the philosophy of ac-
tion developed after Wittgenstein certainly challenged the until then pre-
vailing Humean orthodoxy about the nature of human action and moti-
vation in ways that could suggest an over-intellectualization of behaviour
because of its emphasis on meanings and correct descriptions of actions
and because of its rejection of motives as internal states that cause be-
haviour, the philosophical analysis of action does not necessarily lead to
this conclusion. But it is not necessary to look exclusively to contempo-
rary debates on the philosophy of action to find a richer account of
action. For this task, it is possible to refer directly to Aristotle, Hume or
Kant, whose analysis of action and human agency are insightful and
2 In his view, the subject would have been lost in subsequent social theory, even in
some brands of social theory which, at least in certain respects, identify themselves in
line with the Weberian tradition, beginning with Parsons himself (Campbell, C., The Myth
of Social Action, p. 8), but also with Alfred Schutz, whose phenomenologically inspired
theory of action, would instead be only a theory of meaning. This is also valid, to a cer-
tain extent, in symbolic interactionism, associated with the work of G. H. Mead, or with
Goffman’s dramaturgical perspective. See Campbell, C., The Myth of Social Action, p. 16.
3 See Campbell, C., The Myth of Social Action, pp. 15, 31, 36, 40ff.
The Recovery of Action in Social Theory 81
nuanced enough to help us recognize the general structure and elements

of human behaviour, and provide us with a critical view of Weber’s own
typology of action. This is what I plan to do, after briefly reviewing
Weber’s approach to action, and Campbell’s objections to contemporary
theories of social action.
1. The loss of action in social theory
As we know, the task Weber envisions for sociology as a science is the

understanding of social action, of its meaning, as a precondition for pro-
viding a scientific explanation for it. What counts as a proper explanation
of any given action is something that cannot be decided unless we know
what an action is. This requires us to enter into the ratio or logos defining
the agent’s behaviour,4 for, according to Weber, “we shall speak of ‘ac-
tion’ insofar as the acting individual attaches a subjective meaning to his
behaviour –be it overt or covert, omission or acquiescence”.5
We should notice that this is not always the case: not every behaviour
counts as a meaningful action, since there are reactive behaviours which
lack any precise meaning. Interestingly, the limits between both kinds of
behaviour are not always easy to demarcate, at least not from the outside.
Habitual actions or sudden responses often resemble more reactions
than actions. The basic point, however, is that the connection between
movements and purpose cannot be an accidental one if the behaviour is
to count as an action in the strong sense of the term. For Weber, this is
the case both in individual and in social action: “Action is ‘social’ insofar
as its subjective meaning takes account of the behaviour of others and is
thereby oriented in its course”.6
As Peter Winch likes to say, “all meaningful behaviour must be social
since it can be meaningful only if governed by rules and rules presup-
4 Ultimately, the need for sociological explanations first appears when the action is
not socially intelligible in the first place. See Martin, J. L., The Explanation of Social Action,
Oxford, Oxford University Press, 2011, pp. 13-14.
5 Weber, M., Economy and Society. An outline of interpretive sociology, vol. I, ed. by G. Roth
and C. Wittich, Berkeley – Los Angeles, University of California Press, 1978, p. 4.

6 Weber, M., Economy and Society, vol. I, p. 4.
pose a social setting”.7 However the fact that many actions, and their
meaning, presuppose certain social institutions, such as money, language,
property, and many social relations, does not of itself make the actions,
as such, social. In other words: there is a significant sense in which some
of our actions cannot be defined as social, for society is part of them in
an accidental, not essential, way. Weber’s own example has become fa-
mous:
“Not every type of contact of human beings has a social character;

this is rather confined to cases where the actor’s behaviour is meaning-
fully oriented to that of others. For example, a mere collision of two cy-
clists may be compared to a natural event. On the other hand, their at-
tempt to avoid hitting each other, or whatever insults, blows, or friendly
discussion might follow the collision, would constitute ‘social action’”
(Weber, M., Economy and Society, vol. I, pp. 22-23).
The characterization of any action as “social” does not depend on

external conditions, but rather on the meaning intended by the agent. Of
course, to the extent that we all live in a social context, and have inter-
nalized many social relations,8 it is possibly to argue, with Luckmann,9
that every action is socially conditioned, or even, social by default, very
much as an omission can also be considered an action: because against
7 Winch, P. A., The Idea of a Social Science and its Relation to Philosophy, London,
Routledge, second edition, 1990, p. 116.

8 “I have linked the assertion that social relations are internal with the assertion that
men’s mutual interaction ‘embodies ideas’, suggesting that social interaction can more
profitably be compared to the exchange of ideas in a conversation than to the interaction
of forces in a physical system. This may seem to put me in danger of over-intellec-
tualizing social life”. Winch, P. A., The Idea of a Social Science and its Relation to Philosophy, p.
128.
9 See Luckmann, T., Theorie des sozialen Handelns, Berlin – New York, Walter de
Gruyter, 1992, p. 93ff. See also: “Reine Denkacte verlieren in Anwesenheit anderer ihre
Reinheit. Eine Nicht-Beschäftigung mit anderen in deren Anwesenheit ist mehr oder
minder auffällig und daher bedeutsam. Eine denkende Beschäftigung, die keinen Anteil
an anderen nimmt, sieht wie eine Abwendung aus und kommt adhere einer bestimmte
(‘unhöflichen’) Form gesellschaftlichen Handelns nahe” (Luckmann, T., Theorie des sozialen
Handelns, p. 113).
the background provided by socially shared meanings,10 and consensual

understandings, this particular action appears as lacking something that
would make it socially meaningful.
However, there is a sense in which for an action to qualify as spe-
cifically social it is not enough that it is it is carried out in a social context
and performed by a social being that lives in relationship with others; it is
not even enough that in acting I take others into account, but it is nec-
essary that these generic or individual others enter in the determining
reason of my action.11 For, in acting, I can certainly take others into ac-
count, without these others being the determining reason of my action.
Stressing the difference between proper causes and external condi-
tions of behaviour –a distinction which often gets lost in conventional
sociological analysis12– is important to realize the shift that, according to
Colin Campbell, has taken place in sociological theory in the last decades.
In his account, during the 60s and 70s,
“British sociologists tended to regard action rather than social action

as the central concept… By the end of the 70’s, however, there had
emerged a marked tendency to treat the two terms as if they were inter-
10 See Munch, P. A., “‘Sense’ and ‘Intention’ in Max Weber’s Theory of Social Ac-
tion”, in Hamilton, P. (ed.), Max Weber: Critical Assessments 1, vol. II, London – New York,
Routledge, 1991, pp. 12-22.
11 Raimo Tuomela takes a different approach, both in the definition of social action
and in the underlying philosophy of action: “By a social action we… mean a Joint action
performed by several agents who suitably relate their individual actions to the others’ ac-
tions in pursuing some Joint goal or in following some common rules, practices or the
like. By a single-agent action we mean roughly a performance, viz. something, usually a
change, an agent brings about so that this something has an epistemically public charac-
ter. A singular action is viewed as a process-like sequence of events: (1) an antecedent
mental event (willing in the case of intentional action), (2) a bodily movement, and (3) a
result event, viz. a change in the world”. Tuomela, R., A Theory of Social Action, Dordrecht
– Boston – Lancaster, Reidel Publishing Company, 1984, p. 12. For Tuomela, social ac-
tion is joint-action, and the pre-condition for this is to develop we-attitudes; in our ac-
count, any action that incorporates the considerations of others as a determining ground
for action can be characterized as social action, even if it is performed by a single indi-
vidual.
12 See Martin, J. L., The Explanation of Social Action, p. 66ff. He contrasts common-
sensical approach persisting in the legal tradition, to the causes of behaviour with the
simple counterfactual explanation that is current in sociological theory.
changeable, heralding the beginnings of a reversal of the former order

of significance… By the 80’s the term ‘social action’ has become pre-
dominant and the term ‘action’ is in danger of disappearing altogether”
(Campbell, C., The Myth of Social Action, p. 23).
For Campbell, this move “is indicative of a fundamental change in

the nature of the discipline”.13 Implicit in this revolution of the disci-
pline is the downgrading of the individual as creative social actor, and
the vanishing of subjective meaning –the defining mark of action– into
“inter-subjective” meaning. This new position, largely due to Schutz’s re-
interpretation of Weber, is now the dominant one.14
Campbell particularly complains that contemporary emphasis on the
intellectual dimension of action obscures its psychological and physical
reality, its condition as a worldly event; he thinks that this situation has a
negative impact on sociological analysis, because it directs attention to-
ward the universal or general elements of human interaction, preventing
us from approaching the individualizing elements which give each action
its particular uniqueness.15
Among the individualizing elements that render any action the prop-
erty of a particular subject, Campbell highlights motive and suggests that
its rejection by contemporary social theorists is particularly impover-
ishing.16 Interestingly, Schutz himself highlighted the lack of a theory of
motivation in Parson’s theory of social action,17 suggesting that “only a
theory of motives can deepen the analysis of social action”.18 Campbell
regards the fact that many sociologists tend to reject any talk of motives
in terms of “internal states” as fallacious as a sign of the influence
13 Campbell, C., The Myth of Social Action, p. 24.

14 See Campbell, C., The Myth of Social Action, p. 35.
15 See Campbell, C., The Myth of Social Action, pp. 43-44.
16 See Campbell, C., The Myth of Social Action, p. 67.

17 See Alexander, J. C., “Action and its Environments”, in Alexander, J. C. – Giesen,
B. – Munch, R. – Smelser, N. J. (eds.), The Micro-Macro Link, Berkeley – Los Angeles,

University of California Press, 1987, p. 296.
18 Schutz, A., “Parson’s Theory of Social Action: Critical Review by Alfred Schutz”,
in Grathoff, R. (ed.), The Theory of Social Action. The Correspondence of Alfred Schutz and
Talcott Parsons, Bloomington – London, Indiana University Press, 1978, p. 32.
exerted by contemporary non-naturalist philosophy of action.19 He

seems to think that in order to defend the existence of a subjective
meaning, which is not to be diluted into social or inter-subjective mean-
ing, it is necessary to defend the existence of motives in terms of “inter-
nal states”, which cause our behaviour. This account of motivation, how-
ever, has been criticized by post-wittgensteinian philosophy of action, in-
sofar as it encourages the re-definition of motives in terms of a restrict-
ed notion of “reasons”, as opposed to causes, a manoeuvre which, in
Campbell’s view would have resulted in an over-intellectualizing account
of human agency.20
The point of the philosophy of action developed after Anscombe’s
influential work on Intention, was surely to argue that actions are not just
any sort of causal and caused events, but are meaningful events, whose
meaning cannot properly be assessed by any contingent linkage what-
soever between beliefs and desires, but rather requires an understanding
of the intention which informs them, for this intention explains the pecu-
liar order we discover in the behaviour of animate beings.
19 See Campbell, C., The Myth of Social Action, pp. 46, 53.
20 “This is because the assumption of an extensive parallelism between language and
action is deeply misleading, as there is a fundamental difference between identifying the
meaning of a symbol and establishing the meaning of an event such as an act”
(Campbell, C., The Myth of Social Action, p. 132). Now, in Campbell’s view, while “it is pos-
sible to study texts, ‘dead’ and artificial languages, and indeed linguistic meanings gener-
ally, without necessarily investigating the language users themselves. This cannot be the
case with actions, which are inseparable from actors” (Campbell, C., The Myth of Social
Action, p. 135). Indeed, “all actions, even those such as speech which serve expressive or
communicative purposes, are also physical events… The search for the meaning of a
word is usually ended once its referent… has been successfully identified. An action by
contrast is a physical, biological and psychological event, and hence establishing its mean-
ing involves much more than simply identifying a referent. It involves identifying the
conditions which led to its occurrence, the manner of its accomplishment, as well as its
possible consequences; all of which can be said to comprise its meaning. It follows from
this that whilst the meaning of a word can be studied purely synchronically in relation to
the language system of which it is part, actions must necessarily be understood diachron-
ically as they unfold over time” (Campbell, C., The Myth of Social Action, p. 136). More
generally, while texts can be understood merely in terms of meaning, for the under-
standing of human action other factors have to be considered.
To argue this point, however, there is no need to deny the causal effi-
cacy of the agent’s choice, neither is it necessary to reject the influence of
non-cognitive factors in human behaviour. In fact, we are usually quite
aware of these conditional factors, and often resort to them to explain
why we act in certain ways. For instance, to explain a lack of determina-
tion in pursuing our goals, we might refer to fatigue. Resort to this kind
of explanation, however, does not mean that we think of our own
behaviour simply in causal terms;21 it means, rather, that while we think
of ourselves mostly as rational agents who direct our behaviour accord-
ing to reasons, we are also able to identify some internal and external fac-
tors which influence our behaviour in relevant ways, and which can even
prevent us from attaining our goals.22
Anscombe herself made a distinction between mental causes and rea-
sons and recognized the role of the former in prompting certain reac-
tions and also actions.23 However, she was interested in stressing the dif-
ference between the object of an action –which is the matter of
intention– and what we could call its eventual mental cause, her point
being that intention is neither a form of mental cause nor can it be
reduced to a mental cause. Further, she was keen to point out that a
mental cause need not be a mental event, such as a feeling, a thought, or
an image: it could just as well be a knock on the door, if only it is perceived
by the person affected. And, as such, it cannot be necessarily identified
21 “Explaining purposive behaviour in terms of causal mechanisms is not tantamount
to propounding a causal theory of action. For one thing, the pertinent activity of these
mechanisms is not prior to but concurrent with the movements they guide. But in any
case it is not essential to the purposiveness of a movement that it actually be causally
affected by the mechanism under whose guidance the movement proceeds” (Frankfurt,
H., “The problem of action”, in Frankfurt, H., The Importance of what We Care About,
Cambridge, Cambridge University Press, 1988, p. 75.
22 These reasons often echo some environmental or biological constraints that
condition our agency. Indeed, as Martin notes, “when we account for our actions, it turns
out that we often emphasize the environment and deemphasize the internal states: we
take for granted our (own) reasonable nature, and explain our action as response to prov-
ocations, invitations, or constraints of the environment” (Martin, J. L., The Explanation of
Social Action, p. 21).
23 See Anscombe, G. E. M., “Intention”, in Metaphysics and the Philosophy of Mind. The
Collected Philosophical Papers of G. E. M. Anscombe, II, Oxford, Basil Blackwell, 1981, pp. 75-
82.
with what we usually call “motive”, which in turn, cannot be identified

with the intention.24
Usually philosophers speak of intentions to refer to what an agent in-
tends, to what someone aims at or chooses; and motives to refer to what
determines, or causes, the aim or choice. In recalling these distinctions,
Anscombe was aware that “popularly, ‘motive’ and ‘intention’ are not
treated as so distinct in meaning”, and yet, she noted that even in ordi-
nary language we often make this distinction, for instance, when we say
that someone acted out of love or pity… like a straightforward way to
point at the motive. This, however, is different from saying that motives
“determine” or “cause” our actions: “it is a mistake to think one cannot
choose whether to act from a motive”.25 She suggests that giving a mo-
tive resembles putting one’s own actions in a certain light, and introducing
a distinction, similar to one raised by Schutz,26 between backward look-
ing and forward looking motives; the latter being very much like inten-
tions.27
Now, it seems to me that Campbell wants to retain the idea of motive
as a “mental cause” because of the common-sense notion that voluntary
actions entail some “feeling of effort”, as James would put it, so that this
feeling is the reference point for identifying voluntary acts. However, the
effort we eventually put into action –the feelings Hume associates with the
production of movements, which are perhaps what, following Parsons,
Campbell has in mind when he requires a distinctive “mental act” as
cause of behaviour28–, is not to be confused with the intention, the deci-
sion, or, generally, what we call the will.
24 Parsons’ motivational concepts –drive, drives, need-dispositions– remain at the
causal level. See Parsons, T. – Shils, E. (eds.), Toward a General Theory of Action. Theoretical
Foundations for the Social Sciences, New Brunswick – London, Transaction Publishers, 2001
[first edition 1951], p. 110ff.
25 Anscombe, G. E. M., “Intention”, p. 80.
26 See Schutz, A., “Parson’s Theory of Social Action: Critical Review by Alfred
Schutz”, pp. 33-34.

27 See Anscombe, G. E. M., “Intention”, p. 80.
28 Parsons’s definition of “unit act” is remarkably mechanistic: “The basic unit may
be called the ‘unit act’. Just as the units of a mechanical system in the classical sense, par-
ticles, can be defined only in terms of their properties, mass, velocity, location in space,
direction of motion, etc., so the units of action systems also have certain basic properties
without which it is not possible to conceive of the unit as ‘existing’… In this sense then,
So, while Campbell is right when he notes that “the central question
at issue is whether ‘acting for a reason’ can or cannot be reconciled with
‘behaving from a cause’”,29 it is important to note that such reconcili-
ation depends on 1) understanding the nature of human choice, which,
as Aristotle writes, is never just a mere reason, but rather a reasonably
informed desire, that is, a deliberate desire;30 and 2) further clarifying the
role of emotions in human behaviour. Taken together, both points serve
to provide us with a clearer and richer account of human motivation,31
which can be found relevant to the purpose of restoring a qualified view
of human subjectivity to social theory. As Anscombe writes,
“roughly speaking… the more the action is described as a mere

response, the more inclined one would be to the word ‘cause’; while the
more it is described as a response to something as having a significance
that is dwelt on by the agent, or as a response surrounded with thoughts
and questions, the more inclined one would be to use the word ‘reason’.
But in many cases the distinction would have no point” (Anscombe,
G.E.M., “Intention”, p. 81).
In what follows, however, I am not going to dwell on the debate on

reasons versus causes as conducted in contemporary philosophy of ac-
tion. I will instead turn generally to the issue of motivation as it comes
an ‘act’ involves logically the following: (1) it implies an agent, an ‘actor’. (…) (2) the act
must have an ‘end’, a future state of affairs toward which the process of action is
oriented. (3) It must be initiated in a ‘situation’. (…) Finally (4) there is inherent in the
conception of this unit, in its analytical uses, a certain mode of relationship between
these elements. That is, in the choice of alternative means to the end, in so far as the
situation allows alternatives, there is a ‘normative orientation’ of action” (Parsons, T., The
Structure of Social Action, vol. I, New York, The Free Press, 1968 [first ed. 1937], p. 44.
What Parsons later misses in Pareto is the third element, namely, “an element of ‘effort’
by virtue of which the normative structure becomes more than a mere idea or ideology
without causal relevance” (Parsons, T., The Structure of Social Action, p. 298).
29 Campbell, C., The Myth of Social Action, pp. 64-65.

31 As Martin notes, “if we ignore the issue of motivation altogether –if we go
immediately toward a causal answer– we are taking every shred of personhood from our
actor… Somehow, before we have even commenced our investigation of social life, we
have managed to insult our subjects” (Martin, J. L., The Explanation of Social Action, p. 18).
out in the work of Aristotle, Hume and Kant, to see how their insights
provide us with critical tools to analyze Weber’s typology of action.
2. The philosophers’ approach to human action
I will start by stressing that from both an Aristotelian and a Kantian per-
spective, acting is a rational activity, not only because it is usually behav-
iour guided by the rational pursuit of an end32 –something which is not
necessarily obvious, precisely in the cases of emotional actions or habit-
ual actions– but more fundamentally because it involves, at its very heart,
a reference to reason as a norm,33 which is present even in cases in which
this reference is dismissed.
Stressing the rationality of action in this way is not meant to down-
play the role of the will; on the contrary, the fact that actions are radically
rational, and hence open to opposite results, is precisely why the intro-
duction of a distinctive active power is required. This power is in charge
of realizing any of the options selected by reason, as opposed to a power
which naturally follows whatever impulse we may experience after
sensing certain incentives. Asserting the existence of the will as a
distinctive rational power is compatible with asserting the fact that, for a
rational being, it is natural to follow the dictates of his or her reason.34
Now, as suggested above, the fact that our actions entail a reference
to reason as a norm is not merely related to their adequacy for any end
whatsoever; fundamentally, it entails a requirement about the moral
soundness of those very ends. Further, the fact that every action is
32 Perhaps the sense in which Parsons refers to the “normative value of action”. See
Parsons, T., The Structure of Social Action, p. 49, p. 74ff. I think this is also the sense in
which John Rawls speaks of Kant’s hypothetical imperative as normative. Yet the hypo-
thetical imperative cannot exist apart from the categorical imperative. See Korsgaard, C.
M., “Acting for a Reason”, in Korsgaard, C. M., The Constitution of Agency, Cambridge,
Cambridge University Press, 2008, p. 221, footnote.
33 In Kantian terms, this amounts to saying that the categorical imperative is a con-
stitutive principle of action. See Korsgaard, C. M., Self-Constitution. Agency, Identity, and
Integrity, Oxford University Press, 2009.
34 The eventual effort we may need to put into the realization of those dictates is a
different matter.
reflexively measured against a rational norm is the ground on which we

consider omissions as actions.35 More generally, we could not be made ac-
countable for our behaviours unless we assume the existence of such a
rational norm. This is the case with actions moved by passion or emo-
tion, as well as with habitual actions.
2.1 Acting out of passion according to Aristotle and Kant
Let us take the case of someone who, moved by wrath, hits someone
who offends him. We may say that in this case passion has obscured rea-
son. This is a way of acknowledging that reason has a role to play in that
behaviour, even if reason, in this case, has been overcome or blocked by
passion. Actually, from a classical perspective it would be more accurate
to say that passion –which always involves some sort of bodily, organic
modification– has been kindled by a particular reason –the offence
perceived as unjust–, which at the moment, has overridden the more
universal requirement of moral reason that requires us to refrain from
responding in an intemperate way.
Aristotle points out that while the person acting out of passion
certainly commits an injustice, he is not as such an unjust person, but
only an intemperate one, the difference being that the first harms inten-
tionally, while the second causes harm only as a result of his lack of con-
trol.36 Aristotle doesn’t hesitate to designate decision (prohairesis) as
“cause” of behaviour. This is also the language of Aquinas: “electio”,
e.g. choice, is the (efficient) cause of action. A defining mark of choice is
that it incorporates a ratio, or logos, which points at a certain end-reason.
An agent acts freely when he acts upon deliberative choice. This is also
the case in omissions. Harry Frankfurt has a good example to illustrate
this:
35 Unless there is a rational norm, demanding from us certain behaviour, say paying
taxes, we would lack the reference to interpret a “lack of action” precisely as a particular
sort of action, say fraud.
36 See Aristotle, Nicomachean Ethics, translated by T. Irwin, Indianapolis, Hackett
Publishing Company, 1987, V, 8, 1135 b 20-26.

“a driver whose automobile is coasting downhill in virtue of gravita-

tional forces alone may be entirely satisfied with its speed and direction,
and so he may never intervene to adjust its movement in any way. This
would not show that the movement of the automobile did not occur un-
der his guidance. What counts is that he was prepared to intervene if
necessary and that he was in a position to do so more or less effectively”
(Frankfurt, H., “The problem of action”, p. 75).
By contrast, the fact that the angry reaction is obviously not decided
upon deliberation, but is rather a mechanical consequence of the of-
fender’s provocation, suggests that it lacks what it is necessary to be a
perfect action, namely, deliberative choice, for choice does not merely
require an impulse to follow any particular reason whatsoever, but pre-
supposes a deliberative appraisal of this particular reason in the light of
an end, and ultimately in the light of the broader picture of what it
means to lead a good human life, that is, a life which makes sense as a
whole. Lack of temperance blocks this deliberative process: the agent re-
mains stuck at a particular level and is unable and unwilling to integrate
his/her action into the broad picture of a good human life.
In this description, indeed, it is important to note that lack of tem-
perance causes a specific inability for rational action, but, at the same time,
such inability is considered voluntary, for we expect rational agents to
take universal reason into account, and failure to do so counts as some-
thing reprehensible. In other words, while accepting that pas-
sions/emotions are embodiments of particular reasons, we expect hu-
man beings to introduce order into their actions so that the particular
reasons embodied in emotional reactions do not become all by them-
selves the determining ground of their agency, but are rather reframed in
the light of universal reason, that is, in the light of a broader picture of
what is good and bad for the human being, as a rational being. While
having such a picture does not guarantee the rightness of our particular
choices –we can still deliberately choose what is wrong, which would
result in a worse moral condition– the absence of such a picture hinders
the development of full agency.
Human beings need to anticipate a view of the human good, in order
to make sense of particular actions. In addition, they need to be enabled
to act accordingly. This is why moral philosophers have always found it
necessary to develop an account of the moral virtues in order to provide

a consistent approach to human agency. In order to act, we have to intro-
duce some unity in our passions, and this is the role of the virtues: lack
of virtues results in deficient or inconsistent agency.
Very often we designate this deficient agency as lack of character. By
contrast, full agency requires character. This is implicit in Kant’s words,
when he notes that “the man of principles has character. Of him we
know definitely what to expect. He does not act on the basis of his in-
stinct, but on the basis of his will”.37
From a Kantian perspective, character designates a mode of behave-
iour grown out of the influence of rational principles upon nature. This
is what Aristotle calls moral virtue. As a habit, virtue is a stable dispo-
sition for acting in certain ways; by contrast, sentiment is a transient dis-
position.
This latter distinction marks the difference between an apparently
generous action, performed out of a transient sentiment of generosity,
and a generous action, also performed without much deliberation, but
out of a firmly embedded disposition to act generously, that is, out of
habit or character. If we witness those actions from the outside, we may
not appreciate the difference. However, from the perspective of the
agent they are different.
In an attempt to avoid any confusion between routine and proper
moral virtue, Kant further distinguishes between habitus and habitus
libertatis, to stress the fact that
“virtue cannot be defined as an aptitude for free actions in confor-

mity with law unless there is added ‘to determine oneself to act through
the thought of the law’, and then this aptitude is not a property of
choice but of the will, which is a faculty of desire that, in adopting a
rule, also gives it as a universal law. Only such an aptitude can be counted
as virtue” (Kant, I., The Metaphysics of Morals, edited by M. Gregor,
Cambridge, Cambridge Universty Press, 1996, 6: 407).
37 Kant, I., Anthropology from a Pragmatic Point of View, translated by V.L. Dowdell,
Carbondale & Edwardsville, Southern Illinois University Press, 1978, 7:285.

Kant also distinguishes between passion and habit, but in different

terms from Aristotle. He characterizes passions as a true cancer of prac-
tical reason, whose corruptive power lies in the fact that they become a
sort of principle, alternative to reason, which blocks its ability to con-
sider things in the light of a more general view of the human good and
the universal moral law. Thus, in the Anthropology, Kant defines passion as
an “inclination, which hinders the use of reason to compare, at a par-
ticular moment of choice, a specific inclination against the sum of all in-
clinations”.38
According to Kant, passion introduces an imbalance in our practical
judgments, so that instead of pondering one good against others, in the
light of a general view of what it means to lead a human life, we become
focused on just one particular good, and start judging everything else
from that particular perspective.39 This is clear from the contrast he
draws between emotion and passion:
“The inclination which can hardly, or not at all, be controlled by rea-

son is passion. On the other hand, emotion is the feeling of a pleasure or
displeasure at a particular moment, which does not give rise to reflection
(namely the process of reason whether one should submit to it or reject
it)” (Kant, I., Anthropology, 7: 251-252).40
As it turns out, this hindrance to the comparative work of reason,

which manifestly obstructs the efficacy of the moral law on the deter-
mination of behaviour, is not necessarily linked to violent sensuous im-
pulses, but rather to the inversion of incentives that human beings take
38 Kant, I., Anthropology, 7: 265.

39 See Kant, I., Anthropology, 7: 266.
40 “To be subject to emotions and passions is probably an illness of mind because
both emotion and passion, exclude the sovereignty of reason. Both are also equally
strong according to degree; but in accordance with their quality, emotion and passion are
essentially distinct from one another as to the method of prevention as well as in that of
cure which the physician of souls would have to employ… Emotion is surprise through
sensation, whereby the composure of mind (animus sui compos) is suspended. Emotion
therefore is precipitate, that is, it quickly grows to a degree of feeling which makes
reflection impossible (it is thoughtless)… Passion, however violently it may present itself
(as a frame of mind belonging to the faculty of desire), takes its time, and is deliberative
in order to achieve its purpose” (Kant, I., Anthropology, 7: 251-252).
as determining ground of their will.41 This inversion of the incentives is

at the core of the propensity to evil that, in Kant’s view, affects human
behaviour.42
2.2 Habituation and resolution: Aristotle, Kant and Hume on character
Both Aristotle and Kant stress the need for virtue and character to lead a
good or moral life. Yet, they have different accounts of the genesis of
character in this sense: while Aristotle stresses habituation,43 Kant
stresses resolution. Indeed, according to Kant, a significant feature of
character is that its acquisition is never a matter of habit, but resolution,
not custom, but revolution. Echoing the Christian doctrine of conver-
sion as new birth, Kant writes:
“One may also take it for granted that the establishment of character
is similar to a kind of rebirth, a certain solemn resolution which the per-
son himself makes. This resolution and the moment at which the trans-
formation took place remain unforgettable for him, like the beginning of
a new epoch. This stability and persistence in principles can generally not
be effected by education, examples, and instruction by degrees, but it can
only be done by an explosion which suddenly occurs as a consequence
of our disgust at the unsteady condition of instinct…” (Kant, I., Anthro-
pology, 7: 294-295).
Of course, the differences between the Kantian and the Aristotelian

approaches to character can be qualified if we consider that for Aristotle
virtuous action also relies on deliberate choice.44 Still, Kant’s emphasis
on “conversion”, as the requisite to act well, contrasts sharply with the
weight Aristotle places on habituation to become good.45 Aristotle
41 See Kant, I., Religion within the Boundaries of Mere Reason, ed. by A. Wood and G. di
Giovanni, Cambridge, Cambridge University Press, 1998, 6: 24.

42 According to Kant, this propensity admits of grades. See Kant, I., Religion, 6: 29.
43 See Aristotle, Nichomachean Ethics, II, 1, 1103 a 15-1103 b 15.
44 See Aristotle, Nichomachean Ethics, II, 4, 1105 b 1-12.

45 See Aristotle, Nichomachean Ethics, X, 9, 1179 b 20-33.
further adds that the many “yield to compulsion more than to argument,
and to sanctions more than to what is fine”,46 the reason being that
“feelings seem to yield to force, not to argument”. If his analysis of
virtuous action brings Aristotle closer to Kant, his insistence on habit-
uation and coercion brings him closer to Hume.
As we know, Hume holds reason to be practically inert, and the pas-
sions as the only motives for action. Accordingly, instead of resorting to
the contrast between reason and passion, he develops an alternative ac-
count of character or “strength of mind”, in terms of prevalence of the
calm passions –among them, the general appetite to good and aversion to
evil– above the violent ones. For Hume, what people usually call “rea-
son” and take as the roots of moral character is nothing more than the
calm passions.47 Hume contrasts these calm passions –which often deter-
mine the will– with the violent ones –which also influence behaviour.48
He notes that “men often counter-act a violent passion in prosecution of
their interests and designs”.49
For Hume, passions deeply embedded in our character –passions
which have lost their eventual violence, and are calm, because they have
become custom– can usually override violent passions, moved by short
term desires and interests, although there are exceptional cases in which
the latter overcome the former.50 Accordingly, one could venture that the
whole point for one to acquire character or strength of mind consists in
stimulating the adequate passions, and reinforcing them with custom.
Hume leaves no room for Kant’s emphasis on the freedom of the will as
the basis of character; instead, he insists on the importance of custom.
Hume’s emphasis in the psycho-social conditions of motivation
explains another partial similitude with Aristotle, namely, the fact that, in
order to govern other people, Hume considers it more expedient to work
46 Aristotle, Nichomachean Ethics, X, 9, 1180 a 5.

47 See Hume, D., A Treatise of Human Nature, edited with by L. A. Selby-Bigge, sec-
ond edition, with text revised by P. H. Nidditch, Oxford University Press, 1978, 2.3.3;
SBN, 417.
48 See Hume, D., A Treatise of Human Nature, 2.3.3; SBN, 417-418.
49 See Hume, D., A Treatise of Human Nature, 2.3.3; SBN, 418.
50 See Hume, D. A Treatise of Human Nature, 2.3.3; SBN, 418.

on their violent passions, appealing to their inclinations, than to what is

vulgarly called “their reason”.51
2.3 Acting out of interest and acting morally
While at the practical level, Hume could certainly draw a distinction be-
tween interested and moral actions,52 at the foundational level, his
approach to human agency ultimately depends on his theory of the pas-
sions, and is therefore radically instrumental: the passions are for him the
only movers, whose long-term aims reason is supposed to serve through
the development of character and sound institutions– including, of
course, moral institutions. In this regard Hume seems satisfied with pro-
viding us with a conjectural history of moral institutions, which properly
internalized, would account for moral behaviour.53 From this perspective,
51 See Hume, D., A Treatise of Human Nature, 2.3.4; SBN, 418-419. However, the
similitude with Aristotle is limited, because, unlike Aristotle –who speaks of coercion as a
resource when habituation has failed– Hume adopts this strategy as a general policy. At
any rate, with this view in mind, Hume argues: “We ought to place the object in such par-
ticular situations as are proper to encrease the violence of the passion. For we may ob-
serve, that all depends upon the situation of the object, and that a variation in this par-
ticular will be able to change the calm and the violent passions into each other. Both
these kinds of passions pursue good, and avoid evil; and both of them are encreas’d or
diminish’d by the encrease or diminution of the good or evil. But herein lies the differ-
ence betwixt them: the same good, when near, will cause a violent passion, which, when
remote, produces only a calm one”. Hume, D., A Treatise of Human Nature, 2.3.4; SBN,
418-419. While emotional management has always been at the heart of rhetoric and
politics, explicit recommendation of the empirical management of human passions or
emotions, instead of the promotion of character represents a departure of classical
political philosophy. With this move, Hume is placing instrumental reason at the core of
social and political life.
52 In the sense that, for him, action qualifies as moral arouses the approval of an im-
partial observer, who sympathizes with the pleasure and utility of the agent and those
affected by his or her action.
53 For Hume, character, understood as prevalence of calm passions above the violent
ones, helps us in the individual pursuit of long term interests. Yet, in order to manage the
behaviour of other persons, Hume thinks it is more expedient to focus on the violent
passions. This is in tune with his view about the proper way to arrange institutions so
that people working in them would ultimately promote the common interest while pur-
suing their immediate goals. Indeed: for Hume, the political pursuit of long-term inter-
ests would largely depend on introducing the institutional arrangements necessary to
stimulate and manage violent passions according to a plan. In this way, Hume provides
moral action could be seen as the end-result of successful socialization,

first originated because of the natural desire to meet the ends of our
passions in the most convenient way.54
By contrast, Kant does not think of moral agency as the result of
successful socialization, something that could arise solely from a combi-
nation of nature and pragmatic reason. While there is certainly a natural
history of humanity, which approaches cultural achievements from the
perspective of pragmatic reason,55 this history cannot account for the
specificity of moral action.
Indeed, even if Kant concedes that pragmatic reason plays a role in
the arrangement of political institutions that secure peace, his account of
moral character goes far beyond Hume’s resort to a happy confluence of
calm passions and custom. Kant insists on the specificity of moral reason
and introduces a sharp contrast between acting out of principle and act-
ing out of sentiment (which ultimately amounts to acting out of interest,
in Hume’s sense).56
Thus, while Kant recognizes that “an interest is that by which reason
becomes practical, i.e., becomes a cause determining the will”,57 he intro-
duces a crucial distinction between having an interest and taking an
interest.58 In order to act we certainly need some sort of empirical incen-
tive; however, this is not the same as making that incentive directly the
an argument for basing government not so much on virtue as on balancing interests. See
Hume, D., A Treatise of Human Nature, 3.2.7; SBN, 538.
54 Although violent passions move us noisily to satisfy needs and desires in the short
term, and calm passions move us quietly to pursue long-term interests, generally speak-
ing, everything depends on the balance we can find between both kinds of passions.
Hume says that in the end, both passions refer to the same goods, but the latter are
prepared to look for those goods in the long term. This means that they allow more
room for instrumental reason, to think of ways of best achieving the goals.
55 See Kant, I., Perpetual Peace, in Kant, I., Perpetual Peace and Other Essays, translated by
T. Humphrey, Hackett Publishing Company, 1983, 8: 366.
56 Here, of course, we are confronted again with the duality between reason and
inclination. Thus, in the Anthropology, Kant defines inclination (inclinatio) as “a subject’s

sensuous desire which has become customary (habit)” (Kant, I., Anthropology, 7: 265).
57 Kant, I., Groundwork of the Metaphysics of Morals, translated by M. Gregor and J.
Timmermann, Cambridge, Cambridge University Press, 2012, 4: 460.

58 See Kant, I., Groundwork, 4: 460.
determining ground of our action. We may want to do good business,

and yet we cannot do so with the order of justice marginalized and set
up against the moral law. This means that we are required not to act on
interest but rather to take an interest in a certain way of doing business,
namely, a way that respects the moral requirements. In this account in-
strumental reason is also embedded in moral action. Reason is both the
norm and the ground or motive for moral behaviour.
Like Kant, Aristotle also distinguishes doing the right thing and doing
it for the right reason. Nevertheless, while the Aristotelian notion of vir-
tue entails both right intention and practical competence, Kant’s notion
of virtue is remarkable because of his emphasis on the right intention
above the practical competence required for materializing it. For Kant
this latter aspect could be isolated and explained as a matter of pragmatic
reason.
Again, it is not that Aristotle does not recognize any difference be-
tween moral and pragmatic reason.59 He draws an essential linkage be-
tween practical wisdom and moral virtue, so that, while he recognizes
that practical wisdom entails some sort of cleverness, he never reduces
this cleverness to mere pragmatic knowledge.60 Kant attributes the moral
element of Aristotelian phronesis to practical judgment, whose role is to
check the particularity of actions performed in the sensible world against
the universality of the moral law.61 This requirement –which involves the
reduction of Aristotelian phronesis to the judgment of conscience–, en-
tails the isolation of cleverness from its moral context: there is no longer
place for a cleverness specific to virtue: there is only mundane prudence,
which Kant explains both in terms of pragmatic “knowledge of the
world”, or else “private prudence”.62
59 See Aristotle, Nichomachean Ethics, VI, 13, 1144 a 21-29.

60 So he notes: “Intelligence (phronesis) is not the same as this capacity (cleverness)
though it requires it. Intelligence, this eye of the soul, cannot reach its fully developed
state without virtue… we cannot be intelligent without being good” (Aristotle,
Nichomachean Ethics, VI, 13, 1144 a 30, 35).
61 See Kant, I., Critique of Practical Reason, edited by M. Gregor, Cambridge,
Cambridge University Press, 1997, 5: 69.

62 See Kant, I., Groundwork, 4: 417.
It seems to me that even at this point we can find an interesting anal-

ogy in Aristotle in that he recognizes the existence of an art which is
very easily abstracted from moral regulation, namely, the “art of acqui-
sition”, specifically an “unnatural” form of acquisition. He describes this
unnatural form as a “the product of a certain sort of experience and
skill”,63 no longer connected with the management of the household for
it no longer serves the needs of the household, but rather becomes an
end in itself.64
For Kant, too, pragmatic reason needs to be subordinated to the re-
quirements of universal moral law. Ignoring this subordination, or in-
verting the order of incentives, amounts to immoral action.
3. Weber’s typology of action revisited
In the light of the preceding analysis, it is not difficult to trace continui-

ties and differences between the moral philosophical tradition and
Weber’s own typology of action. As we know, Weber contends that
“Social action, like all action, may be oriented in four ways. It may be:
1) instrumentally rational (zweckrational), that is, determined by expec-
tations as to the behavior of objects in the environment and of other
human beings; these expectations are used as ‘conditions’ or ‘means’ for
the attainments of the actor’s own rationally pursued and calculated
ends.
2) value rational (wertrational), that is, determined by a conscious belief
in the value for its own sake of some ethical, aesthetic, religious, or other
form of behavior, independently of its prospects of success;
3) affectual (especially emotional), that is, determined by the actor’s
specific affects and feeling states:
4) traditional, that is, determined by ingrained habituation” (Weber,
M., Economy and Society, vol. I, pp. 24-25).
63 See Aristotle, Politics, translated by E. Barker, Oxford, Clarendon Press, 1968, I, 9,
1, 1256 b - 1257 a 1.
64 See Aristotle, Politics, I, 9, 1257 b 12 - 1258 a 16.
Weber does not introduce this typology with reference to the philoso-
phical tradition, but rather in conformity to his own notion of “ideal
types –selective reconstructions of the empirical world, for epistemol-
ogical purposes–”, as Alexander puts it.65 At first sight, however, Weber’s
typology could be easily referred back to the motivational structures we
have considered so far. Thus, his description of instrumentally rational
and value rational actions could easily be paralleled with Kant’s own dis-
tinction between pragmatic and moral reason. Things become more
complicated with affectual action and traditional action.66 Still, to a
certain extent, both Aristotle and Hume can provide, in their own ways,
an equivalent account for each of these action-types.
And yet, also against the background provided by the moral philoso-
phers, Webers’ typology of action, as it stands, can be criticized in several
ways. The first criticism would be the very fact of placing “value rational
action” –Weber’s name for moral action– on the same level as the other
action-types, somehow forgetting that every action is moral by default.
Indeed, the moral approach to human agency is not just one approach
among others, but the basic and essential approach; whether we want it
or not, whether we consciously pursue a moral value or not, all our ac-
tions embody a moral meaning.
Jeffrey Alexander noted that Weber was aware of this, and, to a cer-
tain extent, recognized the intrinsic character of moral order67 –no mat-
65 See Alexander, J. C., Theoretical Logic in Sociology, vol. III. The Classical Attempt at
Theoretical Synthesis: Max Weber, London, Routledge & Kegan Paul, 1983, p. 25.
66 “The line between meaningful action and merely reactive behavior to which no
subjective meaning is attached, cannot be sharply drawn empirically. A very considerable

part of all sociologically relevant behavior, especially purely traditional behavior, is mar-
ginal between the two” (Weber, M., Economy and Society, vol. I, p. 4). Peter Winch recalls
Parsons’ critique at this point. See Winch, P. A., The Idea of a Social Science and its Relation to
Philosophy, p. 48ff.
67 For Alexander, Weber’s assertion that “an actor’s decision is ‘more free’ than would
otherwise be the case, to the degree it is ‘based more extensively upon his own ‘delib-
erations’, which are upset neither by ‘external’ constraints nor by irresistible ‘affect’” can
only be sustained “by the actor’s reference to an overarching normative order”. It is the
“constant and intrinsic relation to certain ultimate ‘values’ and ‘meanings’ of life, that
allow will and intentionality to come about: ‘values’ and ‘meanings’… are forged into
purposes (Zwecke) and thereby translated into rational-teleological action” (Alexander, J.
C., Theoretical Logic in Sociology, p. 24).
ter, at this point, that on Weber’s own approach values hold no validity
beyond the cultural system inhabited by the subject–. Nevertheless, this
view is in fact undermined by the rigidity of the typology itself, which in
the last account favours the independence of instrumentally rational ac-
tion. So, Alexander himself, who highlights the many times in which
Weber favours a multidimensional approach to action, is the first to ac-
knowledge that in providing his typology of action in those terms,
“Weber has slipped into an instrumental and materialist framework in
more general, presuppositional terms”.68
Actually, we could further criticize the fact that Weber presents his
division as based on different “orientations” or intended meaning of ac-
tions, rather than on different sources of motivation, when in fact he is
referring to the relative weight that different motivational sources –rea-
son, affect, habit– have in the implementation of action by the subject.
In asserting this we are not giving a “psychological” twist to Weber’s
typology of action, in the sense criticized by Peter Munch, who rightly
observes that “Weber… distinguishes between the intended sense of an
action and the actor’s motivation”, and stresses that
“as a sociologist, he is interested, not in the imputation of motive to

the individual actor, but in understanding social action in its ‘context of
sense’ in relation to ‘typical’ or ‘cross-sectional’ usages based on ‘consen-
sus’, that is, in its ‘cultural significance’” (Munch, P. A., “’Sense’ and ‘In-
tention’ in Max Weber’s Theory of Social Action”, p. 12).
This caution notwithstanding,69 the fact persists that, in distinguishing

actions according to ends, values, affects, and habit, Weber is not distin-
68 Alexander, J. C., Theoretical Logic in Sociology, p. 28. As a result, “he has described
instrumental behaviour not in a multidimensional but in an analytically more reductionist

way. To the degree that his theoretical understanding of action has been so reduced,
Weber will be unable to include normative order in the empirical analysis of zweckrational
behavior, and the voluntaristic element will drop out of this part of his work altogether”
(Alexander, J. C., Theoretical Logic in Sociology, p. 28).
69 Munch further observes that “Weber also attempted to guard against this misin-
terpretation of his ‘verstehende Soziologie’. In the essay on ‘Some categories of com-

prehensive sociology’, a whole section is devoted to the argument that sociology in this
sense is not a part of psychology, and that sociological explanation is not to be confused
with psychological explanation. Even a strictly rational act… –the kind of act that Weber
guishing them merely according to their intended meaning but also ac-
cording to different motivational sources.
Thus, while it is true that in order to get a proper understanding of
social reality we need to focus on the intended meaning of any social
action, it is still the case that 1) irrespective of the agent’s intended
meaning, his or her action has a moral meaning by default; 2) the
differences among action types are not conveyed so much in terms of
intended meanings as in terms of psychological categories.
Taken together, this means that, against the basic moral meaning of
human actions, there are recognizable psychological types. The extent to
which these psychological types can be taken as interpretive cues of so-
cial life is a different kind of problem, although we can get a glimpse of
that by looking at Tocqueville’s work, since Tocqueville, too, often takes a
look at the emerging psychological types to characterize social life as
such.70 Weber thinks of his own categories as defining ideal types, useful
for sociological interpretation, even if they are never to be found un-
mixed in practice; he also recognizes that this classification is not neces-
sarily exhaustive, but only the more relevant for sociological analysis.71
As a matter of fact, many interpreters coincide in pointing out that
one of the most specific elements of Weber’s sociological approach, is
the weight he gives to psychological factors in the explanation of mean-
ingful behaviour. Thus, one salient aspect of his characterization of in-
strumentally rational action is the emphasis on “expectations”. Expec-
describes as ‘the directly ‘most comprehensible kind’– of meaningful action structure– is

not, he claims, explained sociologically by reference to psychic or mental conditions but
exclusively by deriving it from expectations justified by valid experience. Here it seems
clear that the ‘sense’ of an action is inherent in the structure of the action itself, regard-
less of the mental state of the actor, and is directly comprehensible to the recipient as
well as to the observer in terms of established expectations based on verified experience.
At the other extreme is the completely ‘irrational’ conduct, say, a psychotic person, which
may well have a ‘motive’ but does not ‘make sense’ and is absolutely incomprehensible by
any socially established standard and, therefore, unexplainable in terms of comprehen-
sive sociology” (Munch, P. A., “‘Sense’ and ‘Intention’ in Max Weber’s Theory of Social
Action”, pp. 16-17).
70 See Boudon, R., “Individualistic tradition in sociology”, in Alexander, J. C. –
Giesen, B. – Munch, R. – Smelser, N. J. (eds.), The Micro-Macro Link, Berkeley – Los
Angeles, University of California Press, 1987, p. 49.
71 See Weber, M., Economy and Society, vol. I, p. 27.
tations about the behaviour of external objects as well as of other

human beings become the determinant factor of this specific action
type.72 We act rationally according to ends when we take those expec-
tations as the determining reason for our own behaviour aimed at the
consecution of our own ends.
Likewise, rational action according to values would be determined by
conscious beliefs regarding the value –ethical, aesthetic, religious or
otherwise– of any given behaviour, no matter its result:
“The possible (subjective) meaning of social action is not exhausted

by the orientation toward the presumable action of others. In the limit-
ing case, action may be oriented toward the ‘value’ of an act (as a matter
of duty, for example): then it is not oriented towards expectations, but
values” (Weber, M., Economy and Society, vol. II, Appendix I, p. 1376).
Acting because of a desire to further some ends and acting out of

our conviction are thereby not only clearly differentiated, but are also
highlighted as the defining marks of different psychological types. As
such, they are hardly reconcilable. This is part of the reason why in spite
of recognizing that we can articulate both elements in a single action –
we can certainly act rationally according to ends and yet out of convic-
tion or conscious concern for certain values–, Weber also wants to leave
room for that kind of behaviour which exhibits rationality according to
ends while no rationality according to values:
“Choice between alternative and conflicting ends and results may well
be determined in a value-rational manner. In that case, action is instru-
mentally rational only in respect to the choice of means. On the other
hand, the actor may, instead of deciding between alternative and con-
flicting ends in terms of a rational orientation to a system of values,
simply take them as given subjective wants and arrange them in a scale of
consciously assessed relative urgency” (Weber, M., Economy and Society,
vol. I, p. 26).
72 See Weber, M., Economy and Society, vol. II, Appendix I, pp. 1375-1376.
Drawing on the distinction between the objectivity of any given sys-

tem of values and the subjectivity of desire, Weber seems to suggest, in a
very Nietzschean way, that we can choose between being moral or not.
This would crystalize in economic behaviour, as conceptualized with the
subjective theory of value. Jeffrey Alexander reproduces Weber’s sen-
tence in the Protestant Ethic: “The earning of more and more money… is
thought of… purely as an end in itself ”.73 Sure, in this case, Weber is
linking this behaviour to something other than amoral attitudes –he is
linking it to the religious discipline of Calvinist Protestantism; yet, as
such, it represents –just like Aristotle’s unnatural acquisition– a behave-
iour that lacks any notion of intrinsic ends. This is so not because of
anxiety derived from the desire for life, as is the case in Aristotle, but
rather because of anxiety derived from the desire for an afterlife.
This latter consideration, pointing at Weber’s sociology of religion,
already suggests the relevance of emotional dimensions in his social the-
ory, something relevant if we consider that Weber’s approach has often
been criticized for taking rational action as the paradigmatic type of so-
cial action74 and the background for understanding. While the latter is
something difficult to avoid if sociology is to be considered a science in
charge of unveiling the reasons for social action75–, the critique as such
should be qualified. As Jürgen Gerhards notes, the concepts of ratio-
nality and affectivity are used by Weber both at the objectual and the
metatheoretical levels.76 He does not merely speak of rational and affec-
tive action types, but also reserves a role for emotion at the metatheo-
retical level because, in his own words, the interpretation of the meaning
of any action, succeeds at its fullest wherever we find some emotionally
empathic evidence, that is, when we re-live the connection of sentiments
73 Alexander, J. C., Theoretical Logic in Sociology, p. 34.

74 “For the purposes of typological scientific analysis it is convenient to treat all irra-
tional, effectually determined elements of behavior as factors of deviation from a con-
ceptually pure type of rational action”. Weber, M., Economy and Society, vol. I, p. 6.
75 See Weber, M., Economy and Society, vol. I, pp. 6-7.
76 See Gerhards, J., “Affektuelles Handeln – Der Stellenwert von Emotionen in der
Soziologie Max Webers”, in Weiss, J. (Hrsg.), Max Weber heute. Erträge und Probleme der
Forschung, Frankfurt a. M., Suhrkamp, 1989, pp. 335-357. See also Flam, H., Soziologie der
Emotionen. Eine Einführung, Konstanz, UVK Verlagsgesellschaft, 2002, pp. 44-60.
implicit in that action.77 Thus, the more receptive we are to the emotions
involved in certain actions, the better we can revive the meaning involved
in them.78 As Winch notes,
“this point is reflected in such common-sense considerations as the

following: that a historian or sociologist of religion must himself have
some religious feeling if he is to make sense of the religious movement
he is studying and understand the considerations that govern the lives of
its participants” (Winch, P. A., The Idea of a Social Science and its Relation to
Philosophy, p. 88);
this “accounts for the weight which the Idealists attached to concepts
like ‘empathy’ and ‘historical imagination’”,79 but also, more generally, to
Anscombe’s point that certain activities, such as arithmetic, cannot be
understood by an observer unless he himself possess the ability to per-
form that activity.80
At any rate, what is clear is that the distinction between ends-ratio-
nality and value-rationality is not enough to account for the special na-
ture of “emotional action” –determined by present emotional states– or
the nature of “traditional action” –determined by deeply engraved cus-
toms.
Emphasis on present emotional states serves to differentiate emotional
action from action guided by certain expectations or beliefs, that is, from
both instrumentally rational action ends and action according to values.81
Let us notice, however, that emotional action could, at times, resemble
rational action according to values. After all, the fact that an action is
performed out of a present emotional state does not mean that it ex-
cludes certain values. The difference, however, lies in whether this value
represents the determining ground for the action, something which
Weber links to the consciousness of the action.82

79 Winch, P. A., The Idea of a Social Science and its Relation to Philosophy, p. 90.
80 Winch, P. A., The Idea of a Social Science and its Relation to Philosophy, p. 109.
81 See Weber, M., Economy and Society, vol. I, p. 25.
82 See Weber, M., Weber, M., Economy and Society, vol. I, p. 25.
On the other hand, acting according to ends involves a calculating at-

titude which leads to weighing means and ends, ends and consequences,
in ways which preclude both emotional action –determined by present
affect– and traditional action –determined by habit and custom.83 While
we cannot exclude that –as in Hume’s case– custom was originally
grounded in some sort of calculation, the point is that, once the custom
has been established, a different pattern of intentional action is also in
place.84 Weber emphasized the connection between tradition and habit.85
Habitual action is particularly relevant for social analysis, because, as
Berger and Luckmann rightly point out, it is at the basis of the genesis
of social institutions.86
4. The possibilities of interpretive sociology
Of course, while the definition of meaningful action-types according to

their motivational sources represents a suggestive approach to social
analysis, the possibility of identifying those actions-types in practice, as
the social scientist is required to do, is another matter.87
Instrumentally rational, value rational, affective and traditional actions
are all cases of intentional behaviour, actions whose proper meaning, or
83 “Action is instrumentally rational (zweckrational) when the end, the means, and the
secondary results are all rationally taken into account and weighed. This involves rational
consideration of alternative means to the end, of the relations of the end to the sec-
ondary consequences, and finally of the relative importance of different possible ends.
Determination of action either in affectual or in traditional terms is thus incompatible
with this type. Choice between alternative and conflicting ends and results may well be
determined in a value- rational manner” (Weber, M., Economy and Society, vol. I, p. 26)
84 See Berger, P. T., – Luckmann, T., The Social Construction of Reality, New York,
Penguin Books, 1991 [first edition 1966], pp. 70-71.

85 Parsons would have reversed this aspect of Weber, emphasizing, against Weber, the
connection of traditional action and norms. See Cohen, J. – Hazelrig, L. E. – Pope, W.,
“De-Parsonizing Weber: A Critique of Parsons’ Interpretation of Weber’s sociology”, in
Boudon, R. – Cherkaoui, M. – Alexander, J. C. (eds.), The Classical Tradition in Soci-
ology. The European Tradition, vol. II, London, Sage Publications, 1997, pp. 198-215.
86 Berger, P. T., – Luckmann, T., The Social Construction of Reality, pp. 71-72.
87 See Munch, P. A., “‘Sense’ and ‘Intention’ in Max Weber’s Theory of Social Ac-
tion”, pp. 17-18.

lack of it, can only be understood when we ask the agent or analyze the
products of his or her agency88 in light of shared meaning. This certainly
explains the relevance of phenomenology, language and hermeneutics
for sociological analysis. It was perhaps the rigidity of his typology of ac-
tion that prevented Weber from developing this view until to its final
consequences. The fact that Weber himself grants a more privileged
place for the understanding of instrumental action insofar as this lends
itself more easily to observation and generalization according to a law-
like order is a hint of the kind of difficulties we can run into when we
move from the micro to the macro level. For, as Alexander rightly point-
ed out, such a move entailed that voluntarism was eliminated from the
explanation of economic behaviour.89
In spite of this, Weber’s basic points, that action is meaningful behav-
iour, and social action a subcategory of action, remain a source of inspi-
ration for subsequent social theory. If the latter claim was taken for
granted in the philosophical tradition, this was not because of any indi-
vidualistic bias but rather because of the recognition of the individual
subject as the primary social agent. The fact that individual agency can
develop at its fullness only in a social context, does not represent an ob-
jection to its ontological primacy. It only paves the way for the next chal-
lenge that, according to Campbell, action theory should shoulder, name-
ly: providing us with a full account of human agency, which goes beyond
explaining decisions and motivation,90 and dares to explain the problems
surrounding accomplishment of action:
“This is the ‘problem of action’ as ordinary people experience it and

it is the problem of how to ‘carry on’ or ‘keep going’ in the face of the
pressure which threatens to erode all programmes of action. Here the
primary problem is not how to ‘make sense of ’ things or ‘make them in-
88 “A study of social action has to grasp such meanings of action as intended by the
actors, and it has to relate that subjective meaning to the various historical objectivations
in a social situation, say, to a science or some tradition, to a world of ideas” (Grathoff, R.
(ed.), Introduction to The Theory of Social Action. The correspondence of Alfred Schutz and Talcott
Parsons, Bloomington, London, Indiana University Press, 1978, p. xx).
89 See Alexander, J. C., Theoretical Logic in Sociology, p. 31.

telligible’, or even ‘render accountable’, but how to maintain agency”

(Campbell, C., The Myth of Social Action, p. 159).
University of Navarra
Pamplona, Spain
References
Alexander, J. C., “Action and its Environments”, in Alexander, J. C. –

Giesen, B. – Munch, R. – Smelser, N. J. (eds.), The Micro-Macro Link,
Berkeley – Los Angeles, University of California Press, 1987.
——, Theoretical Logic in Sociology, vol. III. The Classical Attempt at Theoretical
Synthesis: Max Weber, London, Routledge & Kegan Paul, 1983.
Anscombe, G. E. M. “Emotion and Will”, in Metaphysics and the Philosophy
of Mind. The Collected Philosophical Papers of G. E. M. Anscombe, II,
Oxford, Basil Blackwell, 1981, pp. 100-107.
——, “Intention”, in Metaphysics and the Philosophy of Mind. The Collected
Philosophical Papers of G. E. M. Anscombe, II, Oxford, Basil Blackwell,
1981, pp. 75-82.
Aristotle, Nicomachean Ethics, translated by T. Irwin, Indianapolis, Hackett
Publishing Company, 1987.
——, Politics, translated by E. Barker, Oxford, Clarendon Press, 1968.
Becker, G., “The economic approach to human behaviour”, in Ramón
Febrero, R. – Schwartz, P. S. (eds.), The Essence of Becker, Stanford,
Hoover Institution Press, Stanford University, 1995.
Berger, P. T., – Luckmann, T., The Social Construction of Reality, New York,
Penguin Books, 1991 [first edition 1966].
Boudon, R., “Individualistic tradition in sociology”, in Alexander, J. C. –
Giesen, B. – Munch, R. – Smelser, N. J. (eds.), The Micro-Macro Link,
Berkeley – Los Angeles, University of California Press, 1987.
Campbell, C., The Myth of Social Action, Cambridge, Cambridge University
Press, 1996.
Cohen, J. – Hazelrig, L. E. – Pope, W., “De-Parsonizing Weber: A
Critique of Parsons’ Interpretation of Weber’s sociology”, in
Boudon, R. – Cherkaoui, M. – Alexander, J. C. (eds.), The Classical
Tradition in Sociology. The European Tradition, vol. II, London, Sage
Publications, 1997, pp. 198-215.
Flam, H., Soziologie der Emotionen. Eine Einführung, Konstanz, UVK

Verlagsgesellschaft, 2002.
Frankfurt, H., “The problem of action”, in Frankfurt, H., The Importance
of what we care about, Cambridge, Cambridge University Press, 1988,
pp. 69-79.
Gerhards, J., “Affektuelles Handeln – Der Stellenwert von Emotionen in
der Soziologie Max Webers”, in Weiss, J. (Hrsg.), Max Weber heute.
Erträge und Probleme der Forschung, Frankfurt a. M., Suhrkamp, 1989, pp.
335-357.
Grathoff, R. (ed.), Introduction to The Theory of Social Action. The
correspondence of Alfred Schutz and Talcott Parsons, Bloomington, London,
Indiana University Press, 1978.
Hume, D., A Treatise of Human Nature, edited by L. A. Selby-Bigge,
second edition, with text revised by P. H. Nidditch, Oxford, Oxford
James, W., “The Feeling of Effort”, in The Works of William James, vol. XI
(Essays in Psychology), edited by F. H. Burkhardt, F. Bowers, I. K.
Skrupskelis, Cambridge, MA, Harvard University Press, 1983, pp. 83-
124.
Kant, I., Anthropology from a Pragmatic Point of View, translated by V. L.
Dowdell, Carbondale & Edwardsville, Southern Illinois University
Press, 1978.
——, Critique of Practical Reason, edited by M. Gregor, Cambridge,
Cambridge University Press, 1997.
——, Perpetual Peace and Other Essays, translated by T. Humphrey,
Indianapolis, Hackett Publishing Company, 1983.
——, Religion within the Boundaries of Mere Reason, edited by A. Wood and
G. di Giovanni, Cambridge, Cambridge University Press, 1998.
——, The Metaphysics of Morals, edited by M. Gregor, Cambridge,
——, Groundwork of the Metaphysics of Morals, translated by M. Gregor
and J. Timmermann, Cambridge, Cambridge University Press, 2012.
Korsgaard, C. M., “Acting for a Reason”, in Korsgaard, C. M., The

Constitution of Agency, Oxford, Oxford University Press, 2008.
——, Self-Constitution. Agency, Identity, and Integrity, Oxford, Oxford
Luckmann, T., Theorie des sozialen Handelns, Berlin – New York, Walter de
Gruyter, 1992.
Martin, J. L., The Explanation of Social Action, Oxford, Oxford University
Press, 2011.
Munch, P. A., “’Sense’ and ‘Intention’ in Max Weber’s Theory of Social
Action”, in Hamilton, P. (ed.), Max Weber: Critical Assessments 1, vol.
II, London – New York, Routledge, 1991.
Parsons, T., The Structure of Social Action, vol. I, The Free Press, New
York, 1968, 1st edition 1937.
Parsons, T. – Shils, E. (eds.), Toward a General Theory of Action. Theoretical
Foundations for the Social Sciences, New Brunswick – London, Trans-
action Publishers, 2001 [first edition 1951].
Schutz, A., “Parson’s Theory of Social Action: A Critical Review by
Alfred Schutz”, in Grathoff, R. (ed.), Introduction to The Theory of Social
Action. The correspondence of Alfred Schutz and Talcott Parsons, Blooming-
ton – London, Indiana University Press, 1978.
Taylor, C., The Explanation of Behaviour, London – New York, Routledge,
1967.
Tuomela, R., A Theory of Social Action, Dordrecht – Boston – Lancaster
Reidel Publishing Company, 1984.
Weber, M., Economy and Society. An outline of interpretive sociology, edited by
G. Roth and C. Wittich, Berkeley – Los Angeles, University of Cali-
fornia Press, 1978.
Winch, P. A., The Idea of a Social Science and its Relation to Philosophy, Rout-
ledge, London, second edition, 1990.
SIMMEL AND RICKERT ON AESTHETICS
AND HISTORICAL EXPLANATION
JOHN LEVI MARTIN
Can there be a science of social action that brackets moral evaluation?

The answer of mainstream sociology has been “yes, I guess”, and it
seems that most of us in the United States, but also in Germany, have re-
lied on the work of Max Weber as the key theoretical contribution dem-
onstrating this possibility. This approach, in which social science is rec-
ognized as being “value related” but still “free from value judgments”, is
considered by its adherents as more sophisticated than the positivism of
French sociology (in which ethics –what we should do– is ultimately a
factual matter akin to the health of an organism). The Weberian solution
was believed to recognize the special binding nature of ethical obliga-
tions –a transcendent “ought” that can orient our lives meaningfully–
without detracting from a sober, indeed, disenchanted recognition that
the concepts we use to grasp the world are constructs produced for their
analytic utility, and cannot be treated as simple reflections of reality. Yet
the results produced by such knowledge must be taken into account by
an ethical actor, at least, any actor who claims that sort of maturity that
is available to those living in the modern world.
The problem with this approach, in a nutshell, is that, if pushed, it
turns out to be fundamentally incoherent. If we produce concepts on
the basis of our analytic interests (what we are trying to do), and thus
adopt a wholly voluntarist theory of concept formation, it is altogether
obscure as to how the results have some sort of binding claim over the
thinkers.
Weber’s methodological and epistemological forays were generally
lengthy, often erudite, critiques of others that tended to begin from the
presumption that Weber had no difficulty determining that other ap-
proaches were fundamentally unsound, but he seemed to lose interest in
giving any clear specification of his own ideas. Indeed, he was not partic-
ularly interested in matters of philosophy in a professional sense, and
114 John Levi Martin
tended to refer readers to others, such as his colleagues and friends

Georg Simmel and Heinrich Rickert, for details.1 In particular, I believe
that Weber considered Rickert to have established the consistency of the
neo-Kantian approach to concept formation that Weber himself saw as
the best way of working. Here I want to give detailed consideration to
this conception, and to Rickert’s attempt to stabilize a neo-Kantian theo-
ry of the sciences, and the cultural sciences in particular.
This stabilization attempt arose because his conception was under
fierce attack. Now, there is a way in which the armies facing one another
were two branches of neo-Kantians: on the one hand, there were those
who emphasized Kant’s theory of categories, and tried to build upon
this formal approach to the preconditions of knowledge. The greatest
opponents of this attempt were really those who, like Schopenhauer and
his disciple Nietzsche, emphasized not Kant’s transcendental analytic but
his transcendental aesthetic: a recognition of the partial, limited, and an-
thropomorphic nature of our most fundamental experience of the
world, and the implication that something “more” lay on the other side
of it. This “more” often connected to emerging “life” theories, and was
used as an attack on all would-be “complete” formal systems, such as
those being constructed by neo-Kantians of the “southwest” school, of
which Rickert became the leader.2
1. Kant’s project
We begin with a brief reprise of the relation of these neo-Kantians to

the master himself. After his monumental Critique of Pure Reason, ex-
ploring the way the human mind can or could know, Kant wrote a sec-
ond Critique of Practical Reason, exploring the nature of our freedom and
our morality. Although these two Critiques corresponded to the two
1 As in the beginning of Weber, M., Economy and Society, edited by G. Roth and C.
Wittich (two volumes), Berkeley, California, The University of California Press, 1978, p.
3.
2 There certainly were figures standing in between these extremes; most important
was Dilthey, as well as the Marburg school most associated with Cohen (who influenced
Simmel).
Simmel and Rickert on Aesthetics and Historical Explanation 115
halves of Kant’s universe, Kant believed that his system would be in-
complete without a “third Critique”, the Critique of Judgment. Here Kant
handled (among other things) the question of how we can judge beauty,
for beauty is a curious sort of predication, as we believe that our state-
ments of aesthetic appreciation are universally valid, although we under-
stand that we cannot prove this to others with concepts. Kant admitted
this immunity to logical proof, yet anchored aesthetics in intersubjective
concordance and in a faculty of reflective judgment.
Reflective judgment is our capacity to attach a universal (such as
“beautiful”) to a particular (such as “this rose”) without having a set of
rules, the way we do when we simply subsume an instance into a general
category (“this rose is a flower”). This was, thought Kant, a key part to
the coherence of his system; indeed, such a capacity was necessary if
free actors are to make use of the lawful knowledge produced by the in-
tellect on the basis of sensory experience. For we must see purposive-
ness in the world if we are to believe that this lawfulness is of some rele-
vance to us. And so we must, Kant argued, interpret aspects of the world
as if they were made for us, even though they (perhaps) weren’t. Thus we
must assume a super-sensible realm from which this orderliness springs.
This capacity to sense “purposiveness without purpose”, then, is key to
our ability to relate our will to our cognitive powers. Although no state-
ment made about Kant’s work will not be contested by someone, I am con-
vinced by Kant that his system was unstable without this capacity for re-
flective judgment.
In any case, Kant’s later work consistently built on this triadic system.
But the neo-Kantians who influenced social science by and large ignored
the “third Critique”, with a few exceptions, most importantly Cassirer.
Rather, they tended to adopt an ethics vaguely along the lines of the sec-
ond critique, and tried to form a science based on a voluntaristic version
of the first. In particular, Dilthey3 had clearly appealed to Kant’s first cri-
tique in proposing his own formulation of the cultural sciences as having
a distinctive theoretical methodology and epistemology: his work should
be seen, he wrote, as a “critique of historical reason”. To Dilthey, history
(the German version, that is) was a general science of the mental life of
3 See Dilthey, W., Introduction to the Human Sciences, translated by R. J. Betanzos,
Detroit, Wayne State University Press, 1988 [1883], p. 146.

willful human beings. It could not proceed along the same lines as a nat-
ural science, which used external causal explanations, but would require
some sort of interpretive understanding of one mind by another. This
formulation became increasingly attractive to the “left” (as we might call
them now) who wanted to put history with the humanities, and unat-
tractive to the “right” who wanted the cultural sciences to have an ex-
planatory capacity. A number of reformulations were put forward at the
end of the century, most importantly, Simmel’s work on the philosophy
of history.
2. Simmel
a) Simmel and the problems of history

Georg Simmel is known to us as a sociologist; even in his own life
Simmel found this definition of his oeuvre solidifying (to his frustration,
as he considered himself a philosopher first and foremost). But Simmel’s
charisma and brilliance were matched by a quality in his philosophical
work that made him harder than others to pin down. It is not that he
lacked core concerns –his work ever returns to issues of objective and
subjective culture, individual and group, wholeness and decomposition
of life– nor was he a loose thinker. But at a time of system-building,
Simmel’s own allegiances were difficult to determine (he might generally
be seen as a Kantian, but in his Philosophy of Money he situates himself
squarely in a Marxian tradition; he appealed to Goethe as one who could
be fused with Kant, and there was something ever Hegelian about his
thought). Further, his own work resisted systemization, as he emphasized
that form could become content, that tendencies could reverse them-
selves, and that there was a second side to everything.
Characteristically, Simmel’s contribution to the question of the meth-
od of cultural sciences, his 1892 Problems of the Philosophy of History,4 was
filled with brilliant insights that popped up later in the work of his col-
4 Geschichtsphilosophie is conventionally translated as “philosophy of history”, but it
could almost with as much justice be translated “historical philosophy”, for the questions
were not merely about how we could know the past, but what the philosophical implica-
tions of the past were, and whether history itself was philosophically significant.
leagues (such as Max Weber) but was not considered a decisive solution.
In this case, Simmel agreed with this verdict, and re-wrote the piece in
1905, saying that he himself had not understood his own point in the
first edition. Here I generally rely on the (translated) second edition,
though making a few references to the first, where differences are nota-
ble.
Simmel accepted the then-current idea that history was fundamentally
a science of subjectivity, of mental processes and their effects. To
Simmel, this implied that history is to some extent “concerned with the
individual, with absolutely unique personalities”. This raises a key prob-
lem of intersubjective access, for “what we call individuality is the pecu-
liar fashion in which ideas –the contents of which are given– are united
in one consciousness”. What does it mean for one person to understand
the mind of another, who has some different mental make up? How
does one individual, one totality, reproduce the structure of a very dif-
ferent one? How does the historian reconstruct the interconnection of
subjective elements in another mind? When the connections are logical
ones between elements, it may not be too difficult, but what about when
these connections are subjective as opposed to objective? We must (as
Weber was later to repeat) to some extent rely on our own experiences
and our own particularity– we must create some sort of empathetic re-
construction.5
This problem can be phrased in a number of ways: first, let us for-
mulate it such that we can envision it as an everyday occurrence. When
we are dealing with some particular person, and we wish to forecast her
behavior, we will need to make some sort of attribution of a mentality to
her, and almost certainly this implied mentality will be richer than what
we might be able to directly support with data from her observable re-
sponses or actions. Thus we make
“presuppositions of the following sort: they are trustworthy in the

sense that they are useful for the purposes of both theory and practice.
But they are not certain in the sense that they can be deduced, with logi-
5 See Simmel, G., The Problems of the Philosophy of History, second edition, translated by
G. Oakes, New York, The Free Press, 1977 [1905], pp. 64f, 72f, 75, 87ff.
cal necessity, from what is actually and empirically given” (Simmel, G.,
The Problems of the Philosophy of History, p. 47).
Note that this is somewhat similar to the problem faced by an neo-

Kantian actor –able to derive principles that are necessary for action but
of an uncertain noumenal status.
This sort of empathetic re-creation, then, is an “imaginative syn-
thesis” that could allow us to make useful deductions, but if we cannot
demonstrate the truth of this synthesis, we have not solved our original
puzzle: “how can a state of mind of one person also be eo ipso repre-
sented as the state of mind of another person?”.6
Let us now re-phrase our question in a way that might seem closer to
the issue for the discipline of history. A historian sets out to make clear
the actions of, say, Themistocles, and to do this, both for his own ana-
lytic purposes and to communicate his conclusions to readers, he must
form some sort of “mental construct” that brings coherence and intelli-
gibility to his data. Is this construct an arbitrary one, one that has more
to do with the individuality of the historian than the individuality of
Themistocles? The historian certainly does not think so; rather, he “feels
that there is a sense in which necessity [n.b.] can be ascribed to this psy-
chological construct”. Indeed, he believes that this mental construct is
objectively and trans-subjectively valid. Thus we have a puzzle. These mental
structures “seem to have a universal, law-like status; however, this is not
actually the case at all”, as each construct must be historically unique.7
Thus although we still see a connection to Kant’s distinction between
the sciences of things and the sciences of persons, for Simmel (and for
others at this time in the emerging modern university) the problem to be
solved no longer seemed to turn on the grander philosophical questions
about how one understands the relation between freedom and necessity,
or between spirit and mind, but rather, how we have can formulate un-
derstandings of the world that have the force of objective law, and how
historians can make statements that are as binding as those that would
come from objectivity. The puzzle resolves itself as one familiar to us –
6 Simmel, G., The Problems of the Philosophy of History, p. 75.

7 Simmel, G., The Problems of the Philosophy of History, pp. 69-71, 76; Die Probleme der
Geschichtsphilosophie, Leipzig, Verlag von Duncker & Humblot, 1892, p. 23, n. 1.
how there can be judgments that are valid and allow us to demand agree-
ment even though they cannot be supported by subsumption into con-
cepts? Simmel’s reformulation of the puzzle of the character of the hu-
man sciences clearly points to Kant’s critique of judgment as the closest
solution and the natural starting place for the formulation of a set of an-
swers.
b) History, aesthetics and the intellect

Yet this is not how Simmel proceeded. Like Dilthey, Simmel assumed
that it was Kant’s first critique that was relevant for the study of history.
Indeed, from his first page onwards, Simmel8 insisted that he was dem-
onstrating a parallelism between history and nature in terms of the “con-
stitutive power of the intellect”, but for the case in which a mind knew
another mind, and not nature. Like others, Simmel9 saw the overall claim
of Kant as being that “in addition to elements of sense perception
which give experience its definitive character, it also includes a priori
forms”. And thus Simmel moved to try to think in analogous terms,
even though this was inappropriate to the problem he formulated.10 (He
was later to use a similar approach to thinking about sociological prob-
lems, where the first critique provided a better template, given the gener-
alizing concepts at issue.)
My argument, then, is that Simmel reproduced the problem that Kant
solved in the “third Critique”, but misidentified it as one pertaining to
the use of the intellect (understanding). If this seems tendentious, there
are two pieces of evidence that offer strong support. The first is that
Simmel11 himself found it necessary to introduce the equivalent of a su-
persensible that could justify the trans-subjective validity of these con-
structions. Noting that he faced a puzzle of how to establish an “inter-
8 See Simmel, G., The Problems of the Philosophy of History, pp. viif, 43, 77, 98, 150; Die
Probleme der Geschichtsphilosophie, p. 21.

9 See Simmel, G., The Problems of the Philosophy of History, p. 132f.
10 This is perhaps even clearer in the first edition, where Simmel (Die Probleme der
Geschichtsphilosophie, p. 2f) suggested that it is possible that there is only a single class of a
prioris –those of the intellect (Verstand), which is also the basis for those of sensibility
and reason.
11 See Simmel, G., The Problems of the Philosophy of History, pp. 94f, 97.
mediate zone” between two consciousnesses, Simmel proposed a collec-

tive inherited unconscious, one modified by successive generations. That
this was an outlandish solution did not escape Simmel. “I am most fully
aware”, he admitted, “that this sort of interpretation has been discred-
ited on the most legitimate grounds”. But he did not want to defend it as
a statement of fact so much as to treat it as “a methodological fiction.
Phenomena occur as if this sort of latent correspondence between our
minds and the minds of completely different persons really obtained”.
This is of course exactly how Kant introduced the complicity of our
minds and the world via the supersensible.
The second type of evidence supporting the argument that Simmel’s
formulation of the problem of history was fundamentally in aesthetic
terms is that Simmel continually used analogies to aesthetics to clarify his
claims. Thus Simmel gave the example of artists making different por-
traits of the same person.
“These differences between the individual artist’s conception of the

aesthetic problem –even though different solutions to this problem may
be equally valuable– establishes the following point: in the solution of
the objective aesthetic problem, it is the personal or subjective factor that
is decisive. The same proof also holds for history” (Simmel, G., The Prob-
lems of the Philosophy of History, p. 86).
Thus it would make a great deal of sense to try to approach the

puzzle of historical explanation from where Kant left off his critique of
judgment. But because the first critique remained the dominant model
for Simmel, he instead attempted to understand the process whereby the
intellect created these historical individuals. This process, Simmel12
argued, turned in large part on the point of view or “problematic” of
the investigator. There were two problems here, though Simmel did not
necessarily put things this way. One was the creation of unified entities
(historical individuals), and the other was the choice of what to examine.
Regarding the latter, Simmel13 noted that “What motivates us to engage
12 See Simmel, G., The Problems of the Philosophy of History, pp. 80, 82.
13 Simmel, G., The Problems of the Philosophy of History, pp. 153-156.
in theoretical activity cannot be something else that also has a theoretical

status. On the contrary, it can only be an impulse of will and a sense or
feeling of significance or value”. Now “significance” can be ambiguous.
In some cases, we may attribute moral or aesthetic significance to an event,
though there is also a third type of significance, an irreducibly historical
one. These three standpoints can be used to categorize any set of events.
Simmel went on to disambiguate “values”, for this term can either be
used to refer to all extra-theoretical considerations that we use in history,
or in a more delimited sense of evaluation.
Simmel did not, in my understanding, dispose of the problem that he
raised with the idea of the historical “significance” of some event –this
sort of extra-theoretical value that seemed outside his main system. Nor
did he tie this issue (of what is significant) to the issue of the formation
of historical individuals (which could have been handled from the per-
spective of an aesthetics). Thus at the same time as he proposed to dis-
tinguish between natural science and history as two different under-
takings, largely on the basis of the conclusion that the more historical
something is, the more individual it is, Simmel14 had an unclear under-
standing of the process whereby aprioris might enter from the intellect
to form the nature of historical individuals. It was this problem that
Heinrich Rickert was (almost) to solve.
3. Rickert
Heinrich Rickert (a friend of Weber’s) was a student of Windelband who

had done his dissertation on the theory of definition.15 The work of his
that became crucial for sociology (On the Limits of Concept Formation in the
Natural Sciences) was in part a response to Simmel’s theory of history. He
published the first edition in 1902. In 1905 Simmel’s second edition
came out, and in 1913 Rickert’s second. In 1921 the third and fourth edi-
tions came out (Weber had died the year before). Here I concentrate on
the first edition, for this is what Weber read (perhaps he glanced at the
14 Simmel, G., The Problems of the Philosophy of History, p. 174f.

15 See Oakes, G., Weber and Rickert: Concept Formation in the Social Sciences, p. 6.
second which is largely identical), although I make references to certain

changes in the later editions. Wherever possible, I use Oakes’s16 partial
translation of the fourth edition, which is mostly the same as the first:
where the two diverge or where no translation exists, I use the first edi-
tion.
a) Concept formation in natural science

As Rickert17 noted, the great problem of his time was the seemingly self-
evident opposition of the psychic and the physical. There were two as-
pects to this –the first, the general problem of how the mind knows the
world, and then within that, the bifurcation of two types of worlds, basi-
cally corresponding to the world of physical phenomena and the world
of conscious human actors, which Rickert assimilated to the problem of
physical as opposed to historical sciences.
Rickert18 began by formulating perhaps the strongest (and soon to be
most influential) version of the neo-Kantian idea of selective abstraction
(as it was to be called later). The essential problem is that as we face the
world, we find an immeasurable multiplicity of phenomena that con-
found any epistemic theory of “knowing” as a simple intromission of
what is “out there”. Not only is there an extensive multiplicity (in that
there are an infinite number of things), but each thing itself has intensive
multiplicity in having an infinite number of properties or aspects that
can be of interest.
For this reason, we must use concepts: mental structures that seize
upon some aspects of objects and ignore others. Generalizing concepts
(the kind most often envisioned by ideas of selective abstraction) unite
concrete instances through the common possession of certain aspects
selected. Such generalizing concepts are the fundamental units of natural
science. Because these concepts are formed by suppressing particular-
ities, as we ascend in generality, we necessarily leave the concrete reality
16 Rickert, H., The Limits of Concept Formation in Natural Science, Fifth Edition, trans-
lated by G. Oakes, Cambridge, Cambridge University Press, 1986 [1929].
17 See Rickert, H., The Limits of Concept Formation in Natural Science, p. 13.
18 See Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, Tübingen, Verlag
von J. C. B. Mohr, 1902, p. 36.
of individual phenomena behind. Even more, our concepts develop that

sort of intrinsically hypothetical nature that we associate with nominal-
ism.19 Still, the concepts so produced are not arbitrary, but rather have a
general validity (to which we shall return in greater detail when consider-
ing the objectivity of historical concepts), allowing the human mind to
successfully make use of a limited set of limited concepts to deal with
the unlimited nature of reality20.
b) The problem of history and of individuals

The problem enters when science attempts to grasp individual con-
figurations as such, as opposed to considering them only as members of
classes. This is, argued Rickert, key for the discipline of history, for we
call “historical” that which for logical reasons can never be subsumed
under a natural scientific concept.21 Just as natural science’s job is to find
the general in the real, historical science has the task of bringing out the
individual in the real.22
This sort of individuality –most simply, things that only happen
once– has long been understood as posing a problem to the scientific
status of history (see, for example, objections discussed by Spencer),23
and thus a challenge to the project of rationalism as a whole.24 Now, in
nineteenth century thought, one common way of dealing with the prob-
lem posed by the unique nature of the material of historical science was
to claim that history was a science of an entire realm of phenomena as-
sociated with psychological processes, and which perhaps could be stud-
ied via direct introspection or intuition.
Rickert admitted this as an empirical coincidence –“Most historical
sciences are predominantly concerned with mental processes”.25 If, for
20 See Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, p. 73.
22 See Rickert, H., The Limits of Concept Formation in Natural Science, p. 54, cf. 107.
23 See Spencer, H., The Study of Sociology, New York, Appleton, 1896 [1873].
24 See Rickert, H., The Limits of Concept Formation in Natural Science, p. 214f.
this reason, one wishes to call history a “human science” (Geisteswissen-

schaft), then that is all well and good as far as it goes, but it was, em-
phasized Rickert, important not to imagine that this substantive distinction
had a direct methodological implication.26 First, Rickert27 denied that (as
Dilthey and Wundt imagined), we have some sort of immediate access to
mental life. Indeed, if we are to think rigorously, we realize that we can’t
claim that there is a psychology that is simply the study of psychic pro-
cesses, because all the sciences begin with and manipulate data that are
fundamentally psychic processes, namely sense impressions.28 There is
nothing particularly unmediated in those psychic processes that we use as
avenues for the investigation of mind in contrast to those used in the in-
vestigation of physical reality.
Thus there is no intuitive fast track to knowledge of mental life, for
the world of the psychological, like the world of objects, has a multiplic-
ity and hence also poses problem of intensive and extensive infinity, and
so must be “worked up”.29 This is one reason to reject the division be-
tween natural sciences and history as being fundamentally about the sub-
stance of investigation. But second, we know that we can have a general-
izing/natural-scientific approach to mental phenomena, for that is just
what we see in the new science of psychology.30
Thus the central issue is a formal one of grasping the individual, not a
substantive one of knowing the psychological. A natural-scientific ap-
proach to history that fails to recognize the special challenge of studying
the individual would be in effect to re-derive sociology, which as we all
can see, said Rickert,31 has not really accomplished much. This failure
comes because a natural scientific approach implies the development of
laws, and the subject of social life is one that is so historical by its very
nature that we can expect that such laws will be few and far between.
29 See Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, pp. 185, 188.
30 See Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, p. 211f.
31 See Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, pp. 287f, 294,
590; The Limits of Concept Formation in Natural Science, p. 133.

The core difference between history and the natural sciences, then, is
a methodological one, one pertaining to the difference between general-
izing and individualizing concept formation. However, it is not accidental
that the individualizing sciences tend to be more associated with the
realms of human action, with life, and with thought. As we array the sci-
ences from physics to chemistry to biology to anatomy to history we find
more and more historical elements, and the ultimately historical science
is not that of the “mental” but that of culture –culture in opposition to
nature.32
This fundamental relation between history and culture, thought
Rickert, explains the frequent understanding that there is some of the
aesthetic in the craft of the historian. A painter, overwhelmed by the
multiplicity of the object, tries to simplify it in such a way so as to best
communicate his essential intuition as to the nature of this object. In
contrast, the natural scientist cares nothing for this aspect of the object
and is happy to be rid of it in the process of concept formation.33 His-
tory sits in between: for the historian, the aesthetic elements are only a
means to an end, communicating “how it actually was”, and not an end
in themselves, as in the arts.34
We must see that Rickert, like Simmel, has basically approached the
central problem of aesthetic communication in Kant, namely, how to
establish intersubjectively valid presentations or statements without con-
cepts. But like Simmel, Rickert turned away from this path; indeed,
Rickert argued that Simmel was incorrect to emphasize the issue of in-
tersubjective concordance in the first place. Although Rickert35 was quite
complimentary regarding Simmel’s insights as to the differences between
narrative and lawful sciences, he believed that Simmel had made a funda-
mental error in attempting to begin from psychological presupposi-
tions.36

36 Simmel seems to have agreed, as he replaced this approach in the first chapter of
the first edition with a more Rickertian emphasis on immanent limits in the second edi-
tion.
Thus to Rickert, the essential problem was of the individual historian

forming concepts in a non-arbitrary way, and not the establishment of
intersubjective concord (which might or might not use concepts). Thus
when turning to the key issue raised by Simmel of not knowing interio-
rity of others, Rickert37 emphasized that this issue of the inaccessibility
of foreign mental life [fremden Seelenlebens, translated by Oakes as “third
person mental life”] is no different from the more general problem of
just not having enough data.
Yet as we shall see, Rickert was forced to recapitulate Simmel’s
thoughts along these lines; although able to push the problem outside
the empirical, Rickert acknowledged the need to account for agreement
as to nonreal meaning configurations in a way that made much of his
post- and contra-Simmelian work superfluous. But to make this point,
we will need to follow Rickert as he defined a way for historical science
to grasp individuals.
c) Values and the formation of individuals

Because they too partake in infinite multiplicity, individuals must be
grasped conceptually, yet we cannot do so using the process of selective
abstraction, for this requires suppressing particularity. But the signify-
cance of an individual (say, an individual human being, such as Goethe)
is his particularity. “There is no general concept under which he can be
subsumed”.38
How, then, can we form individuals in a way that they can be mani-
pulated by the intellect? Indeed, what is a true individual (as opposed to
an instantiation of a class, which can be subordinated without loss in the
more general concept)? Consider a nondescript dog that you happen to
see on the street; to you, this may be simply “a” dog, a member of the
larger class, and not an individual at all, but to me, this is (let us say)
“Scout”, a unique and moderately irreplaceable animal. This difference,
argued Rickert39 comes because the individual name indicates the value of

38 Rickert, H., The Limits of Concept Formation in Natural Science, p. 89.
the pet to the owner. And this turns out to be more generally the case –
individuality is correlative to the value that the candidate individual has
for us. Thus the value solves the same problem for an individualizing sci-
ence that the suppression of concrete particularities does for a general-
izing one.40
For it is this focus on values and individuality that gives us, thought
Rickert, the proper formulation of the division between history and nat-
ural science. It has to do not with the substantive issue of there being a
class of “cultural” phenomena that elude the grasp of the natural sci-
ences, but that these sciences, in their quest towards a pure model of
determinations connecting wholly qualityless elements (the ultimate units
of physics, perhaps quarks), necessarily attempt to shake off all intu-
itively accessible aspects of actuality. But many of the questions that are
of interest to us –including some about inorganic processes– cannot be
answered in such a way, and we find a need for a complementary
approach that does not attempt to sacrifice empirical individuality.41
“The problem of concept formation in history, therefore, is whether a scientific

analysis and reduction of perceptual reality is possible that does not at the same time
–as in the concepts of natural science– forfeit individuality, and yet also does not
produce a mere ‘description’ of facts that cannot yet be regarded as a sci-
entific representation” (Rickert, H., The Limits of Concept Formation in Nat-
ural Science, p. 78).
We have seen with the case of the pet that individuality seems to re-
quire a value relation connecting observer and observed, precisely what
most analysts would assume must completely undermine the possibility
of a science. Rickert’s argument was that, first, such a value relation was
a necessary part of the formation of the individualized concepts required
by history and, second, that the objectivity of history came through this
value relation, not in spite of it.
40 See Rickert, H., The Limits of Concept Formation in Natural Science, pp. 83, 101.
41 See Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, pp. 264, 266.
The problem, then, is how there can be a science that develops such
individualizing concepts given that (as any historian knows) it is no more
the case that all observers from all perspectives would formulate their in-
dividual concepts in the same way than that we would really value each
other’s pets.42 Does this mean that individualizing concepts lack objectiv-
ity?
d) Objectivity and individual concepts

As a good neo-Kantian, Rickert43 rejected any idea that the objectivity of
science comes via the agreement of the contents of its concepts with ul-
timate reality, because natural science is not about reproduction of the
actual world [wirklichkeit], but production of a conceptual world (with an
unbridgeable gap between us and actuality). Instead, we must demon-
strate the possibility for a non-arbitrary character to our investigation (most
importantly, our principles of selection44) as we “work up” reality via
concepts.
Now there is, admitted Rickert,45 a way in which dependence upon
the selection of values pushes history away from the objectivity pos-
sessed by natural science. However, since the importance of certain val-
ues to members of a particular community is an empirical question, one
that can be answered via empirical methods, there is the possibility of
having a value consensus that links not only the investigator with others
of his breed, but with his subjects of study.46 Thus whether we are con-
sidering Alexander Rodchenko alone, or the entire Russian Productivist
School, we form our concept as an individual that has a relation to the
42 See Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, p. 504; also The
Limits of Concept Formation in Natural Science, p. 100.
43 Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, p. 656.
46 The historian “surmounts all caprice when, for example, he related the develop-
ment of art to aesthetic cultural values and the development of a state to political cultur-
al values. In this way, he produces a representation that –insofar as it avoids unhistorical
value judgments– is valid for everyone who acknowledges aesthetic or political values as
normatively general for the members of his community” (Rickert, H., The Limits of Con-
cept Formation in Natural Science, p. 199).
values that bring us to our historical investigations, most obviously (in

this case) aesthetic values. Since the Productivists themselves were
oriented to such aesthetic values, the sharing of values between the
historian and the object of study, often understood as threatening objec-
tivity, actually ensures it.47
But what if there is no overlap in values between the historian and the
“mental beings that belong to the most comprehensive historical nexus”?
In such a case, Rickert48 argued, “to understand these beings, [the histo-
rian] must at least be able to ‘get the feel’ of their values”. If we cannot
empathize with the values of actors, we will see them as “things”, in
Kantian terms, as objects fit for physical science, not anthropology. So
rather than the historian being able to arbitrarily construct individuals
based upon his own values, he must attempt to relate to the values of his
subjects to the fullest extent possible, even if this is only a theoretical re-
lation.
In sum, Rickert’s approach to formalizing the process of concept for-
mation required a bold maneuver –to recognize that precisely because his-
tory was, above all else, a science of concrete reality (Wirklichkeit), it
could not be separated from valuation. Although this in no way meant
that historical science should incorporate value judgments,49 it meant that
the division between “is” and “ought” was not the same for history as
for natural science. When it comes to human life oughtness is, and it is nec-
essarily, where individuality exists as such.
e) Generality and validity

Yet we have, in a way, only pushed off the problem of objectivity: we
have defined individuals by values, and focused on the values that are
non-arbitrarily shared (between the investigator, his interlocutors, and his
subjects). But the thing about values is that they aren’t merely things that
“are” but things that “should be”. Values that do not, inherently, possess
validity as values are no values at all. In other words, they are an “ought”
that ought to be an ought. Thus we have not solved the problem of ob-
jectivity until we have established not only the existence of the value re-
lation, nor even its commonality, but also its validity. Rickert thus ac-
cepted that “the value with reference to which objects become historical
individuals must be a general value: in other words, a value that is valid for
everyone”.50
Hence Rickert51 concluded that for us to derive “the objectivity of
history in the highest sense” we have to (first) “suppose that at least some
values or other are absolutely valid” and (second) suppose that some
“substantively embodied and normative general human values objectively
approximate them more or less closely”. Then our search for axiological
development has a valid grounding. Although we do not have to assume
that the values used in concept formation are themselves unconditionally
valid, we need to assume that they are related to such valid values.52
Every historian worth his or her salt, however, certainly knows that
values change; so what can it mean to require general validity? It is not
that the values must be timelessly generally valid, but that they be valid for
a community. If this is so, there is a unification of the “normatively gen-
eral” validity of a value (everyone should value this) and its “factually gen-
eral” validity (people do value this), “for under these conditions factually
general values must also appear as requirements for all members of the
community”.53 It is important that Rickert did not mean here a commu-
nity of investigators, but an actual human community of interdependent
lives. The values of this community are then what we call cultural values.54
Now by “cultural values” Rickert did not mean any values that hap-
pen to be shared among the members of some group (such shared val-
ues may be trivial or accidental). Rather, cultural values are those values
“that the members of a community take seriously”. Further, these values
are cultural in a sense that returns us to the root of the word– these are
values whose cultivation we may reasonably expect.55 Thus they get at
what is essential to a human community, and what is seen as “oughting”
50 See Rickert, H., The Limits of Concept Formation in Natural Science, p. 89, cf. 130, 105.
53 Rickert, H., The Limits of Concept Formation in Natural Science, pp. 130f, 134.

to be essential to them. This implied that cultural values are those that
permit, indeed demand, a sort of development. “In other words, real
culture exists only where value-related or historical-teleological develop-
ment either exists or has existed. Thus we see an even more intimate con-
nection between culture and history”.56
f) Civilization and progress

Taken seriously, the logic suggests there can be no history of a place and
time that is oriented to invalid values, and such persons can do no historio-
graphy. Rickert seems to have fallen into Hegel’s form of history in
which most of the world’s continents simply have no history at all.
Rickert recognized and was disturbed by the ethnocentric implications,
and tried to grapple with them.
Rickert began by recapitulating the distinction between what Ger-
mans called Naturvölkern and Kulturvölkern. Although this corresponds to
the English distinction between “primitive” and “civilized”, the emphasis
in English pertains to position on a developmental continuum –the
“early” peoples who are not yet in “cities” (on the one hand) with “civil”
states, laws and forms of behavior (on the other). In contrast, the dis-
tinction in the German terms returns to the issue highlighted by Rickert
as the distinction between nature and culture, and has more of a sense
of those who are stuck in a world with no linear time as opposed to
those who are part of a developmental story. The very implication is that
only Kulturvölkern have a history. The others, if they are to be studied at
all, are properly the subject of generalizing sciences, no different from
animals, trees, rocks and waters.
Rickert’s57 response was to make two careful (and related) qualifica-
tions. The first was to point out that just as the true Calvinist church is
an invisible one, so too is the church of culture. We may accept the dis-
tinction between primitive and civilized as a formal one without being
confident that we can ever tell where any particular people belongs.58

58 “We’ve certainly already seen that the division between primitives and historical or
civilized can be precisely determined by our considerations, because if in the course of

time a people shows no fundamental alternations with regard to normative general cul-
The second was to emphasize that when we believe that a society lacks
cultural values, all we are really able to determine is that they are not con-
necting to our values. We cannot say that any culture really is, in itself, a
natural society. The only societies that are without history are those of
ants and bees.59
We can appreciate Rickert’s efforts to shield us from the harsher im-
plications of his work, yet we cannot be wholly comforted. Rickert’s de-
termined effort to find the grounds of objective historical concept for-
mation has necessarily led him to emphasize cultural values of progres-
sive civilizations –not in the triumphalist version of Hegel’s history, but
because uncultivatable values cannot be good, and we cannot demand
that others respect invalid values, and if we cannot find universally valid
values we cannot have an objective history. History then requires the
possibility of infinitely progressive ascension in terms of civilizational
values.
Rickert’s key second edition was published in 1913. A year later, the
Great War began, and in short order, no commentator on intellectual life
in Europe would be sanguine at the prospect of grounding a philosophy
in continuous value progress. There could not have been a less prom-
ising choice of a basis on which to construct a historical science.
g) Non-real meaning configurations

But it gets worse. Thinking through his logic, Rickert60 concluded that
both logic and ethics depends on the following: “That the world is so
arranged [eingerichtet] that in it the goal of cognition actually can be
reached”. Here he could have followed Kant and simply declared this a
transcendental condition and been done with it. Indeed, relying on the
hypothesis of a supersensible realm, as did Kant in his “third Critique”,
tural values, we bring it only under natural-scientific general concepts and say that it has
no historical development. But this division is formal; that is, if cultural and historical de-
velopment is existent in some people, we could always make our conclusions only with a
perspective on the normatively general values that are known to us, and so long as we
thus are not secure that we are aware of all these values, we must be careful not to deny
any people the title of ‘Cultural-Folk’”.
60 Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, p. 737.

would have sewn up the argument nicely in logical terms, even if it were
unsatisfactory as a scientific statement. And it would have highlighted
the connection of the cultural sciences to aesthetics. Instead, Rickert had
tried to specify the non-arbitrariness of these individualizing concepts
without such a hypothesis, and instead, to base his argument on the
empirical (even if not completely knowable) facts of cultural civiliza-
tions.
But Rickert himself later concluded that he had not solved the prob-
lem of agreement in the constitution of individual concepts after all.
Hence he returned to the issue from a new perspective in a section on
value consensus added in later editions. Rickert61 recognized that values
are part of a domain of configurations that aren’t “real” –like the mean-
ing of a word, they are neither corporeal nor psychic. Thus when we
think about “generality”, we must acknowledge the possible generality
that arises when all the subjects have a similar experience of something
that isn’t real, like a value. Rickert maintained that meaning “can be di-
rectly grasped by us in common with other persons as the same”, and this
happens when we grasp this meaning “in its individuality, in the same way
that we perceive the same body as an individual”62 (but unlike a historical
individual, this meaning is not empirical). Rickert suggested that we use
the word “understanding” to refer to these meaning configurations
which we try to grasp as wholes.63
This grasping of the individual meaning by different persons had
strong implications for the issue of how a historian can transcend a mere
conceptual representation of another’s mentality and attain a real grasp
and a re-creation of the other –and hence if, and if so, how, we can un-
derstand an alien mentality.64
63 See Rickert, H., The Limits of Concept Formation in Natural Science, pp. 158f, 164.
Rickert (Die Grenzen der naturwissenschaftlichen Begriffsbildung, Fünfte Auflage, Tübingen,

Verlag von J. C. B. Mohr, 1929, p. 546f) suggested aesthetics as an analogue whereby we
can understand a compound having a unitary meaning not reducible to a sum of mean-
ing of its parts; a second example was a facial expression –in all cases, we cannot go from
the parts to the whole, but can only interpret parts from the perspective of the whole.
Rickert’s example was not a member of a different culture, but contrarily, that we hear a
Rickert’s answer as to the possibility of “the re-creation of the mean-

ingful mental life of another person” was to acknowledge that (contra
Scheler) we have no direct understanding, but to argue that
“because we have understood the meaning that really lies in this men-
tal life as nonreal meaning –on the basis of our knowledge of mental life
in general, which is grounded in our own mental life– we can construct
the mental life of another person in such a way that we acquire mediated
knowledge of it as an interpenetration of nonreal meaning and real
mental life that can be re-created” (Rickert, H., The Limits of Concept For-
mation in Natural Science, p. 170).
That is, we actually have an experience of this meaning (even though

this is not an experience of something in the real world), and we use this
as the basis of our re-creation of the other’s mental life. “Nonreal
meaning, therefore, can form the bridge between our own real mental
life and the real mental life of another person. This is because, as nonreal
meaning, it can be ascribed neither to us nor to another person”.65
Rickert66 understood that this capacity to grasp a meaning that does
not actually dwell in our own mental life, but “swims freely” about, was
fundamental for his project, and indeed for science in general, for if we
lacked a faculty for such understanding, we could only learn what we al-
ready knew. Thus he created a third realm, to mediate between the two
that seemed again threatening to separate.67
German express satisfaction with the treaty of Versailles! What could his mind be like?!
67 This is literally true. “The central historical individuals are in their totality thus
composed [zusammengesetzt] out of three factors: out of physical bodies [Körper], out of
minds [Seele], and out of non real meaning” (Rickert, H., Die Grenzen der naturwissen-
schaftlichen Begriffsbildung, p. 596). At the same time, Rickert (Die Grenzen der naturwissen-
schaftlichen Begriffsbildung, p. 608) then found himself recapitulating the dualism that this
third realm was intended to solve, as any historical understanding using this non-real
meaning was a double-relation: on the one hand, the re-experience of the real mental life of
the past, and, on the other, the understanding of the non-real meaning configuration, two
layers laminated so tightly together that the practicing historian would have no idea of
their conceptual distinction.
In other words, after all this work, Rickert, upon re-reading his own
solution, has precisely recapitulated the claims of Simmel’s with which
he originally took such strong issue!68 Simmel of course more flagrantly
proposed that the non-real mental constructs that allow of a non-arbi-
trary re-creation of others’ mentalities were an inherited collective un-
conscious, but he did not mean to defend this literally –he simply needed
to posit some substrate, and he did so in a way that was logically strong if
empirically weak.
Rickert, on the contrary, posited a different substrate that was empir-
ically less implausible but logically of dubious status –simply claiming
that there is a reality of meaning and that we experience the meaning of
these values in their own realm, completely begging the question of how
we know that our “experience” of a value (say, honor) is in fact the same
as that of another. Rickert69 seemed to only admit the possibility of dif-
ferences in the degree to which persons grasped the complete meaning
configuration –not differences of meaning itself.70
Even more, the alert reader will no doubt seize upon the similarity of
this expression –that these absolute meanings “swim, so to speak, ‘freely’
between the [historical] individual, in whom they actually live, and
ourselves”71– to Kant’s use of the same idea of “swimming between”
both in the first critique when accounting for the role of the imagination
in the production of evaluative standards and in the “third Critique” to
account for our capacity to establish intersubjective validity of ideals of
beauty by appeal to some shared archetypes.72 We have clearly returned
to a structural weakness inherent in the philosophy of the subject.
68 Further, Rickert (Die Grenzen der naturwissenschaftlichen Begriffsbildung, p. 592), return-
ing to the issue of “missing data”, now admitted that in many cases there was nothing to
be done but for the historian to bring into play his “intuitive imagination” [anschauliche
Phantasie] so as to complete a meaningful construction of the individual in whom this in-
dividual meaning once dwelt.
70 Rickert (The Limits of Concept Formation in Natural Science, p. 173) hurried to leave
this issue once he had come to this formulation: “We need not consider any further the
particulars of how this part of historical activity as the transconceptual, perceptual, or ‘in-
tuitive’ grasp of the factual material of history takes place, and the extent to which, in
every such case, more than mere perception or mere ‘intuition’ is involved”.
71 Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, p. 583f.
72 In his Critique of Judgment, Kant (Critique of Judgment, translated by W. S. Pluhar,
We must also note how Rickert’s solution parallels that which was
being formulated at around the same time in France by Saussure and
Durkheim (each drawing influence from the other). Saussure73 found
that any attempt to begin by deriving language as an empirical phenom-
enon of communication between persons failed –“We are left inside the
vicious circle”– and hence argued that empirical speaking (parole) had to
be derived from a posited language (langue). This idealized language was
social, unmodifiable by the individual, unitary and homogeneous. Quite
similarly, Durkheim74 argued that “The totality of beliefs and sentiments
common to the average members of a society forms a determinate
system with a life of its own”, the “collective consciousness”. Like
Saussure’s language, this was a unitary and homogeneous entity: “It is the
same in north and south, in large towns and small, and in different pro-
fessions”. This explains how different consciousnesses can “cleave” to
one another –there is a single system outside of all that has objectivity
for each.
Rickert too was forced to solve the problem of intersubjectivity
through a deus ex machina –that we must posit a realm of meaning that is
fixed and identical for all. Of course, were this true, it would indeed
simplify matters greatly. And it may often be a reasonable approximation
Indianapolis, Hackett Publishing Company, 1987 [1790], p. 84; Kritik der Urteilskraft,
Frankfurt am Main, Suhrkamp, 1974 [1790], p. 153) struggled to sketch out the psychol-
ogy of beauty, and suggested not only that the average is the most beautiful (a point
common to many late eighteenth century theorists), but that there is a single image that
“swims between” (or hovers between) the singular and various intuitions of the in-
dividuals [Sie ist das zwischen allen einzelnen, auf mancherlei Weise verschiedenen, Anschauungen der
Indviduen schwebende Bild für die ganze Gattung…”]. As Makkreel (Imagination and Interpretation
in Kant, Chicago, University of Chicago Press, 1990, p. 115) has pointed out, this lan-
guage refers back to that in the First Critique regarding the way the archetypes of the
imagination can produce “monograms” that serve as ideals of sensibility, which serve as
an orientation “more as a sketch that swims in the midst of different experiences, than a
determinate image…” […welche mehr eine im Mittel verschiedener Erfahrungen gleichsam schwe-
bende Zeichnung, als ein bestimmtes Bild ausmachen…] (Kant, I., Kritik der reinen Vernunft,
Leipzig, Verlag von Felix Meiner, 1944 [1787], p. 551; A570/B598). In both places, Kant
opposes the lack of a determinate rule here to the obscurity of the process involved.
73 de Saussure, F., Course in General Linguistics, translated by W. Baskin, New York,
Philosophical Library, 1959 [1915], p. 9.

74 Durkheim, E., The Division of Labor In Society, translated by W. D. Halls, Glencoe,
Illinois, Free Press, 1984 [1893], p. 38f.

to treat this as if it were true –certainly we do so in our daily lives. But

philosophy is not the place to confuse a determinate statement, a useful
approximation, and a transcendental one, and I think this is exactly what
Rickert did. Further, Rickert undercut the ways in which others had sal-
vaged these ideas. We have seen his scornful rejection of the average as
any type of basis for the grounding of the ideal; further, despite his ob-
viously similar language, he took nothing from Kant’s “third Critique”.
Although Kant too posited a trans-individual realm that could explain
intersubjective concordance in terms of qualitative perception, he noted
that this was a transcendental proof of a hypothetical statement: we
must treat the world as if this were true. The advantage of an approach
that says (for example) that speakers need to act as if words have the
same meaning for all persons as opposed to insisting that they really do
have this one meaning is that unlike the latter, the former does not ex-
plode when faced with the undeniable evidence of polysemy and prag-
matic failure in everyday speech.
It is easy for us to wonder at how deep thinkers like Simmel and
Rickert could be led to proposing something as fantastic as an imaginary
realm of ideas or values. But the problem of how to comprehend inter-
subjective comprehension has been fundamental to almost all post-medi-
eval thought, with most “solutions” being little other than the claim to
having a solution (most famously, Heidegger’s75 pronouncement that Da-
sein is primordially Mitsein). The puzzle is still live enough that the dis-
covery of the mirror neuron system has been hailed as the solution
(“Thanks to this mechanism, actions done by other individuals become
messages that are understood by an observer without any cognitive me-
diation”76), a claim taken seriously by contemporary sociological theorists
(e.g., Lizardo77). Rickert and Simmel, it seems, were right to think that
they needed something here, even if it appeared fanciful. The problem
75 Heidegger, M., Being and Time, translated by J. Macquarrie and E. Robinson, New
York, Harper Collins, 1962 [1926].
76 Rizzolatti, G. and Craighero L., “The Mirror-Neuron System”, Annual Review of
Neuroscience, 27 (2004), p. 183.

77 Lizardo, O., “‘Mirror neurons’, Collective Objects and the Problem of Transmis-
sion: Reconsidering Stephen Turner’s Critique of Practice Theory”, Journal for the Theory
of Social Behaviour, 37 (2007), pp. 319-350.
with Rickert’s solution, however, was not simply that it was fanciful, but
that it had a poor anchor.
For Rickert’s main solution was to wed the objectivity of the cultur-
al/historical sciences to the validity of civilizational values, which we
have seen was a disastrous choice. But he realized that establishing the
uniform experience of values required a further lemma, that values are
not subjective constructs at all, but rather are, in and of themselves, in-
tersubjectively stable “things” of which different persons have experi-
ences.78 Although Rickert’s work has large stretches of impressively care-
ful analysis, the overall solution is a failure.
4. Weber
Max Weber did not seem to understand this: he had only read Rickert’s
early edition, before the acceptance of nonreal meaning configurations.
Weber seemed convinced that the details of the problems that con-
fronted him were solved by Rickert’s approach (although he himself did
not seem to use Rickert’s system in any disciplined or coherent way).
Thus Weber had a misplaced confidence that the stability of a voluntarist
neo-Kantian approach to cultural values had been proven, even as he –
Weber– redefined “cultural values” in a way that made no sense for the
Rickertian project, basically depriving them of their unconditional val-
idity, which is key for the establishment of history as a science in
Rickert’s work. Still, Weber was sure that there could be a neo-Kantian
approach to concept formation that could produce a social science
whose only relation to ethics would be that it could provide a means to in-
dependently and exogenously chosen values. Indeed, he (and his follow-
ers) acted as if anyone who denied was a fool or a “big child”. In Ger-
many Weber had been able to brow-beat his opponents into relative
quietude, and in America, those opponents and critics rarely were impor-
ted.
78 Further, as Oakes (Weber and Rickert: Concept Formation in the Social Sciences, p. 142)
has argued, in his work on the System of Philosophy (published in the same year as the third
and fourth editions of Die Grenzen), Rickert derived very different claims about values
which completely undermine their status as trans-individually experienceable.
But the distorted version of “value relevance” that Weber seems to

have adopted led to a fundamentally solipsistic vision of the work of a
cultural scientist –just as Weber’s Puritan faced his angry God alone, so
Weber’s scientist (or his politician) had only his transcendent value to
reckon with when forming his arsenal of concepts and his explanations.
The famous “methodological individualism” of Weber’s work came not
in his rooting things in acting individuals, but in his uprooting of the
notion of truth from the lived interactions of members of some com-
munity (which Rickert still found crucial).
5. Back to Simmel
It may well be that this approach to formalizing a neo-Kantian vocabu-

lary is simply a dead end. But perhaps if it can be salvaged, the way to re-
start would be to return to Simmel’s work. Whatever one thinks about
Weber, no one will argue that he was a playful man (and one might well
doubt that he truly understood what play was all about). Simmel, how-
ever, understood play and playfulness as one form of the more general
dynamic interchange between form and content. Weber remained locked
into his unshakable conviction that there was little hope on the horizon
to save us from a completely formally rational existence, despite a lack of
empirical evidence then or now to support such a hypothesis. Simmel, on
the other hand, understood that it is not simply that the form can be-
come all-powerful, in the way that bureaucratic rules lead to their own
formal rationality at the expense of substantive rationality. Rather, form
can become a content, actually become this, as it does in play or in sociabil-
ity.
This understanding suggests the possibility of an intersubjective con-
nection based on substantive terms, even if this resists formal exposi-
tion. In the idle chatter, the flirting, the banter of playful subjectivity, we
use forms to communicate, but refuse to be constrained by them. Indeed,
the content, the substance, of our interaction is nothing other than the
mutual recognition that the forms have no power over us –that for this
protected time and space, we master them. If this sort of communi-
cation is possible between flirting youth at a ball, I do not understand
why it is not possible between two aged historians!
Although Simmel never solved the problem of intersubjectivity, if we

were to pick up the pieces in his work to start anew, we would start with
his writings on love and on women, especially his 1911 essay on “Female
Culture”. This might seem implausible, but it is here that Simmel returns
to the themes of historical epistemology, and he devotes a large portion
of this text to a direct re-statement of many of the epistemic issues dis-
cussed in his philosophy of history. Simmel emphasizes our need to find,
behind the empirical data we have in others’ actions, “thoughts, feelings,
and intentions that can never be directly ascertained but can only be con-
jectured on the basis of the intuitive imagination”.79 Although we know
that in everyday life we make such psychic attributions with great con-
fidence, as scientists, we must recognize that “facts of observation are
consistent with a number of different psychological infrastructures that
are in principle unlimited…”.
Why return to this basic issue of the indeterminacy of explanations
for historical action? This essay was an attempt to understand what
letting women into the world of the mind might mean. Simmel realized
that this might shake any subjective concord about the social world that
male scholars may have had, given that “the psychological interpretation
of men by women may be fundamentally different in many respects
from the manner in which women psychologically interpret one another
–and the converse is also the case”. “Women as such not only have a
different mix of identity with and difference from historical objects than
men do, and thus the possibility of seeing things men do not see; by
virtue of their distinctive psychic structure, they also have the possibility
of seeing in a different way” –seeing in a different way “from the a priori
of their nature…”.80
Now Simmel is not proposing that sort of solution that Mannheim
would have reached for –that we need both men’s and women’s views.
Something more weighty is being uncovered, something not to be solved
by academic manipulations. For Simmel, the haunting idea is that just as
perhaps we men (I will follow Simmel in assuming a male speaker and
79 Simmel, G., “Female Culture”, translated by G. Oakes, in Georg Simmel: On Women,
Sexuality and Love, New Haven, Yale University Press, 1984 [1911], p. 78f.
80 Simmel, G., “Female Culture”, p. 80.
male audience) can never really see things the same way as women, so
too we may never have that spiritual unity with woman that we want.
This theme is a recurring one for Simmel who, in his Soziologie,81 elo-
quently speaks of the fundamental incompleteness of our relations with
others due to our limited capacity to truly know another’s individuality,
which reaches its depth in the passionate craving for union between the
sexes: “something essentially unattainable”. The epistemic problem
raised in the philosophy of history is, for Simmel, also an abiding per-
sonal, existential problem. For we always do injustice to others’ individ-
uality by seeing them via general concepts. Even where we seek the
greatest fusion with another and believe that we have gained it, we find
that a woman’s physical surrender “does not eliminate a final secret re-
serve of her soul”. “This,” says Simmel,82 “is the purest image –but per-
haps also the crucially decisive original form– for the loneliness of the
human being, who is ultimately an alien, not only in relation to the things
of the world, but also in relation to those to whom he is closest”.
The same epistemic problems in history thus return to frustrate our
eternal attempt to “seek one another, complement one another” across
“the deepest metaphysical chasm”. These somewhat depressing thoughts
actually come from an essay on flirtation, written for a popular publi-
cation.83 Why this discussion of the hiatus irrationalis between men and
women on an essay on this lightest of topics? Because flirtation is the
way in which we come to deal with the fact that love is, as Plato said,
having and non-having (also discussed in Simmel’s84 work on Schopen-
hauer and his85 “Religion and the Contradictions of Life”, where he
81 See Simmel, G., “How is Society Possible?”, translated by Kurt H. Wolff from
Soziologie, in Georg Simmel on Individuality and Social Forms, edited by D. N. Levine, Chicago,
University of Chicago Press, 1971 [1908], p. 10; Soziologie, third edition, in The Sociology of
Georg Simmel, translated and edited by K. H. Wolff, Glencoe, Illinois, The Free Press,
1950 [1923], p. 128.
82 Simmel, G., “Flirtation”, translated by G. Oakes, in Georg Simmel: On Women,
Sexuality and Love, p. 149f.

83 First published in Der Tag, May 1909, then included in Philosophical Culture, Collected
Essays, along with “Female Culture”.

84 See Simmel, G., Schopenhauer and Nietzsche, translated by H. Loiskandl, D. Weinstein,
and M. Weinstein, Amherst, Mass., University of Massachusetts Press, 1986 [1907], p. 56.
85 Simmel, G., Essays on Religion, edited and translated by H. J. Helle in collaboration
with L. Niedler, New Haven, Yale University Press, 1997 [1909], p. 38.
argues that this tension reaches its acme in our feeling for God).
Simmel’s take on this is that flirtation is the unique relation between the
sexes that has that fusion of form and content that Schiller identified as
the heart of aesthetics. It is, says Simmel, the best example of “pur-
posiveness without purpose” (Kant’s formula from the “third Critique”)
that we have.
In other words, Simmel accentuates the problem of intersubjectivity
that is intrinsic to history and proposes at least one type of resolution –
flirtation– as relying on the same “indifference to its object” that charac-
terizes the pure aesthetic. But it is not that flirtation provides the answer
for history, because flirtation, by definition, does not matter. My point so
far is not that Simmel has solved the problem, but that he has identified
it and made it unmistakable. Even a Rickert could not argue that the so-
lution to the lover’s anguish would be “more data”.
When thinking about historiography, one is free to imagine asympto-
tically approaching that dreary and monotonous agreement that would
make us finally a science; when thinking about love, such an eleatic pro-
gression is absurd. We both know and doubt, we utterly fuse then draw
back with the realization that we have no idea who the beloved is, and
again, we know in fact that we do. But it seems that Simmel –without
denying this uncertainty– was beginning to recognize a key type of
certainty here. For in a posthumously published fragment on love he
writes that
“with the exception of religious feelings, love is more intimately and

unconditionally linked with its object than is any other feeling. The clear-
cut fashion in which love develops from the subject corresponds to the
same clearcut fashion with which it is oriented to the object. The crucial
point here is that no instance of a general sort is inserted between them”
(Simmel, G., “On Love (a fragment)”, translated by G. Oakes, in Georg
Simmel: On Women, Sexuality and Love, p. 64).86
86 Also see earlier “How is Society Possible?”, p. 13. William James had made the
same point: “Every Jack sees in his own particular Jill charms and perfection to the en-
chantment of which we stolid on-lookers are stone-cold. And which has the superior
view of the absolute truth, he or we? Which has the more vital insight into the nature of
Jill’s existence, as a fact? Is he in excess, being in this matter a maniac? or are we in defect,
being victims of a pathological anesthesia as regards Jill’s magical importance. Surely the
That is, here individual meets individual without the general concepts
that Simmel had argued in other works always intervene between us and
others. Simmel may be wrong, but he is serious that there is a different
epistemic quality to the mental states involved. “Thus love exists as an
intention, directly focused on this object” –here he is using “intention”
in the scholastic sense of the connection between a mental state and a
physical object. Were that so –and it is not obvious that it cannot be so–
this would have serious implications for the resolution of the question
of historiography, and perhaps for other cultural sciences, including gen-
eralizing ones.
6. Conclusion
This emphasis on a fundamentally second person approach to social knowl-

edge –one that was neither the “first person” knowledge that Scheler or
even Dilthey imagined the historian must have, or the “third person”
knowledge (Oakes’s translation of fremden Seelenleben, we recall) that the
“generalizing” sciences seemed to claim –was the basis for the theo-
logical approach of Simmel’s famous student, Martin Buber. In Buber’s87
words, “the relation to the Thou is direct”. Although we easily remember
Buber as only discussing the relation of Man to God, this is incorrect; al-
though theological through and through, Buber’s approach is a pheno-
menology of the face-to-face encounter, and the difference between the
relation of (on the one hand) I to It (or even He or She), and (on the
other), of that of I to Thou. Although we may now emphasize the con-
nection to Heidegger’s work (clearly inspired by Simmel, whom Hei-
degger considered one of the few living philosophers worthy of notice),
we cannot take Buber’s work out of the context of the triad of persons
–not, as in most western philosophy, the three persons of the Christian
godhead, but instead, the triad of grammatical persons.
latter... For Jack realizes Jill concretely, and we do not...” (James, W., “What Makes a Life
Significant”, in Matthiessen, F. O., The James Family: A Group Biography, New York, Alfred
A. Knopf, 1947 [1900], p. 404f).
87 Buber, M., I and Thou, translated by R. G. Smith, Edinburgh, T. & T. Clark, 1937
[1923], p. 11.
This triad had indeed been a key structuring element for Simmel –the
structural differences between the planes of the Individual, the Social,
and the Objective. (Although we cannot go into this here, we see a close
relation to the work of the central figure of the Marburg school of neo-
Kantians, Herman Cohen, who also emphasized this triad and drew
upon it in his relational theology; we further cannot understand Simmel’s
use of this triad without understanding his relation to Schelling, but
again, we must leave this to the side for now.) In his sociology, Simmel
had tended to assume that the “social” was a collective –a second person
plural, if a second person at all– and one destined to be transcended by
objectivity. Just as arbitrary group norms and rituals are the scaffolding
from which an objective, defensible enlightenment ethics can develop, so
too more generally, the “social” is a stepping stone on the way to objec-
tivity.
But Simmel’s theory of love as intention –if we take it more seriously
perhaps than he did– suggests that our best social knowledge may not
start from such second person plural knowledge, and push it towards ob-
jectivity, but from a second person singular, a compilation of value-satu-
rated embracings of particularity. As Simmel says, in “Female Culture”88
among other places, one does not need to be Caesar in order to know
Caesar. But it seems that the question is whether in order to know
Caesar, one has to love him.89
University of Chicago
Chicago, United States
88 See Simmel, G., “Female Culture”, p. 79.

89
An earlier version of this was presented at the conference Georg Simmel: Life, Self,
Culture, Society, at the Franke Center for the Humanities, University of Chicago, Novem-
ber 12, 2011; I thank the participants, especially Donald Levine and Hans Joas, as well as
Monica Lee, for their comments.
References
Buber, M., I and Thou, translated by R. G. Smith, Edinburgh, T. & T.

Clark, 1937 [1923].
Dilthey, W., Introduction to the Human Sciences, translated by R. J. Betanzos,
Detroit, Wayne State University Press, 1988 [1883].
Durkheim, E., The Division of Labor In Society, translated by W. D. Halls,
Glencoe, Illinois, Free Press, 1984 [1893].
Heidegger, M., Being and Time, translated by J. Macquarrie and E.
Robinson, New York, Harper Collins, 1962 [1926].
Kant, I., Kritik der reinen Vernunft, Leipzig, Verlag von Felix Meiner, 1944
[1787].
——, Kritik der Urteilskraft, Frankfurt am Main, Suhrkamp, 1974 [1790].
——, Critique of Judgment, translated by W. S. Pluhar, Indianapolis,
Hackett Publishing Company, 1987 [1790].
James, W., “What Makes a Life Significant”, in Matthiessen, F. O., The
James Family: A Group Biography, New York, Alfred A. Knopf, 1947
[1900], pp. 404-419.
Lizardo, O., “‘Mirror neurons’, Collective Objects and the Problem of
Transmission: Reconsidering Stephen Turner’s Critique of Practice
Theory”, Journal for the Theory of Social Behaviour, 37 (2007), pp. 319-
350.
Makkreel, R. A., Imagination and Interpretation in Kant, Chicago, University
of Chicago Press, 1990.
Oakes, G., Weber and Rickert: Concept Formation in the Social Sciences,
Cambridge, Mass., MIT Press, 1988.
Rickert, H., Die Grenzen der naturwissenschaftlichen Begriffsbildung, Tübingen,
Verlag von J. C. B. Mohr, 1902.
——, Die Grenzen der naturwissenschaftlichen Begriffsbildung, Fünfte Auflage,
Tübingen, Verlag von J. C. B. Mohr, 1929.
——, The Limits of Concept Formation in Natural Science, Fifth Edition,

translated by G. Oakes, Cambridge, Cambridge University Press, 1986
[1929].
Rizzolatti, G. and Craighero L., “The Mirror-Neuron System”, Annual
Review of Neuroscience, 27 (2004), pp. 169-192.
de Saussure, F., Course in General Linguistics, translated by W. Baskin, New
York, Philosophical Library, 1959 [1915].
Simmel, G., Die Probleme der Geschichtsphilosophie, Leipzig, Verlag von
Duncker & Humblot, 1892.
——, “How is Society Possible?”, translated by Kurt H. Wolff from
Soziologie, in Georg Simmel on Individuality and Social Forms, edited by D.
N. Levine, Chicago, University of Chicago Press, 1971 [1908], pp. 6-
22.
——, The Problems of the Philosophy of History, second edition, translated
by G. Oakes, New York, The Free Press, 1977, pp. 1905.
——, Schopenhauer and Nietzsche, translated by H. Loiskandl, D. Weinstein,
and M. Weinstein, Amherst, Mass., University of Massachusetts Press,
1986 [1907].
——, “Flirtation”, translated by G. Oakes, in Georg Simmel: On Women,
Sexuality and Love, New Haven, Yale University Press, 1984 [1909], pp.
133-152.
——, Essays on Religion, edited and translated by H. J. Helle in
collaboration with L. Niedler, New Haven, Yale University Press,
1997 [1909].
——, “Female Culture”, translated by G. Oakes, in Georg Simmel: On
Women, Sexuality and Love, New Haven, Yale University Press, 1984
[1911], pp. 65-101.
——, “On Love (a fragment)”, translated by G. Oakes, in Georg Simmel:
On Women, Sexuality and Love, New Haven, Yale University Press, 1984
[1921-2], pp. 153-192.
——, Soziologie, third edition, in The Sociology of Georg Simmel, translated

and edited by K. H. Wolff, Glencoe, Illinois, The Free Press, 1950
[1923], pp. 87-408.
Spencer, H., The Study of Sociology, New York, Appleton, 1896 [1873].
Weber, M., Economy and Society, edited by G. Roth and C. Wittich (two
volumes), Berkeley, California, The University of California Press,
1978.
THE REALIST SOCIOLOGICAL
APPROACH TO ACTION
ALEJANDRO N. GARCÍA MARTÍNEZ
Introduction
One of the key areas of debate –if not the main one– in sociological re-
flection continues to be that associated with the problem of defining the
autonomic aspects and interrelationships related to, on the one hand,
socio-cultural structures and on the other, with agency or human freedom.
This is a key question given the compelling need to adopt a position on
these issues before we can make any observation on mankind’s place in
society. Indeed, when research of any importance is carried out, we have
to position ourselves at some point along the continuum of voluntarism
or social determinism, between the extremes of subjectivism and objec-
tivism. Essentially, it is a matter of adopting a basic stance with regard to
the question of the nature of human being and his action in a social
context: we can understand and think of an individual as a reality con-
strained –to a greater or lesser extent– by the social structures in which
he lives, or conceive him as a free subject who basically acts by virtue of
subjective intentionalities.
The methodology, themes and results that can be obtained in sociolo-
gical research will depend on whether we adopt an extreme or balanced
position with respect to these anthropological concepts. So, if we adopt
an extreme form of determinism, a human being and the action he is ca-
pable of performing will be reduced to a mere reproduction of the so-
cial system in which he finds himself, and man becomes nothing more
than a puppet in the hands of impersonal forces that forcefully constrain
his activity. If, on the other hand, we subscribe to a purely voluntarist vi-
sion of human action, we end up ignoring the social constraints of ac-
tion and “the social” becomes reduced to a more or less random com-
bination of individual wills, with no consistency or characteristics of its
own. In short, there is a strong undercurrent of anthropological reflec-
tion about the efficacy of human action that underpins sociological re-
150 Alejandro N. García Martínez
search, acting as an inevitable frame of reference. It is a reflection that,

as Margaret Archer sums up, arises from our daily experiences:
“For it is part and parcel of daily experience to feel both free and en-
chained, capable of shaping our own future and yet confronted by tow-
ering, seemingly impersonal, constraints. Those whose reflection leads
them to reject the grandiose delusion of being puppet-masters but also
to resist the supine conclusion that they are mere marionettes then have
the same task of reconciling this experiential bivalence, and must do so
if their moral choice is not to become inert or their ‘political’ action inef-
fectual. Consequently in facing up to the problem of structure and agen-
cy social theorists are not just addressing crucial technical problems in
the study of society, they are also confronting the most pressing social
problem of the human condition” (Archer, M. S., Culture and Agency, re-
vised ed., Cambridge, Cambridge University Press, 1996, p. xii).
This study aims to show how two recent and complementary sociolo-
gical approaches try to resolve the problem of achieving an adequate
concept of human action in society: the morphogenetic approach of
Margaret Archer and the relational theory of Pierpaolo Donati.
Although certain differences exist in their assumptions, the two ap-
proaches share many basic criteria, which means we can label them both
as realist conceptions of man in society. They are realist in the sense that,
contrary to empiricism or positivism, they believe that human action and
social reality cannot be reduced to experiences. Although sociology re-
flects on socially constructed reality, the main goal of research in that
field is not limited to events experienced directly, or to mere abstract
concepts or socially constructed linguistic resources.1 In contrast to these
postulates, sociological realism emphasizes the existence of a social reali-
ty that is independent of our understanding of it, and that is accessible
through the use of an appropriate scientific methodology. Precisely be-
cause of their realist defence of social order, both approaches –the mor-
phogenetic and the relational– continually strive to highlight the
emergent nature of social reality, a reality sui generis that can be studied
1 See Lewis, P. A., “Metaphor and Critical Realism”, Review of Social Economy no. 54
(1996), pp 487-506.
The Realist Sociological Approach to Action 151
scientifically using certain specific criteria for observation and interpre-

tation. As Donati sums it up: “Critical realism vindicates the possibility
of knowing social reality, meant as a ‘fact’, which is not observer-depen-
dent and which is not built on quicksand”.2
To explain the basic questions in each of these two sociological ap-
proaches and to clarify their way of explaining human action in society,
this study is divided into three parts: the first part serves as an introduc-
tion to the essential analytical distinctions proposed by the morphogene-
tic approach of Margaret Archer, which is followed by an outline of the
key elements of Donati’s relational sociology, placing special emphasis
on his concept of social relations as a basic element of society. Finally, in
the third section, which also serves as a conclusion, some of the conver-
gences and similarities of the two perspectives are discussed.
1. Structure, Culture and Agency: the Morphogenetic Approach of

Margaret Archer
The morphogenetic approach proposed by Margaret Archer in her aca-

demic work takes, as its frame of reference, the philosophy and realist
epistemology developed during recent decades in the English-speaking
world by Roy Bhaskar.3
There are three main pillars on which Archer reflects in her writings:
a) First of all, in keeping with realist epistemology, she stresses the

need to include the stratified nature of social reality as a starting point
for sociological research, employing a series of analytical distinctions that
enable us to perceive not only the interrelationships and interdependen-
cies between socio-cultural structures and individual actions but also,
2 Donati, P., Relational Sociology: A New Paradigm for the Social Sciences, London,
Routledge, 2011, p. 97.
3 See Bhaskar, R., The Possibility of Naturalism, third ed., London, Routledge, 1998; A
Realist Theory of Science, London, Routledge, 2008; Reclaiming Reality: A Critical Introduction
to Contemporary Philosophy, London, Routledge, 2011; Critical Realism: A Brief Introduction,
London, Routledge, 2013.
their autonomous and emergent properties that make them mutually ir-
reducible. She dedicates some of her most representative works to this
fundamental objective, such as Culture and Agency4 or Realist Social Theory:
the morphogenetic approach.5
b) On a more transversal note, although she also engages in specific
reflections (for example, her work Being Human6), another of the con-
cerns of this sociologist is the recovery of a rehumanizing vision of
man. Man’s dehumanization, his structural submission in many post-
modern currents of thought, both those that dissolve the human condi-
tion into the mere interplay of insuperable structures or into all-encom-
passing external signs, is a constant feature of her work. Thus, time and
again, her attempt to recover a non-reductionist vision of man reappears,
with the ultimate goal of re-establishing in him the capacity to transform
and articulate –although this capacity will vary, depending on the struc-
tural conditions of the specific moment in which he acts– the actual
structures within which he finds himself; and which, in turn, will also im-
pose constraints or limitations on the individual.
c) Finally, a third significant aspect of Archer’s sociological proposal
is the inclusion of her analysis as to how people subjectively deal with
the social constraints in which they are immersed. The question of reflex-
ivity or the internal conversation by means of which we interpret and give
meaning to our social action has been the subject of some of her latest
works, such as Structure, Agency and the Internal Conversation, Making Our
Way Through the World: Human Reflexivity and Social Mobility or Conversations
About Reflexivity,7 in which she shows her increasing interest in this ques-
tion.
4 Archer, M. S., Culture and Agency, revised ed., Cambridge, Cambridge University
Press, 1996 [1988].
5 Archer, M. S., Realist Social Theory: The Morphogenetic Approach, Cambridge,

6 Archer, M. S., Being Human: The Problem of Agency, Cambridge, Cambridge University
Press, 2000.
7 Archer, M. S., Structure, Agency and the Internal Conversation, Cambridge, Cambridge
University Press, 2003; Making Our Way Through the World: Human Reflexivity and Social
Mobility, Cambridge, Cambridge University Press, 2007; Conversations About Reflexivity,
In the detailed and complex sociological proposal presented by

Margaret Archer, the description of the key aspects of her morphogene-
tic analysis is particularly relevant to this article. This analysis takes the
form of a discussion of the alternative proposals of individualist and
collectivist approaches. For the former, human action in society can be
explained exclusively by referring to subjective intentionalities. For the
latter, subjective intentionality is nothing other than an apparent freedom
that masks the determination effect social structures exert on individual
consciences. Ultimately, the different exponents and theories corre-
sponding to these two broad approaches favour one of the two conflic-
ting elements (socio-cultural structures and human agency) making its
counterpart a purely dependent element.
a) Analytical Dualism versus Conflationist Formulations

Unlike genuinely collectivist or individualist viewpoints, the morpho-
genetic approach takes as its starting point a balanced reflection on the
constraint that socio-cultural elements exert on social action, and peo-
ple’s ability to control and even transform such structures: To what ex-
tent does structure or culture constrain human action? To what degree
can human action influence and transform the structure or culture of a
society?
In order to answer these questions, the methodological proposal of
Margaret Archer begins by establishing certain analytical distinctions that
are intertwined in empirical societies but that can be differentiated ana-
lytically: culture, structure and agency. They also have specific emergent prop-
erties, so that it is possible to analyse the reciprocal influences and inter-
dependencies that develop between these dimensions. However, instead
of differentiating and distinguishing them, these concepts have often
been fused, making no distinction between them. Usually, one of these
elements is considered autonomous, while the other is purely dependent:
or the values are controlled by an unrivalled dominant group, or the ac-
tors simply reproduce values or culture.
By contrast, adopting analytical dualism as the methodology pro-
posed for the morphogenetic approach enables us to perform a rich ana-
lysis of the cultural dynamic, in which different elements are analysed in
terms of their interdependencies but also, their own autonomies and
emergent characteristics. To this end, Archer proposes a fundamental

distinction between the Cultural System, which refers to the logical rela-
tionships between propositions (this has to do with the logical consis-
tency of the world of ideas), and Socio-Cultural Interaction, which refers
to causal relations (causal consensus) in the sphere of interpersonal in-
fluences (persuasion, power struggles, etc.). It is precisely in the interplay
of reciprocal influences, and in the interrelations between the Cultural
System and Socio-cultural Interaction, where the social and cultural
change of a society is configured (morphogenesis) or its stability (mor-
phostasis) is promoted.
This basic distinction of analytical dualism enables us to understand
the extent to which the internal relationships of the Cultural System
(logical coherence) affect agency (Socio-cultural Interaction), and the ex-
tent to which interpersonal influences (Socio-cultural Interaction) in turn
configure the Cultural System:
“The Cultural System and Socio-Cultural life do not exist or operate

independently of one another; they overlap, intertwine and are mutually
influential. But this is precisely the point, for I am not asserting dualism
but rather the utility of an analytically dualistic approach, the main rec-
ommendation for which is the very fact that it allows this interplay to be
explored” (Archer, M. S., Culture and Agency, p. xix).
With these analytical distinctions it is possible to understand and shed

light on social change. And it is precisely at this point that the concept of
the morphogenesis of culture appears: in reference to the complex
changes that produce the variation in form, structure or the given states
of a system (“morphostasis” is the inverse), and in which the end pro-
duct is called “Elaboration”. In this way, it is possible to analyse the re-
ciprocal interplay between culture, structure and agency over time.
The basic conception underlying any form of morphogenesis is a
firm commitment to the inclusion of the historical perspective: the start-
ing point for any socio-cultural interaction is a set of predetermined con-
ditions. In turn, the final result of this interaction will be the new cultural
configuration or Elaboration, or rather the maintenance of the foregoing
conditions:
Cultural Constraint – Socio-cultural Interaction – Cultural Elaboration
These analytical categories proposed by the morphogenetic approach

are completely at odds with what Archer calls the “myth of cultural inte-
gration”, the root cause of certain conflationist formulations, which ulti-
mately reduce human action to a variable that depends on one of its pre-
dominant dimensions. The general acceptance of this myth is due to cer-
tain deficiencies in the concept of culture and the way it has been viewed
in different disciplines: at the descriptive level, the notion of culture con-
tinues to be excessively vague and poorly developed; and at the explan-
atory level it oscillates between being an absolutely independent variable
and occupying a position of supine dependence.8 It is from these dif-
ficulties that the myth of cultural integration arises, with the a priori premise
that in every culture there is always a coherence that can be discovered.
There are two fundamental deficiencies in the myth of cultural inte-
gration: the negation of cultural inconsistencies, on the one hand; and
the conflation (non-differentiation) between integration of the Cultural
System and Socio-cultural Interaction. There are three types of con-
flation: upwards, downwards or central. Downwards Conflation9 denies in-
consistency at the level of the Cultural System, which is always regarded
as perfectly integrated, and it strongly proposes the subordination and
dependence of the Socio-cultural level to the level of the Cultural Sys-
tem. In this way, it leaves interpersonal and reciprocal influences (which
we could conceptualize as power) suspended and ineffectual, and accord-
ingly they fail to fulfil any role whatsoever in the establishment of the
domination of the central system of values. By contrast, Upwards Confla-
tion10 concedes an essential role to power in the imposition of culture
(resembling the “manipulated consensus” found in some Marxist per-
spectives). Basically, from their inception both these types of conflation
maintain a position which is characteristic of individualism or collecti-
vism:
8 See Archer, M. S., Culture and Agency, p. 1.

9 See Archer, M. S., Culture and Agency, chapter 2.
10 See Archer, M. S., Culture and Agency, chapter 3.
“In the heritage of Individualism it was ‘structure’ which became the

inert and dependent element, whilst Collectivism fostered instead the
subordination or neglect of ‘agency’, thus respectively perpetuating the
two forms of social theorizing which I have termed the fallacies of ‘up-
wards conflation’ and ‘downwards conflation’” (Archer, M. S., Realist So-
cial Theory: The Morphogenetic Approach, p. 33).
The third type is central conflation, proposed by sociologists such as

Giddens or Bauman, who seem to solve the problem of the preceding
types of conflation by coinciding in the contention that structure and
agency cannot be explained by dualism –be it upwards or downwards.
Yet they ultimately make the same mistake, which is none other than
eliding or suppressing the differences between both analytical levels; for
example by means of non-specific concepts, such as that of “social prac-
tices” used by Giddens, which remove specificity from each of the ana-
lytical distinctions, reconverting them into a single concept that masks
their mutual relationships and interdependencies.11
b) The Analysis of Cultural Dynamics

The analytical distinction between the Cultural System and Socio-cultural
Interaction provides an approach that enables cultural dynamics to be
analysed. To this end, Archer sets out to test the following propositions:
A) There are logical relationships between the components of the

Cultural System;
B) There are causal influences that are exercised by the Cultural Sys-
tem at the Socio-cultural level;
C) There are causal relations between groups and individuals at the
Socio-cultural level;
D) There is an Elaboration of the Cultural System due to the fact
that the Socio-cultural level modifies the current logical relationships
and introduces new ones.
11 See Archer, M. S., Culture and Agency, Chapter 4; Realist Social Theory: The Morphogene-
tic Approach, p. 93ff.

The aim of the first proposition is to establish the existence of objec-

tive contradictions and complementarities within a Cultural System with-
out any reference to the level of Socio-cultural Interaction. To maintain
this proposition it is necessary, at the same time, to challenge it on vari-
ous levels: ontologically, it means we have to show that there are objec-
tive (i.e. independent of the observer) relationships of contradiction and
complementarity (or they are simply independent); epistemologically,
such that relationships can be understood by referring to logical and in-
variable principles, i.e. they are not relative; and, methodologically,
whereby the problems implicated in their transcultural identification are
not impossible to tackle.12 To put this another way: if we wish to defend
the position that there are logical relationships between the components
of the Cultural System, it implies confronting the strong thesis of relati-
vism. This is why the morphogenetic approach is decidedly non-relati-
vist.13
Once we accept the possibility that the cultural system does not al-
ways present itself as integrated and can harbour contradictory, com-
plementary or simply independent ideas, we come to the second propo-
sition of the morphogenetic approach, which maintains that there are
causal influences exerted by the Cultural System at the Socio-cultural lev-
el. Men create culture but the Cultural System also acquires some degree
of autonomy and, up to a point, constrains Socio-cultural Interaction.
Thus, the logical properties of the Cultural System (contradiction, com-
plementarity or independence) create different situational logics that
mould or define the context of the Socio-cultural Interaction.14
Archer explains that there are two types of situational logic derived
from the level of integration, which the Cultural System presents at a
given moment in time or for a specific context of action: we may have a
scenario of Constraining Contradictions (proposition A and B are logi-
cally inconsistent, the situational logic thereby constrains those who want
to maintain A towards correction manoeuvres); or a scenario of Conco-
mitant Complementarities (A and B are complementary, by invoking A
we evoke B, and, as B is consistent, its adhesion to A is reinforced).
12 See Archer, M. S., Culture and Agency, p. 107ff.

Depending on whether we are dealing with one or other type of situa-

tional logic, we can expect certain operative effects,15 a directionality fa-
voured by this same logic –detached, in turn, from the level of inte-
gration in the Cultural System dimension. The result of which could be a
scenario of constraints or concomitant complementarities. In the for-
mer, the contradictions of the Cultural System favour a dynamic that is
conducive to “ideational syncretism”, a syncretism that corrects one or
both of the positions that are in contradiction. In the latter, the case of
complementarity in the world of ideas, the favoured dynamic is that of
an “ideational systematization”, a strengthening and development of the
relationships between the parts of the Cultural System, and as such,
there is an impetus towards the reproduction and perpetuation of the
Cultural System.
However, the situational logic that derives from this stronger or
weaker integration at the level of the Cultural System does not determine
(although up to a point it does constrain) Socio-cultural Interaction (oth-
erwise we would be in a position of downwards conflation), and we
should pay attention to the dynamic that characterizes Socio-cultural In-
teraction:
“It was never suggested that the logical state of affairs in the Cultural
System causally determined the extent of Socio-Cultural integration –a
proposition only a downwards conflationist could endorse. On the con-
trary it is maintained that orderly or conflictual relations at the S-C level
can show a significant degree of independent variation from those charac-
terizing the CS at any time. In short, S-C integration does not mirror CS
integration. Although the former is indeed conditioned by the latter, it al-
so has its own dynamics which need to be examined” (Archer, M. S., Cul-
ture and Agency, pp. 185-186).
So, taking as our starting point the conditioning of the situational

logic derived from the Cultural System (which will be coherent or contra-
dictory), the different possibilities for the development at the level of

Socio-cultural Interaction refer, in turn, to the integration or conflict

present at the Socio-cultural level:
a) When the situational logic is one of constraining contradiction

(poor integration of the Cultural System, contradiction between the pro-
posals), it may be the case that the level of Socio-cultural Interaction is
highly integrated or disordered. If it is orderly and integrated, the differ-
ent groups or structures can contain the contradiction that exists at the
cultural level (for example, through an authoritarian containment of the
ideas in conflict by means of the “official truth”), so that the final result
is morphostasis. Conversely, if the Socio-cultural Interaction level is not
so integrated and it is possible for conflicting groups to activate cultural
contradictions, then there is a possibility of some kind of cultural change
(morphogenesis).
b) When the situational logic is one of concomitant complementarity,

we likewise need to observe the level of integration within Socio-cultural
Interaction to account for the cultural Elaboration or Morphostasis.
Thus, in a scenario of cultural complementarity and well-integrated So-
cio-cultural Interaction, the social group can provide support to this
complementarity and the result is morphostasis. By contrast, in the case
of a disorganized or poorly integrated Socio-cultural level, what Archer
calls contingent complementarities may appear: new routes of explora-
tion and recombination of familiar elements are forged, which are trans-
formed into a new form that rebuffs familiarity (innovation emerges).
This makes it easier to understand the last proposition (D) indicated

above, which attempted to show that there is elaboration of the Cultural
System, due to the fact that the Socio-cultural level modifies the current
logical relations and introduces new ones. In any case, and as already
explained –both in the case of contradictions and complementarities at
the Cultural System level–, it is also necessary to see what is happening at
the Socio-cultural level. Whether there is complementarity or contradic-
tion between two elements is a systemic question (it refers to the Cultural
System), yet the contingency of their conjunction, of the directionality
towards complementarity –or contradiction–, and the associated effects
on Cultural Elaboration, are due to action at the Socio-cultural level. In
short, there is an activation that depends on human agency, understood

to be individual choice in this relational framework.
In sum, although it is true –as downwards conflationists maintain–
that the logical relationships in the Cultural System causally constrain hu-
man action in Socio-cultural Interaction, this is not the whole truth. We
need to take into account the effects of the competitive contradictions
and contingent complementarities activated at the level of Socio-cultural
Interaction by the selective accentuation (human agency) of certain
logical relationships of the Cultural System. In other words, social action
chooses and determines which logical relationships have cultural promi-
nence in society (and this is part of the truth that upwards conflation in-
troduces). However, to successfully analyse cultural dynamics requires us
to analytically distinguish the Cultural System level and that of Socio-cul-
tural Interaction (Culture and Agency), in order to study their autono-
mous development and the reciprocal influences that can give rise to in-
novative Cultural Elaboration.
c) Human Agency and Reflexivity

This summary of the morphogenetic approach as an analytical model to
supersede other reductionist explanations of human action in society has
shown us one of the main interests of Archer: that which concerns the
ontological status of culture, social structures and agents, with their spe-
cific properties and reciprocal irreducibility. The individual, social and
cultural spheres are interdependent and interrelated, but their causal in-
fluences are expressed in specific properties that must be conceptualized
differently and in the light of their mutual interdependencies. Therefore
this is a sociological proposal, the basic premise of which is the emer-
gent nature of the social.16
With such analytical distinctions, this social realist theory manages to
conceptualize the way in which the emergent properties of structure and
culture are related to human action. They constitute the situations and
logics in which human action unfolds. At the same time they serve as
constraints and enablements for practical action. Whatever the case, both
16 See Chernilo, D., “Introducción”, in Archer, M. S., Teoría social realista: el enfoque mor-
fogenético, Santiago de Chile, Universidad Alberto Hurtado, 2009, p. 13.

dimensions require activation on the part of concrete agents situated in a

relational framework. In other words, the causal efficacy of socio-cul-
tural elements for human action is intrinsically related to the individual
aspirations, preoccupations and interpretations that each agent develops
in relation to himself and the social context in which he is immersed.
This is why, as a necessary continuation of the morphogenetic approach
it becomes necessary to analyse the active capacity of the agents im-
plicated in social action. Who are these agents and how do they acquire
their personal or social sense of identity? How do they interpret their re-
lationship with the socio-cultural situation in which they act? These are
the answers that Archer tries to provide in some of her most recent
works,17 through a systematic investigation of human agency and reflex-
ivity.
It is precisely the process by which an agency mediates in the struc-
ture that she proposes to elucidate through the concept of “internal con-
versation”; in other words, how agents deliberate about their social situa-
tion and relate it to their personal interests and aspirations. Through this
internal conversation, agents interpret and make sense (although only
partially and fallibly, but with “real” practical consequences) of their so-
cial position in relation to their vital projects, concerns and desires; and
they generally adjust their practical choice to the conclusions of this in-
ner deliberation (which, once again, is fallible).
The concept of reflexivity explores this question, although it cannot be
developed in detail in this overview of Archer’s morphogenetic ap-
proach. Suffice to say at this point that this refers to the way in which
agents anticipate how their plans for action can modify the circum-
stances due to their socio-cultural context:
“Because they [agents] possess personal identity, as defined by their

individual configuration of concerns, they know what they care about
most and what they seek to realize in society. Because they are capable of
internally deliberating about themselves in relation to their social circum-
stances, they are the authors of projects that they (fallibly) believe will
17 See Archer, M. S., Structure, Agency and the Internal Conversation; Making Our Way
Through the World: Human Reflexivity and Social Mobility; Conversations About Reflexivity.
achieve something of what they want from and in society. Because pur-
suit of a social project generally spells an encounter with social powers,
in the form of constraints and enablements, then the ongoing ‘internal
conversation’ will mediate agents’ receptions of these structural and cul-
tural influences. In other words, our personal powers are exercised
through reflexive interior dialogue and are causally accountable for the
delineation of our concerns, the definitions of our projects, the diag-
nosis of our circumstances and, ultimately, the determination of our
practices in society. Reflexive deliberation constitutes the mediatory pro-
cess between ‘structure and agency’; they represent the subjective ele-
ment which is always in interplay with the causal powers of objective so-
cial forms” (Archer, M. S, Structure, Agency and the Internal Conversation, p.
130).
2. The Relational Sociology of Pierpaolo Donati
Using more genuinely sociological references, the relational approach

proposed by Donati also presents itself as a methodological approach,
and as a substantial alternative to the programme of individualism and
collectivism or their reductionist attempts to explain the role of human
beings in society. For the relational approach, the explanation of the hu-
man being and his action in society cannot be drawn from a merely
nominalist supposition about society (as individualism suggests), nor
from the social determination of individual behaviour (as more collec-
tivist and structuralist positions propose). On the contrary, a fruitful and
honest understanding of human action in society must take as its starting
point people in relation to one another, and the relationship itself as an
emergent reality. This is how the author himself explains it:
“Individualists claim that society has no ‘reality’ […] Methodological

individualism implies a nominalist philosophical position to understand
social issues […] Proponents of holism remind us that there is an order
to reality that must be taken into account or individual behaviour is in-
comprehensible. These authors emphasize the decisive role of culture as
a condition of individual behaviour, and as a key point in the explanation
and comprehension of social phenomena […] The position of relational
sociology is neither on the side of individualists nor of proponents of

holism. It asserts that there is an order to reality –sociological reality–
that both classical and modern philosophy have yet to understand. So-
ciety is not an organic body and neither is it a sum of individuals. It is
more like a relational configuration that goes beyond the simple sum of
individuals and it never goes as far as being an organic body, which
means it never exhausts its own possibilities. […] In other words, the so-
cial is not just found in individuals but also, prior to and beyond indi-
viduals. It is the emergence of their relationships” (Donati, P., Repensar la
Sociedad: El enfoque relacional, Madrid, EIUNSA, 2006, pp. 58-59).
The relational approach attempts to offer an innovative methodology

for studying man in society. To achieve this, it is necessary to redefine
and accurately verify what the object of study is in this discipline. And
what is the object of study in sociology? What is what we call society, in
which we suppose that people act and are constrained in their action to
varying degrees? The answer the relational approach affords is clear: so-
ciety consists of relations between people. It is a reality that many ap-
proaches often implicitly take for granted, and which they do not really
make a fundamental and methodological analysis of: “My key point is
that what is common to the classical sociological tradition is implicit and
not yet still well understood today: i.e. that the nature of ‘social facts’ is a relational
matter”.18
So, social reality is relational. Aristotle said that our being is also rela-
tional19 and, although this has often been conceived as a secondary form
of being with respect to our being as substance, for example, it is cer-
tainly as important as our substantial or material nature. This is what the
“relational twist” began to postulate from the different perspectives of
authors like Weber, Simmel, Husserl, Buber, etc. And it is what the rela-
tional approach eventually intends to consolidate as the object of study
in sociology: social science studies social facts, not in the Durkheimian
18 Donati, P., Relational Sociology: A New Paradigm for the Social Sciences, London,
Routledge, 2011, p. 1.
19 See Aristotle, Categories 7:6, in The complete works of Aristotle: The revised Oxford trans-
lation, ed. by J. Barnes, 2 vols., Princeton, Princeton University Press, 1984.

sense as “things” but as relational realities. In other words, sociology

studies social relations. As Donati sums it up:
“In brief, for relational sociology, social reality is social relationality.

Society does not ‘have’ or ‘include’ social relations, but ‘consists of ’ so-
cial relations. The social relation/ship is the ‘ultimate entity’, that is, the
irreducible element or ‘molecule’ of social reality (that which gives the
social its qualities, properties and powers)” (Donati, P., Relational Sociology,
p. 98).
The implications of understanding society as relationships are varied.

For a start, we need to clarify what these relationships consist of, which
are conceived as the core and essence of society. For Donati, relation-
ships are specific forms of interaction between people, which generate
forms of interdependence that are also specific.20 This reciprocity and
interdependence in the relationship between two or more people can
adopt varied and contingent forms: every relationship implies give and
take, but each type of relationship admits different forms of giving and
receiving. For example, a commercial relationship is not the same as a
friendship, and the forms of reciprocal exchange in such relationships
are not the same either. Yet, furthermore, no two friendships are the
same either as the specific way in which two individuals develop a friend-
ship is not entirely decided by the structure of the relationship but rath-
er, the particular way in which each of the two individuals expresses their
friendship is also important. This is where the contingent nature lies,
which is also present in the relationship, although we can simultaneously
talk about a basic structure or ideal type for this specific relationship.
This is why it is not sufficient to describe the different types of relational
structure. To understand reciprocal action we also need to be aware of
the inherent contingency of every relationship and the singularity of the
action realized intentionally by the members who are linked through it:
“Social networks are constructed out of freedom and structural con-
20 See García Ruiz, P., “Presentación”, in Donati, P., Repensar la Sociedad, Madrid,
EIUNSA, 2006, p. 12.

straints: this is why they are the perfect expression of the relationship –
which is always problematic– between freedom and determinism”.21
However, although relationships assume various forms and include
the agent as an element, which is characteristic of people who are recip-
rocally linked, it does not imply that a relationship can be examined just
by looking at its structure or the interaction of its members. The rela-
tionship itself becomes an emergent reality which goes beyond its mem-
bers in reciprocal action. The specific connection which is created be-
tween the ego and alter ego depends on each of them but it also exceeds
them: it gives rise to a reality sui generis… So what type of reality is that
which links people but, at the same time, goes beyond them?
a) The Reality of Relationships

For Donati, to say that a phenomenon is “social” is the same as claiming
that it “exists in relation”; in other words it is a reality that emerges from
a context of concrete relationships between concrete subjects.22 A rela-
tionship is therefore a relational reality (but not for that reason ontologi-
cally secondary) which exists “between” individuals and their reciprocal
action:
“In general terms, a social relationship should be understood as the

immaterial reality of the interhuman, situated in space and time. It is be-
tween the subjects who act as agents, and as such, it constitutes their
orientation and reciprocal action. It is different to everything that is in
the individual or collective actors, who are regarded as poles or end-
points of the relationship. This ‘in-between reality’, made up of objective
and subjective elements, is the sphere in which both the distance and in-
tegration of individuals with respect to society is defined: it decides
whether, and in what manner, degree and capacity, an individual can dis-
tance or commit himself to other more-or-less proximal subjects and
institutions, and in general, to the dynamics of social life” (Donati, P. Re-
pensar la Sociedad: El enfoque relacional, p. 55).
21 Donati, P. Repensar la Sociedad: El enfoque relacional, p. 155.

22 See Donati, P. (ed.), Lezione di sociologia. Le categorie fondamentali per la comprensione della
società, Padua, Cedam, 1998, p. 363.
According to all the above, understanding society as emergent rela-

tionships does not come down to understanding the inner psychology of
individuals or identifying “mechanical laws” that govern their conduct.
On the contrary, society is a configuration of relationships that implies,
as we have seen in the presentation of the morphogenetic approach of
Archer, three dimensions: “the pre-existing structural socio-cultural
forms; the actions of the subjects, agents constrained by these forms;
and the results of their interactions that actualize (morphostasis) or
modify (morphogenesis) the initial socio-cultural structures”.23
This characterization of the social relationship, which is neither
merely structural nor exclusively psychological, leads Donati to concep-
tualize it using what he calls three semantics: referential, structural and
generative . Given that the social relationship is at the same time a point
of reference, link and emergent reality, a relation “refers, i.e. makes sym-
bolic references; connects or structurally binds, and in being a reciprocal
action (the Italian rel-azione), emerges out of mutual interaction”.24
According to the first meaning or referential semantic, the social rela-
tionship is refero (reference), a significant intentionality by which one re-
fers to something or somebody in accordance with a meaning which is
shared to a greater or lesser degree. It is a good idea to clarify that in this
sense as refero, the social relationship continues to include all the ele-
ments which are implicated in it. Although there may be significant in-
tentionalities, the referential meaning of the concrete relationship does
not depend on any of the subjective intentionalities, taken in isolation or
individually. Once again, the significant reference of the relationship as
such is something more than, and different to, the subjective intentio-
nalities or the subjectively intended meanings, given that the socio-cul-
tural mediation in which the reciprocal relationship is produced also has
a part to play.
According to structural semantics, the relationship is also religo, a tie
or link and reciprocal conditioning, which is at the same time “constrain-
ing and enabling”.25 The specific interdependence created by each type
of relationship links the agents in different ways and in this connection,
23 Donati, P., Repensar la Sociedad: El enfoque relacional, pp. 55-56.

24 Donati, P., Relational Sociology, p. 124.
it generates a series of reciprocal conditioning factors, which should not

be regarded only as limits for reciprocal action but also, as the channel
through which it unfolds.
Finally, the relation-ship also brings with it a third type of semantic,
which we can call “generative” or, simply, emergent, according to which
the different agents and components present in the relationship (means,
norms, values…) interact, producing an effect which cannot be explained
solely by the properties of each of the components or actors that are
brought together but rather, as something which goes beyond them. This
is the emergent reality and causal aspect of the relationship.26
Being emergent realities, social relationships have specific properties
that cannot be reduced to that of their components. They only appear
when there is a relationship and they can influence –and in fact do in-
fluence– the subjects implicated, whose behaviour is orientated by a par-
ticular guiding principle. This guiding principle can be considered as a
specific form of exchange that is established in accordance with certain
values, rules, goals and means which are available to it.27
In other words, the social aspect emerges from the components that
lie in the relationship, yet it is the relationship itself and not the compo-
nents that constitute it. This is why for Donati social reality “does not re-
side either in individuals or structures or a mixture of both. On the con-
trary, it resides in their relationship”.28 It is a relationship that cannot be
reduced to aggregate phenomena but rather, to a reality that is different,
relational and emergent, which only appears as such when this reference
and reciprocal link is established, when the reciprocal action is actualized
between the parties involved. This is why social relationships cannot be
observed or understood solely by paying attention to their components
or the subjects that are implicated: it is necessary to be aware of the rela-
tional reality that emerges from the connection and reciprocal reference
established therein.
Donati understands social structure as a real form of social organiza-
tion which has its own reality and dynamism, given that:
26 See Donati, P., Relational Sociology, p. 88.

27 García Ruiz, P., “Presentación”, in Donati, P., Repensar la Sociedad, p. 21.
28 Donati, P., Repensar la Sociedad: El enfoque relacional, p. 57.
“a) it is configured as an emergent property of relationships that pre-

cede the occupants of status-roles; b) it demonstrates a power of its own
because it supports certain tendencies and potentialities by means of
links, incentives, rewards and sanctions that condition the actions of the
occupants of different positions; c) it depends on relationships with
causal properties that are independent of human action and not simple
abstractions of the repetitive or routine behaviour of agents” (Donati, P.
(ed.), Lezione di sociologia, p. 367).
It is precisely because of this dynamism and contingency of social re-

lationships that it is necessary to adopt a way of observing and con-
ceiving society that is capable of embracing all the important dimensions
of the social and the emergent properties involved.29 This is the aim of
relational methodology, which, in order to study human action in society,
claims that
“to place society on a scientific footing means understanding and ex-

plaining why certain relationships exist between some human subjects
and not others, and also why they act in one way and not another if and
when they relate to one another, and why they relate to one another in a
certain way” (Donati, P., Repensar la Sociedad: El enfoque relacional, p. 55).
3. Conclusion: Complementarities between Critical Realism and

Relational Sociology
The critical realism represented by Archer and the relational sociology

consolidated by Donati are the two main forms of humanistic social
theory today. Their ways of conceptualizing human action are comple-
mentary and there are many more convergences than differences in their
approaches. The most important convergence is their attempt to offer a
sociological explanation of human action in society that is not reduc-
tionist, and they manage to move beyond sterile debates between individ-
ualism and collectivism, between objectivism and subjectivism. As a re-
29 See Donati, P., Teoria relazionale della società, Milan, Franco Angeli, 1991, pp. 314-
323.
sult, they offer a sociological perspective that directly addresses one of

the most basic problems of every social theory, namely the conceptual-
ization of human action so that it can be regarded as effective but with-
out neglecting the constraints it is subjected to by socio-cultural struc-
tures:
“The sociological problem of conceptualizing the person is how to

capture someone who is both partly formed by their sociality, but also
has the capacity to transform their society in some part. The difficulty is
that social theorizing has oscillated between these two extremes. On the
one hand, Enlightenment thought promoted an ‘undersocialized’ view of
man, one whose human constitution owed nothing to society and was
thus a self-sufficient ‘outsider’ who simply operated in a social environ-
ment. On the other hand, there is a later but pervasive ‘oversocialized’
view of man, whose every feature, beyond his biology, is shaped and
moulded by his social context. He thus becomes such a dependent ‘in-
sider’ that he has no capacity to transform his social environment”
(Archer, M. S., “Persons and ultimate concerns: who we are is what we
care about”, in Glendon, M. A. – Malinvaud, E. (eds.), Conceptualization of
the Human Person in Social Sciences, The Pontificial Academy of Social Sci-
ences: Vatican City Press, 2006 p. 261).
Nevertheless, although the underlying problem is the same, the tradi-

tions on which they are based and the referents for the respective lines
of argument of one or other approach are not homogeneous, as Archer
herself recognizes.30 Critical realism seems to be more concerned with
finding the link between structure and human agency in a non-reified
manner, whereas the relational approach has paid more attention to the
way these structures are configured through a predominant and funda-
mental focus on the constitution of the social aspect itself, through the
relationship as its basic reality. Despite these and other differences in the
focal point of their respective research, there are many points of agree-
ment and complementarities that link both perspectives, as shown by the
30 Archer, M. S., “Critical Realism and Relational Sociology: Complementarity and
Sinergy”, Journal of Critical Realism, no. 9, 2 (2010), pp. 200-201.

mutual interest they have manifested and explored in recent years.31

Thus, critical realism has gradually and decisively incorporated the rela-
tional perspective and the methodological characterization of the relation-
ship as presented by Donati in his own studies, and, for its part, relational
sociology resorts more and more explicitly to the concepts and metho-
dological tools proposed by the morphogenetic approach to research
socio-cultural changes and tendencies. In relation to the conceptualiza-
tion of human action in more genuinely subjective terms, there has also
been a rapprochement of the two positions, to the point that even
Donati has recognized that Margaret Archer probably proposes a better
conception of man, from the perspective of social realism:
“This re-conceptualisation grants humankind (i) temporal priority, (ii)

relative autonomy, and (iii) causal efficacy, in relation to the social beings
that they become and the powers of transformative reflection and action
which they bring to their social context, powers that are independent of
social mediation” (Archer, M. S., “Persons and ultimate concerns: who
we are is what we care about”, p. 262).
Also, in relation to the latest developments concerning the question

of reflexivity, we can see this convergence and mutual interest: for
Donati “reflexivity is synonymous with internal conversation, and there-
fore pertains to human persons and not to social networks and organiza-
tions. It can, however, be extended to social groups, in so far as they can
express a collective mode of reflexivity”.32 Conversely, in reference to
her latest work, The Reflexive Imperative,33 Archer recognizes that “the syn-
ergy works both ways, and I am currently casting the theory of the
making and breaking of modes of reflexivity in relational terms”.34
31 See Archer, M. S., Making Our Way Through the World: Human Reflexivity and Social
Mobility; Donati, P., “Il ruolo della riflessività nell’agire sociale: quale ‘modernizzazione ri-
flessiva?’”, preface in Archer, M. S. (ed.), Riflessività umana e percorsi di vita, Trento,
Erickson, 2009, and also Donati, P., Relational Sociology, pp. 97-119.
33 Archer, M. S., The Reflexive Imperative in Late Modernity, Cambridge, Cambridge Uni-
versity Press, 2012.

34 Archer “Critical Realism and Relational Sociology: Complementarity and Sinergy”,
p. 206.
In conclusion, the complementarity and convergence in terms of

subject matter and methodology of critical realism and relational socio-
logy mean they approach human action in society in a non-reductionist
manner. From these sociological perspectives human action is conceived
as an action in-situ, a reciprocal action that to be understood must take
into consideration the significant references it includes, the interpersonal
or social links it generates, and the emergent reality which in itself con-
stitutes the relationship as such. This is how, through the mediation of
reflexivity and internal conversation, agency and socio-cultural structures
can join forces and not only lead the way to a sociologically fruitful un-
derstanding of human action but also, to cultural change and social dy-
namics.
University of Navarra
Pamplona, Spain
References
Archer, M. S., Realist Social Theory: The Morphogenetic Approach, Cambridge,

——, Culture and Agency, revised ed., Cambridge, Cambridge University
Press, 1996.
——, Being Human: The Problem of Agency, Cambridge, Cambridge
——, Structure, Agency and the Internal Conversation, Cambridge, Cambridge
——, “Persons and ultimate concerns: who we are is what we care
about”, in Glendon, M. A. – Malinvaud, E. (eds.), Conceptualization of
the Human Person in Social Sciences, The Pontificial Academy of Social
Sciences: Vatican City Press, 2006, pp. 261-283.
——, Making Our Way Through the World: Human Reflexivity and Social
Mobility, Cambridge, Cambridge University Press, 2007.
——, Conversations About Reflexivity, London, Routledge, 2010.
——, “Critical Realism and Relational Sociology: Complementarity and
Sinergy”, Journal of Critical Realism, no. 9, 2 (2010), pp. 199-207.
——, The Reflexive Imperative in Late Modernity, Cambridge, Cambridge
Aristotle, The complete works of Aristotle: The revised Oxford translation, ed. by
J. Barnes, 2 vols., Princeton, Princeton University Press, 1984.
Bhaskar, R., The Possibility of Naturalism, third ed., London, Routledge,
1998.
——, A Realist Theory of Science, London, Routledge, 2008.
——, Reclaiming Reality: A Critical Introduction to Contemporary Philosophy,
——, Critical Realism: A Brief Introduction, London, Routledge, 2013.
Chernilo, D., “Introducción”, in Archer, M. S., Teoría social realista: el

enfoque morfogenético, Santiago de Chile, Universidad Alberto Hurtado,
2009, pp. 11-20.
Donati, P., Teoria relazionale della società, Milan, Franco Angeli, 1991.
—— (ed.), Lezione di sociologia. Le categorie fondamentali per la comprensione
della società, Padua, Cedam, 1998.
——, Repensar la Sociedad: El enfoque relacional, Madrid, EIUNSA, 2006.
——, “Il ruolo della riflessività nell’agire sociale: quale ‘modernizzazione
riflessiva?’”, preface in Archer, M. S. (ed.), Riflessività umana e percorsi di
vita, Trento, Erickson, 2009.
——, Relational Sociology: A New Paradigm for the Social Sciences, London,
Routledge, 2011.
García Ruiz, P., “Presentación”, in Donati, P., Repensar la Sociedad,
Madrid, EIUNSA, 2006, pp. 9-47.
Lewis, P. A., “Metaphor and Critical Realism”, Review of Social Economy
no. 54 (1996), pp 487-506.
LEVERAGE AND TRUTH
SOPHIE DJIGO
How is it that a person might have a reason to do something and not do

it? In recent years, this old question has received a lot of attention from
contemporary philosophy of action and ethics. I think that two challeng-
ing significant interpretations of this problem can be drawn from the
various contributions of philosophers: the first one focuses on leverage
as the evidence for having a reason. Having a reason is grounded in moti-
vation and a lack of motivation means that the agent has purely and sim-
ply no reason to do what he was supposed to want. This view corre-
sponds to a certain version of reasons internalism offered by philoso-
phers like Bernard Williams. A second interpretation opposes the idea
that leverage would be the final criterion of the truth of a reason state-
ment. Having a reason would rather appeal to normative claims which
are independent of the agent’s motives. For the agent to have a reason
means that he should do something, and this, because it is a moral truth.
This second position has been supported by philosophers like Derek
Parfit.1
In this paper, I shall discuss the articulation of both truth and lever-
age and try to find an explanation of the discrepancy between them. Sec-
tion 1 examines the story of Augustine’s conversion and the difficulty of
resolution. It seems that if he does not change his life while he claimed
he would, it is because he is not really motivated to do it, and this throws
suspicion on will. The “half-wounded will” is the expression of the a-
gent’s questioning his reason for acting, of his skeptical attitude towards
his reason for conversion. Such a doubt might weaken this reason to the
point of dissolving it. Then, in section 2, I will focus on reasons interna-
lism, a philosophical view about the truth of reason statements and how
the agent appropriates a reason to make it his own. I will be also discus-
sing the role and limits of deliberation in this process of appropriation
and the type of rationality engaged in practical reasoning. Moral truth
1 Parfit, D., Reasons and Persons, Oxford, Clarendon, 1984.

176 Sophie Djigo
plays an important role in practical reasoning since the truth of a reason

statement is independent of the agent and gives him a reason to act. In
the last two sections, my concern will be to consider the agent’s relation
to moral truth (hence to the truth of a reason statement) and the pos-
sibility of truth-blindness. There is a crucial distinction to be made be-
tween the theoretical perspective, from which the agent ascribes a reason
to him and considers his having a reason as a psychological fact, and the
practical perspective, from which the agent avows the truth of the rea-
son statement and assumes the responsibility for carrying out his resolu-
tion, that is, for leverage.
I am not concerned here with deliberative questions about the more
convenient or economic way to cook tonight, or facts which would give
me a “reason” to order a steak rare instead of medium rare. Natural or
technical questions are outside my field of reflection. Therefore, my pur-
pose will be limited to voluntary actions involving moral values, whose con-
sequences are crucial to human life. I wish to defend a tight link between
theory of action and morality which is grounded in the fact that moral
truths give us reasons for action.
1. The half-wounded will
In his Confessions, Augustine tells the story of his conversion and decision
to devote his life to God. In a couple of striking chapters, he wonders
how one might want to do something (living a Christian life) while not
doing it and he considers such a phenomenon as “monstrous”.2 Indeed,
that sounds paradoxical: if I want to do Φ, I am supposed to do Φ if
there is no obstacle to my action. Of course, we need to distinguish two
types of cases: when a person’s will is stifled by exterior factors, which
are independent from him, like other persons’ wills, natural phenomena
or facts, and when his will has a totally free hand. Augustine’s perplexity
is about the latter, when the free will does not meet any resistance except
from its own. Augustine, like most of ourselves, endorses the common
thought that “if there’s a will, there’s a way”. In this view, will is the con-
2 Augustine, Confessions, VIII, 9, tr. by E. B. Pusey.

Leverage and Truth 177
dition of human action, so that action should naturally ensue from a vol-
untary decision.
We have to admit that this is not the case and Augustine is the first
one to experience the painful hiatus between will and action. We can ex-
press this first problem as such:
(A) There is a gap between motivation (will) and action.

This statement can be analyzed in two different ways:
(A1) The agent has the motivation to do Φ and he does not do Φ.
(A2) The agent has a reason to do Φ and he does not have the corre-
sponding motivation.
Both cases have the same practical result: the agent does not do Φ.
Let’s begin with the first statement (A1), obviously problematic since
we usually tend to juxtapose motivation and action to the point that it is
inconsistent to be motivated to do Φ while not doing it. The only way to
remove this inconsistency is to question motivation: if the agent does
not do Φ, it is because he does not really want it. This is how Augustine
suggests to construe the “monstrous thing” of a will without a corre-
sponding action. Since the causal link between motivation and action
can’t be questioned, the problem must lie in the motivation itself. A mo-
tivation from which no action ensues is a partial and weak one.
Augustine conceptualizes this difficulty by the distinction of two kinds
of will: the entire will and the half-wounded will. These two kinds are
neither coexisting in one’s mind nor clashing; the argument is that for
him to be moved, the agent needs a one and only will. As soon as a sec-
ond will appears, a conflict emerges between the agent’s motivations
which helps him from acting according to one of them. So, the very dif-
ficulty is not the balance of power between a partial will and an entire
one, but the heterogeneity or homogeneity of one’s will.
This is very close to the Kantian conception of the autonomy / het-
eronomy of the will, except that in Augustine’s context, the entire will is
also led by a divine power. But the question of being entire or fragment-
ed appears in both philosophers as a means to explain the possibility of
internal disagreement. A part of Augustine is motivated to be converted
and another part of him prefers to go on in his own past way. Like a cer-
tain amount of energy, will becomes weaker if it is divided. The causal
178 Sophie Djigo
link between motivation and action needs a strong will to be guaranteed,

that is, an entire and undivided will. In the case of Augustine, his motiva-
tion of being converted means a great change in his life and his internal
conflict leaves him in a state of stagnation. To a certain extent, he leads
his usual life, thus following his second motivation. So, we can’t say that
Augustine does not act at all because of his clashing wishes; but he wants
something passionately without being able to do it.
The gap between motivation and action is not due to the weakness of
the will, but to the multiplicity of wills, which reveals the fragmented
psychology of the agent. Augustine is torn between two very different
motivations: a spiritual aspiration and habits.
“The mind commands the mind to will, and yet, though it be itself, it
obeys not. Whence this monstrous thing? And why is it? I repeat, it com-
mands itself to will, and would not give the command unless it willed;
yet is not that done which it commands. But it wills not entirely; there-
fore it commands not entirely. For so far forth it commands, as it wills;
and so far forth is the thing commanded not done, as it wills not. For the
will commands that there be a will; –not another, but itself. But it does
not command entirely, therefore that is not which it commands. For were
it entire, it would not even command it to be, because it would already
be. It is, therefore, no monstrous thing partly to will, partly to be unwil-
ling, but an infirmity of the mind, that it does not wholly rise, sustained
by truth, pressed down by custom. And so there are two wills, because
one of them is not entire; and the one is supplied with what the other
needs” (Augustine, Confessions, VIII, 9).
According to this passage, will is clearly suspected: if I don’t do Φ, it

is because I don’t really want it. The problem is to understand what it
means to really want something? Augustine’s analysis is an attempt to deal
with this question: a real motivation means an entire and unique will.
Will can’t be thought as an isolated attitude, it has to be replaced in the
general context of the person’s various desires. Indeed, for the agent to
act, his will need not to be strong: Augustine insists on his “great will”
and his “unequalled desire” for conversion which are not sufficient to
push him to change. Even if a motivation is passionate, its power de-
pends on the number of other wishes and motivations and is shared
between them.3 As far as the agent has other competing motivations, the
action cannot be done. However, if we follow Augustine’s line of
thought, we should conclude that among the half-wounded wills, none
of them is able to motivate the action. Then, Augustine should be para-
lyzed, his action being suspended. As we know it, this is not the case. For
instance, Augustine compares his own internal fight with a person delib-
erating if he should go to the theater or to the church. If the will, even
passionate, to go the church is clashing with the will to go to the theater,
the result of this conflict should be that this person neither goes to the
theater nor to the church. But he might also go to the theater, even if his
will to go to the church is very strong. Similarly, Augustine is not in a sit-
uation where action would be suspended, he rather continues to lead his
past life, that is, to live according to his old habits, instead of living the
Christian life he aspires to.
This is a striking point: obviously, the two conflicting wills are not e-
qual and the agent finally obeys to one of them. However, we must
imagine that while going to the theater, Augustine is full of remorse, de-
spair and self-contempt. Even if he is acting voluntarily, he is not satis-
fied because something is spoiling his enjoyment. That is what we mean
when we speak of a mixed pleasure or decision. On the one hand, the
person enjoys going to the theater and we can suppose that he indulges
himself. On the other hand, he thinks he should rather have gone to the
church, because he believes in God and thinks that it is a good thing to
do. The consequence of a divided will is not the blocking of action, but
“grievous perplexities”. On the contrary, when the agent is determined
to act and goes to the theater without feeling any hesitation or remorse,
his resolution entails happiness.
To sum up these first elements, we can say that for Augustine, the
problem of not doing Φ while wanting to do Φ depends on a half-
wounded will, which means that the agent does not really want Φ be-
cause his will is mixed by other considerations and when he deliberates,
these other considerations command conflicting wills, one of them
having the upper hand even if his will to do Φ is strong and passionate.
3 On this issue, see also Jon Elster and his understanding of the weakness of will in
terms of a reversal of the agent’s temporary preferences, in Elster, J., Agir contre soi, Paris,
Odile Jacob, 2007.
180 Sophie Djigo
But for the winning will, the price to be paid is hesitation, regrets, grief.
In this view, if I don’t do Φ, it means that I don’t want it entirely. The
necessary and sufficient condition for acting appears to be an entire will.
Then, (A1) would be construed as follows:
(A1') The agent has the half-wounded motivation to do Φ and he
does not do Φ.
So, if we take motivation as an entire, unique will, we shall admit that
to be entirely motivated to do Φ is to do Φ (“to will was to do”4). If I
don’t do Φ, it means that I am not entirely motivated to do it. Thus, an
entire will is not something I can assert, since I could claim that I am
completely motivated to go to the church without going. An entire will
shows in action and resolution (absence of regrets). I cannot say hon-
estly that “I entirely want to go” while not doing it. My action is a neces-
sary consequence of my will. Then, if I am free (no external obstacles)
and if my will is entire,
(A3) Being motivated to do Φ is doing Φ.
Augustine adopts a psychological point of view in his analysis of will
and he refrains from moral considerations in most of his case. Indeed,
he comments as well on a crook deliberating if he should kill a man by
poison or by the sword as on a person deliberating if he should read the
apostle or the psalm. He is making a case to show the plurality of clash-
ing wills, whatever their morality. However, the point of his argumenta-
tion is to explain why he does not manage to convert. Here, the problem
is no more the half-wounded will than the gap between reason and mo-
tivation. Augustine’s problem is not his failure to act according to his
motivation, but his failure to want entirely what he should do. How is it
that some of our wills are so partial, even if we are passionate? It would
be easier to understand that one is not entirely motivated to jump into a
frozen river, but it is quite difficult to explain why the agent is not en-
tirely motivated to do what he should do, that is, what he has a reason to
do.
Now, we must consider (A2):
4 Augustine, Confessions, IX, 8.

(A2) The agent has a reason to do Φ and he does not have the corre-
sponding motivation.
In the same chapter, Augustine’s account of his tormented mind and
partial will shows that the problem is no more that he has clashing wills,
but conflicting reasons to want such or such a thing. Part of him wants
to be converted because his mind has been “sustained by truth” while
another part wants to lead a life of sin because of custom. Being at war
with himself, Augustine is torn between these two reasons: truth or cus-
tom. Since he has not changed his life yet, we can admit that custom
constitutes an explanatory reason for his behavior. But his torment
comes from his considering that he should be converted, since now, it ap-
pears to him to be true that it is better to live a Christian life. However,
knowing the truth is not enough to provide him the corresponding mo-
tivation.
Furthermore, we can distinguish two moments in Augustine’s life: in
his past life, he describes himself as blind, unable to see the truth and
following only worldly hopes. Then comes the turning point when he ad-
mits the truth:
“Where are you, O my tongue? You said, verily, that for an uncertain
truth you were not willing to cast off the baggage of vanity. Behold, now
it is certain, and yet does that burden still oppress you; whereas they who
neither have so worn themselves out with searching after it, nor yet have
spent ten years and more in thinking thereon, have had their shoulders
unburdened, and gotten wings to fly away” (Augustine, Confessions, VIII,
7, my underlining).
It is striking that Augustine refers to his motivation as something

linked to truth. He believes the truth of the statement that he has a rea-
son to convert and this truth is not up to him, it is something objective,
even metaphysical in his own case. From the moment he sees and admits
the truth, he cannot live as before and moreover, his past life appears to
be unbearable. Recognizing the truth is the condition of a new motiva-
tion to change his life. It is only because of this revelation of truth that
Augustine decides to live another life and that he feels compelled to
change. But the conversion is not immediate, since the new motivation
clashes with the old one. Hence his grief: he knows that he should
182 Sophie Djigo
change, and he is ashamed of his own reluctance to this new life. He is

deeply dissatisfied with what he is doing while in the same time, he is un-
able to do what he should. What he should do is something difficult, that
is why he is delaying it. Neither can he turn back now that he has accept-
ed the truth. He is even confessing that he has always known the truth
but could not admit it: “unwilling to exercise self-scrutiny”.
From this confession can be drawn at least two remarks. First, the
knowledge of truth may indicate what should be done. If, as Augustine
believes it, God exists, then, it is true he should change his life and live
according to his faith. The statement about reason implies distinct levels
of truth: (i) the metaphysical truth on which the reason for conversion
relies (normative claim) and (ii) the truth of the statement that Augustine
has a reason to change his life. When it is about truth, a second element
immediately follows: belief. Belief is then supposed to play the role of a
psychological link between truth and action. It is because I believe that
the house is on fire (that God exists) that I jump into the canal (that I
live a Christian life). Belief is what motivates the agent to do what he
thinks he should do. But what if the agent has a false belief ? What if
Augustine, who gambles on God’s existence, is mistaken? If I want truth
to constitute a reason for me to act, then, I am supposed to believe that
it gives me such a reason. My motivation depends on my belief and the
rightness of my action depends on the truth of my belief. In that sense,
if my belief is false, the reason for my action will be an explanatory one
and if it is true, the reason will also be a justifying one.
Secondly, knowing the truth is not a sufficient condition for motiva-
tion. What if I consider that truth does not give me a reason for acting?
This leads us to the problem of indifference or resistance to truth. This
is clearly the problem that Augustine encounters. He thinks that it is true
that he should change his life and he avows that he has always known the
truth even if he winked at it. It means that the statement that he should
change his life is true whatever his motivation. Its truth does not depend
on its leverage on Augustine. On the contrary, Augustine thinks that he is
in the wrong because of his lack of motivation. He fails in doing what
he thinks he should do and this failure does not question the truth of the
reason statement. Furthermore, he confesses that he deliberately ignored
that he should change and that he resisted to the statement that he knew
was true. This confession poses a serious problem: how to explain that
someone believes that he should do Φ while not being motivated to do

it?
The controversy opposing reasons internalism and externalism fo-
cuses on the truth of the statement about having a reason or not. And
the question is to determine whether this truth depends or not on the a-
gent’s belief. Thus, according to Bernard Williams’ internalist view, for it
to be true that the agent has a reason to do Φ, the agent must be motiva-
ted to do Φ after a deliberation based on his motivational set. If doing Φ
might contribute to the realization of one of his desires or general moti-
vations, then he has a reason to do it. Such an internal reason is related
to the agent. If we try to explain Augustine’s example in the light of in-
ternalism, we shall say that the young Augustine did not have a reason to
change his life because the desire for spiritual happiness was absent from
his motivational set. It is only from the moment he saw this type of hap-
piness as valuable and preferred it to the pleasures of the body that he
had a reason to change his life. Having a reason is the consequence of a
modification in the motivational set.
However, this shift in Augustine’s motivations is not the result of a
deliberation on the basis of preexisting motivations. This is a movement
of the heart, a modification of his vision, not only the result of a ra-
tional procedure. That’s why John McDowell uses the case of conversion
to defend an externalist approach. As he says,
“The idea of conversion would function here as the idea of an intelli-

gible shift in motivational orientation that is exactly not effected by in-
ducing a person to discover, by practical reasoning controlled by existing
motivations, some internal reasons that he did not previously realize he
had. But if its upshot is a case of considering matters aright, why should
such a process not count as someone’s being made aware of some ex-
ternal reasons, reasons that he had all along, for acting in the relevant
ways?” (McDowell, J. H., “Might There Be External Reasons?”, in
McDowell, J., Mind, Value and Reality, Cambridge, Mass., Harvard Univer-
sity Press, 1998, p. 102).
Augustine’s conversion is neither motivated by a deliberation nor by a

new external reason. What happened in his case? According to Augus-
tine, his conversion is due to the dissolution of other wills and the turn
184 Sophie Djigo
to an entire will. This seems to be the resolution of a conflict but this

very conflict is based on the resistance to truth. As he said, Augustine
does not discover a new reason since he has always known he should
convert, he rather becomes sensitive to this reason: he finally gives it
enough importance to count for him as a reason. McDowell says that
conversion has something to do with the capacity to change one’s way of
considering some matter, of becoming aware of an external reason. If
we want to clarify what is at stake here, and in the internalist / externalist
argument, we need a clearer view of the meaning of having a reason and
of the question whether having a reason is the same thing as endorsing a
reason.
2. Reasons and motivations
In the previous section, I tried to sketch a picture of Augustine’s expe-

rience of a gap between his reasons for conversion and his motivation to
do it. Although he believes he should change his life, he needs a lot of
time to move from his partial will to resolution and action. The problem
of having a reason for action without the corresponding motivation is a
classical one, but it has been revived by Bernard Williams in his paper
Internal and external reasons,5 where he defends what he calls an internalist
position. The difference between reasons internalism and externalism is
above all a difference in the meaning of “having a reason”. According to
internalism, an agent A has an internal reason to do Φ iff a rational de-
liberation and his knowledge of the relevant facts would show him that
doing Φ might help fulfill one or several elements from his motivational
set. On the contrary, the externalist interpretation considers that A has a
reason to do Φ whatever his motivational set. Williams’ theory asserts
that in the external sense, it is unclear what it means to “have a reason”
and he challenges the externalist to clarify this meaning, otherwise the
agent’s reasons would be reduced to internal reasons. Reasons inter-
nalism is not a descriptive theory about the nature of psychological mo-
tives. It is rather a normative claim which offers to determine the agent’s
5 Williams, B., “Internal and external reasons”, in Williams, B., Moral Luck,
Cambridge, Cambridge University Press, 1981, pp. 101-113.

reasons to do Φ from his motivations under certain conditions of ration-

ality.
I won’t reproduce the whole argument, since it is well-known and has
been so much written about. I’ll focus on a few aspects which I think are
crucial in both internalist and externalist views. Williams is interested in
the way we come to have a reason and his main argument is that having a
reason is the result of a deliberative process about how to achieve desires
and aims belonging to one’s motivational set. A motivational element al-
ways precedes the deliberation, so that there cannot be a gap between
reason and motivation: having a reason means being motivated by it
since this very reason aims at doing something which is supposed to ful-
fill my desire. If we think however that there is such a gap, that is be-
cause we use the expression “having a reason” in an externalist interpre-
tation, which is misleading. Thus, Williams writes:
“If an agent really is uninterested in pursuing what he needs; and this

is not the product of false belief; and he could not reach any such mo-
tive from motives he has by the kind of deliberative processes we have
discussed; then I think we do have to say that in the internal sense he in-
deed has no reason to pursue these things” (Williams, B., “Internal and
external reasons”, p. 105).
Here the accent falls on leverage. In fact, Williams’ reasons internal-

ism can be read as a criticism against a range of moral theories which
take to be obvious that moral considerations should be motivating. This
optimistic view is the target of Williams’ argumentation and that sheds a
light on the importance he attaches to leverage as a condition on which a
proposition may count as a reason. This might be called an explanatory
constraint: for it to be true that A has a reason R to Φ, R must explain
that A would do Φ in certain conditions. The fact that R is related to the
agent’s motivational set explains why he is doing Φ: because a rational
deliberation is supposed to have shown him that doing Φ would help
fulfill his desires. In this schema, the agent is first of all endowed with a
set of desires, then, he uses a rational procedure of deliberation in order
to fulfill these existing desires, after which we can say that he is mo-
tivated to do Φ or that he has a reason to do it, as far as it is a means to
reach his aims. Put in that way, we could think that an internal reason is
186 Sophie Djigo
nothing more than the cause of action. But Williams wants a normative
conception of reasons, according to which the agent would also be justi-
fied to follow this reason. Indeed, the agent has a reason to do Φ only if
he would be likely to do it after a rational deliberation and his knowledge
of the relevant facts. Internal reasons are thus normative since they cor-
respond to motivations ensuing from a rational procedure.
In other words, a moral consideration can’t be considered to give a
reason for acting unless the agent could be motivated to act because he
has become aware that this action might help fulfill his desires. The force
of Williams’ argument is to reject the prejudice in favor of an automatic
link between morality and practical power. And of course, it is more real-
istic to say that morality has not the compelling force some philosophers
might sometimes attribute to it. In this perspective, “having a reason” is
construed as having leverage, being influential. The internalist view
stands against a definition of reasons in terms of moral considerations
(what would be better to do / what should I do) and offers a practical
definition of what a reason is (a motivating consideration). The rejected
definition could be that of externalism, so that Williams says:
“What is the difference supposed to be between saying that the agent

has a reason to act more considerately, and saying one of the many other
things we can say to people whose behaviour does not accord with what
we think it should be? As, for instance, that it would be better if they
acted otherwise?” (Williams, B., “Replies”, in Altham, J. E. J. – Harrison,
R. (eds.), World, Mind, and Ethics: Essays on the ethical philosophy of Bernard
Williams, Cambridge, Cambridge University Press, 1995, p. 191).
Williams’ point is that it is superfluous to use the word “reason” to

designate a moral consideration. Not only is it useless, but it is also dark-
ening our understanding of what it means to have a reason. Therefore
we need a distinction between the following two kinds of statements: the
reasons statement that “the agent has a reason to do Φ” and the moral
statement that “the agent should do Φ”.
The internalist view is relevant to the questioning of the link between
morality and leverage, insofar as we are ready to admit the lack of lever-
age of many moral statements. Indeed, it might be very helpful to con-
front morality with a practical requirement, or to understand phenomena
like weak will and irresolution or how we delay things we ought to do.
Furthermore, leverage is also a requirement which helps confounding an
hypocritical agent or understanding cases of bad faith. If we take the
case of a man who maltreats his wife, telling him that he has a reason to
treat her better might not have any leverage, even if he says that he
agrees that it is better to treat her good. He can be perfectly aware of the
moral statement while not obeying it. According to Williams, he has no
(internal) reason to treat his wife better, because he thinks he wouldn’t
fulfill one of his desires if he were. Hence, we should say that he has no
reason to treat his wife better even if it would be moral to do it, even if
he ought to do it.
The psychological link between reason and motivation is belief, a link
the externalist is accused to be lacking. That the agent believes he has a
reason is a condition for being motivated to act. Such beliefs may be true
or false and if the agent’s motivation involves a false belief, then, we can-
not say that he has an internal reason to do Φ, since such a reason must
be provided by a rational deliberation, which might rectify false beliefs.
But in the case of the man who maltreats his wife, Williams precises that
his behavior is not altered by a rational deliberation. This sounds para-
doxical. How can a man want to maltreat his wife if he is perfectly
rational? The problem of Williams’ argumentation appears when we put
his example in the negative form. Instead of saying that A has no reason
to treat his wife better although it would be better (moral statement), we
may say that A has a reason to maltreat his wife although it is completely
immoral. This reversal shows a striking asymmetry. Does it ensue from
this man having no reason to treat his wife better that he has a reason to
maltreat her? On the one hand, the internalist criticism was helpful to
point the distance between what the agent should do and what he actu-
ally does. It is not because the agent ought to do Φ that he has a reason
to do it (morality’s lack of leverage). But on the other hand, internalism
leads to the conclusion that A has a reason to do wrong things, to act
with great immorality.
Let’s analyze another of Williams’ example, to see how he deals with
this unacceptable consequence of the internalist definition of a reason:
the case of a person who is not motivated to do what he needs. Needs
should be part of our existing motivational set. But Williams imagines a
person with a lack of interest which is neither due to a wrong deliber-
188 Sophie Djigo
ation, nor to a false belief. We must admit that a person might be

unmotivated to pursue what he needs. Williams describes this case as not
having an (internal) reason at all. He offers the example of a person who
does not take medicine he needs because of a fundamental lack of inter-
est in preserving his health. We could think of this borderline case as
bringing elements in favor of the externalist view, according to which
this person has a reason to take medicine even if he does not (want to)
take it. Williams’ argumentation against it is that the internal interpre-
tation is more intuitive and also that we simply don’t need it. Indeed, we
might still say that this person has an internal reason to take medicine
since “at some level he must want to be well”.6
We have then two possible statements: (1) the person has no reason
to take medicine he needs since he is totally deprived of any interest in
his health or (2) the person has a reason to take medicine in the sense
that he must take care of his health at some level. This last complement
shows how improbable it is that this person does not want to be well.
These assumption relies on the fact that it is part of the human motiva-
tional set to want to be healthy and it is one of our most basic needs, so
it would be very strange and unusual to refuse “rationally” to protect
one’s health. The difficulty resides in the “level” at which he must want
to be well. We can understand that level in at least three ways: in the first
one, Williams could have in mind a natural level, which refers to our
needs as living creatures and which is the most basic one compared to
more cultural, social and intellectual levels. In the second way, this could
be an unconscious level, by contrast with conscious wills. In the third
way, we could imagine a scale with various degrees of wills, some of
them being stronger, the others, weakened by the Augustinian principle
of being supplied with what the other needs. The desire to be well would
be very weak in this case, and his leverage would be countered by other
stronger wills, such as an immediate pleasure incompatible with his treat-
ment or the cost of the medicine.
What is the difference with the externalist view? Williams’ strategy is
not so much to prove externalism to be false as to show its being super-
fluous. We were wondering what difference externalism would make. I
6 Williams, B., “Internal and external reasons”, p. 106.

think that there is an important difference between both views which can
be explained by the interpretation of “must” in Williams’ previous state-
ment. In fact, in the internal sense, “must” means that it would be highly
probable and it has an empirical meaning. The existence of such a moti-
vation concerning one’s health would be probable given our knowledge
of human needs. In the external sense, “must” has a completely different
meaning, much closer to the modal “should”. It is rather a normative
meaning, saying that for a person to be normally rational, he must /
should take care of his health. In the empirical perspective, we make an
assumption based on our knowledge of human usual motivations, and
this assumption might be false. There might be human persons lacking
interest in their health even if it is very improbable. Were it to happen,
the only thing we could say would be to note the absence of motivation,
and hence, of the corresponding reason. But in the normative perspec-
tive, we could say that this person has still a reason to take medicine,
even if he does not want to be well. And we shall add to this first state-
ment that he also should care about his health.
In the internalist view, the criterion of having a reason is potential le-
verage: I have a reason to act if it might motivate me to act (in the ade-
quate conditions, if I don’t have a stronger reason not to act, etc.). Lever-
age is the expression of my interest in the action which I conceive as a
means to get closer to my aims and ends. Then we saw that leverage was
surrounded by certain conditions of rationality, such as knowledge of
the relevant facts, rational deliberation and consequently, rational motiva-
tions ensuing from the previous deliberation. If we want to understand
how it is possible for the internalist to assert that the man who maltreats
his wife has a reason to act as such or that the person who refuses to
take medicine he needs has no reason to do it, we need a clearer appre-
ciation of the internalist conception of rationality.
Among the various kinds of internalism and externalism, Derek
Parfit spots the fault line that divides them. Each views has a completely
different conception of rationality and of the conditions of rationality
applying to reason statements. As he writes:
“According to these Externalists, if

(R) we have a reason to do something,
that entails that
190 Sophie Djigo
(E) if we knew the relevant facts, and were fully substantively ratio-
nal, we would be motivated to do this thing.
To be substantively rational, we must care about certain things, such
as our own well-being. If Williams’s imagined person were fully rational,
these Externalists would claim, he would be motivated to take the medi-
cine that he knows he needs. That could be true even if, because he is
not fully substantively rational, no amount of informed deliberation
would in fact motivate him.
Internalists hold a different view. On their view, more fully stated, for
it to be true that
(R) we have a reason to do something,
it must be true that
(M) if we knew the relevant facts, and deliberated in a way that was
procedurally rational, we would be motivated to do this thing.
To be procedurally rational, we must deliberate in certain ways, but
we are not required to have any particular desires or aims, such as con-
cern about our own well-being. If Internalists allowed such further re-
quirements, then, as Williams writes, ‘there would be no significant dif-
ference between the internalist and externalist accounts’, since Internal-
ism would allow ‘anything the externalist could want’” (Parfit, D. –
Broome, J., “Reasons and Motivation”, Proceedings of the Aristotelian Society,
Supplementary Volumes, vol. 71 (1997), p. 101).
We already said that for Williams, an internal reason cannot be re-

duced to an explanation of the action, but is also concerned with justifi-
cation, that is, with the agent’s rationality. If Parfit is right to attribute a
“procedural rationality” to internalism, we need to go further with the
very concept of deliberation. Indeed, we can suppose that Williams’ con-
cept of deliberation is very close to Aristotle’s practical syllogism. The
specificity of practical deliberation is that its conclusion is not a belief
but an action, a result which is difficult to conceive of given its propo-
sitional premises. Aristotle underlines this difficulty, since his account of
deliberation comprises two apparently incompatible aspects: intelligence
and the appetitive aspect. In deliberation, these contradictory functions
of the soul are mixed to work together. The point is that for a deliber-
ation to be possible, at least one premise must be a desire which will
move the agent. The agent has a desire D and according to his reasoning,
doing Φ appears to be the best way of satisfying D. If the deliberation

was not an inference from such a desire, the reason statement provided
by intelligence could not be directed towards a practical efficiency.
This model of deliberation has nonetheless a blind spot: desires
(what Williams calls the agent’s motivational set) are not provided by the
deliberation itself. Deliberation presupposes these desires, which are all
the simplest given. According to Aristotle, the agent does not deliberate
on aims or on what is desirable. Thinking of what is desirable is outside
the field of deliberation. I think that Williams’s conception of deliber-
ation is similar to Aristotle’s definition of practical deliberation and con-
sequently, if Williams wants to avoid implications in which an agent has
an internal reason for acting wrongly, he has to endorse to a certain ex-
tent an Aristotelian normative conception of human nature, claiming
that the rational agent’s motivational set is composed of human needs
and aims. This would help understand why the person who does not take
medicine which he needs “must at some level” want to be well. But if
the internalist admits this idea of human nature and of correlated de-
sires, what would be his difference with externalism?
The motivational set described by Williams entails desires as well as
dispositions of evaluations, patterns of emotional reaction, personal loy-
alties, projects, a lot of things he calls “commitments of the agent”.7 His
description is more refined, in particular because he refuses to think of it
as a “statically given” or as a set of egoistic desires. However, in the in-
ternalist view, there is no constraints on the agent’s aims and values. An
element of the agent’s motivational set will give a reason for doing Φ
only if the agent’s belief in the relevance of doing Φ to satisfy this ele-
ment is true. This proposition does not take into account the relevance
and the truth of this very motivational element. If we think of the previ-
ous example of the person who refuses to take medicine he needs, in the
internal sense, he has no reason to take medicine he needs in order to be
well since he does not aim at being well.
At this point, the externalist view may have something to add.
Indeed, in Williams, such aims (being well) must be part of our motiva-

192 Sophie Djigo
tional set, but they might happen to be lacking. Considering this, we may
only notice the presence or the lack of interest in health in the agent’s
motivational set. In the final analysis, saying that the agent has a reason
to Φ means that he would be motivated after informed and procedurally
rational deliberation, that is, he would be willing to take his medicine if
he aimed at being well and believed that he needs to take some medicine
to protect his health. Taking the medicine is what appears to be the best
means, after deliberation, to satisfy the desire to be well. If a person is
uninterested in being well (and this is not subject to a deliberation), then,
he has no reason to take the medicine. Moral considerations will not
have any leverage on him if he does not already have the will to be
healthy. Not only does the internalist interpretation appeal to aims which
are outside the field of deliberation and then, of its definition of rationa-
lity, but it also implies a purely psychological fact: motivation as an empi-
rical and contingent fact.
According to Parfit’s externalist interpretation, saying that the agent
has a reason is not an empirical inference, but a normative claim, that the
agent should be motivated to Φ. In this perspective, despite the fact that
our agent is not motivated to protect his health, we might say that he has
a reason to do it, and by this, we mean that he should be motivated to do
it if he were substantially rational (if he cared about his health). When
the externalist makes such an external reason statement, he appeals to a
normative fact. Of course, internal reasons correspond to psychological
facts which may have a normative significance, since they correspond to
the model of the rational agent, well informed and after a perfectly cor-
rect deliberation. There are norms of what an agent would do after an
informed and rational deliberation. But they are not normative facts, as
Parfit uses this term. In Parfit’s sense, a normative fact is for instance the
fact that I give a normative significance to psychological facts. In our
previous example, the psychological fact would be:
the fact that the agent is not motivated to take medicine he needs
after rationally deliberating;
and the normative fact would be:
the fact that believing the truth that he needs medicine to protect his
health gives him a reason to take it.
That’s why the most important criterion for having an external reason
is not leverage but truth: saying that the man has a reason to treat his
wife better means that it is a moral truth that it is better not to maltreat
his wife and that believing this truth gives him a reason to treat his wife
better. But this kind of truth does not depend on the agent, they are
normative truths, and even, moral truths. Among such truths, we can
find this one: “It is true that it is worth being well”. The person refusing
to take medicine he needs would not only be said to have a reason to
take it, whatever his motivational state, but he would also be said to be
irrational if he persists in his refusal. In the externalist perspective, one
can have a reason for acting even if one is irrational or unmotivated.
Having a reason for acting depends on true beliefs about normative
facts. Considering that kind of facts, the question arises whether we are
not swinging to a metaphysical account. However, normative facts are
not necessary metaphysical facts. If we want to avoid this direction, we
could stick to a descriptive position, and observe how action and moral-
ity are governed by norms. Where the internalist describes psychological
facts about one’s motives, the externalist might also describe the fact that
we take it to be true that it is worth protecting one’s health. When we
provide external reasons for acting, we obey norms of rationality con-
cerning the role played by moral truths (and not only factual or empirical
truths) in our life.
In what precedes, we saw how rationality was limited when consider-
ed as procedural and reduced to a correct deliberation. I want to evoke
two more reasons to reject such a conception of rationality. The first one
concerns the link between motivation and a preexisting motive and it has
something to do with Augustine’s story of conversion. Indeed, as far as a
motivation ensues from a rational deliberation which is itself oriented by
a motive (whatever it is), conversion appears as a means to fulfill this
motive. Apparently, internalism might offer an adequate account of
Augustine’s case, by describing his inner torment in terms of clashing
desires. On the one hand, Augustine aims at the pleasures of the body
and on the other hand, he wants spiritual happiness. How will the inter-
nalist explain this situation? He must say that as long as Augustine fails
to deliberate rationally or is mistaken on facts, he would not be motiva-
ted to convert. It is only at the moment when he succeeds in deliberating
rationally that he is finally motivated to change. Or we could follow
McDowell’s line of thought and conceive of conversion as a non rational
phenomenon.
194 Sophie Djigo
Or else, we might consider that as Williams briefly says, “the process

of deliberation can have all sorts of effects on S (the agent’s motivational
set)”.8 This point needs further clarification, since it is the expression of
a fine position which keeps the motivational set outside the scope of de-
liberation while leaving open the possibility that deliberation modifies
indirectly this set. Let’s go back to Augustine and the question how did
he come to prefer spiritual happiness to the pleasures of the body? I
think that his motivation could change after a rational deliberation, not
because of modification in the deliberative process, but in the “relevant
facts” he is supposed to know. It might be, for instance, that Augustine
has to take into account a new fact like his older age or his lassitude. Be-
cause of these new facts, the result of the deliberation will not be the
same as before: given these relevant facts, Augustine’s deliberation might
motivate him to change. This conversion would not have been possible
before and the deliberation provided by these new facts consequently
alters Augustine’s desires and strengthens his desire for spiritual happi-
ness. If this desire is stronger than the desire for terrestrial pleasures, the
will to change will no longer be half-wounded because of other clashing
wills.
This argumentation seems to diminish the difference between exter-
nalism and internalism, so we must not lose sight of reason statements.
In the internalist view, Augustine had no reason to be converted until he
was motivated to do it by virtue of his strongest desire (entire will) for
spiritual happiness. In the externalist view, Augustine, as he avows,
should change his life and had a reason to change even when his will was
half-wounded. That is because of his being aware of what he should do
that his incapacity to change was so painful. If conversion appeared to
be the best means to reach spiritual happiness, however, he had to
choose between clashing aims. The increased valuation of spiritual hap-
piness above all gave him a reason for conversion. It does not follow
from this that Augustine was motivated to convert and his purpose in
chapter VIII was precisely to report how difficult it was to do what he
knew he should do. The desire for spiritual happiness had no more lever-
age than the normative truth of what he should do and as he confessed,

Augustine might have delayed his conversion again and again and he
might never have changed his life. His will to change his life could have
remained a pious hope.
My second remark is that both internalism and externalism stand on
equal terms concerning the question “What counts as a reason?”. In fact,
their aim is not to take care of this problem, but to settle the question
about reason statements. For the internalist, if the agent is not motivated
to do Φ, then, we have no grounds for saying that he has a reason for Φ.
The externalist shall say that moral truth gives a reason for doing Φ,
whatever the agent’s motivation. How does a proposition come to count
as a reason for the agent? The controversy between internalism and ex-
ternalism must not conceal the crucial problem revealed by the gap be-
tween leverage and truth: the difficulty for the agent to recognize a rea-
son as such, in other words, to internalize a reason.
If we try to sum up very briefly the difference between internalism
and externalism, we shall say that the internalist is broadly concerned
with the question “What counts for me as a reason?”. And for a propo-
sition to count for me as a reason, it must be such as it would motivate
me after an informed rational deliberation. As to the externalist, his
problem is rather “What should count for me as a reason?”, and the an-
swer is that normative truth gives me a reason for acting. The existence
of such a difference between both views shows that there is a gap
between a reason for acting and my acting for a reason. How do I come
to consider R as a reason? What if, after an informed and rational de-
liberation, I still do not see R as a reason? How can we explain such a
resistance to normative truths, and consequently, to the external reasons
based on it?
3. Reasons avowal
The debate between reasons internalism and externalism was strictly

about the truth conditions of reason statements. For the internalist, it is
false to say that an agent has a reason for acting, while he is lacking the
corresponding motivation; the externalist maintains that the statement is
true. My point in the previous section was to accentuate a certain num-
ber of remarks which I think are very helpful in the reflection on the
196 Sophie Djigo
reasons for action. And philosophers like Williams or Parfit have shown
the advantages of both positions. Let us go beyond the question of the
truth of reason statements and see the implications for a thinking on ac-
tion and morality.
From what has been said, I think that we need to maintain the exis-
tence and the truth of external reasons. We saw that deliberation was not
enough to reach moral truth, since it consists in giving reasons which are
desire-based. However, Williams is right to point the gap between truth
and leverage and the ineffectualness of moral truths. Given these re-
marks, my starting point will be that there are external reasons and that
the problem is for the agent to “internalize” those reasons. Given that it
is true that the person has a reason to take medicine he needs, we need
to focus on the articulation between external reasons and leverage, in
other terms: how to make an internal reason of an external one?
In fact, when the agent is thinking of his reason for acting, he is not
only deliberating what would be the most appropriate way to fulfill his
desire. He is also engaged in a normative estimation either of his beliefs
or of elements of his motivational set. The reflective agent is appraising
what might give him a reason for action and in his assessment normative
facts play an undeniable role. If I have the desire to maltreat my wife, I
am in the position to choose to give it my approval or not. Of course, no
internalist would say that this man has a reason to maltreat his wife just
because he feels like doing it. The justifying condition is meant to con-
nect internal reasons with the agent’s rationality, in such a way as “what
we can correctly ascribe to him in a third-personal internal reason state-
ment is also what he can ascribe to himself as a result of deliberation”.9
The problem is that deliberation has limited result and does not always
succeed in harmonizing both first-person and third-person perspectives.
However, this is the challenge: that a third-personal reason statement (an
external reason) corresponds to a first-personal (internal) one.
In the case of the person refusing medicine, the problem is that he
considers that protecting his health does not give him a reason to take
medicine he needs. So, he accepts his lack of interest in health, he makes
do with it. If we refer to an external reason, we would say that he has a

reason for taking medicine because protecting his health should count
for him as a reason. The difference it will make is that in this view, this
person will be seen as responsible for his inaction, since it is up to him to
approve of protecting his health. We have to consider that the agent has
the capacity for assessing what counts as a reason and he also has au-
thority on his own attitudes, so that the result of his normative assess-
ment modifies his own motivational state, not in terms of desires but in
terms of commitments. If the agent decides that the protection of his
health counts as a reason, he is then under pressure and committed to
act.
We must distinguish two different points here: (1) the specific posi-
tion of the agent who is in a situation of choice concerning what counts
for him as a reason (2) the responsibility for ignoring an external reason,
or failing to make “the good choice” and to appropriate what should
count for him as a reason. The first question introduces the authority of
the agent as a fundamental element in practical reasoning. The second
one concerns the moral constraint on reasons for action and how nor-
mative claims play a role in the agent’s deciding what counts for him as a
reason.
Beginning with the question of the agent’s situation of choice, I
would like to comment on a passage from Hegel’s Philosophical Propaedeu-
tic:
“15. Such expressions as these are often used: ‘My will has been de-
termined by these motives, circumstances, incitements, or inducements’.
This expression implies that I have stood in a passive relation [to these
motives, etc.]. In truth, however, the Ego did not stand in a merely
passive relation but was essentially active therein. The Will, that is, ac-
cepted these circumstances as motives and allowed them validity as mo-
tives. The causal relation here does not apply. The circumstances do not
stand in the relation of cause nor my Will in that of effect. In the causal
relation the effect follows necessarily when the cause is given. As reflec-
tion, however, I can transcend each and every determination which is
posited by the circumstances. Insofar as a man pleads in his defence that
he was led astray through circumstances, incitements, etc. and, by this
plea, [hopes] to rid himself of the consequences of his deed, he lowers
himself to the state of an unfree, natural being; while, in truth, his deed
198 Sophie Djigo
is always his own and not that of another or the effect of something
outside himself. Circumstances or motives have only so much control
over man as he himself gives to them” (Hegel, G. W. F., The Philosophical
Propaedeutic, partially translated by W. T. Harris, revised by A.V. Miller,
translated by A. V. Miller, edited by M. George and A. Vincent, New
York, Basil Blackwell, 1986, §15).
Hegel’s point here is very helpful for my purpose, since he describes

the agent (the Ego) as radically indeterminate, that is, a free will. Natural
and individual determinations are set aside in practical reasoning and the
agent is himself in a certain distance from them which enables him to
evaluate them, to decide whether he is going to make them his own or
not. It belongs to practical reasoning (“resolving”, in Hegelian terms)
that the agent’s existing motives, impulses, beliefs or character come to
be alien, something he has to question in order to determine if he will
follow them. The particularities of the individual are not influential like
natural constraints would be: they count only insofar as I have made
them my own. An impulse, a motive cannot give me in itself a reason for
acting: I rather estimate if it gives me a reason. A reason cannot be im-
posed on me by someone else, neither another agent nor an element
from my motivational set, without my appropriating. If I have a desire to
do something or if someone tells me that I should do that thing, in both
cases, I have to rule on this desire or this normative claim, to determine
whether it gives me a reason or not.
My having a reason is thus independent from my motives and I can-
not put them forward as an excuse for what I have done. Hegel is chal-
lenging the discrepancy between the explanatory condition and the jus-
tifying one: his criticism has a real target here, which is the attempt to
justify an action by its explanatory reason. The agent might say that the
reason why he maltreats his wife is a strong impulse of anger. And we
could say that we understand why he acted like that, however, his action
is not justified. Saying this presupposes, as Hegel tries to put it, that the
agent had a choice. And his strategy of exoneration (defence proof) is to
show he had not. His motivation was reduced to his impulse or con-
strained by circumstances. In other words, he was in a passive position
which restrained the power of his will. His room for maneuver was ex-
tremely reduced (or non-existent). When he deliberated, he took into ac-
count such determinations which were strongly influential on his final

motivation to act.
The source of what we have called an internal reason is internal in
the agent and cannot come from an external source like the others or the
law. On the contrary, external reasons are provided from external
sources, like custom, rules, etc. When Williams speaks of the elements
of the agent’s motivational set, he considers them internal. When Hegel
speaks of impulses and motives, he also treats them like something inter-
nal which is independent of the agent: he simply has them. But thanks to
reflection, he can “transcend” them. It seems to me that Hegel makes an
interesting distinction between “to be internal” and “to be one’s own”: I
can have internal desires while not having made them my own, a pos-
sibility that requires a reflection on these impulses. The major point is
whether deliberation is enough for the agent to make his motives his
own.
Deliberation might be conceived of as a rational process intending to
settle the question whether a motive gives me a reason for acting and this
definition would be more general than the restrictive idea of deliberating
to find the best way to fulfill a desire. This possibility is included in
Williams’ argumentation when he briefly depicts different modes of de-
liberation, like “where there is some irresoluble conflict among the ele-
ments of S, considering which one attaches most weight to”.10 Even un-
derstood in this wider sense, deliberation remains a reflection on the
motives which would give me a reason, and not a normative estimation
of which motive I should make my own or not (still the same purpose
of leverage). Where Hegel speaks of “transcending one’s natural deter-
minations”, I prefer to use the concept of normative truths playing the
role of moral constraints on my resolution.
My capacity as a rational being to decide what should count for me as
a reason contributes to my responsibility for what I do. In the mentioned
text, Hegel offers a very exigent notion of responsibility which neither
motives nor external circumstances would lighten. If the agent is passive
in respect of his motives, he might be accused of criminal negligence. It
means that passivity can be in itself a fault because the agent refuses the

200 Sophie Djigo
responsibility which is incumbent upon him. That is also why the truth
of a reason statement does not depend on the agent and is not a purely
empirical statement provided by the fact that he would be motivated to
act after deliberating. I am responsible for following a motive I should
have rejected as not giving me a reason. These remarks, however illu-
minating, are relevant in cases where a preexisting motive incites the
agent to act. Thus, it is a matter to restrain motivational leverage by ratio-
nal normative claims.
However, it is not helping us much concerning the case where we
should do something while we are not motivated to do it. In such cases,
there are no impulses to select. Can we then be responsible for not being
motivated, for not having any desire which should give us a reason for
some action? For instance, can we consider the man who does not take
medicine he needs as responsible for his lacking of interest in his health?
This is the very problem of internalizing an external reason. We can
interpret such lack of interest as a situation where the person theore-
tically approves the importance of health and does not follow it in prac-
tice. One can approve of an external reason, an “I should-statement”
only theoretically. In general, most people agree that racism is an un-
founded attitude and a barbaric behavior until the day when they are
confronted to, say, their beloved daughter planning to marry a foreigner.
Or else a wealthy man who thinks that poverty is an unacceptable injus-
tice and who is reluctant to pay more taxes. Such cases show how the
theoretical point of view and the practical one might be inconsistent. We
can easily admit certain moral truths as long as they are not constraining,
as we are not really concerned by them. This is again the gap between
truth and leverage.
So, if we take the person who is not motivated to follow an external
reason, how come that he can be motivated by this reason? It seems that
the practical reasoning is vain if it is not already rooted in a preexisting
motivation. Saying that means to take the agent’s reasoning as purely
theoretical, that is, he stands like a mere bystander of his own behavior,
inferring what he would probably do in a certain context. As Williams
said, “what we can correctly ascribe to him in a third-personal internal
reason statement is also what he can ascribe to himself as a result of de-
liberation”. The difference is precisely that the agent’s position is a first-
personal perspective which is not symmetrical or similar with the third-
personal one. The external third-personal perspective focuses either on

the normative question: “What ought the agent do?” or on the empirical
question “What will he do?”. The first-personal question is rather “What
shall I do?” and this question concerns only the agent and determines his
future action. This question is at the same time evaluative as far as it
takes into account a normative assessment, and practical since the agent
is deciding what he is going to do.
Indeed, deliberation or if we want another term, practical reasoning,
is not reduced to an inferential reflection whose conclusion would be an
empirical statement (a prediction of what the agent would do). The con-
clusion of this reasoning is action and coming to action is not the result
of a psychological investigation on one’s motives, but of a normative as-
sessment of guiding elements. So, if we want to take external reasons se-
riously, we need not only to approve them, but to avow them. I borrow
this concept of avowal from Richard Moran,11 who applies it to beliefs
and other attitudes and who had taken it before from Georges Rey.12
Commenting on the difference between reporting on an attitude of mine
and avowing it, he says: “An avowal of one’s belief, by contrast, is not
made on any psychologically explanatory basis, and is rather the expres-
sion of one’s own present commitment to the truth of the proposition
in question”.13
I cannot use this notion of avowal in the same way as Rey and Moran
since my concern is not self-knowledge of one’s attitudes, but action and
reasons for action. My point is that there is a difference between the
agent’s way of ascribing to himself a reason and his avowing it. “As-
cribing” sounds like a predictive statement grounded on motivational
psychological evidence and provided by a deliberation on this evidence
and the relevant context. To a certain extent, we could imagine the agent
betting on what he will himself do! “Avowing” a reason means to con-
11 Moran, R., Authority and Estrangement. An Essay on Self-Knowledge, Princeton and
Oxford, Princeton University Press, 2001.

12 Rey, G., “Towards a Computational Account of Akrasia and Self-Deception”, in
Rorty A. O. – McLaughlin, B. (eds.), Perspectives on Self-Deception, Berkeley – Los Angeles
University of California Press, 1989.
13 Moran, R., Authority and Estrangement. An Essay on Self-Knowledge, p. 86.
202 Sophie Djigo
sider the constraining power of it and to commit oneself to its truth, not
like in the case of a theoretical approval, but as a practical engagement.
In this view, admitting the truth of an external reason involves that the
agent makes up his mind and acts in accordance with his resolution.
Hence, the action does not ensue from an inferential deliberation on the
agent’s motives, habits and circumstances, but from his avowal of one’s
reason.14
The problem is that the agent can adopt both perspectives on him-
self: a first-personal and a third-personal.15 From the second one, he
might be a mere bystander of his own reasoning, looking for psycholo-
gical evidence, beliefs, past habits and predicting his own future action.
Like in the position of a spectator, he might be mistaken and his predic-
tion, be false, since there is no constraint on action, it is only a matter of
probability. He might also claim that he will do Φ and not do it, and the
fact that his intention has not been realized would mean nothing else
than a contingent fact. By contrast, the position of avowing consists in
endorsing the reason statement and assuming the responsibility for its
being true or not. The result is not a prediction which might be false, but
an action I am expected to do. If the person avows that he needs medi-
cine to protect his health, it does not mean that he would be motivated
to take it after deliberation, (or not, if he shows a lack of interest in his
health), but that he has intention to take it (he shall take it). Avowal is the
internalization of an external reason: the importance of protecting one’s
14 This is also very close to Christine Korsgaard’s position on the morally con-
straining aspect and normativity of most of the agent’s identities, which give him reasons
for acting. See for instance Korsgaard, C. M., Self-Constitution. Agency, Identity and Integrity,
Oxford, Oxford University Press, 2009: “So I am these things –this country’s citizen,
these people’s daughter, this person’s old friend– perforce, and not because I chose to be
them. And yet these identities give rise to reasons and obligations, as much as the ones
that I do more plainly choose, like profession or an office or a friendship quite deliber-
ately sought out. But I want to argue that while that is true in one way, in another way it
is not. For whenever I act in accordance with these roles and identities, whenever I allow
them to govern my will, I endorse them, I embrace them, I affirm once again that I am
them. In choosing in accordance with these forms of identity, I make them my own” (pp.
42-43).
15 An impersonal or agent-neutral perspective, like in Nagel, T., The Possibility of Al-
truism, Princeton, Princeton University Press, 1970.
health, which was lacking in our example, needs to be avowed for this
man to come to take medicine. Taking medicine is not the consequence
of a new interest in health (motivational element) but of a rational com-
mitment to the normative truth concerning the value of health.
Considered as an empirical phenomenon, Augustine’s decision to
change his life appears to be only a fact about his psychological life, the
fact that he believes he has a reason to change his life which is different
from the fact that he does have this reason. On the one hand, Augustine
describes his resolution to change and having made such decision is to
be committed to conversion. It is now up to him to carry his resolution
through as an active agent who chose what was to be made his own. In
this perspective, for him, conversion is not a theoretical expectation, just
as if he was in the passive position of observing what he would do. It is
rather something he intends to do, the practical conclusion of his earlier
reasoning. On the other hand, he can treat himself as an empirical ob-
ject, with a precise history made of custom, attraction to carnal pleasures
and a tendency to delay his resolution.
All this past evidence might instill doubt in Augustine’s mind about
the firmness of his resolution which is now seen as a psychological fact
about him. From this theoretical perspective, his will shows to be weak,
half-wounded, so inconstant that he is no longer motivated or resolute to
change. What Moran called a “tactical substitution of the theoretical
point of view for the practical one”16 must be for Augustine to consider
his resolution as an empirical object and to estimate its strength “from
the outside” as if it was independent of him. Indeed, when he takes into
account his will to keep his old habits and his past incapacity to change,
he focuses on how his will to change is fragile and he thinks that there
might be a long way to go. But in reflecting as such, he forgets that his
resolution is strong insofar as he wants to follow it and that it is still up
to him: his decision to act for a reason is grounded on his being com-
mitted to its truth. If he still thinks that he should change his life, he
does not need any other (empirical) import to secure the success of his
future conduct.
16 Moran, R., Authority and Estrangement. An Essay on Self-Knowledge, p. 81.

204 Sophie Djigo
With respect to reason statements, the theoretical point of view con-

sists in deliberating what I (the agent) would do in the context, given my
motivational set. If the agent who needs medicine treats his lack of in-
terest in health as a psychological given, he would not be motivated for
action and he would remain indifferent to the others’ advice that he
should care about his health. Even if he agrees that he should take it, his
taking into account his lack of interest might show him his incapacity to
act hence he does not feel committed to the truth he admitted intellec-
tually. But if he avows that he should take the medicine, his lack of
interest should not be a barrier, since it is up to him to want to protect
his health or to resolve to do it because he is committed to the truth of
the reason statement (“Protecting his health gives a reason to take the
medicine”). Instead of speaking of internal and external reasons, I shall
speak of a theoretical and a practical point of view on one’s reason for
action, or of ascribing to me a reason for acting and avowing a reason
for acting, assuming that ascribing does not commit me to the truth of
the reason statement (if my motivation is too weak, it means that I have
no reason for acting) whereas avowing does (I am responsible for carry-
ing through my resolution, I am committed to the truth of the reason
statement). In other words, saying that “I have a reason for doing Φ”
compels me to act if I avow it, not if I only ascribe it to me like a mere
bystander would do.
4. Indifference to truth
It was argued in the previous section that the link between leverage and
truth depends on the agent’s capacity of avowal, that is, to commit one-
self to the truth of the reason statement. Cases where the agent ap-
proves the truth of a reason statement while not following it are explain-
ed by his forgetting his responsibility for following this reason, his com-
mitment to its truth. Leverage is thus rather thought as the expression of
one’s meeting his commitments than as the result of a deliberation in-
volving a motivation. In the last section, I want to make a few remarks
on the difficulty to meet such commitments to truth. I have already of-
fered a first explanation with Moran’s idea of the “tactical substitution”,
but I think there are two more elements that need to be considered. As I
tried to articulate truth and leverage of the reason statement, it is no

longer possible to reject a reason statement as “too far from me and my
motivational set”, as “not concerning me”. The truth of this reason does
not depend on me; what is up to me is my avowing it or not (refusing it
as a perfectly irrational agent or approving it only theoretically). Now, if
true reason statements have no leverage on me, that is because of a
certain indifference to truth.
Let us imagine a person similar to Wittgenstein’s meaning-blind,17
who suffers from, say, “truth-blindness”. He knows the truth, he may ad-
mit it, but he remains insensible (uncommitted) to it. If moral truths
were as much compelling as language rules, it would not be a matter: this
person would do what he has a reason to do, even if he cannot avow it.
He would obey external reasons. However, we have agreed with Williams
that for the agent to do what he has a reason to do, his resolution has to
come from him, he must be convinced by its truth. And furthermore, he
is also committed to normative truth. A person with truth-blindness
would conclude after rational deliberation that he has a reason to do Φ
but he will not be motivated to do it, because he does not feel engaged
by truth. How might such indifference occur?
If an agent is not interested in following a normative truth, we can
take it as the result of a conflict of interest: he purely and simply has as
much reason not to do Φ event if it is true that he has a certain reason to
do it. For instance, if a person needs medicine to be cured of an illness
and if the only existing medicine involves high risks of being killed by
side effects, he will be divided between a reason to take medicine to pro-
tect his health and the reason not to take them to avoid the risks of
dying promptly.
However, except these well-known cases of conflict between reasons,
truth-blindness implies another type of conflict, between the Ego-
perspective (self-interest) and moral truth.18 Such conflict has something
to do with the so-called internal reasons: in fact, the superiority of in-
ternal reasons lies in their personal character and it is not surprising that
17 Wittgenstein, L., Remarks on the Philosophy of Psychology, vol. I, edited by G. E. M.

Anscombe and G. H. von Wright, translated by G. E. M. Anscombe, Oxford, Basil
Blackwell, 1980, §247.
18 On this issue, see also Parfit, D., Reasons and Persons.
206 Sophie Djigo
Williams attaches a great importance to them since it is part of his more

general reflection on moral individualism and his criticism against agency
as not self-directed enough. For Williams, when the agent cultivates mor-
al virtues, he focuses on the way in which the others might judge him in-
stead of turning his attention to the significance of his resolution. What
is important is the distinction between self-directed and selfish resolu-
tions.
The theoretical approval of a moral truth is not hard for the agent
insofar as it is not committing him to act. The link between my approval
of a normative truth and my committing to it is broken here and I be-
have as if I was not concerned by this normative claim in practice. In
fact, I agree with moral truth until it clashes with self-interest so that I
do not make it my own. It is quite different to consider a self-directed
stance of agency, which is about the agent’s making up his mind without
taking into account the point of view of the other. In this sense, “in-
ternal” means that the reason I have for acting is rooted in my convic-
tion and that I am not only conforming to external standards of behav-
ior.
Concerning the problematic gap between leverage and truth, we can
sum up the most important elements I tried to advance. It is not up to
the agent if he has a reason to do Φ or not: the truth of the reason state-
ment is a normative truth, for instance a moral truth such as protecting
my health gives me a reason to take medicine I need. Denying such a
normative fact would be purely and simply irrational. However, the exe-
cution of my resolution depends on me and it is not because I have a
reason to do Φ that I will really do it. Does it mean that I disagree with
the normative truth? Not necessarily: I can be perfectly rational and ad-
mit of my reason to take medicine after deliberation while not taking it.
This inconsistency does not concern the truth of the reason statement,
but my relationship to this truth, that is, if I am committed to it or if I
simply consider my belief in it as a psychological fact. What is up to me
then, is leverage: it is to meet my engagement to the truth of the reason
statement. I think that I should do Φ and it is now up to me to follow
the reason I think I have to act. If I avow that I should do it, I am as-
suming the responsibility for following this reason or not. If I just take it
to be an empirical given about me (a belief of mine), I might question
my own resolution and finally fail to carry it through. After all, it is the
agent’s business to give leverage to what he should do because of its

binding truth.
STL
Lille, France
208 Sophie Djigo
References
Augustine, Confessions, tr. by E. B. Pusey.

Elster, J., Agir contre soi, Paris, Odile Jacob, 2007.
Hegel, G. W. F., The Philosophical Propaedeutic, partially translated by W. T.
Harris, revised by A.V. Miller, translated by A. V. Miller, edited by M.
George and A. Vincent, New York, Basil Blackwell, 1986.
Korsgaard, C. M., Self-Constitution. Agency, Identity and Integrity, Oxford,
Oxford University Press, 2009.
McDowell, J., “Might There Be External Reasons?”, in McDowell, J.,
Mind, Value and Reality, Cambridge, Mass., Harvard University Press,
1998.
Moran, R., Authority and Estrangement. An Essay on Self-Knowledge,
Princeton and Oxford, Princeton University Press, 2001.
Parfit, D., Reasons and Persons, Oxford, Clarendon, 1984.
Parfit, D. – Broome, J., “Reasons and Motivation”, Proceedings of the
Aristotelian Society, Supplementary Volumes, vol. 71 (1997), pp. 99-130.
Rey, G., “Towards a Computational Account of Akrasia and Self-
Deception”, in Rorty A. O. – McLaughlin, B. (eds.), Perspectives on Self-
Deception, Berkeley – Los Angeles University of California Press,
1989, pp. 264-296.
Williams, B., “Internal and external reasons”, in Williams, B., Moral Luck,
Cambridge, Cambridge University Press, 1981, pp. 101-113.
——, “Replies”, in Altham, J. E. J. – Harrison, R. (eds.), World, Mind, and
Ethics: Essays on the ethical philosophy of Bernard Williams, Cambridge,
Wittgenstein, L., Remarks on the Philosophy of Psychology, vol. I, edited by G.
E. M. Anscombe and G. H. von Wright, translated by G. E. M.
Anscombe, Oxford, Basil Blackwell, 1980.
MIXED ACTIONS IN THE WORK OF
HARRY FRANKFURT
TERESA ENRÍQUEZ
The purpose of this paper is to provide an overview of the theory of ac-

tion articulated in the work of Harry G. Frankfurt based on a specific
keynote: the Aristotelian concept of “mixed action”. Distinctive to the
degree that it encompasses the will’s capacity for reflection on its own
desires, Frankfurt’s approach discloses a wide range of ways in which the
voluntary and the involuntary are mixed. The first section of this paper
outlines a number of key ideas; the second addresses the nature of such
admixture in the threefold typology of conflicts of desires set out by
Frankfurt in the essay “Three concepts of free action” (1975); and in line
with the radical approach taken by Frankfurt himself, the third and final
section describes a special kind of conflict of desires –“ambivalence”–
as discussed in the essay “Identification and wholeheartedness” (1987).
No particular thesis is advanced here; rather, the objective is to trace the
development of Frankfurt’s thought, so as to discern some of the mean-
ings of the notion of “mixed action”.
Any analysis of mixed actions requires a refined focus on two fundamen-

tal dimensions of human action: the voluntary and the involuntary. In his
account of human action, which also frames his study of personal iden-
tity, Frankfurt describes a clear framework shaped by a set of acute ob-
servations. His contribution has received close critical attention from
specialists in the field,1 and is a standard reference in debates concerning
1 See for instance Buss, S. – Overton, L. (eds.), Contours of Agency: Essays on Themes
from Harry Frankfurt, Cambridge, MIT Press, 2002; International Workshop on Belief, Respon-
sibility and Action, Valencia, 12-14/XI/2008, in Philosophical Explorations, 12/2 (2009);
Kane, R., (ed.), The Oxford handbook of free will, Oxford-New York, Oxford University
Press, 2011.
210 Teresa Enríquez
the freedom of the will and moral responsibility.2 A single-volume col-

lection3 of essays on the philosophy of mind and the philosophy of ac-
tion written over a twenty-year period (1969-1988) furnishes sufficient
material for the purposes of this paper, given that, as Frankfurt himself
avers in the preface to the collection: “The inner organization of the will
and what that implies for us are, in any case, the subjects I have mainly
attempted to explore”.4 The question underlying this exploration of
Frankfurt’s essays relates to the ways in which the voluntary and involun-
tary dimensions are mixed in human actions.
The term “mixed actions” was coined by Aristotle5 as part of an at-
tempt to draw a clear distinction between the voluntary and the involun-
tary: “mixed” actions are those concerning which “it may be debated
whether such actions are voluntary or involuntary”.6 The issue of mixed
actions is dealt with as an instance of the problematic of voluntary ac-
tion in the section of Nicomachean Ethics that addresses the principles of
moral virtue.7
The existing research that asserts a connection between Frankfurt’s
thought and Aristotle’s work points to the similarity in the way they un-
derstand the relationship between means and ends,8 an issue that also has
a certain bearing on the question of mixed actions. Frankfurt does not
2 See Widerker. D., – Kenna, M. Mc., (eds.), Moral Responsibility and Alternative Possibili-
ties: essays on the importance of alternative possibilities, Burlington, VT, Ashgate, 2003.
3 Frankfurt, H., The Importance of What We Care About. Philosophical Essays, New York,
Cambridge University Press, 1988. The text is cited hereafter as F-IWCA, followed by the
relevant page number(s).
4 F-IWCA, viii.
5 See Aristotle, Nichomachean Ethics, trans. W. D. Ross, Kitchener, Batoche Books,
1999, III, 1, 1109b 30-1111b 3. The text is cited hereafter as NE.

6 NE, III, 1, 1110a 7.
7 See Thomas Aquinas, In NE, III, trans. C. I. Litzinger, Chicago, Henry Regnery
Company, 1964, Lecture 1, no. 1: “After the Philosopher has treated virtue in general, he
treats here certain principles of virtuous acts. In defining virtue, he said (305) that virtue
is a habit of correct choosing because virtue works by means of choice. Now he logically
discusses choice together with the voluntary and ‘willing’. The voluntary is common to
these three: for the voluntary is anything that is freely done, choice however concerns the
things that are for the end, and willing considers the end itself ”.
8 See Schumacher, C., Zwecke und Mittel bei Aristoteles und Harry Frankfurt, Hamburg,
GRIN Verlag, 2007.

Mixed Actions in the Work of Harry Frankfurt 211
use the term “mixed actions” often and when he does use the adjective
“mixed”, he does so in a different sense.9 However, his innovative ap-
proach to the consideration of desire, which encompasses a detailed
analysis of first and second-order conflicts of desire, may be enriched by
a reading from the perspective of the Aristotelian concept of “mixed ac-
tions”.
This connection between Frankfurt’s work and Aristotle’s thought is
not contrived; indeed, the former opens his essay entitled “Coercion and
moral responsibility”10 with an epigraph cited from the latter: “On some
actions praise indeed is not bestowed, but pardon is, when one does
what he ought not under pressure which overstrains human nature”.11
In a certain sense, this quotation implies that the individual who car-
ries out the (second) type of mixed actions can be morally responsible.
According to Aristotle, the first type of mixed action may deserve blame
or praise depending on the greatness of the good lost or gained;12 and
the third type of mixed action is always blameworthy because the evil of
the action is so reproachable that the individual should choose death
rather than commit it.13 Neither of the other two types of mixed action
may be consented to –only undue action, which is not too serious,
caused by a threat that outstrips ordinary “human strength”.
Fear is one of the factors that may cause some kinds of mixed action.
Although the fearful individual acts for his own purposes –that is, his ac-
tion is voluntary to a certain extent– he acknowledges that he would not
do so if it were not for the threat to which he is subject; thus, his action
is also, at least in part, involuntary. So as to clarify in what sense an ac-
tion may be voluntary and/or involuntary, Aristotle discusses actions
9 See, for instance, F-IWCA, 53, 65.

10 F-IWCA, 26-46.
11 NE, III, 1, 1110a 23-25.
12 NE, III, 1, 1110a 19-23: “For such actions men are sometimes even praised, when
they endure something base or painful in return for great and noble objects gained; in the
opposite case they are blamed, since to endure the greatest indignities for no noble end
or for a trifling end is the mark of an inferior person”.
13 NE, III, 1, 1110a 26-28: “But some acts, perhaps, we cannot be forced to do, but
ought rather to face death after the most fearful sufferings; for the things that ‘forced’
Euripides’ Alcmaeon to slay his mother seem absurd”. Alcmaeon killed his mother
Eriphyle to escape the curse of his dead father, Amphiaraus.
carried out “from fear of greater evils or for some noble object”;14 these
cases exemplify two causes of fear:
“For example, if a tyrant were to order one to do something base,

having one’s parents or children in his power, and if one did the action
they were to be saved, but otherwise would be put to death […] Some-
thing of the sort happens also with regard to the throwing of goods
overboard in a storm; for in the abstract no one throws goods away vol-
untarily, but on condition of its securing the safety of himself and his
crew any sensible man does so. Such actions, then, are mixed, but are
more like voluntary actions; for they are worthy of choice at the time
when they are done, and the end of an action is relative to the occasion”
(NE, III, 1, 1110a 8-13).
To Aristotle’s mind, the end for which mixed actions are carried out
renders them voluntary as such. Frankfurt is less categorical in this re-
gard: first, he divorces freedom from such external factors as tyranny or
stormy weather. Frankfurt also refers to the twofold origin of the exter-
nal factors that may prompt a free agent to carry out an action which –in
other, better circumstances– (s)he would refuse to do, so as to under-
score the significance of internal desire for moral responsibility, as op-
posed to external conditions, whatever they may be. The example
Frankfurt gives concerns two men who come to a fork in a mountain
road; both choose the path to the right: one, because another individual
threatens that he will cause an avalanche if he takes the path to the left;
and the other, because he realizes that the natural conditions along the
left-hand path will lead to an avalanche. Frankfurt draws the following
conclusion:
“There are interesting differences between these situations, to be

sure, but there is no basis for regarding the man as acting more or less
freely or of his own free will in the one case than in the other. Whether
he is morally responsible for his decision or action in each case depends
14 NE, III, 1, 1110a 4.

not on the source of the injury he is motivated to avoid but on the way
in which his desire to avoid it operates within him” (F-IWCA, 45).
In Frankfurt’s view, threatening external factors can provide an agent

with a reason to pursue a particular course of action, including an action
that the individual might have rejected in other circumstances; however,
such a situation undermines neither freedom nor moral responsibility.
The latter depends to a greater extent on an internal factor, “on the way
in which his desire to avoid it [injury] operates within him”. In the
agent’s response to a threat, both Aristotle and Frankfurt discern a form
of agency that remains open to freedom: the free action of a subject
who endeavors to find a way of avoiding harm.
In contrast to the approach taken by other analytic philosophers,
Frankfurt proceeds to a more detailed account of desire; and he lights on
Aristotle’s definition of the voluntary as that which stems from an inner
moving principle:
“Aristotle maintained that behavior is voluntary only when its moving

principle is inside the agent. This cannot be correct if ‘inside’ is con-
strued in its literal sense: The movements of an epileptic seizure are not
voluntary, but their moving principle or cause is spatially internal to the
agent. The location of a moving principle with respect to the agent’s
body is plainly less relevant than its ‘location’ with respect to the agent’s
volition” (F-IWCA, 171).
Such binary pairs are difficult to understand, though inextricably link-

ed with one another: voluntary and involuntary; internal and external; ac-
tive and passive.15 The Aristotelian text to which Frankfurt alludes, with-
out explicitly citing it, comes from the chapter concerning mixed actions:
“Those things, then, are thought involuntary, which take place under
compulsion or owing to ignorance; and that is compulsory of which the
moving principle is outside, being a principle in which nothing is con-
tributed by the person who is acting or feeling the passion, for example,
15 See F-IWCA, 59.

if he were to be carried somewhere by a wind, or by men who had him

in their power” (NE, III, 1, 1110a 1-3).
An action motivated by a wholly external principle is not a mixed ac-

tion; rather, it is clearly an involuntary action. As opposed to the internal
disposition of the obligated agent, the nature of the external principle is
irrelevant; it makes no difference whether it is a natural condition like the
wind or a human cause such as kidnapping. In the case of physical coer-
cion, the “externality” of the principle is in fact literal; however, such lit-
eral externality is not required to give rise to a clearly non-voluntary ac-
tion, as is the case of ignorance.16 Frankfurt’s position appears to be
more extreme than Aristotle’s as regards physical coercion; rather than
being an “involuntary action”, he regards it as an absolute absence of ac-
tion:
“A person is sometimes said to have been coerced even when he has

performed no action at all. Suppose that one man applies intense pres-
sure to another man’s wrist, forcing him to drop the knife in his hand. In
this case, which involves what may be called ‘physical coercion’, the vic-
tim is not made to act; what happens is that his fingers are made to open
by the pressure applied to his wrist […] We might say that in instances
of physical coercion the victim’s body is used as an instrument, whose
movements are made subject to another person’s will” (F-IWCA, 26-27).
In Frankfurt’s view, physical coercion does not even amount to an

“involuntary act” since causing a “person’s simple movement” does not
16 Thomas Aquinas, In NE, III, Lecture 3, no. 5: “The involuntary is a privation of
the voluntary. But the voluntary implies a movement of the appetitive power presup-
posing a knowledge via sense or reason because a good perceived moves the appetitive
power. A thing is involuntary on two accounts: one, because the movement of the
appetitive power is excluded –this is the involuntary resulting from violence– the other,
because a mental awareness is excluded –this is the involuntary resulting from igno-
rance”. Aristotle refers to Merope (wife of Cresphontes, in Euripides’ eponymous trage-
dy), who killed her own son because she wrongly believed him to be an enemy (see NE,
III, 1, 1110b 18-1111a 21). Is this really not an action? Perhaps. However, besides this ex-
treme example as dramatized by Euripides, there is a wide range of actions carried out in
and through ignorance that are not encompassed by the scope of this paper.
involve causing an “action” as such.17 In the essay entitled “The problem

of action”,18 Frankfurt holds that this is more than a “mere happening”,
marking a clear break with the causal theories he describes as “implau-
sible” due to the confusion generated by “counterexamples of a well-
known type”: for instance, the man who, despite his intention to spill the
contents of his glass as a sign, upends it at the wrong time (involuntarily)
because his hand is trembling with anxiety.19 Action –and with all the
more reason, intentional action20– is not a bodily movement but one of
the “ways in which a person may be related to the movements of his
body”.21
An action is an action before it is voluntary; that is, an action may be
traced from the distinction between “what an agent does and what
merely happens to him, or between the bodily movements that he makes
and those that occur without his making them”.22 The idea that every ac-
tion involves a movement of the body is open to question: are our own
sincere thoughts and desires not to be regarded as real actions?23 The
agent’s being is more than bodily; and the agent’s actions are more than
bodily movements. Whatever the case may be, the definition of action
connotes an internal moving principle.
Nevertheless, the case of actions undertaken so as to avoid punish-
ment is problematic. The (external) threat produces an (internal) state of
17 See F-IWCA, 69.

18 F-IWCA, 69-79.
19 See F-IWCA, 70-71.
20 Intentional action is defined in terms of the third meaning of the concept of “in-
tentional”. Intentional denotes (a) purposive movement, whether or not it is under the
guidance or control of the person (e.g. pupil dilation has a purpose, but it is not control-
led by the individual); (b) a pleonastic description of action, as enacted by the person;
and (c) “in a more appropriate usage, it refers to actions which are undertaken more or
less deliberately or self-consciously –that is, to actions which the agent intends to per-
form. In this sense, actions are not necessarily intentional” (F-IWCA, 73).
21 See F-IWCA, 71.
22 F-IWCA, 69.
23 To take the issue a step further: does the formation of volition in itself not meet
all the criteria that define action as such? Volition does not simply “happen” in the per-
son’s internal state; rather, as Frankfurt avers, the person produces such volition. Need
volition involve a movement of the body? The argument suggests that actions primarily
entail praxis, rather than a principle of bodily movement (kinesis).
desires in the victim. A series of questions arises in this regard: Is the

victim’s action –that is, the action of an agent under threat– caused by
the external threat or the internal desire? Is such an action voluntary or
involuntary? Frankfurt is emphatic that the “internal” and “external” are
descriptive terms for moving principles; rather than read them literally,
they should be interpreted in relation to the will.24 Frankfurt is also in-
trigued by the Aristotelian identification of the voluntary with the inter-
nal and the non-voluntary with the external; a whole essay is devoted to
exploring the implications of the agent’s capacity to render “internal” a
desire that was experienced, initially at least, as “external”. To a signifi-
cant extent, such a shift depends on the person’s (in)capacity to identify
with the desire as such; the title of the essay alludes to this point: “Iden-
tification and externality”.25
An accurate conception of Frankfurt’s explanatory framework
concerning the degrees to which the individual person may be identified
with his/her own desires requires a clear understanding of three closely
related concepts: action, will and person. The earliest account of these
three ideas is provided in the landmark essay, “Freedom of the will and
the concept of a person”.26
The will may be distinguished from other similar concepts,27 because
it specifically denotes “an effective desire –one that moves […] a person all
the way to action”.28 The will is not merely an “intention”, no matter
how firm such an intention may be. The will is not simply another desire;
rather, it is a desire that leads to action. Hence, the will is expressed in
the performance of the action. If every subject with a will is an agent –
that is, capable of prompting actions based on desires– then human
beings are not the sole agents, nor are we the only beings with a will.29
24See F-IWCA, 59, 61, 171.

25F-IWCA, 58-68.
26 F-IWCA, 11-25.
27 Such as, for example, first and second-order desires and volitions, intention, free-
dom to act, the freedom of the will and moral responsibility.

28 F-IWCA, 14.
29 See F-IWCA, 78.

Frankfurt explicitly accepts that children and animals may carry out
voluntary actions;30 in this respect, his view parallels Aristotle’s thought,31
as well as the work of other, contemporary thinkers.32 In this regard,
Thomas Aquinas likewise offers five reasons33 to show that action stem-
ming from the sense appetite is voluntary.
None of the philosophers referred to above sees the acknowledg-
ment of animal agency as undermining the specific nature of human
agency.34 Indeed, to Frankfurt’s mind, the person as such may be defined
as a particular way of being an agent –in other words, as a being capable
of relating in a specific way to his own action. Thomas Aquinas also de-
fines the person in terms of control over his own acts.35 Frankfurt is un-
questionably right to dismiss as inadequate any description of the person
as a subject comprised of mental and physical attributes, given that such
features are not the exclusive prerogative of human being.36 The concept
of the person is bound up with the will or, to be more precise, with the
“structure of the will”, as outlined in the draft definition of the person
formulated as follows:
“It is my view that one essential difference between persons and

other creatures is to be found in the structure of a person’s will. Human
beings are not alone in having desires and motives, or in making choices.
They share these things with the members of certain other species, […] it
seems to be peculiarly characteristic of human, however, that they are
30 See F-IWCA, 12.

31 See NE, III, 1, 1111a 25-1111b 3.
32 See González, A. M., “Action in a Narrow and in a Broad Sense”, in González, A.
M. – Vigo, A. G. (eds.), Practical Rationality. Scope and Structures of Human Agency,

Hildesheim – New York, Olms, 2010, pp. 126-128.
33 Thomas Aquinas, In NE, III, Lecture 4, no. 3.
34 See Idem, pp. 123-168.

35 Thomas Aquinas, Summa Theologiae, trans. Fathers of the English Dominican Prov-
ince, K. Knight, Online Edition, 2008, I, q. 29, a. 1, co. “The particular and the individual
are found in the rational substances which have dominion over their own actions; and
which are not only made to act, like others; but which can act of themselves; for actions
belong to singulars. Therefore also the individuals of the rational nature have a special
name even among other substances; and this name is ‘person’”.
36 See F-IWCA, 12.
able to form what I shall call ‘second-order desires’ or ‘desires of the sec-
ond order’” (F-IWCA, 12).
A refined understanding of the specific nature of human action rests

on the distinction between different orders of desires. Desires of the
first order are desires to take (or not take) one particular course of action
or another. Animals only have first-order desires: they desire to get
something, to move in one direction or another, to run, sleep, eat, etc.
Given our capacity for reflective self-analysis, however, human beings
may desire to have desires –or even desire not to have desires. Desires
concerning first-order desires are desires of the second order.37 The ex-
pression “desire of the second-order” denotes higher orders (second,
third, fourth, and so on indefinitely) in general terms, which are based on
reflection on preceding orders of desires.38
Insofar as desires prompt action in an effective way they comprise the
will of the agent. The complexity of personal agency enriches the defi-
nition of the will as a form of effective desire. The person’s will is not re-
duced to a first-order desire that proved effective; the will is more than
that:
“Someone has a desire of the second order either when he wants sim-
ply to have a certain desire or when he wants a certain desire to be his
will. In situations of the latter kind, I shall call his second-order desires
‘second-order volitions’ or ‘volitions of the second order’. Now it is
having second-order volitions, and not having second-order desires gen-
erally, that I regard as essential to being a person” (F-IWCA, 16).
The idea of “second-order desires” is narrowed down further to the

concept of “second-order volitions”, whereby the individual “wants a
certain desire to be his will”. The following questions may arise internally
in light of this more refined definition: Do I want this desire to lead to
37 See F-IWCA, 12-13. Frankfurt states that: “No animal other than man, however,
appears to have the capacity for reflective self-evaluation that is manifested in the forma-
tion of second-desires” (12). In a footnote to this observation, he says “I propose to use
the verbs ‘to want’ and ‘to desire’ interchangeably” (12-13).
38 See F-IWCA, 21, 64-66.
action? Do I want this desire to be “effective”? Do I want this desire to

be my “will”? For a more in-depth account of the idea of will, the fact
that second-order volitions are wholly personal, and related to the (much
more complex) distinction between “free will” and “freedom of the
will”, should be borne in mind.39
The nature of purely personal action is made clear by contrast with
the behavior of the “wanton”. The wanton is not concerned with the
type of desires that ought to be turned into actions: “he does not care
about his will”.40 Since he does not formulate second-order volitions, his
actions spring from more intense desires.41 The behavior of an unreflec-
tive agent may be glossed in the meanings encompassed by the word
“wanton” itself: “Wanton, adjective and noun. Of persons: Undisci-
plined, ungoverned; not amenable to control, unmanageable, rebellious.
Of children: Naughty, unruly”.42
Nevertheless, despite any possible defect, including the case of the
wanton, the will is “effective desire”. Thus, action and will remain bound
up with one another, for persons as well as for unthinking agents. Action
based on a desire discloses the will, irrespective of whether or not the
merits of such action are taken into consideration by the agent. How-
ever, the will is not a simple concept; it does not simply denote an “ef-
fective desire”. The complexity of the will is evinced by the so-called
“structure of the will of the person” which gives rise to different types
of mixed actions.
II
Reflection on desire attests to the complexity of personal experience of

action itself; it also enables a more detailed account of the different ways
in which the voluntary and involuntary are mixed in action. Frankfurt
39 See F-IWCA, 16-25.

40 F-IWCA, 16.
41 F-IWCA, 16-17
42 Oxford English Dictionary, (2012).
opens his essay entitled “Three concepts of free action”43 by framing the
issue in the following terms:
“There are many situations in which a person performs an action be-

cause he prefers it to any other among those he thinks are available to
him, or because he is drawn more strongly to it than to any other, and
yet is reluctant nonetheless to describe himself without qualifications as
having acted willingly” (F-IWCA, 47).
To address the contrasts between “what he did” or what “he wanted to

do”44 and what “he really wanted to do”, to use his own emphases,45
Frankfurt presents three possible scenarios of action: A, B and C. The
definition of each situation is outlined below, so as to facilitate an ana-
lysis of the meaning of mixed action in each case.
Type A situations are occasions of “worldly misfortune”; that is, situ-
ations in which external circumstances run counter to the individual’s de-
sires. Thus, the Type A agent,
“Given the alternatives he confronted, he preferred without reserva-

tion the one he pursued […] This discrepancy, between the world as it
was and as he wanted it to be, that supports his claim that he did not act
altogether willingly” (F-IWCA, 47).
43 F-IWCA, 47-57. In fact, the essay focuses on a single concept, Frankfurt’s idea of
action; the title of the essay stems from its purpose: it was written in response to another
paper: Locke, D., “Three concepts of free action”, Proceedings of the Aristotelian Society,
XLIX (1975), pp. 95-112.
44 The same expression is used in a later essay, which focuses on other aspects and
nuances of the matter: “We are particularly concerned with our own motives. It maters
greatly to us whether the desires by which we are moved to act as we do motivate us be-
cause we want them to be effective in moving us or whether they move us regardless of
ourselves or even despite ourselves. […] This means, moreover, that we are to some de-
gree passive with respect to the action we perform. For in virtue of the fact that we do
not unequivocally endorse or support our own motive, it can appropriately be said that
what we want […] is in a certain ordinary sense not something we really want” (F-IWCA,
163).
45 See F-IWCA, 47.
Type B situations involve “internal conflict”; that is, the circum-

stances contrary to the individual’s desires are internal –in other words,
they are desires with which the person does not identify but which,
nevertheless, prompt the action:
“This person’s denial that he has acted altogether willingly reflects his
sense that in the conflict from which his action emerged he was defeated
by a force with which, although it issued from inside of him, he did not
identify himself” (F-IWCA, 48).
Finally, Type C situations are defined in contrast to the other two

types of scenario:
“The situation of a person who succumbs to a threat because he is

unable to defy it is not, accordingly, of type A. Nor is it necessarily of
type B either, since there is no reason to assume that a person who acts
because of the irresistibility of a desire would prefer, given his alterna-
tives, to be motivated by some different desire. The situation is thus of a
distinguishable type, C, whose special characteristic is that the agent acts
because of the irresistibility of a desire without attempting to prevent
that desire from determining his action” (F-IWCA, 49).
In light of this description of the three situations of action,

Frankfurt argues that B and C be excluded from W, the category of free
actions46 and, as a result, relieved of moral responsibility.47 It may be
inferred from such criteria of classification that free action must meet
two conditions: first, that it contain a second-order volition (Type C situ-
ations are excluded on the grounds that they lack volition of the second
order); and second, that this second-order volition cause the action (this
condition precludes Type B situations). Type A, B and C actions reveal a
“will” in the sense of an “effective desire in action”; but only Type A ac-
tions are “free”.48 Hence, Type B and C actions are both (i) expressions
46 See F-IWCA, 50.

47 See F-IWCA, 57.
48 In his commentary on Aristotle’s work, Thomas Aquinas discerns three possible
principles of action: the voluntary, choice, and the will (In NE, III, Lecture 1, no. 1. Vid
of “will” and, at the same time, (ii) actions which are not free. What is
meant by conditions (i) and (ii)? Or to re-frame the question in other
terms: In what ways may these three types of action be regarded as
mixed actions?
Type C may be the starting-point in this context as it comprises ac-
tion carried out at the most basic level: the field of first-order desires.
The individual acts in accordance with a desire he regards as “irresist-
ible”. Incompatible desires –for instance, one’s money or one’s life– face
off as diametrically opposed forces before which the agent is merely a
spectator, who acts in the end on the desire that holds most sway within
him. The Type C agent is determined as such by his “inability” to control
his own action. Given that he does not reflect on the irresistible desire,
his behavior is similar to that of a wanton; not only does he act on irre-
sistible desire, he acts because of the irresistible desire.49 “[N]o second-
order volition plays a role in the economy of his desires”.50 The “will” of
such an agent is determined by the intensity of the desire, not by the in-
dividual himself. Since it does not encompass second-order desire, Type
C action may be described as voluntary only in a very vague sense. Once
the most intense desire asserts itself, the conflict of desires is resolved:
the will (effective desire) corresponds to the strongest desire. Rather than
a mixture of the voluntary and the involuntary, the action of a wanton
of this kind evinces a displacement of one desire by another, and con-
cerns only desires of the first order. If the agent were to say, “I didn’t do
what I wanted”, this is because there never was a “what I wanted”, as no
second-order volition was ever conceived. There was “will” –an effective
desire arose, prevailed and was enacted– but there was no personal par-
ticipation –that is, no act of identification or rejection on the agent’s
part.
The other two types of situation involve some reflection on desire,
albeit in opposed ways. The action of a Type A agent encompasses voli-
supra note 7). The distinction Frankfurt draws between the will (effective desire) and free
action may mirror the first two principles: the voluntary (including movements caused by
the sense appetite) and choice (which requires deliberation). The preliminary nature of
this paper precludes any more detailed discussion of this point.
49 See F-IWCA, 49.
50 F-IWCA, 50.
tion of the second order. The second-order volition is formed by the in-
dividual’s identification with a desire, and his wanting this desire to be
enacted. Unlike the Type C agent, the Type A agent sees the alternative
to the threat as a choice for which he may opt. The example Frankfurt
gives is of a bank clerk who decides to hand over the money to an
armed robber because he would rather lose the money than lose his
life.51 Aristotle furnishes a similar example of mixed action: the individ-
ual who throws his cargo overboard so as to survive the storm. Under
such circumstances, the (first-order) desire to save one’s life is incompati-
ble with the (likewise first-order) desire to safeguard one’s wealth. Which
of these desires ought to determine action? According to Frankfurt, if
the action is determined by the most intense desire, it is the action of a
“wanton” or Type C agent. However, given the fact that not every threat
is “irresistible”,52 it is possible that the Type A agent may choose which
desire to act on and, as a consequence, reject one desire in favor of an-
other. Thus, the Type A agent is not coerced by the threat;53 rather, he
acts (albeit within certain limits) in an autonomous way.54
In this regard, the Type A agent acts in a wholly personal way. He
makes a choice –in the second order of volition– by identifying himself
with a given desire. Throwing the cargo overboard is a mixed action
which shows that the individual’s life was more important to him than
his cargo. He acknowledges that throwing the cargo overboard was “not
what he wanted”, but he also understands that he did desire to do so at
51 See F-IWCA, 55. Frankfurt borrows this example from Don Locke, who also ac-
knowledges the choice made on the bank clerk’s part.
52 The distinction between resistible and irresistible threats, as well as the difference
between non-coercive and coercive conditions, is dealt with in detail in the essay
“Coercion and moral responsibility” (F-IWCA, 26-46). The essay also addresses other
forms of coercion: coercive offers.
53 See F-IWCA, 49-50. The Type A situation is also explored in other essays; see, for
instance, F-IWCA, 38: “Suppose P threatens to take from Q something […]. The choice
between the alternatives with which P’s threat confronts him is entirely up to him. He
must, of course, choose between them; he must decide whether to do what P demands
and escape the penalty; or whether to refuse to do it and incur the penalty. He is free,
however, to make either decision. […] The choice is his own, and there is no basis for
claiming that he bears anything less than full moral responsibility for whichever decision
he makes”.
54 See F-IWCA, 50.
that time because, otherwise, he would have died. His mode of action
was personal because he enacted his deepest desire, what he most
wanted. The involuntary dimension in the first order of desire –throwing
the cargo overboard– was rendered voluntary in the second order of
desire. Aristotle makes the same point: although it involves the involun-
tary, this “mixed action” would appear to be more voluntary. Thus, the
Type A agent is morally responsible for what he does; he carries out a
voluntary action by establishing a hierarchy of desires and choosing
which one is to be “his will”.
The third situation of mixed action entails the action of the Type B
agent. The “circumstances” running counter to his desire are not exter-
nal –as in Type A situations– but internal:
“In situations of Type B, it is the inner circumstances of his action

that are discordant with the agent’s desires. What motivates his action is
a desire by which, given the alternatives he confronts, he does not want
to be moved to act. There is a conflict within him, between a first-order
desire to do what he actually does and a second-order volition that this
first-order desire not be effective in determining his action. In other
words, he wants to be motivated effectively […] by some desire other
than the one that actually moves him to act as he does” (F-IWCA, 48).
Although this issue lies beyond the scope of this paper, Frankfurt’s
characterization of “effective desires” as “inner circumstances” is strik-
ing. The situation is exemplified by the drug addict who starts taking
drugs again even though his desire is not to do so. In what way are the
voluntary and the involuntary mixed in this action? The action of taking
drugs again corresponds wholly to a desire of the first order; it does not
defer to second-order desire –or, to be more precise, it runs counter to
the addict’s “second-order volition”. The conflict of desires emerges be-
tween him and the desire to take drugs that arises within him; this
conflict entails orders of desire –the thinking agent has reflected– but it
remains unresolved by the agent, who feels “unable”55 and defeated: “he
did not do what he wanted”. To a certain extent, therefore, the individ-
ual’s own action is alien to him:
55 F-IWCA, 48.
“In virtue of the discrepancy between the desire that motivates his
action and the desire by which he wants to be motivated, the agent in a
situation of Type B may not be morally responsible for what he does.
The desire that moves him is in one way, to be sure, indisputably his. But
it moves him to act against his own will, or against the will he wants. In
this respect it is alien to him, which may justify regarding him as having
been moved passively to do what he did by a force for which he cannot
be held morally responsible” (F-IWCA, 48).
Hence, can the addict’s decision to take drugs be described as his

“will”? If the notion of will is defined as effective desire, then the an-
swer to the question must be in the affirmative. However, Frankfurt’s po-
sition is clear: the desire “moves him to act against his own will”. The
Type B agent is similar to an individual subjected to coercion in at least
three ways: (i) the subject acts against his own second-order volition56
and, since (ii) he acts because he cannot avoid doing so, he is likewise (iii)
exempt from moral responsibility.57
In fact, the Type A agent and the coerced subject act in the way they
do because “they could not have done otherwise”.58 The impossibility of
doing otherwise accounts for the action. In the second part of the essay
“Three concepts of free action”, Frankfurt’s purpose is to argue –against
Locke and other proponents of the Principle of Alternative Possibili-
ties– that there is a profound difference between two drug addicts who
engage very different forms of voluntary disposition when they take
drugs:59 one is a willing addict; the other is an addict against his will: an
56 F-IWCA, 41-42: “An offer is coercive […] when the person who receives it is
moved into compliance by a desire which is not only irresistible but which he would
overcome if he could. In that case the desire which drives the person is a desire by which
he does not want to be driven. When he loses the conflict within himself, the result is
that he is motivated against his own will to do what he does. […] For his will when he
acts is a will he does not want to be his own. He acts under a compulsion which violates
his own desires”.
57 F-IWCA, 39: “If the victim’s desire or motive to avoid the penalty with which he is
threatened is […] so powerful that he cannot prevent it from leading him to submit to
the threat […]. He cannot effectively choose to do otherwise. It is only then that it may be
proper to regard him as bearing no moral responsibility for his submission”.
58 F-IWCA, 9.
59 See F-IWCA, 51-52.

unwilling addict. As a Type B agent, the latter finds within himself a de-
sire that moves him to act “against the will he wants”. “Neither [of the
addicts] can refrain from taking the drug”; however, the act of drug-
taking may be performed with an inner disposition of consent or rejec-
tion.60 The moral responsibility of the willing addict who takes drugs
freely is the same as that for a non-addict who takes drugs because he
enjoys them. Frankfurt notes that the unwilling addict, on the other
hands, does not act freely:
“The actions of the willing addict and of the non-addict both belong
to W [free actions], while that of the unwilling addict does not. In evalu-
ating the moral responsibility of the willing and unwilling addicts, Locke
tends to ignore the distinction between performing an action one is un-
able to avoid performing and performing an action because one is unable
to avoid performing it” (F-IWCA, 51-52).
The condition of “because one is unable” should be underscored;

without such a condition, it would be impossible to distinguish between
the moral responsibility of the willing drug addict and the moral signifi-
cance of the action of an addict who hates his addiction but, never-
theless, takes drugs. Frankfurt faults proponents of the Principle of Al-
ternative Possibilities for failing to acknowledge this distinction.61
The drug addict who wants to give up drugs but, nevertheless, con-
tinues to take drugs is not the master of his own will. If he takes drugs
“because he cannot” refrain from doing so, his drug-taking is not a free
action. In Frankfurt’s view, the Type B agent is not morally responsible
for what he does because he “acts against the will he wants”.62
60 Frankfurt is referring to this interior disposition when he notes that: “it is far from
apparent that a person who is not free to refrain from performing a certain action cannot
be free to perform it, or that he cannot perform it freely” (F-IWCA, 51).
61 See F-IWCA, 1-10.
62 F-IWCA, 48. In this regard, the limitations of the definition of will as “effective
desire” become clear: is drug-taking (an effective desire) truly not the “will” of the un-
willing drug addict? For him, drug-taking is an act that both (a) reflects the will of the ad-
dict, and (b) is carried out against the will the addict wants to have. This situation
prompts the following question: what is more constitutive of a person’s “will” –the reali-
zation of a desire or identification with the desire?
The desire not to take drugs remains within the Type B agent even
when he is taking drugs. The desire he wants to have –not to take drugs–
operates at a less effective, more internal level. The Type B agent retains
a capacity for identification that separates him, as a person, from his
first-order desires. He has a desire at a more internal (second order) level,
but it is ineffective. An action carried out on such a basis may be re-
garded as a “mixed action”: an effective desire with which the agent does
not identify is mixed with an ineffective desire –or to be more precise, a
desire whose intention for effectiveness has been defeated– with which
the individual does identify. According to Frankfurt, the action of drug-
taking for a drug addict who does not want such activity to be his will is
a mixed action in Aristotelian terms, but it is involuntary because the un-
willing addict is not morally responsible for the act.
In short, only Type A actions belong to the category of free actions.
Frankfurt summarizes the foregoing account of the three types of
agents in a comment on the unwillingness involved in Type A situations,
as compared with the involuntary status of Type B and Type C scenarios:
“The agent […] of Type A endorses the desire that moves him to act.
Despite his dissatisfaction with a state of affairs in which he finds the de-
sire to merit his endorsement, he is satisfied, given that state of affairs, to
be moved by the desire. Thus his action is in accordance with a second-
order volition, and the unwillingness with which he acts is of a different
character than the unwillingness with which agents act in situations of
Types B and C. This difference is reflected in the fact that he may be
morally responsible for what he does, whereas they are not” (F-IWCA,
50).
Frankfurt draws a distinction between two types of involuntary con-

duct: (i) the involuntary action of Type A agents (his rejection of exter-
nal circumstances) that does not disrupt the harmony of his internal de-
sires because his action enacts the will (effective desire) he wants; and (ii)
the more radical form of involuntary action carried out by Type B and
Type C agents, for whom the lack or rejection of identification with the
effective desire renders freedom and moral responsibility void.
Effective desires that lack the approval of individual agents render
the latter passive spectators to their own action; an inability to exercise
control that gives rise to an absence of moral responsibility: such (in)ca-

pacity may vary considerably from person to person, and even from one
situation to another.63
In a given situation, a person may feel “incapable of acting other-
wise”; however, in another, more serious situation, the same individual
may discover unexpected capacities within himself. Frankfurt gives the
example of a father facing a stark choice (his money or his life) who feels
unable to bear the suffering of violent death; but who in the face of an-
other terrible choice (his own life or the life of his child) proves capable
of “mobiliz[ing] his potential strength. [...] Knowing that what would be
lost is too precious to lose, on the other hand, may enable him to find re-
sources within himself upon which he is incapable of drawing in less
portentous context”.64 Frankfurt discusses this sense of control or com-
mand in terms that echo Aristotle:
“Sometimes we speak of threats as coercive even when we have no

particular evidence that their victims are incapable of defying them. This
is because there are certain penalties which we do not expect anyone to
be able to choose to incur. A person who surpasses this expectation
thereby performs not merely right or wrongly but with a certain heroic
quality” (F-IWCA, 40).
Aristotle likewise acknowledged the existence of punishments be-

yond the ordinary strength of human nature, and held that those who
yielded so as to avoid the threat –thus carrying out an action they should
not– were deserving of forgiveness. Frankfurt expresses no view regard-
ing the treatment the agent may deserve from others,65 perhaps because
what others say has no bearing on the internal state of the agent, who
acts in accordance with his own judgment.
63 See F-IWCA, 39.

64 F-IWCA, 40.
65 F-IWCA, 40: “We think that the person, although he was in fact quite unable to
control a desire, ought to have been able to control it. […] We may believe that the per-
son is morally responsible for his own ability […] It is fundamentally a matter of our lack
of respect for the person who has been coerced”.
Frankfurt explores the individual’s ability to control his own desires in

an essay entitled, “Identification and externality”,66 which details the
problems that arise in attributing passions to the person. An analysis of
conflicts of desires may at least help clarify what comes between the per-
son and his passions: decision.67 Thus, in the end, Frankfurt concedes
that a power stronger than the passions in the individual’s internal form
must be acknowledged: the person himself.68
Might not this relation of the person to his passions arise in the case
of Type B and Type C agents –in Type B agents, by recognizing the pos-
sibility of acting voluntarily while not acting; and in Type C agents, by
turning to some external support (to check into a drug rehabilitation
center, for instance), acknowledging one’s inability as an individual to
overcome addiction?
The strong person fulfills Cervantes’ dictum: “mishaps, which have
not strength to end our lives, have not power to finish our patience”.69 A
person who is able to control his own will has what Frankfurt calls “free-
dom of the will”. That Frankfurt refers to the three types of agents in
his description of such freedom is noteworthy:
“It is in securing the conformity of his will to his second-order voli-

tions, then, that a person exercises freedom of the will. And it is in the
discrepancy between his will and his second-order volitions, or in his
awareness that their coincidence is not his own doing but only a happy
chance, that a person who does not have his freedom feels its lack. The
unwilling addict’s will is not free [Type B]. […] It is also true, though in a
different way, that the will of the wanton addict [Type C] is not free” (F-
IWCA, 20-21).
Type A agents exercise freedom of the will; neither of the other two
types of agent (Type B, exemplified by the unwilling addict, and Type C,
66 F-IWCA, 58-68.
67 See F-IWCA, 68.
68 See F-IWCA, 68.
69 M. de Cervantes, Los trabajos de Persiles y Segismunda, Madrid, Calpe, 1920, Book 2,
Chapter 7. English translation: The Travels of Persiles and Sigismunda accessed at

http://www.ems.kcl.ac.uk/content/etext/e006.html#d0e3463.
by the wanton addict) has the ability to secure “the conformity of his
will to his second-order volitions”. However, there is a further, more
fundamental conflict that has not been discussed thus far.
III
Although the action of Type A agents is still marked by a conflict of de-

sires, a certain consensus also abides. Since he acts on a desire with
which he identifies, the Type A agent has at least chosen for himself
what he wants to do, and takes responsibility for it. Harmony reigns be-
tween the first and second orders of desire. Is the harmony in which the
Type A agent acts susceptible to disturbance? Frankfurt takes into con-
sideration the possibility of a more deep-seated conflict of desires: a
conflict that affects second-order desires. This is the kind of inner divi-
sion experienced by individuals who have not decided on the type of will
they want for themselves. The response to this conflict is related to the
wholeheartedness of identification, as Frankfurt argues in the essay
“Identification and wholeheartedness”.70
To Frankfurt’s mind, the freedom of the will is not only vulnerable to
attack ‘from below’: the rebellion of first-order desires, as in Type B
situations; it is also susceptible to disruption ‘from above’: by a lack of
clear definition in second-order volitions. Frankfurt probes the inner
depths of the person in his analysis of the second-order conflict of de-
sires, observing that:
“If there is an unresolved conflict among someone’s second order

desires, then he is in danger of having no second-order volition; for un-
less this conflict is resolved, he has no preference concerning which of
his first-order desires is to be his will. This condition, if it is so severe
that it prevents him from identifying himself in a sufficiently decisive
way with any of his conflicting first-order desires, destroys him as a per-
son” (F-IWCA, 21).
70 F-IWCA, 159-176.
The person is defined by the freedom of the will; a critical absence

of freedom would undermine the person as such.71 This phenomenon is
referred to as “ambivalence”, and it has been addressed by expert schol-
ars in the field.72 It is framed as the flip-side of the conflict experienced
by Type B agents73 which arises when a general consensus is not reached
regarding what the person wants at the most profound level:
“Another sort of inner division occurs when there is a lack of coher-

ence within the realm of the person’s higher-order volitions themselves.
This does not concern the relation between volitions and will. It is a mat-
ter not of volitional strength but of whether the highest-order prefer-
ences concerning some volitional issue are wholehearted” (F-IWCA,
165).
The clarification Frankfurt offers here suggests that the conflict does
not pertain to the field or sphere of “volitional strength”, weakness in
which may coerce Type B agents. The division is not due to a lack of co-
herence between the orders of desire; rather, it stems from a lack of clar-
ity in identification within the higher order:
“It has to do with the possibility that there is no unequivocal answer

to the question of what the person really wants, even though its desires
do form a complex and extensive hierarchical structure. There might be
no unequivocal answer, because the person is ambivalent with respect to
the object he comes closest to really wanting: In other words, because,
with respect to that object, he is drawn not only toward it but away from
it too” (F-IWCA, 165).
How may the conflict that arises between second-order volitions be

resolved? The response counseled here is deliberate decision: “One thing
a deliberate decision accomplishes […] is to establish a constraint by
71 See F-IWCA, 19.

72 Regarding this concept in Frankfurt’s work, see Swindell, J. S., “Ambivalence”, Phi-
losophical Explorations, vol. 13, no. 1 (2010), pp. 23-34.
73 See F-IWCA, 164-165.
which other preferences and decisions are to be guided”.74 Decision-

making requires a capacity for reflection, which is lacking in such agents
as children and animals. To decide is to decide for oneself because, to a
certain extent, decisions reflect and shape personal identity.75
A decision creates an intention and, thus, a rule to govern action.76
The hierarchical structuring of the orders of desire evinces an acknowl-
edgment of the authority pertaining to the highest order. A unique au-
thority is affirmed when the person decides in an unconditional way.
However, the fact that it exists need not mean that such authority is
insurmountable, as noted above. Although it may be “defeated by outlaw
forces” (such as the desire to take drugs that overwhelms the unwilling
addict; or any first-order desire that prevails over second-order volition
in Type B agency), there is no conflict within the authority itself.77
At the same time, the one who is “unconditional” in his decisions
knows what he wants and that he wants it. Nevertheless, he may still be
susceptible to a special kind of involuntary act. Even the unconditional
may experience an “inability to do”. In no sense is this “inability” a sign
of weakness. On the contrary, once the decision has been taken, the per-
son is unable to do anything that runs counter to what he identifies with;
Frankfurt cites the example of Luther in this regard, especially his re-
nowned assertion: “Here I stand; I can do no other”.78 Frankfurt draws on
the context of this example to ground the meaning of the term “volitio-
nal necessity”:
74 F-IWCA, 175.
75 See F-IWCA, 172.
76 F-IWCA, 175: “A function of decision is to integrate the person both dynamically
and statically. Dynamically insofar as it provides […] for coherence and unity of purpose
over time; statically insofar as it establishes […] a reflexive or hierarchical structure by
which the person’s identity may be in part constituted”.
77 See F-IWCA, 175.
78 F-IWCA, 86. Luther’s statement –“Here I stand. I can do no other” is usually
quoted in the following context: “Unless I am convinced by the testimony of the Scrip-
tures or by clear reason (for I do not trust either in the pope or in councils alone, since it
is well known that they have often erred and contradicted themselves), I am bound by
the Scriptures I have quoted and my conscience is captive to the Word of God. I cannot
and will not recant anything, since it is neither safe nor right to go against conscience.
May God help me. Amen” (Brecht, M., Martin Luther, trans. J. L. Schaaf, Philadelphia,
Fortress Press, 1985, vol. 1, p. 460).
“What he was unable to muster was not the ‘power’ to forbear, but
the ‘will’. I shall use the term ‘volitional necessity’ to refer to constraint
of the kind to which he declared he was subject […] A person who is
subject to volitional necessity finds that he must act as he does […] gen-
erally does not construe the fact that he is subject to volitional necessity
as entailing that he is passive at all […]. People are generally quite far
from considering that volitional necessity renders them helpless bystand-
ers to their own behavior. Indeed they may even tend to regard it as actu-
ally enhancing both their autonomy and their strength of will” (F-IWCA,
86-87).
The definition of volitional necessity79 is at the heart of one of

Frankfurt’s most important essays: “The importance of what we care
about”.80 This form of necessity enhances the freedom of the personal
will. Volitional necessity lies at the far end of the spectrum from the in-
capacity of the individual who sees himself as subject to coercion. Such
necessity is a freely self-imposed form of subjection:
“Moreover the necessity is to a certain extent self-imposed. It is gen-

erated when someone requires himself to avoid being guided in what he
does by any forces other than those by which he most deeply wants to be
guided […] On the other hand, even if volitional necessity is self-im-
posed there must be some respect in which it is imposed or maintained
involuntarily. The condition that it be self-imposed helps to account for
the fact that it is liberating rather than coercive” (F-IWCA, 88).
An action carried out in accordance with a volitional necessity com-

prises a mixture of the voluntary and the involuntary. However, the in-
voluntary dimension of volitional necessity receives the person’s consent.
79 There is a distinction here that should not be overlooked –between free and forced
volitional necessities, depending on whether or not the desire from which it stems arises
within the person: “free volitional needs have too little necessity in them. […] In order to
be morally interesting, on the other hand, a need must be radically distinct from a desire”
(F-IWCA, 108).
80 F-IWCA, 80-94.
Indeed, given that they prevent him from doing “unthinkable” actions,
such volitional necessities define the essence of the person as such:
“Under no circumstances, similarly, will a person will to do what is

for him unthinkable. As the set of its essential characteristics specifies
the limits of what a triangle can be, so does the set of actions that are
unthinkable for a person specify the limits of what the person can will to
do. It defines his essence as a volitional creature” (F-IWCA, 188).
An “unthinkable” action is rendered impossible by a non-physical in-

ability. A person would be physically unable to leave a room whose exits
are blocked or barred, for instance; but an unthinkable action is due to a
volitional incapacity. This incapacity prevails even when the person
thinks “he should perform the action, and that he has the desire to per-
form it […] because he cannot will to perform it”.81 Such actions recall the
third type of mixed actions described by Aristotle: “some acts, perhaps,
we cannot be forced to do, but ought rather to face death after the most
fearful sufferings”.82 In contrast to Aristotle, however, Frankfurt holds
that the “unthinkable is not essentially a moral category”.83 In this re-
gard, he gives an example of an unthinkable action from literature: Lord
Fawn’s failure to share a matter of intimate concern with a person infe-
rior to himself in breeding and class.84 Frankfurt is careful to distinguish
clearly between an unthinkable action as such and Type B and Type C
situations: an unthinkable action is not like an addiction, nor like an irre-
sistible impulse.85
Indeed, Frankfurt goes so far as to say that a subject who lacks voli-
tional necessities, or whose volitional necessities are wholly accidental is
not a person:
81 F-IWCA, 181.
82 NE, III, 1, 1110a 26-28.
83 F-IWCA, 182.
84 See F-IWCA, 183.
85 See F-IWCA, 182.

“There are individuals who are prepared to do anything, if the conse-

quences are sufficiently desirable –that is, if the price is right. What they
will to do is therefore never determined exclusively by their own nature
but is always a function of their circumstances. This means that as far as
their wills are concerned, they have only accidental characteristics. And
to the extent that a person is a volitional entity, such an individual is a
person with no essential nature at all” (F-IWCA, 188).
These statements are striking, and Frankfurt makes them in the con-
text of an argument against a view of action in which personal identity is
attenuated for pragmatic purposes. Such issues comprise the topic of the
essay entitled “Necessity and desire”.86 Frankfurt notes the existence of
“false needs”, created solely by (more or less completely fickle) desires,
which are therefore “gratuitous or perverse”.87 This account of the false
and perverse encompasses the critique of bullshit as hot air devoid of
any meaningful content,88 and a defense of the truth89 as a reality of
concern to everyone, thus functioning as a guide to human conduct.90
The assent prompted by the truth enables a more refined understand-
ing of the acceptance at the heart of love. In a series of lectures com-
piled in the book entitled The Reasons of Love, Frankfurt provides a wide-
ranging overview of love and the role it plays in the life of every person;
he traces a clear parallel between reason and love:
“It is in some degree precisely because loving does bind our wills that
we value it as we do […] The necessities with which love binds the will
are themselves liberating. There is a striking and instructive resemblance
in this matter that between love and reason […] The former guides us
most authoritatively in the use of our minds, while the latter provides us
with the most compelling motivation in our personal and social conduct
[…] While each imposes upon us a commanding necessity, neither entails
86 F-IWCA, 104-116.
87 F-IWCA, 115.
88 See F-IWCA, 117-133.
89 See Frankfurt, H., On Truth, New York, Alfred Knopf, 2006. This essay also draws
on the parallel between good and truth.

90 See F-IWCA, 82.
for us any sense of impotence or restriction. On the contrary, each char-

acteristically brings with it an experience of liberation and enhancement”
(Frankfurt, H., The Reasons of Love, New Jersey, Princeton University
Press, 2004, p. 64).91
At its heart, love is the source of need, of a volitional necessity. As

noted above, because it springs from the will itself, such necessity cannot
be regarded as being in any way coercive.92 The power of love “does not
rest upon the reliability of arguments or of evidence. It rests upon confi-
dence in ourselves”.93 Love conditions the voluntary and is, in turn, con-
ditioned by personal identity. Hence, it may be concluded that love –“the
controlling tendencies and responses”– expresses the identity of the in-
dividual.94 Personal identity is related to the needs of the will.95 A break-
down in love is a sign of a broken identity, which would also lead to a
rupture in the field of the voluntary:
“The psychic integrity in which self-confidence consists can be rup-

tured by the pressure of unresolved discrepancies and conflicts among
the various things that we love. Disorders of that sort undermine the
unity of the will and put us at odds with ourselves. The opposition with-
in the scope of what we love means that we are subject to requirements
that are both unconditional and incompatible. That makes it impossible
for us to plot a steady volitional course. If our love of one thing clashes
unavoidably with our love of another, we may well find it impossible to
accept ourselves as we are” (F-RL, 50).
91 The text is cited hereafter as F-RL, followed by the relevant page number(s).
92 See F-RL, 46-47.
93 F-RL, 49.
94 See F-RL, 50.
95 F-RL, 50: “The necessities of a person’s will guide and limit its agency. They deter-
mine what he may be willing to do, what he cannot help doing, and what he cannot bring
himself to do. They determine as well what he may be willing to accept as a reason for
acting, and what he cannot bring himself to count as a reason for acting. In these ways,
they set the boundaries of his practical life; and thus they fix his shape as an active
being”.
Reflection on mixed actions may gradually lead to an understanding

of the type of divide that may be discerned at the heart of every person
as such. As categories, the voluntary and the involuntary have an impact,
each in its own particular way, on the various areas of life that are open
to the action of the will: action, first-order desires, second-order voli-
tions, the act of decision and, ultimately, love and personal identity. This
close reading of Frankfurt’s work from the perspective of mixed action
has disclosed a number of features concerning the unity underlying the
agency of the human person.
Universidad Panamericana
Aguascalientes, México
References
International Workshop on Belief, Responsibility and Action, Valencia, 12-

14/XI/2008, in Philosophical Explorations, 12/2 (2009).
Aristotle, Nichomachean Ethics, trans. W. D. Ross, Kitchener, Batoche.
(Cited as NE).
Brecht, M., Martin Luther, trans. J. L. Schaaf, Philadelphia, Fortress Press,
1985.
Buss, S. – Overton, L. (eds.), Contours of Agency: Essays on Themes from
Harry Frankfurt, Cambridge, MIT Press, 2002;
Cervantes, M. de, Los trabajos de Persiles y Segismunda, Madrid, Calpe, 1920.
Frankfurt, H., The Importance of What We Care About. Philosophical Essays,
New York, Cambridge University Press, 1988. (Cited as F-IWCA).
——, The Reasons of Love, New Jersey, Princeton University Press, 2004.
(Cited as F-RL)
——, On Truth, New York, Alfred Knopf, 2006.
González, A. M., “Action in a Narrow and in a Broad Sense”, in
González, A. M. – Vigo, A. G. (eds.), Practical Rationality. Scope and
Structures of Human Agency, Hildesheim – New York, Olms, 2010, pp.
123-168.
Kane, R., (ed.), The Oxford handbook of free will, Oxford – New York,
Locke, D., “Three concepts of free action”, Proceedings of the Aristotelian
Society, XLIX (1975), pp. 95-112.
Schumacher, C., Zwecke und Mittel bei Aristoteles und Harry Frankfurt,
Hamburg, GRIN Verlag, 2007.
Swindell, J. S., “Ambivalence”, Philosophical Explorations, vol. 13, no. 1
(2010), pp. 23-34.
Thomas Aquinas, Commentary on the Nicomachean Ethics, trans. C. I.
Litzinger, Chicago, Henry Regnery Company, 1964.
——, Summa Theologiae, trans. Fathers of the English Dominican

Province, K. Knight, Online Edition, 2008.
Widerker. D., – Kenna, M. Mc., (eds.), Moral Responsibility and Alternative
Possibilities: essays on the importance of alternative possibilities, Burlington,
VT, Ashgate, 2003.
PRACTICAL KNOWLEDGE AND
PERCEPTION1
EVGENIA MYLONAKI
In this paper, I examine the relation between intentional action and mo-
rality from the perspective of their epistemology. In particular, I study
the relation between the knowledge one has when one knows what one
is doing in acting intentionally (knowledge in acting, for short) and the
knowledge one has when one knows what one ought to do in the partic-
ular circumstances one finds oneself and not in general (knowledge in the
circumstances, for short); and I focus on a problem concerning the role of
perception in the Anscombean conception of knowledge in acting and
the Murdochean conception of knowledge in the circumstances.2
1 For invaluable help with earlier drafts of this paper I would like to thank John
McDowell, Kieran Setiya, Karl Schaffer, Matt Boyle, James Pearson, Greg Strom, Aristei-
des Baltas, Alexandra Newton, and Steven Kyle. I would also like to give special thanks to
Andrea Kern for a series of extremely helpful and exciting discussions on almost all the
issues touched on in this paper, to Patricio Fernández for his sharp criticism at the con-
ference on Theories of Action and Morality at the University of Navarra, and to Konstandi-
nos Sfinarolis for his support and understanding.
2 For a brilliant discussion of the relevance of Murdoch’s book The Sovereignty of Good
to contemporary discussions see Setiya, K., Murdoch on the Sovereignty of Good, (unpublish-
ed). For explicit appropriations of Murdoch’s view in discussions of practical knowledge
see Blum, L., “Moral Perception and Particularity”, Ethics, 101/4 (1991), pp. 701-725;
Bagnoli, C., “Moral Perception and Knowledge by Principles”, in Hernandez, J. (ed.),
New Intuitionism, London, Continuum, 2011, pp. 89-106; Clarke, B., “Imagination and
Politics in Iris Murdoch’s Moral Philosophy”, Philosophical Papers, 35/3 (2006), pp. 387-
411; McDowell, J. H., “What is the Content of an Intention in Action?”, Ratio, 23/4
(2010), pp. 415-432. For discussions on Anscombe’s book Intention and practical knowl-
edge see for instance Moran, R., “Anscombe on ‘Practical Knowledge’”, in Hyman, J. –
Steward, H. (eds.), Agency and Action (Royal Institute of Philosophy Suppl. 55), Cambridge,
Cambridge University Press, 2004; Falvey, K., “Knowledge in Intention”, Philosophical
Studies, 99/1 (2000), pp. 21-44; Setiya, K., “Knowledge of Intention”, in Ford, A. –
Hornsby, J. – Stoutland, F. (eds.), Essays on Anscombe’s Intention, Cambridge, MA, Harvard
University Press, 2011; McDowell, J. H., “How Receptive Knowledge relates to Practical
Knowledge”, (unpublished); etc. For more elaborate renderings of the Anscombean
tradition in the philosophy of action see Thompson, M., Life and Action: Elementary Struc-
tures of Practice and Practical Thought, Cambridge, MA, Harvard University Press, 2008;
Rödl, R., Self-Consciousness, Cambridge, MA, Harvard University Press, 2007; etc.
242 Evgenia Mylonaki
Before I motivate the problem, a few introductory words are in order

about the Anscombean conception of knowledge in acting and the
Murdochean conception of knowledge in the circumstances. On certain
so-called neo-Aristotelian accounts,3 such as the one Murdoch lays out in
her book The Sovereignty of the Good and her paper Vision and Choice in Mo-
rality,4 knowledge in the circumstances is grounded in a certain sort of
perception. Say I’m on the phone with an old friend, Mark, who is too
chatty for my taste but who happens to be going through a rough phase
in his life. And now say someone sees that I haven’t uttered a word for
over an hour and asks, Why don’t you just hang up the phone on him? And I re-
ply, I can’t hang up the phone on him now, I can hear that he’s vulnerable. The
thought is simple. Knowledge in the circumstances is not the knowledge
of what one ought to do in general; for instance the knowledge that one
ought not to hang up the phone on one’s friends. In our example, knowl-
edge in the circumstances is the knowledge I have of not hanging up the
phone on Mark as what I ought to do right here, right now. And now the
neo-Aristotelian twist is that I know this because I hear that Mark is
vulnerable; or else that this knowledge of mine is grounded in my per-
ception of Mark’s vulnerability. To put it more formally, my knowledge
of not hanging up the phone on Mark counts as my knowledge of what
I ought to do in the circumstances in virtue of my perception of Mark’s
vulnerability. From now on I will be calling the perception that is taken
to constitute the ground of knowledge in the circumstances moral percep-
tion.
Now one might think that knowledge in acting is also grounded in
perception, because an intentional action is also a happening and hence
3See for instance McDowell, J. H., “Virtue and reason”, in McDowell, J. H., Mind,
Value, and Reality, Cambridge, MA, Harvard University Press, 1979; McNaughton, D. A. –
Rawling, P., “Unprincipled Ethics”, in Hooker, B. – Little, M. O. (eds.), Moral Particularism,
Oxford, Oxford University Press, 2000, pp. 256-75; Little, M., “Moral Generalities
Revisited”, in Hooker, B. and Little, M. O. (eds.), Moral Particularism, pp. 276-304; Lance,
M. – Little, M., “Defending Moral Particularism”, in Dreier, J. (ed.), Contemporary Debates
in Moral Theory, Oxford, Blackwell, 2006, pp. 305-321; Dancy, J., Ethics without Principles,
Oxford, Clarendon Press, 2004; Nussbaum, M., Love’s Knowledge, Oxford, Oxford Univer-
sity Press, 1990; etc.
4 Murdoch, I., The Sovereignty of Good, London, Routledge and Kegan Paul, 1970;
Murdoch, I., “Vision and Choice in Morality”, Proceedings of the Aristotelian Society, Supple-
mentary Volume, 30 (1956), pp. 32-58.
Practical Knowledge and Perception 243
an observable thing. Say that I’m cutting basil because I’m making pesto.
And now say that someone walks into the kitchen and upon seeing me at
the cutting board asks: What are you doing there? And I reply, I’m cutting basil
because I’m making pesto. Now, on this line of thought, one might say that I
know what I am doing by observing what is happening in the world as a re-
sult of my intending or trying (or by observing what is happening in the
world and having introspective knowledge of what I intend or try to do).
In other words, one might say that this knowledge, knowledge in acting,
is a case of common or garden observational knowledge. But Anscombe
claims in her Intention5 that the topic of intentional action is non-circu-
larly captured by specifying the agent’s distinctive knowledge of what she
is doing when she is acting intentionally (knowledge in acting). And, she
argues that we should specify the agent’s knowledge in acting by distin-
guishing it from common or garden observational knowledge. To put it
as generally as possible, it seems that for Anscombe the intentionality of
intentional actions goes hand in hand with the capacity of the agent to
know non-observationally what she is doing.
A first formulation of the problem concerning the role of perception
in the grounding of these two forms of knowledge can now be given.
Both knowledge in acting and knowledge in the circumstances are forms
of practical knowledge: knowledge that brings about what it under-
stands, as opposed to receptive knowledge, which derives from what it
understands. In its simplest formulation this is the claim that if I hadn’t
known I was cutting basil in order to make pesto I simply wouldn’t have
been cutting basil in order to make pesto; and if I hadn’t known I ought
not to hang up on Mark because he was vulnerable, I simply wouldn’t
have been doing what I ought to do in not hanging up the phone on
him. But if moral perception –perception that, as we said before,
grounds knowledge in the circumstances– is to really deserve the name
(i.e. if it is not perception only so-called), then it has got to be under-
stood as sensible affection and thus as a form of receptivity. To perceive
Mark’s vulnerability and thus know what to do in the circumstances
ought to be no more mysterious and no less receptive than seeing your
5 Anscombe, G. E. M., Intention, second edition, Ithaca, Cornell University Press,

1957.
face is bruised and thus know that you’ve been beaten up. But now
Anscombe seems to suggest that it is the mark of practical knowledge
that it does not rest on sensible affection and thus receptivity. For
Anscombe, to say that I see what is happening in the world and thus
know what I am doing is to say that my knowledge is not practical and
my doing not intentional. The idea seems to be this: sensible affection
seems to amount, when things go well, to that knowledge of things in
the world which is derived from what is known, receptive knowledge.
But we said above that practical knowledge is knowledge that brings
about what it understands. Thus, knowledge grounded in sensible recep-
tivity would simply not be of the right kind to be practical. In the face of
this then, how can the neo-Aristotelians or Murdoch want to explain a
species of practical knowledge (knowledge in the circumstances) as
knowledge grounded in perception? How can one form of practical
knowledge (knowledge in the circumstances) be identified as grounded
in moral perception, while the other form of practical knowledge
(knowledge in acting) is identified as contrasting with knowledge
grounded in sensible receptivity (knowledge by observation)? What I
plan to do over the course of this paper is address this tension.
In the first part I will argue against a tempting two-factor view of
knowledge in the circumstances. On this view, knowledge in the circum-
stances is the result of the joint exercise of two capacities for knowledge,
a receptive capacity through which we know what happens to be the case
in the agent’s environment and a practical capacity through which we
know what to do in general. On this account then, knowledge in the cir-
cumstances is the knowledge we have when these two capacities for
knowledge are exercised jointly.
In the second part I will argue against the attempt to posit moral per-
ception as a special form of sensible receptivity such that it may be the
ground of practical, i.e. non receptive, knowledge. On this picture, it is
ordinary sensible affection that constitutes the ground of receptive
knowledge and this leaves it entirely open that a special sort of sensible
affection may constitute the ground of practical knowledge.
In the third part I will suggest the beginnings of an alternative ac-
count. On my view we should distinguish moral from ordinary perception by
distinguishing between distinct forms of knowledge. On my account
moral perception is shorthand for a form of knowledge through which
we may know practically what is already available to one’s receptive

knowledge.
In the fourth part I will deal with a problem that my view seems to
face. If knowledge in the circumstances is the knowledge which knows
practically exactly what is available through one’s receptive knowledge,
then we have to say that one and the same thing could be known in two
ways; both practically and receptively. But, if there are two ways of
knowledge shouldn’t there also be two objects known? If we answer this
question in the affirmative, then how can I say that in practical knowl-
edge we may know practically what is already available to us in our recep-
tive knowledge?
In the fifth and final part I will suggest that Anscombe in her Intention
answers the same question with regard to knowledge in acting and I will
claim that we could give a similar answer to our own question with
regard to knowledge in the circumstances. Knowledge in the circum-
stances and knowledge in acting may thus be seen to constitute two spe-
cies of one and the same capacity for knowledge: the capacity for self-
knowledge.
2. Moral perception as ordinary perception
The most immediately appealing way to address the tension between

Anscombe’s and Murdoch’s accounts is the following. If practical knowl-
edge is knowledge that brings about what it understands, then it is to be
contrasted with receptive knowledge, knowledge that is derived from
what it understands. And if this is so, Anscombe is simply right to ex-
clude from the realm of the species of practical knowledge she considers
(knowledge in acting) perception or any form of receptivity for that mat-
ter. Murdoch on the other hand is simply confused on this front. The
form of practical knowledge she considers, knowledge in the circum-
stances, cannot qua practical knowledge be grounded in perception, or
any form of receptivity for that matter. On this view, knowledge in the
circumstances is the result of the joint exercise of two capacities for
knowledge: a receptive capacity through which we know particular facts
that happen to obtain in the agent’s environment and a practical capacity
through which we know general rules for the setting of ends, or what to
do in general. On this view then moral perception is nothing more than

ordinary (i.e., no more mysterious and no less sensible than) receptive
knowledge of what happens to be the case in the agent’s environment.
And practical knowledge is knowledge of what the agent’s general rules
prescribe for her. Knowledge in the circumstances then is the result of
the joint exercise of these two capacities for knowledge. And the object
of knowledge in the circumstances is partly known receptively and partly
known practically. From now on I will be calling this view revisionist for it
attempts to resolve the tension between Murdoch’s and Anscombe’s ac-
count by revising Murdoch’s account.
It will help to understand this revisionist account to present knowl-
edge in the circumstances in terms of the reasoning structure it is con-
stituted by. Knowledge in the circumstances may be understood in terms
of the reasoning that takes us from premises stating general rules about
what one ought to do to conclusions about what we ought to do in par-
ticular circumstances, via premises that concern the nature of these
circumstances. And something about this reasoning, it is commonly
thought, is what qualifies knowledge in the circumstances as practical
knowledge; I will call it practical deliberation or deliberation for short.6 On
this widely accepted view, deliberation involves the application (or the
specification, etc.) of general rules about what one ought to do, in circum-
stances that are particular. The form these rules may take varies accord-
ing to different theories. On some, the rules are general precepts of the
form “Do x”; on others, they are general truths of the form “X is
good”; etc. But the idea is simple. There are general rules that determine
what one ought to do or what it is good to do, etc., in general. And the
morally trained agent is the one who has been inculcated with these prin-
ciples in such a way as to be able to apply (or specify, understand, in-
terpret, etc.) these principles in particular circumstances, whenever the
agent judges that the situation demands it.
Within the context of this widely accepted idea of how we come to
have knowledge in the circumstances we may formulate two models of
how perception figures in deliberation, both of which deny that there is
any distinctive cognitive activity we might want to call moral perception.
6 I will come back to this idea at the end of the paper.

On the first model, deliberation proceeds as follows: the agent has at her
disposal a set of general rules pertaining to what she ought to do. Say
that a given agent has been inculcated with the principle that children
ought to be protected from abuse. Now, on this model, the agent ob-
serves her environment, i.e. registers what happens to be the case in the
circumstances she finds herself in. And now say that in this context the
agent also registers that a child happens to be abused. Given the exis-
tence of this particular perception the agent is in a next phase able to ap-
ply her general rule to the particular case presented in her overall percep-
tion of her circumstances: If I ought to protect children from abuse and
this child here is, as I among other things perceive, being abused, then it
follows that I ought to protect this child here from abuse. (If I ought to
x in circumstances y and I perceive that my circumstances here and now include y,
then I ought to here and now x).
Alternatively, on the second model, one determines a most general
end; i.e. adopts a most general rule. Say one has adopted the rule that
one ought to help one’s fellow human beings as one’s most general end.
To carry out this end, one needs to further figure out the means to this
end. And to do this, one needs to have a certain sort of technical experi-
ence-knowledge of how to do things that is based on observation of
means-end connections of the form “y is the best/easiest/necessary/etc.
way to do x”. On this model, the agent determines how an end is to be
achieved either by retrieving or building up from her repository of pre-
viously perceived means-end connections one particular means-end con-
nection. This is what in a next phase enables her to specify the end she
adopts: If I ought to do x, and –as I have perceived– the means to x is y,
then I ought to do y; with certain qualifications of course, such as that y
not be immoral, etc.
Now in a world in which we could not observe our environment
(sensibly receive various features of the circumstances we find ourselves
in) it is doubtful that we could ever act. And in a world in which we
could not gather (technical) experience through our affection by observ-
able means-end connections it is doubtful that we could ever act well. But
the question now is, is this observation and technical experience suffi-
cient to fit the bill of what we originally wanted to call moral perception?
The answer begins to appear negative when we appreciate a certain
feature of the way we ordinarily assess others for their actions. The
thought is not new. Positively assessing others for their actions presup-
poses not merely their alignment with what we take to be the right gener-
al rules or principles, but also a certain sort of sensitivity that allows
them on each occasion to privilege some of the features and aspects of
the situations with which they are confronted. We praise our child not
just for giving half of her sandwich to her hungry classmate, but also for
noticing the hunger. We are moved when a new acquaintance drops a line
to see how we are doing in a time of crisis not just for doing what is the
friendly thing to do, but also for seeing us as a real friend would. In fact, if
we are told that our child did something generous upon being told that
this was an occasion for generosity and that our new acquaintance gave
us a call in order to do what is appropriate by the standards of friend-
ship, we may get disappointed or even unsettled.
But now the problem with the aforementioned accounts of moral
perception is that this sort of sensitivity –that which we are most in-
clined to praise on at least some occasions– cannot be presented as an
instance of simply registering (receptively knowing) what happens to be the
case (in one’s environment or with regard to how to do things) or ordinary
perception, as I will be calling it from now on. Ordinary perception may be
presupposed for the noticing of the hunger above, but it does not
amount to that noticing. This kind of noticing is more aptly described as
a sort of singling out of a feature of the agent’s situation in the agent’s
perception of what happens to be the case as meriting a certain kind of
response. Perhaps, it is more aptly described as a kind of perceiving
which presents a feature of the agent’s circumstances not just as what
happens to be the case but as what sticks out in a certain way; in partic-
ular as what calls for action. And it is this singling out or distinctive
perceiving which we would like to single out by the name of moral percep-
tion.
Now, one might object that this singling out of a feature of the
agent’s situation as meriting a certain kind of response is not a sort of
perception to be distinguished from ordinary perception, but is an activi-
ty performed by the rule that figures in the major premise of the prac-
tical syllogism. On this view, ordinary perception or receptive knowledge
is that through which the agent knows what happens to be the case in
her circumstances, and practical knowledge is that through which the
agent knows what general rule to apply in the circumstances. In our
abuse example, the agent has at her disposal all the facts of the case
through an exercise of her capacity for receptive knowledge. And then
the general rule that one ought to save children from abuse picks out one
of these facts –that a child is being abused– as what the rule is to be ap-
plied to; i.e. as what commands the agent to adopt this general rule as
her end in the circumstances.
Now this may seem entirely unproblematic, but it is not. If it is the
general rule itself that provides the criteria for its own applicability (it is
after all what picks out the relevant feature of the circumstances as rele-
vant) then in cases of conflict between rules, there will be no way of de-
termining what general rule to set as an end. For each general rule will be
privileging that feature of the circumstances that rightly (by the lights of
each rule) calls out for each such rule’s applicability. And even if there is
no actual conflict, the question still remains: out of the large repository
of practical rules that may each time determine the agent’s end, what de-
termines which general rule is each time to be activated; i.e. which rule is
each time to be applied to the agent’s circumstances? The distinctive role
of moral perception was reserved by the neo-Aristotelian accounts as
precisely a way to solve this problem. To do away with the distinctive
character of moral perception is to leave this problem pending.
But one might object that this problem could be solved without
abandoning the revisionist view. Thus one might insist that the agent
perceives various features of the circumstances in which she finds her-
self through an exercise of her capacity for receptive knowledge; say that
there is a child which is getting abused, that the person who abuses the
child holds a gun, etc. But then one might add that the agent is the kind
of person who has as her most general end the rule that one ought to help
others in need. And now, one could say that it is this most general end
that picks out one of the deliverances of the exercise of our capacity for
receptive knowledge (ordinary perception) –say that a child is being
abused– as what calls for the activation of a less general rule as her end-
say to save children from abuse. In this way then, one might still seem
able to insist that knowledge in the circumstances is the result of the
joint exercise of two knowledges: a receptive knowledge through which
the agent knows the particular facts that happen to obtain in her en-
vironment and a practical one through which the agent knows what less
general rule to set as an end for herself.
But now the problem is this: in general a capacity is identified as the

specific capacity that it is by reference to its active exercise. We learn
what a capacity to see is by examining instances of seeing things. And we
learn what a capacity to judge is by looking at cases of judging that x. If
something like this is true in general and we take it that we have the
capacity for practical knowledge, then this too would need to be speci-
fied as such in terms of its active exercise. Knowledge of what one
ought to do falls, as we have assumed in this paper, under the genus of
practical knowledge. So it would have to be the case that in the active ex-
ercise of the capacity to know what one ought to do, this knowledge
would be such as to bring about its object in understanding it. Now if
the object that is brought about in the understanding of knowledge of
what one ought to do is the individual doing that falls under the general
rule,7 then this object is only knowable in knowledge in the circum-
stances. And so the active exercise of our capacity for practical knowl-
edge would have to be the knowledge of what the agent ought to do in
the circumstances she finds herself in and not in general. But this is by
the lights of the view under consideration the outcome of a joined ex-
ercise of two forms of knowledge –one receptive and one practical–
which are responsible for knowing different parts of what is known in
knowledge in the circumstances. For what is known in knowledge in the
circumstances is on this view known partly receptively and partly
practically. But then the definition of our capacity for moral practical
knowledge in terms of its active exercise would be a definition that
would essentially involve the capacity for another form of knowledge,
the receptive form of knowledge. For the object of knowledge in the cir-
cumstances is by the very lights of this view not knowable practically in
its entirety. Thus, the two factor conceptions of knowledge in the cir-
cumstances are bound to be misguided. The perception we want to call
7 One could here object that the object brought about and that is understood in
knowledge of what one ought to do is “the highest good” or “happiness”, or something

like that, and that the individual action is “merely” connected to that object as a means,
which is known through another species of practical knowledge. I cannot argue against
this objection here; I can only point out the absurdity of conceiving of knowledge of
what one ought to do as practical without conceiving the individual doing that it is pro-
ductive of as the (result of the) exercise of this knowledge. I owe this objection to Greg
Strom.
moral must play a role in the constitution of knowledge in the circum-

stances when we take this knowledge to be practical; and not when we
take this knowledge to be receptive. But then, must we not assume that
moral perception must be a special form of receptivity such that it may
ground or be involved in knowledge in the circumstances qua practical
knowledge? It is this view that I will take up and argue against in the next
section.
3. Moral perception as a special sort of sensible receptivity
In the previous section we suggested that we should define moral per-

ception as what delivers insight into what feature of the situation with
which the agent is each time confronted rightly calls out for the application of
a general rule. It is, I believe, this perception that Murdoch explains as the
loving attention or gaze to the particularities of the situation with which the
agent is confronted.8 And it is the object of this perception that, I be-
lieve, David Wiggins explains as “what strikes the person as the in the si-
tuation most salient feature of the context in which he has to act”;9 in other words
as the sensitivity not merely to features of one’s situation but to features
of one’s situation as the practically salient features in the situation. If this is
right, then the sort of perception that is involved in practical deliberation
cannot be ordinary perception in the sense given by the view described
in the previous section; i.e., an instance of receptive knowledge. For to
say that this sort of sensitivity (moral perception) is sensitivity to salience
just is to say that the (salient) feature of the situation is not perceived in
the same way as the rest of the features of the situation. It may then
begin to seem as if moral perception is a species of receptivity all right,
only of a very special sort: one that allows it to ground knowledge in the
circumstances qua practical knowledge. For how else, might one wonder,
can we explain a sort of perception that is a species of receptivity all right,
but is such as to ground practical knowledge?
8See for instance Murdoch, I., The Sovereignty of Good, pp. 17-23.
9Wiggins, D., Needs, Values, Truth: Essays in the Philosophy of Value, Oxford, Oxford
University Press, 1998, p. 234.
But just what might be this special way in which the salient feature of
the situation is perceived? One family of views posits moral perception
as a special receptive faculty alongside our other (perhaps visual, tactile,
etc.) receptive faculties, or as a special activation of our ordinary receptive
faculties, or as the result of the co-operation between perceptual and
non-perceptual capacities. A careful and thorough examination of each
of the views that assume that the distinctness of moral perception lies in
its character as a special receptive faculty would take me far afield. What
I shall do instead is draw a taxonomy of these views that will allow me to
give a few promissory notes about their dubiousness and a final reason
for their failure.
So, if on the one hand we assume that moral perception is to be
specified as a mere sensible receptive faculty among or alongside our or-
dinary sensible receptive faculties (say sight, touch, etc.), we can suggest
that (1) The objects of the moral receptive faculty are out there and im-
pinge on our sense organs in more or less the way that the ordinary ob-
jects of our other receptive faculties are out there and affect our senses.
The only difference is that they are very special objects. They are (or are
such as to give rise to) moral sensibles as opposed to ordinary sensibles
(however exactly these sensibles might be conceived in the non-moral
case).10 Now the epistemology of these moral sensibles might be con-
ceived in a variety of ways: (1a) One could suggest that moral sensibles
are received by a special receptive faculty. The immediate problem with
this view is that it does not make much (non-figurative) sense to posit a
special, i.e., moral receptive faculty without also positing the existence of
a moral faculty that can be contrasted with our sensible faculties as
such.11 As John McDowell has argued, this would constitute solving a
mysterious problem by positing an even more mysterious and occult fac-
ulty.12 And it would be to ignore the plain fact that moral perception
10 By the term sensibles I simply refer to what the perception of an object presents us
with. I don’t mean to take a side in the various debates on just what this thing might be.
11 Intuitionism has been the paradigmatic form of this account. See for instance
Moore, G. E., Principia Ethica, Cambridge, Cambridge University Press, 1903; Ross, D.,
The Right and the Good, Oxford, Clarendon Press, 1930.
12 See McDowell, J. H., “Values and Secondary Qualities”, in McDowell, J. H., Mind,
Value, and Reality.

ought to be no more mysterious and no less sensible than ordinary sensi-

ble affection. (1b) Alternatively, one might suggest that moral sensibles
are perceived as a result of the co-operation between ordinary perceptual
and non-perceptual (say intellectual) faculties.13 But this would face a
problem analogous to McDowell’s problem above: just as (1a) above
merely posits a mysterious and occult faculty so (1b) posits some myste-
rious and occult way of co-operating for two ordinary faculties.14 (1c)
Last, one could suggest that moral sensibles are received by our ordinary
receptive faculties. (1c.1) On a weak reading of moral sensibles, moral
perception ought to be construed on the model of chess perception. On
this view, moral sensibles are like chess sensibles in that they do not dif-
fer from other sensibles in their form, but in their subject-matter.15 Plau-
sible though it may be, this view lacks explanatory power. For moral sen-
sibles were posited exactly in order to explain the possibility of per-
ceiving moral objects in the way one perceives chess pawns. (1c.2) On a
stronger reading, moral sensibles are exactly like size or mass sensibles.
On this view, we can make room for moral perception only if we devel-
op a less chauvinistic view of the senses. A less chauvinistic view of the
senses would allow us for instance to assume that the eye is hit by light in
more or less the way it is hit by injustice. But this view would commit us
to a whole bunch of counter-intuitive claims, such as saying that failure
to perceive injustice might be correctable by a visit to the ophthalmolo-
gist.
On the other hand one could suggest that (2) The objects of moral
perception are moral super-sensibles that supervene on non-moral sen-
sibles in such a way as to be perceived indirectly.16 Talk of moral percep-
13 For a refined view along these lines on which the perceptual capacity in question is
an intellectualized perceptual ability, see Watkins, M., – Jolley, K., “Polyanna Realism:
Moral Perception and Moral Properties”, Australasian Journal of Philosophy, 80 (2002), pp.
75-85.
14 I owe this point to Greg Strom.
15 Wright, J. C., “The Role of Moral Perception in Mature Moral Agency”, in
Wisnewski, J. J. (ed.), Moral Perception, Newcastle, Cambridge Scholars Publishing, 2008,

pp. 1-24.
16 For variants of these views see Audi, R., “Moral Perception and Moral Knowl-
edge”, Proceedings of the Aristotelian Society, Supplementary Volume, 84 (2010), pp. 79-97;
Cullison, A., “Moral Perception”, European Journal of Philosophy, 18/2 (2010), pp. 159-175,
Norman, R., “Making sense of Moral Realism”, Philosophical Investigations, 20 (1997), pp.
tion would in this case be used to signal the existence of a distinctive

epistemological connection: say immediate perception of non-moral
sensibles plus inference to the existence of moral properties, or immedi-
ate perception of non-moral sensibles plus non-inferential judgment that
the moral properties constituted by the non-moral properties do in fact
exist, etc. But this move makes talk of moral perception figurative at
worst and redundant at best. For the point of calling moral perception
perception was to acknowledge the (seeming at least) fact that the objects
of moral perception are “objects of sensible cognition” –i.e., no less
available to our receptive faculty through their affection of our sensibility
than the objects of ordinary perception. To say that the objects of moral
perception are available to our cognition by way of the affection of our
receptive faculty but not by way of their affecting our receptive faculty is
to simply ignore the intuition that moral perception is a form of recep-
tivity and simply collapse into the view I argued against in section two.
Last, but not least, one could argue that (3) Value perception should be
construed on the model of the immediate perception of secondary and
not primary properties.17 But the problem with this suggestion is that the
epistemology of the secondary properties themselves is an even more
problematic issue.
It seems then that these views can hardly deal with the task of giving
an adequate account of moral perception. The reason they can’t is that
they cannot provide an adequate solution to the task of identifying what
a capacity is a capacity for by referring to its active exercise. For, by their
lights too, the active exercise of the capacity for moral practical knowl-
edge would be a composite of a practical cognitive and another cognitive
(albeit special or distinctly receptive) act. In other words, by their lights
too, reference to the active exercise of the capacity for moral practical
knowledge would not allow us to identify the capacity as practical. The
problem with the two-factor views examined in the previous section was
not that they presented knowledge in the circumstances as the joint exer-
cise of one practical and one ordinary receptive knowledge, but that they
presented it as the joint exercise of a practical and a non-practical form
117-135; McGrath, S., “Moral Knowledge by Perception”, Philosophical Perspectives, 18/1

(2004), pp. 209-228.
17 See McDowell, J. H., “Values and Secondary Qualities”.
of knowledge. And the views examined in this section have not done
anything to solve this problem.
4. Moral perception as a form of knowledge
But we don’t have to accept any of the above views in order to talk non-
figuratively of moral perception. We can instead deny one driving as-
sumption of all the accounts presented so far. This is the assumption
that if what affects our senses can at all be known, it may only be known
receptively. We may thus distinguish between different sorts of percep-
tion not in accordance with a prior distinction between species of sensi-
ble receptivity, but in accordance with a prior distinction between distinct
cognitive activities. To do this, one ought to look not to the objects of these active-
ities qua operations of the receptivity, but to their credentials as cognitive activities.
On this line of thought, we should take ordinary perception and moral
perception to be short hands for two distinct forms of knowledge. Thus,
by ordinary perception we might now understand that knowledge, whose
credentials include all that sensibly testifies to whether things really are as
the agent takes them to be; e.g., the proper function of the sense organs
and the truth-attaining reasoning capacities of the perceiver and the per-
ceiving conditions that allow for this proper function (e.g., natural light
conditions, the position of the perceiver with regard to the perceived
object, availability of background knowledge about things, etc.). And we
might also say that the nature of these credentials –that they all pertain
to whether the agent’s grasp of how things are is as it ought to be given
how things are– qualifies the knowledge as receptive: knowledge that is
derived from what is known. In contrast, by moral perception we could
understand that knowledge whose credentials qua knowledge include all
that testifies to whether the agent’s grasp of how things are in her situa-
tion is as it ought to be given what must matter to her; e.g., the proper
shaping of one’s character (the form of the (habitually) shaped interests
and concerns of a mature human individual) and the conditions that do
not hinder the proper expression of this character (e.g., not being drunk
or devastated). And we might also say that the nature of the credentials –
that they all pertain to whether the agent’s grasp of how things are in her
situation is as it ought to be given what must matter to her– qualifies the rele-
vant knowledge as practical: knowledge that brings about what it under-

stands.
On this account, we can distinguish between moral and ordinary per-
ception without having to assume that when one knows one and the
same thing by way of moral and ordinary perception one knows two dif-
ferent objects or species of objects. The reason is that moral perception
does not constitute a form of receptive knowledge at all. It constitutes a
form of practical knowledge: knowledge that is in understanding pro-
ductive of this feature of the situation’s mattering to it as it should. And
so when she knows a feature of her situation practically she does not
thus know an object over and above the object that she knows in recep-
tive knowledge.
But now the question arises: How is it possible to know non-recep-
tively a thing that is also known receptively? We do after all talk of moral
perception and not merely of moral knowledge. We do talk of hearing
Mark’s vulnerability (being affected by Mark’s vulnerability) when we talk
of knowing what we ought to do in the relevant cases, and we do not
want to take this talk metaphorically. Receptivity must somehow feature
in the picture. But if moral perception is shorthand for practical knowl-
edge, then must we not fit receptivity together with practicality? And
does this not bring us back to the views discussed in the previous two
sections?
This question may also be posed in the following way: when one
knows what one ought to do in moral perception, one also has receptive
knowledge of what one knows in moral perception. To illustrate this let
us take a case in which what is perceived is only known receptively and a
case in which what is perceived is also known practically. Take the first
case first: Say I help administer a psychological experiment in which I am
told by the researcher to register all of the participant’s reactions. And
now say that in this context I perceive Mark’s vulnerability and in this
perceiving receptively know it. I can if asked tell you how things stand
with regard to Mark’s psychological condition; and how things are will be
the judge of whether my claim is knowledge or not. But now let us think
of a different scenario and say that while I’m administering the psy-
chological experiment on Mark I decide to stop the experiment because I
register Mark’s vulnerability. Now in this latter case my knowledge of
Mark’s vulnerability is distinctive in a very straightforward sense; it is
what prompts me to do something, as opposed to what simply acquaints

me with how things stand in the world. But to have this distinctive
knowledge at all I must be acquainted with how things stand in the
world; I must have receptive knowledge of how Mark’s psychology is
faring. In similar cases then we may speak of two ways of knowing being
operative. But now the question arises: if we may speak of two ways of
knowing (of one thing) ought we also not speak of two objects known?
And if we do speak of two objects known are we not again supposing
that there must be a very special and queer sort of receptive faculty that
allows us to know practically (in knowledge in the circumstances) one of
the two objects? This I want to say in the following section is the ques-
tion that Anscombe deals with towards the end of her Intention. And it is
by looking at Anscombe’s answer to this question that we may find a way
out of the current predicament. Doing so will help us better understand
how it is possible to know non-receptively a thing that is available to us
in receptive knowledge. And this will help us appreciate what knowledge
in the circumstances and knowledge in acting share in common.
5. Anscombe’s problem and practical knowledge as self-knowledge
As I said in the introduction, Anscombe argues in Intention that the topic

of intentional action is non-circularly captured by specifying the agent’s
distinctive knowledge of what she is doing when she is acting intention-
ally (knowledge in acting). And, she suggests, the agent’s knowledge in
acting is known without observation. To put it as generally as possible, it
seems that for Anscombe the intentionality of intentional actions goes
hand in hand with the capacity of the agent to know what she is doing
non-observationally. It is useful to consider Anscombe’s worry about
knowledge in acting here, because the object of this knowledge, i.e. the
intentional action, is also a happening that can be the object of a dif-
ferent species of knowledge; observational or receptive knowledge. As
Michael Thompson observes, there is tons of stuff we know non-obser-
vationally –i.e. thoughts, intentions, mathematical truths, etc.18 But as
18 See Thompson, M., “Anscombe’s Intention and Practical Knowledge”, in Ford, A. –
Hornsby, J. – Stoutland, F. (eds.), Essays on Anscombe’s Intention.

Thompson also observes, intentional actions are epistemologically dis-

tinct from mere thoughts or even intentions in that, no matter whether
the agent has non-observational epistemological access to them –they are
also available (at least to a third party) as merely observable.19 And,
Anscombe asks: if in the case of a particular intentional action –say the
opening of a window– we may speak of two ways of knowing it, one
observational and one not, then must we not also speak of two objects?
One being what I’m doing and the other being what happens as a result
of what I’m doing. If not, then how is it possible to speak of two ways
of knowing one and the same object? But if we speak of two ways of
knowing one and the same object, she says, ought we not to speak of
two objects really?20
But one may here object that the case of knowledge in acting is not
analogous to the case of knowledge in the circumstances, because
knowledge in acting is for Anscombe explicitly defined as knowledge
without observation; whereas, on the Murdochean view, knowledge in the
circumstances explicitly involves perception. Consequently, in the case of
knowledge in the circumstances we have reason to suppose that the
problem of the possibility of two knowledges of one and the same thing
is a significant problem. It is after all explicit that in this sort of non-re-
ceptive knowledge receptivity must be involved. Whereas in the case of
knowledge in acting, the sort knowledge Anscombe considers is explic-
itly defined as not involving any observation and so any operation of our
receptivity. That what is known by the agent in knowledge in acting may
also be known observationally from a third person’s perspective should
not be particularly alarming. There is after all nothing too problematic
about the claim that one and the same thing might be known in different
ways from different perspectives.
But this hesitation rests upon a mistaken interpretation of
Anscombe. In the rest of this section I will try to clear out the mistake.
As McDowell stresses,21 some of the most interesting cases of
See Thompson, M., “Anscombe’s Intention and Practical Knowledge”.

19
See Anscombe, G. E. M., Intention, p. 51.

20
21 McDowell, J. H., “How Receptive Knowledge relates to Practical Knowledge”,
(unpublished). I owe a lot of what I understand about Anscombe’s conception of the

role of receptivity in Anscombe’s account of knowledge in acting to this paper.
intentional action that Anscombe considers are cases in which one must
look at what one is doing to know in action that one is doing it. To bring
this point home McDowell focuses on her example of writing on the
board. To know that I am writing ‘I am a fool’ on the board, I must be
looking to see whether I’m indeed writing at all. This and similar cases
might make it seem as if knowledge in acting must be knowledge by ob-
servation in some sense of observation. And so, to keep with Anscombe’s
thesis that knowledge in acting is knowledge without observation, inter-
preters have struggled to specify exactly that species of observation that
Anscombe must mean to include in knowledge in acting. So, for instance,
interpreters have suggested that the observation Anscombe means to
include in knowledge in acting is the observation of inner mental items –
such as intentions– from which (together with knowledge of facts of
causal or reliable connections with events in the world) the agent infers –
and thus comes to know– what she is doing.22 Others have suggested
that the observation Anscombe means to include in knowledge in acting
is some sort of introspective or proprioceptive perception, such that it
suffices on its own to constitute non-inferential knowledge of what one
is doing.23 Yet others have suggested that what Anscombe means to ex-
clude from knowledge in acting is some kind of active or self-controlling
perception, thus excluding passive perception.24 But no matter how ap-
pealing these interpretations may seem, they are revisionist. They simply
fail to take Anscombe’s crystal clear injunction to heart: knowledge in
acting ought to be understood, if at all, as knowledge without observa-
tion; and not as knowledge without this or that kind of observation. And
yet, there are cases, as McDowell insists, in which the agent herself has
22 The core of such views is usually called ‘two factor thesis’. See Falvey, K. T.,
“Knowledge in Intention”, Philosophical Studies. For a recent version of it see for instance
Paul, S. K., “How We Know What We’re Doing”, Philosophers’ Imprint, 9/11 (2009), pp. 1-
24.
23 See Velleman, D., “What Good is a Will?”, in Leist, A. – Baumann, H. (eds.), Action
in Context, Berlin – New York, de Gruyter – Mouton, 2007; O’Brien, L. F., “On Knowing
One’s Own Actions”, in Roessler, J. – Eilan, N. (eds.), Agency and Self-Awareness, Oxford,
Oxford University Press, 2003, pp 358-382.
24 Schwenkler, J., “Perception and Practical Knowledge”, Philosophical Explorations,
14/2 (2011), pp. 137-152; Grunbaum, T., “Anscombe and Practical Knowledge of What
Is Happening”, Grazer Philosophische Studien, 78 (2009), pp. 41-67.
to look to what is happening if she is to know in action what she is

doing. How are we to explain these cases given Anscombe’s clear thesis?
One could suggest that knowledge in acting merely presupposes some
receptive knowledge of what is happening but it can never be exactly of
what this receptive knowledge is knowledge of. But this is to return to a
two-factor view of knowledge. In other words, on this line of thinking,
one would have to claim that to know in action (at least sometimes) one
needs to both observe what is happening and “in some other way” know
what one is doing where what is happening is not what one is doing. But
given Anscombe’s insistence that knowledge in acting is knowledge with-
out observation, knowledge in acting would have to be the latter form of
knowledge, and this knowledge would not on its own be able to yield
knowledge of what one is actually doing.25 And this is a consequence
that Anscombe does not want to be left with.
We can explain the cases McDowell insists on without revising the
Anscombean thesis that knowledge in acting is knowledge without ob-
servation and without reverting to a two-factor view. What one knows in
knowledge in acting one does not know by observation. But what one
knows in knowledge in acting may nevertheless be of what is knowable
receptively. The sense in which knowledge in acting involves knowledge
by observation is the sense in which knowledge in acting may be of what
is known receptively that it is what the agent herself is doing. For instance,
of the writing of the phrase “I am a fool” on the blackboard I know that
it is what I am doing, and I know this non-observationally. But the ap-
pearing of the phrase on the board is available to me in receptive knowl-
edge. And now the question is exactly parallel to the question we asked
above in the case of knowledge in the circumstances: how is it possible
to have two knowledges of one and the same thing (the writing on the
board) without having two objects known?
We are now in a position to see how Anscombe’s answer to this ques-
tion show us the way to answer the equivalent question in the case of
knowledge in the circumstances. Ultimately, the fact that we can answer
25 This is roughly equivalent to McDowell’s argument against the two factor views in
the case of theoretical knowledge. See McDowell, J. H., “Knowledge and the Internal”,
Philosophy and Phenomenological Research, 55/4 (1995), pp. 877-893.
this question in the same way in both cases of knowledge points to the
unity of the two forms of knowledge.
In Anscombe’s example, I’m writing “I’m a fool” on the board. And
to know (in acting) that I am writing “I am a fool” on the board, I need
to look to see what is happening; i.e., I need to look to see that what I
am writing gets written. As I urged above, we must take this to mean that
what is happening there –that what I’m writing gets written– is known in
two ways: by observation and in acting. And then it seems natural to ask
the question raised at the end of the previous section: how is it possible
to have two knowledges without also having two objects known? In oth-
er words, how is it possible to claim that we can know one thing in two
ways without assuming that what is thus known is two different objects?
In answering this question previously, Anscombe reports that she had
said that there are no two different objects known because in this case “I
do what happens”. Or else, because “…when the description of what
happens is the very thing which I should say I was doing, then there is
no distinction between my doing and the thing’s happening”.26 But what
is this response if not a mere insistence that what is known in the two ways
is one and the same thing? Everyone who heard the phrase, Anscombe
says, thought it was extremely paradoxical and obscure.27 And she notes
later on that this puzzlement may have been caused by something mod-
ern philosophers have blankly misunderstood: “namely what the ancient
and medieval philosophers meant by practical knowledge”.28 Anscombe’s
suggestion here is that the inability to understand the phrase “I do what
happens” and the inability to accept the possibility of knowing one and
the same thing in two ways is due to the inability to understand just what
one of these two ways of knowing might be; i.e., what practical knowl-
edge really is.
By Anscombe’s lights it is not puzzling at all to say that in the cases
where one has to look to see if what one is writing gets written when
what I’m writing gets written, I do what happens. The reason that it is
not puzzling is that there is a form of knowledge which in itself is not
26 Anscombe, G. E. M. Intention, pp. 52-53.

27 See Anscombe, G. E. M. Intention, p. 53.
28 Anscombe, G. E. M. Intention, p. 57.
receptive but may be of what receptive knowledge is of, that it is what

oneself is doing. In simpler words, when what is happening is what I
should say I was doing, I may know one and the same thing in two ways:
I may know what is happening receptively, and I may know of what is
happening, which is available to me receptively, that it is what I am doing.
And it is this, that what is happening is what I am doing, that I know
non-receptively or non-observationally. So, in knowing in acting what is
happening I know no new or different thing over and above the one I
know in receptively knowing –looking at– what is happening, because in
this case my knowledge is not mere knowledge of an object (a thing
known receptively) as such. It is knowledge of an object (a thing known
receptively) that it is what I am doing. And it is in this sense an instance
of self-knowledge. Knowledge that what is happening is what I am
doing. Of a happening then I may have two knowledges: knowledge of
the happening as other, and thus receptive knowledge, and knowledge of the
happening as what I do and thus self-knowledge.
But now one may object: if practical knowledge may be of what is
happening in this sense, then how is it that it is the cause of what it
understands? For if it is knowledge of something that is known recep-
tively, it must be knowledge of something that is there already, must it
not? In what sense then can knowledge of what is happening be knowledge
that is practical; i.e., that causes what it understands? Or in other words, what
is that which is caused in practically knowing what is happening? It is, as
Anscombe argues in the very last part of her Intention, the action form of
what is happening; this distinctive order of the world which, for
Anscombe, is specified by two things: 1) that it is what the burden of
correction falls on when there is a mismatch between the knowledge
claims and what the knowledge claims are of29 and 2) that this order is
the conclusion of one’s practical reasoning; or the structure that answers to
a special question why.30
From this we see that what modern philosophers have blankly misunder-
stood is that when the medieval and the ancient philosophers talked of
29 See Anscombe, G. E. M., Intention, pp. 56-57.

30 For more on this see Anscombe, G. E. M., Intention, p. 57.
practical knowledge they talked of self-knowledge of the kind that un-

problematically extends all the way to what is happening –and not to part
of what is happening at best– and thereby gives it its action form. In this
sense then, Anscombe’s phrase “I do what happens” is to be taken liter-
ally. And practical knowledge is the knowledge of what happens as what
I do.
From this we can now see how to answer the equivalent question for
knowledge in the circumstances. We said in section four that we can take
ordinary and moral perception to be shorthands for two forms of
knowledge that may be of one and the same thing, because moral
perception need not be taken to constitute a species of receptive knowl-
edge. We are now in a position to see how moral perception may be un-
derstood as constituting a species of practical knowledge, which is to be
contrasted with receptive knowledge. When we say that I perceive Mark’s
vulnerability morally we are saying that I know of Mark’s vulnerability –
which is rendered available to me in receptive knowledge– that it is what
I ought to do something about. The emphasis is on the “I”; when I
morally perceive something I ipso facto know what I ought to do; or at the
very least I know I (and not one or any human being or any rational being,
etc.) ought to do something. Think of walking by a beggar on the street
on your way to work. If you happen to (rightly) believe in philanthropy,
in seeing the extended hand you may know of it that is what you ought
to leave some of your change in. Or, if you happen to (rightly) believe in
change and not philanthropy, you may know of it that it is what you
ought to change the world for. Or if you happen to (rightly) believe that
change is necessary but impossible, you may know of it that it is what
you ought to do the impossible for. This self-knowledge is not the dis-
oriented attention to oneself that Murdoch warns us against in urging us
to see that the true moral task that lies ahead of us is the loving surren-
der to the endlessness of the other. It is not the neurotic focus on one-
self in the presence of moral demand. On the contrary; it is the recog-
nition of the simple truth that what we ought to do is out there; we see it.
We don’t need to infer what it is that we ought to do from any general
principles detailing what rational or human or any other beings ought to
do and what our receptively knowable circumstances are. And we don’t
need to infer what it is that we ought to do from any general desires of
ours (together with the above mentioned general principles) and what our
receptively knowable circumstances are.31 What it is that we ought to do,

we do not know by seeing in a queer way nor by seeing queer stuff, but
by seeing Mark’s vulnerability and the extended hand, and knowing of
them that they are what we ought to do something for or about. But if
this is practical knowledge we are talking of, we need to specify the sense
in which it is the cause of what it understands. It is after all dubious to
even imply that the demandingness of a moral situation is only there if
the agent thinks it is. But the species of practical knowledge we are
talking about here is the cause of what it understands. And what it un-
derstands is not the demandingness of a moral situation conceived as ex-
ternal to the capacity of the agent to fall under it. What our species of
practical knowledge understands is the demandingness of a moral situa-
tion as internal to the capacity of the agent to fall under it. What it un-
derstands (and thus brings about) is the thing known (Mark’s vulnerabili-
ty let’s say) as actually operative on the agent’s capacity to fall under the
demandingness of moral situations. In simpler words, if the agent didn’t
know of Mark’s vulnerability that it was what she ought to do something
about, then in the circumstances the agent wouldn’t be thinking or doing
what she ought to in whatever it was that she was thinking or doing. And
it is in this sense that knowledge in the circumstances brings about what
it understands.
The account of knowledge in acting and knowledge in the circum-
stances presented in this paper does not aspire to constitute a full ac-
count of practical knowledge. But it is the beginning, I believe, of an ac-
count of practical knowledge that does away with the tension I started
this paper with: the tension between the role of perception in the
Anscombean conception of knowledge in acting and the Murdochean
conception of knowledge in the circumstances. That is, if we follow the
lead of this account, we may see that Anscombe never meant to deny
that what is perceived (receptively known) may not be that of which the
agent knows that she does it. And that Murdoch never meant to affirm
that what is morally perceived (practically known in the circumstances) is
known receptively in knowledge in the circumstances. On the contrary;
31 This is not to say that the order that our moral action or thinking reflects may not
be represented in a syllogistic structure. It is to say though that we do not need to think

syllogistically to be morally necessitated, to fully fall under a moral principle.
both Anscombe and Murdoch meant to explain knowledge in acting and

knowledge in the circumstances as species of self-knowledge: knowledge
concerning something –which may or may not be receptively known–
that it is either what the agent herself does or what the agent herself ought
to do something for. And they both meant to make sure that we under-
stand that this knowledge is spontaneous as an instance of self-knowl-
edge. Now it may be that both Anscombe and Murdoch are wrong about
what they take knowledge in acting and knowledge in the circumstances
to be. Or at least I’ve said nothing in this paper to show that their ac-
counts must be right. What I hope to have done though is to say some-
thing about why their accounts could and should be defended together.
And this, that their views allow for a unitary account of a species of
moral and a species of action knowledge, must be a sign; a sign of truth.
University of Patras
Patras, Greece
References
Anscombe, G. E. M., Intention, second edition, Ithaca, Cornell University

Press, 1957.
Audi, R., “Moral Perception and Moral Knowledge”, Proceedings of the
Aristotelian Society, Supplementary Volume, 84 (2010), pp. 79-97.
Bagnoli, C., “Moral Perception and Knowledge by Principles”, in
Hernandez, J. (ed.), New Intuitionism, London, Continuum, 2011, pp.
84-105.
Blum, L., “Moral Perception and Particularity”, Ethics, 101/4 (1991), pp.
701-725
Clarke, B., “Imagination and Politics in Iris Murdoch’s Moral
Philosophy”, Philosophical Papers, 35/3 (2006), pp. 387-411
Cullison, A., “Moral Perception”, European Journal of Philosophy, 18/2
(2010), pp. 159-175.
Dancy, J., Ethics without Principles, Oxford, Clarendon Press, 2004
Falvey, K. T., “Knowledge in Intention”, Philosophical Studies, 99/1 (2000),
pp. 21-44.
Grunbaum, T., “Anscombe and Practical Knowledge of What Is
Happening”, Grazer Philosophische Studien, 78 (2009), pp. 41-67.
Lance, M. – Little, M., “Defending Moral Particularism”, in Dreier, J.
(ed.), Contemporary Debates in Moral Theory, Oxford, Blackwell, 2006,
pp. 305-321.
Little, M., “Moral Generalities Revisited”, in Hooker, B. – Little, M. O.
(eds.), Moral Particularism, pp. 276-304
McDowell, J. H., “Virtue and reason”, in McDowell, J. H., Mind, Value,
and Reality, Cambridge, MA, Harvard University Press, 1979
——, “Values and Secondary Qualities”, in McDowell, J. H., Mind, Value,
and Reality.
——, “Knowledge and the Internal”, Philosophy and Phenomenological
Research, 55/4 (1995), pp. 877-893.
——, “What is the Content of an Intention in Action?”, Ratio, 23/4

(2010), pp. 415-432.
——, “How Receptive Knowledge relates to Practical Knowledge”,
(unpublished)
McGrath, S., “Moral Knowledge by Perception”, Philosophical Perspectives,
18/1 (2004), pp. 209-228.
McNaughton, D. A. – Rawling, P., “Unprincipled Ethics”, in Hooker, B.
– Little, M. O. (eds.), Moral Particularism, Oxford, Oxford University
Press, 2000, pp. 256-275
Moore, G. E., Principia Ethica, Cambridge, Cambridge University Press,
1903.
Moran, R., “Anscombe on ‘Practical Knowledge’”, in Hyman, J. –
Steward, H. (eds.), Agency and Action (Royal Institute of Philosophy Suppl.
55), Cambridge, Cambridge University Press, 2004
Murdoch, I., “Vision and Choice in Morality”, Proceedings of the Aristote-
lian Society, Supplementary Volume, 30 (1956), pp. 32-58.
——, The Sovereignty of Good, London, Routledge and Kegan Paul, 1970.
Norman, R., “Making sense of Moral Realism”, Philosophical Investigations,
20 (1997), pp. 117-135.
Nussbaum, M., Love’s Knowledge, Oxford, Oxford University Press, 1990.
O’Brien, L. F., “On Knowing One’s Own Actions”, in Roessler, J. –
Eilan, N. (eds.), Agency and Self-Awareness, Oxford, Oxford University
Press, 2003, pp 358-382.
Paul, S. K., “How We Know What We’re Doing”, Philosophers’ Imprint,
9/11 (2009), pp. 1-24.
Rödl, R., Self-Consciousness, Cambridge, MA, Harvard University Press,
2007.
Ross, D., The Right and the Good, Oxford, Clarendon Press, 1930.
Schwenkler, J., “Perception and Practical Knowledge”, Philosophical
Explorations, 14/2 (2011), pp. 137-152.
Setiya, K., “Knowledge of Intention”, in Ford, A. – Hornsby, J. –

Stoutland, F. (eds.), Essays on Anscombe’s Intention, Cambridge, MA,
Harvard University Press, 2011.
——, Murdoch on the Sovereignty of Good, (unpublished).
Thompson, M., “Anscombe’s Intention and Practical Knowledge”, in
Ford, A. – Hornsby, J. – Stoutland, F. (eds.), Essays on Anscombe’s
Intention, Cambridge, MA, Harvard University Press, 2011.
Thompson, M., Life and Action: Elementary Structures of Practice and Practical
Thought, Cambridge, MA, Harvard University Press, 2008
Velleman, D., “What Good is a Will?”, in Leist, A. – Baumann, H. (eds.),
Action in Context, Berlin – New York, de Gruyter – Mouton, 2007.
Watkins, M., – Jolley, K., “Polyanna Realism: Moral Perception and
Moral Properties”, Australasian Journal of Philosophy, 80 (2002), pp. 75-
85.
Wiggins, D., Needs, Values, Truth: Essays in the Philosophy of Value, Oxford,
Wright, J. C., “The Role of Moral Perception in Mature Moral Agency”,
in Wisnewski, J. J. (ed.), Moral Perception, Newcastle, Cambridge
Scholars Publishing, 2008, pp. 1-24.
NAME INDEX
Alexander, J. C. 8417, 100 f., 104 Cervantes, M. de 229

Anscombe, G. E. M. 13 f., 85-88, Chernilo, D. 16016
105, 2412, 243, 244 ff., 257 f., Clarke, B. 2412
25821, 259-265 Cohen, H. 144
Aquinas, Th. 90, 2107, 21416, 217, Cohen, J. 10685, 1142
22148 Craighero, L. 13776
Archer, M. S. 12, 150-160, 16016, Cullison, A. 25316
161 f., 166, 168 ff. Damschen, G. 589
Aristotle 7-11, 13, 384, 406, 43, 46, Dancy, J. 2423
48, 50, 80, 88 ff., 92-95, 9651, Dilthey, W. 1142, 115, 119, 124,
98, 9963, 9964, 100, 104, 163, 143
190 f., 210 ff., 21416, 217, 22148, Djigo, S. 12 f.
223 f., 228, 234 Donati, P. 12, 150 f., 162 f., 16318,
Audi, R. 25316 164 f., 16521, 166, 16623, 16625,
Augustine 7, 13, 175-180, 1804, 167, 16726, 16727, 168, 16829,
181-184, 193 ff., 203 170, 17031
Bagnoli, C. 2412 Durkheim, E. 136
Bauman, Z. 156 Elster, J. 1793
Berger, P. T. 106, 10684 Enríquez, T. 13
Bhaskar, R. 151 Euripides 21113, 21416
Blum, L. 2412 Falvey, K. T. 2412, 25922
Boudon, R. 10270 Fichte, J. G. 24
Brandom, R. 199, 1910, 3228 Flam, H. 10476
Brecht, M. 23278 Förster, E. 2315
Broome, J. 190 Frankfurt, H. 13, 8621, 90 f., 209-
Buber, M. 143, 163 215, 21523, 216 f., 21837, 219
Campbell, C. 10, 79 ff., 83 ff., 87 ff., 22148, 223-231, 23172, 232-
f., 107 f. 237
Cassirer, E. 115 García Martínez, A. N. 12
García Ruiz, P. 16420, 16727 Hazelrig, L. E. 10685
Gerhards, J. 104 Hegel, G.W. F. 7 ff., 17 ff., 21, 23,
Giddens, A. 156 25 f., 2618, 27 ff., 2925, 30, 3026,
Goethe, W. 2315, 116, 126 33, 131 f., 197 ff.
Goffman, E. 802 Heidegger, M. 17, 137, 143
González, A. M. 10 f., 21732 Henrich, D. 562
Grunbaum, T. 25924 Horstmann, R.-P. 3026
270 Name Index
Hume, D. 11, 80, 87, 89, 94 ff., 9651, O’Brien, L. F. 25923

97, 9754, 100, 106 Oakes, G. 12115, 13878
Husserl, E. 6017, 163 Pareto, V. 8728
James, W. 87, 14286 Parfit, D. 175, 189 f., 192, 196, 20518
Jolley, K. 25313 Parsons, T. 802, 87, 8724, 8932, 10066,
Kant, I. 7, 9 ff., 17, 23 ff., 27, 384, 41, 10685
418, 43, 5015, 51, 55 f., 562, 57 ff., Paul, S. K. 25922
6018, 61-68, 6839, 69-75, 80, 89, Pinkard, T. P. 8 f., 174, 176, 2315
8932, 90, 92 ff., 9441, 95, 97-100, Pippin, R. B. 163, 174, 188, 199
114 ff., 118 ff., 125, 132, 135, 137, Plato 141
142 Pope, W. 10685
Korsgaard, C. M. 8932, 8933, 20214 Puls, H. 562
Kosman, A. 38, 39, 43 Rawling, P. 2423,
Lance, M. 2423 Rawls, J. 8932
Lewis, P. A. 1501 Rey, G. 201
Little, M. 2423 Rickert, H. 11 f., 114, 121-139, 142
Lizardo, O. 137 Ridge, M. 371
Locke, D. 22043, 22351, 225 f. Rizzolatti, G. 13776
Luckmann, T. 82, 106, 10684 Rodchenko, A. 128
Luther, M. 232 Rödl, S. 9 f., 4913, 2412
Makkreel, R. A. 13572 Rorty, R. 3127
Mannheim, K. 140 Ross, D. 25211
Martin, J. L. 11 f., 814, 8312, 8622, 8831 Rousseau, J. J. 151
Marx, K. 33 Saussure, F. de 136
Matthiessen, F. O. 14286 Scheler, M. 134, 143
McDowell, J. H. 163, 17, 175, 187, 3228, Schelling, F. 23-26, 28, 144
183 f., 193, 2411, 2412, 2423, 252 Schiller, F. 142
f., 25417, 258 ff., 26025 Schönecker, D. 10, 563, 589, 6433, 6535,
McGrath, S. 25316 7464
McNaughton, D. A. 2423 Schopenhauer, A. 114, 141
Mead, G. H. 802 Schumacher, C. 2108
Moore, G. E. 25211 Schutz, A. 80, 802, 84, 87
Moran, R. 201, 20113, 203, 204, 2412 Schwenkler, J. 25924
Müller, A. W. 5216, 5317 Sellars, W. 8
Munch, P. A 8310, 101, 10169, 10687 Setiya, K. 2412
Murdoch, I. 13, 2412, 242, 244 ff., Simmel, G. 12, 114, 1142, 116-121,
251, 263 ff. 125 f., 134 f., 137, 139-144, 163
Mylonaki, E. 13 Spencer, H. 123
Nagel, T. 20215 Spinoza, B. 2618
Nietzsche, F. 114 Strom, G. 2507, 25314
Norman, R. 25316 Swindell, J. S. 23172
Nussbaum, M. 2423 Themistocles 118
Name Index 271
Thompson, M. 2412, 257 f. Wiggins, D. 251

Tocqueville, Alexis de 102 Williams, B. 175, 183-191, 194,
Tuomela, R. 8311 196, 1969, 199 f., 205 f.
Velleman, D. 25923 Winch, P. A. 81, 827, 828, 10066,
Watkins, M. 25313 105
Weber, M. 11 f., 80 ff., 84, 89, 99 Wittgenstein, L. 19, 384, 80, 205
ff., 10169, 102-106, 10687, 107, Wright Mills, C. 80
113 f., 1141, 117, 121, 138 f., Wright, J. C. 25315
163 Wundt, W. 124

Reason and Normativity: Theories of Action and Morality

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

Reason and Normativity: Theories of Action and Morality

Hochgeladen von

Copyright:

Verfügbare Formate

ALZNAUER / TORRALBA (EDS.

) | Theories of Action and Morality 12 REASON AND NORMATIVITY

ALZNAUER / TORRALBA (EDS.) | Theories of Action and Morality

ISBN 978-3-487-15387-2 OLMS

cyan = Sonderfarbe 2-fbg (tiefe + Pantone 7473)

Reason and Normativity

edited by / editado por / herausgegeben von

Georg Olms Verlag

Theories of Action and Morality

Georg Olms Verlag

Die Deutsche Nationalbibliothek lists this publication in the

Die Deutsche Nationalbibliothek verzeichnet diese Publikation

© Georg Olms Verlag AG, Hildesheim 2016

Umschlaggestaltung: Inga Günther, Hildesheim

Ana Marta González

John Levi Martin

Alejandro N. García Martínez

Name Index.................................................................................................. 269

figures in this tradition, the central or paradigmatic form of action is

The first essay in the volume, Terry Pinkard’s “Contrasting Concepts of

Sellars called the “space of reasons”. According to the static conception,

of the problem, according to González, is an inadequate understanding

ments were not understood as completely arbitrary, they were under-

is a matter of understanding certain normative claims and does not re-

be, despite appearances to the contrary, a species of practical knowledge

“To put law over man is a problem in politics

Agency has come to occupy a special and central place in contemporary

literature seeking to clarify what an intention is and what kinds of rea-

Obituary of Marie Smith in The Economist, Feb 7th 2008.

This is a version of an argument McDowell makes against Pippin in “Towards a

Reading of Hegel on Action in the ‘Reason’ Chapter of the Phenomenology”, in

such. One way of taking the intersubjectivist position is to see agency as

University Press, 1994, p. vii, p. 451.

Reason? Autonomie der Vernunft?: Proceedings of the V Meeting Italian-American Philosophy,

and Recognition: Self-Consciousness and Self-Constitution”, in Grøn, A. – Raffnsøe-

Recognition: Self-Consciousness and Self-Constitution”. If no individual has any author-

Reason develops, and as it develops, so do we.

Naturalism to Hegel’s Naturalism”, in Haag, J. – Wild, M. (Hrsg.), Übergänge, diskursiv oder

1987, § 40, p. 159f.

stances– that could not be explained simply by electricity and magnetism

when he characterized that conception of the absolute as the night in

dicate this, Hegel even calls it an act of the imagination representing

20 “...so ist er vorstellende Einbildungskraft überhaupt” (Hegel, G. W. F., Jenaer Real-

philosophie, pp. 179-180).

anything other than itself. Or, to put it in Hegel’s preferred terminology,

view of divine rationality is Rolf-Peter Horstmann’s conception of Hegel’s “monism of

27 Richard Rorty’s conception of “changing vocabularies” in history is a contem-

porary pragmatist version of the idealist conception of history as driven by changes in

despairs of the space of reason at all and instead is more “skeptical”:

Brandom, R., “The Structure of Desire and Recognition: Self-

——, “Autonomy and Its Burdens”, in Dottori, R. (ed.), Autonomy of

In contemporary literature, the concept of an end is often used to distin-

The concept of an end determines what falls under it according to its

form of predication, in which the progressive and the perfective aspect

An avalanche is sliding down the slope. An avalanche has slid down

A progressive predication represents something as moving towards

this, it differs from “being”.

DNA replicates itself as the double helix unwinds by means of Heli-

Here A and A’ differ in their conceptual determination. And when

predication. In the case of the concept of an end, it is a unity of judg-

7 Kant, I., Critique of Judgment, trans. by J. C. Meredith, ed. by N. Walker, Oxford,

Oxford University Press, 2007, §10, 5:220, p. 51.

We have developed the concept of an end by describing a logical