The Vienna Circle in Hungary/Der Wiener Kreis in Ungarn

W
Veröffentlichungen des
Instituts Wiener Kreis
Band 16
Hrsg. Friedrich Stadler
András Máté
Miklós Rédei
Friedrich Stadler (Hrsg.)
Der Wiener Kreis in Ungarn

The Vienna Circle in Hungary
SpringerWienNewYork
András Máté, Department of Logic, Eötvös Loránd University,
Budapest, Ungarn
Miklós Rédei, London School of Economics, London,
United Kingdom
Friedrich Stadler, Universität Wien, Institut Wiener Kreis,
Wien, Österreich
Gedruckt mit Unterstützung der Aktion Österreich-Ungarn
Das Werk ist urheberrechtlich geschützt.

Die dadurch begründeten Rechte, insbesondere die der Übersetzung, des Nach-
druckes, der Entnahme von Abbildungen, der Funksendung, der Wiedergabe auf
photomechanischem oder ähnlichem Wege und der Speicherung in Datenverarbei-
tungsanlagen, bleiben, auch bei nur auszugsweiser Verwertung, vorbehalten. Die
Wiedergabe von Gebrauchsnamen, Handelsnamen, Warenbezeichnungen usw. in
diesem Buch berechtigt auch ohne besondere Kennzeichnung nicht zu der Annahme,
dass solche Namen im Sinne der Warenzeichen- und Markenschutz-Gesetzgebung als
frei zu betrachten wären und daher von jedermann benutzt werden dürfen.
Produkthaftung: Sämtliche Angaben in diesem wissenschaftlichen Werk erfolgen trotz

sorgfältiger Bearbeitung und Kontrolle ohne Gewähr.
© 2011 Springer-Verlag/Wien
Printed in Germany
SpringerWienNewYork ist ein Unternehmen von

Springer Science + Business Media
springer.at
Satz/Layout: Reproduktionsfertige Vorlage des Instituts Wiener Kreis (Robert Kaller)

Druck: Strauss GmbH, D-69509 Mörlenbach
Gedruckt auf säurefreiem, chlorfrei gebleichtem Papier

SPIN 12594641
Bibliografische Information der Deutschen Nationalbibliothek

Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen
Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über
http://dnb.d-nb.de abrufbar.
ISBN 978-3-7091-0176-6 SpringerWienNewYork

Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Miklós Rédei and Friedrich Stadler

Austria-Hungary in Philosophy and Science:
A Search for the Evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
History of Philosophy
Károly Kókai
Anton Fischers philosophische Grundlagen
der wissenschaftlichen Erkenntnis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Gábor Á. Zemplén

Early 20th Century Conventionalism in Hungary.
A Case for Győző Zemplén’s Reappreciation . . . . . . . . . . . . . . . . . . . . . . 45
Wolfgang L. Reiter
Wer war Béla Juhos? Eine biographische Annäherung . . . . . . . . . . . . . . . . 65
Gergely Ambrus
Juhos’ Antiphysicalism and his Views on the
Psychophysical Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
Physics
Ladislav Kvasz
Classical Mechanics between History and Philosophy . . . . . . . . . . . . . . . . 129
Judit X. Madarász and Gergely Székely

Comparing Relativistic and Newtonian Dynamics
in First-Order Logic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .155
Gergely Székely
On Why-Questions in Physics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
László E. Szabó
Lorentzian Theories vs. Einsteinian Special Relativity –
A Logico-empiricist Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
Logic
András Máté
Die Rezeption der neuen Logik in Ungarn . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
H. Andréka, J. X. Madarász, I. Németi, P. Németi, G. Székely

Vienna Circle and Logical Analysis of Relativity Theory . . . . . . . . . . . . 247
László Kalmár
The Development of Mathematical Rigor from Intuition to
Axiomatic Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
Index of Names / Namenregister . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289
Die Autorinnen und Autoren / The Authors . . . . . . . . . . . . . . . . . 295

Preface
This publication brings together the articles contributed to a bilateral

project on ‘The Vienna Circle and Hungary’ that was carried out jointly by
the Vienna Circle Institute and the Department of Philosophy and Logic
of Eötvös Lóránd University in Budapest. This project was generously
financed in 2007/8 by the Aktion Österreich-Ungarn (Action Austria-
Hungary) Foundation. The concluding conference took place at the end
of May 2008 in Vienna’s Collegium Hungaricum under the title ‘The Vi-
enna Circle in Hungary. The History and Philosophy of Science – A New
Assessment of the Vienna Circle in Hungary’. Happily this was not the
first scholarly cooperation between Budapest and Vienna on this theme:
there had already been two prior bilateral projects (through the Aus-
trian Agency for International Cooperation in Education and Research
(OeAD-GmbH)), which led to the publication of two books on the out-
standing Hungarian scientists and philosophers John von Neumann and
Imre Lakatos, in the form of the ‘Vienna Circle Institute Yearbooks’: John
von Neumann and the Foundation of Quantum Physics. Ed by Miklós
Rédei and Michael Stoeltzner (2001) and Appraising Lakatos. Mathemat-
ics, Methodology and the Man. Ed. by George Kampis, Ladislav Kvasz
and Michael Stöltzner (2002).
This productive collaboration was continued with further meetings
in Budapest and Vienna, and on the Hungarian side was coordinated
predominantly by András Máté and his colleagues. The editors would
like to express their thanks to him and to all those colleagues – especially
to Károly Kókai from the Institute Vienna Circle – who have collaborated
on this project and who have made their contributions towards this book
as token of a more long-term collaboration.
Vienna, Budapest and London February 2011
Miklós Rédei
(London School of Economics)
Friedrich Stadler
(University of Vienna and Institute Vienna Circle)
Miklós Rédei and Friedrich Stadler
Austria-Hungary in Philosophy and Science:

A Search for the Evidence
Contrary to the well-known joke in football concerning teams on the match

schedule (‘Austria – Hungary? Who are they playing against?’) within
the field of science and philosophy we are dealing with a topic that has
been little explored. We see isolated instances of activity taking place
and there are a few publications already available to us. However, due to
the asymmetrical linguistic relationship between German and Hungarian
as well as the dominance of English, articles written in Hungarian have
unfortunately received little consideration.
Nevertheless the current research in this area is remarkable: a volume
on ‘Hungary and the Brentano School’ (Baumgartner et al. 1994) presents
a comprehensive reception in the cases of logical objectivism, descrip-
tive psychology, and phenomenology. The influence of Franz Brentano
on Géza Révész – especially on his Zur Grundlegung der Tonpsycholo-
gie (1913) – the founder of experimental psychology in Hungary, must
here be mentioned in connection with Carl Stumpf and Ernst Mach. It
is therefore not surprising that Moritz Schlick, the founder of the Vienna
Circle, recommended Révész to his colleague Hans Reichenbach. (Schlick
to Reichenbach, 26.11.1920. Reichenbach to Schlick, 17.10.1920. Wiener
Kreis Archiv Haarlem, NL)
Stronger historical roots are evident in the reception of Ernst Mach
in Hungary (Stadler 1982; Haller/Stadler 1988): as a student of law
and philosophy, Karl Polányi (along with his brother Michael Polányi)
was a member of a predominantly Jewish intellectual circle, the so-called
‘Galilei Circle’ (from 1908 to 1919), which, as part of its aim of enlightened
education, viewed empirical science as the means of correcting society’s
10 Miklós Rédei and Friedrich Stadler
shortcomings. In the centre stood the economist and physiologist Gyula

(Julius) Pikler whose anti-metaphysical stance was an expression of a ‘rev-
olutionary’ rationalism. Other well-known members were Karl Mannheim
(whose works Otto Neurath would continue to engage with when in exile
in England) and Georg Lukács, who was to continue to represent Marxist
philosophy internationally until long after the failure of the Hungarian
communist revolution. In terms of institutions, the ‘Hungarian Society
of Free-thinkers/Freemasons’ constituted a platform which elected Karl
Polányi president. He was to continue these activities following his em-
igration to Vienna with the ‘Ernst Mach Society’, the society dedicated
to the popularisation of the ‘scientific conception of the world’ of the
Vienna Circle, until 1934. Therefore one could say that the Hungarian
‘late Enlightenment’ of the Galilei Circle – which incidentally paralleled
the ‘first Vienna circle’ of c. 1907–12 – was an anticipation of the later
public-relations work in ‘Red Vienna’. In terms of content, in Budapest in
1910 Mach’s The Analysis of Sensations was partly translated with the
aim of making his methodology applicable to the social sciences. Forming
the backdrop to this was, apparently, Lenin’s attack on Mach as a rep-
resentative of ‘bourgeois’ science and philosophy. (Lenin’s Materialism
and Empirio-Criticism. Critical Comments on a Reactionary Philosophy.
Russian edition 1909; German translation 1927.)
As Péter Szegedi has already described on the Galilei-Circle (Szegedi
1997), there was in the Budapest of that period an intellectual alliance of
conservatives and left-wing intelligentsia in the ‘Society of Social Sciences’.
Here as in the Galilei circle, the topic of social reform was on the agenda,
put forward by impoverished self-taught students. The initiator, Gyula
(Julius) Pikler (Professor in Budapest) was a positivistic philosopher of
law and at the same time President of the Society of Social Sciences, amid
a vehemently conducted cultural struggle involving Károly and Mihály
Polányi. One is compelled to see an analogy here with the life and work of
the legal positivist Hans Kelsen in Vienna, Cologne, and Prague until his
emigration. (A recent overview und the reception of Kelsen in Hungary:
András Jakab 2010).
In Budapest, influenced by progressiveness and course- and lecture-
based education, topics were embedded in an atheist, rational, and hu-
manist world view with the motto ‘Teaching and Learning!’. In the area
of publications there existed the periodical Free Thought (Szabadgondo-
Austria-Hungary in Philosophy and Science: A Search for the Evidence 11
lat ), a Library of the Galilei Circle (a Galilei Kör könyvtára), as well as

the Galilei Magazines (Galilei Füzetek ). Among the proponents could
be found György Pólya, Zsigmond Kende, Mór Korach (professor at the
Technical University of Budapest); and acting as the first president was
Karl Polányi (who, like Pólya, studied the leading figures of Marx and
Mach in Budapest and Vienna at the time (who together determined the
particular character of ‘Austrian Marxism’)). Corresponding to this is
the fate of the Galilei Circle, which in 1919 ceased to exist as an intellec-
tual phenomenon spanning science and literature, positioned politically
between the extreme left and right.
Some of these ambitions found a continuation in Viennese exile: Karl
Polányi demanded a form of socialisation based on the model of English
guild socialism, and, working for the Vienna Chamber of Labour, he was,
among other things, the editor of the journal Der Österreichische Volks-
wirt (‘The Austrian Economist’ 1924–1933). He married the left-wing
activist Ilona Duczynska, who wrote a most original book on Theodor
Körner, the adviser of the Republikanischer Schutzbund and later Presi-
dent of Austria after 1945. (Duczynska 1975) (Dalos 1984).
Karl Polányi presented to the Mach Society on ‘Economic Statistics’
(1930), before writing his masterwork The Great Transformation (1944)
during English exile, which like his later historical studies in Ökonomie
und Gesellschaft (1968) presented a comparative history of economics
between the market and planned economy. The concept propagated by
him of the ‘cultural embeddedness of economies’, and his ‘methodological
nominalism’, set out in opposition to Othmar Spann’s universalism, and
which, incidentally, inspired Karl Popper, constituted original contribu-
tions towards a methodology and theory of modern science. The existence
of an international ‘Polányi Society’ documents, finally, the present-day
relevance of Polányi – and likewise of his brother Michael in the field of
the sociology of science; he ended his career working in North America
(New York and Toronto).
It is surely here that we must locate Karl Popper’s memory of a ‘hut’
(Baracke) in post-revolutionary Vienna, where he was involved in the
discussion and translation of Lenin’s Materialism and Empirio-Criticism
and where, right until the end, from a philosophical and epistemological
perspective, he fought against Mach on the side of Lenin (Stadler 2001,
447ff.).
In considering the Hungarian emigration to Austria after the First

World War, in spite of well-known names such as Arnold Hauser, Béla
Balázs, Johannes Wilde, Béni Ferenczy, Alexander Korda or Georg Lukács
(Kókai 2002), the assessment seems right that Vienna constituted a kind of
periphery in comparison to Berlin and England for natural scientists. The
Viennese scientists and intellectuals had themselves collectively turned
into a kind of exiled science, and the waves of emigration that were to
follow in 1934, with a negative culmination after the Anschluss in 1938,
meant the enforced end of any free scientific work following the National-
Socialist takeover. Also contributing to this was the fact that already in
the period between 1933 to 1938 Austria was not a country welcoming
immigration. (Heiss/Rathkolb 1995; Seeber 2003) A few exceptions, such
as the artistic avant-garde centred around the MA group, constitute a
phenomenon of the Hungarian exile that has been little explored until
recently (Péter 2010).
In a letter to Moritz Schlick, the founder of the Vienna Circle, dated
nd
2 April 1929, his pupils and supporters called on him to stay in Vienna
after he had obtained a chair at the University of Bonn. Among the signa-
tories are to be found Hans Hahn, Rudolf Carnap, Viktor Kraft, Friedrich
Waismann, Herbert Feigl, Felix Kaufmann, Karl Menger, Maria Kasper,
Rose Rand, Gustav Bergmann, Heinrich Neider, Kurt Gödel, Gustav Beer,
Theodor Radakovic, Amalie Rosenblüth, Otto Neurath, and Olga Hahn-
Neurath, but also the Hungarian mathematician and statistician Eugen
Lukács as well as the renowned John von Neumann.
At the second conference for the epistemology of the exact sciences
on the foundational issues in mathematics and quantum mechanics, held
in Königsberg in 1930 and organized by Kurt Reidemeister, John von
Neumann was an active participant and discussant in connection with the
first public presentation of Kurt Gödel’s results, which were so to shake
Hilbert’s programme. (Stadler 2001, 350-356).
The continuation of this exchange took place in exile in the US: in
the Harvard Circles from 1940 onwards and at the ‘Institute for the Unity
of Science’ of Philipp Frank, with Gerald Holton as his assistant: there
was the ‘Science of Science Discussion Group’ in 1940-1, and from 1944,
the ‘Interscientific Discussion Group’ in the tradition of the Vienna Circle
and the Ernst Mach Society, involving, among others, R. Carnap, H.
Feigl, W. V. O. Quine, R. von Mises, A. Tarski, N. Goodman, P. W.
Bridgman, E. G. Boring, S. S. Stevens, J. Schumpeter, and the Hungarian

mathematician George Pólya. One of the events was a presentation on
February 28, 1946, by Oskar Morgenstern and John von Neumann on
their pioneering book Theory of Games and Economic Behavior (1944),
and also one by Dénes Gábor (Nobel Prize winner for physics in 1971) on
information theory and holography (Holton 1993; Stadler 2007).
II
Von Neumann, perhaps the most prominent member of the group of bril-
liant scientist that emerged and then left Hungary in the first part of the
20th century, was not a philosopher but a mathematician and mathemat-
ical physicist but his relation to the philosophers of the Vienna Circle has
both historical and theoretical significance.
After finishing his formal education in 1926 in Zurich and Budapest,
von Neumann went to Göttingen to work as Hilbert’s assistant. He was
working on the Hilbert program, he published a paper on it (von Neu-
mann 1927) and he was hoping that the program can be carried through.
But his participation in the famous Second Conference for Epistemology
of the Exact Sciences organized in 1930 jointly by the “Gesellschaft für
Empirische Philosophie” and by the Vienna Circle, a similar group of
philosophers based in Vienna. Participants of the Königsberg meeting
included R. Carnap, A. Heyting, J. von Neumann and K. Gödel. Carnap
gave a talk on the foundations of logicism, Heyting on the intuitionistic
foundations of mathematics and von Neumann on the formalist founda-
tions of mathematics. (For a detailed account of the events related to this
conference see (B. Gyenis and M. Rédei 2004, 135, 137) and (Gödel 1986,
327-330.) It was in the discussion session on September 7 of this con-
ference that Gödel announced the first version of what became known as
Gödel’s first incompleteness theorem. Von Neumann grasped immediately
the significance of Gödel’s result for the axiomatic foundation of mathe-
matics and, as his letter to Gödel (November 20, 1930) (Rédei (ed.) 2005,
123-124) shows, shortly after the Königsberg conference, he obtained what
is known as Gödel’s second incompleteness theorem, which he communi-
cated to Gödel in that letter. As can be inferred from their correspondence
Gödel told von Neumann that he had already established the second in-
completeness theorem. Von Neumann therefore decided not to publish

on the topic (von Neumann to Gödel November 29, 1930) (Rédei 2005,
124-125). From the second incompleteness theorem von Neumann had
drawn a very strong conclusion for the Hilbert program: “. . . there is no
rigorous justification for classical mathematics” (von Neumann to Gödel,
November 29, 1930), (Rédei 2005, 124) a position which von Neumann
also re-iterated in his letter to Carnap (June 7, 1931) (Rédei 2005, 85-86):
Thus I am today of the opinion that

Gödel has shown the unrealizability of Hilbert’s program.
There is no more reason to reject intuitionism (if one disregards the aesthetic
issue, which in practice will also for me be the decisive factor).
Therefore I consider the state of the foundational discussion in Königsberg
to be outdated, for Gödel’s fundamental discoveries have brought the question
to a completely different level. (I know that Gödel is much more careful in the
evaluation of his results, but in my opinion on this point he does not see the
connections correctly).
Indeed Gödel was more careful:
I wish to note expressly that Theorem XI [the second incompleteness theorem]

does not contradict Hilbert’s formalistic viewpoint. For this viewpoint presup-
poses only the existence of a consistency proof in which nothing but finitary
means of proof is used, and it is conceivable that there exists finitary proofs
that cannot be expressed in the formalism of P [Russell’s Principia plus the
Peano axioms]. (Gödel 2003, 195)
Strongly formulated disagreements with Gödel can also be found in

von Neumann’s letter to Gödel (January 12, 1931) (Rédei 2005, 125-127).
The source and nature of disagreement between Gödel and von Neumann
on the implications of the second incompleteness theorem is still an issue
that would warrant a careful historical study and which also has system-
atic interest. It is clear that von Neumann interpreted the notion of “finite
methods” in connection with Hilbert’s program in a much more restricted
sense than did Gödel; however, the precise nature and their disagreement
is less clear than it should be.
In spite of fundamental professional disagreements between von Neu-
mann and Gödel about the ramifications of the second incompleteness
theorem von Neumann had the highest respect for Gödel and his work:
he compared Gödel to Aristotle (Gödel 2003, 8 and references there), and
in his letter to A. Flexner (September 27, 1939), in which von Neumann

tries to help Gödel to get a visa to the U.S.A., von Neumann writes:
The claim may be made with perfect justification that Gödel is unreplaceable
for our educational program. Indeed Gödel is absolutely irreplaceable; he is the
only mathematician alive about whom I would dare to make this statement.
[. . . ] I am convinced that salvaging him from the wreck of Europe is one of the
great single contributions anyone could make to science at this moment. (von
Neumann to Flexner, September 27, 1939) (Rédei 2005, 117)
While the von Neumann-Gödel debate is perhaps the most fascinating

aspect of the relation of von Neumann and the Vienna Circle, there are
other, philosophically and historically important issues worth of attention:
the relation of Carnap’s syntactic concept of empirical science to von
Neumann’s view on the axiomatic approach in the sciences, especially
in physics; von Neumann’s analysis of the status of causality in light of
quantum mechanics and the logical positivists concept of causality have
hardly been analyzed. More generally, much more historical work could
and should be done on the impact of the logical positivist’s philosophical
analysis on von Neumann’s treatment of foundational-interpretational-
philosophical problems related to the sciences.
III
After the Second World War the status in Hungary of the philosophical
position represented by logical empirism (and by the members of the Vi-
enna Circle in particular) became very weak. This was mainly due to the
fact that the characteristic feature of logical positivist’s attitude towards
science, philosophy and philosophy of science, a position which empha-
sized the value-freeness of science, its independence of ideology and the
sharp distinction between history and philosophy of science was not in line
with the ideology based on Marxism. But logical positivism was present
to some extent in the philosophical life as subject of research in history
of philosophy. Prominent philosophers also made accessible some of the
works of members of the Vienna Circle in Hungarian translation: Gen-
erations of students of philosophy learned about logical empirism from
the collection edited by Ferenc Altrichter (Altrichter 1972) and form his
accompanying historical essay on the topic.
The impact of the Vienna Circle is still to be felt in some current
trends in philosophy of science in Hungary. Not in the sense that philoso-
phers of science would subscribe uncritically to the doctrines of this tra-
dition. But the choice of problems to deal with and the methods of ana-
lyzing them are obviously influenced by the spirit of logical empirism in
the works of some scholars: the problems the philosophical analysis fo-
cuses on are raised by the sciences, and the analysis is technically explicit,
rigorous, involving both the concepts of sciences and relying heavily on
the techniques of modern logic – the tradition of scientific philosophy is
clearly alive.
This is shown by some of the papers in the current volume:
The logical analysis of the theory of relativity undertaken by the group
of algebraic logicians in the Rényi Institute of Mathematics of the Hungar-
ian Academy of Sciences exemplified by their contribution to this volume
refers back explicitly and consciously to Reichenbach’s work as forerun-
ner. (Members of the group include I. Németi, H. Andréka, G. Székely,
J. Madarász). This logical analysis of the theory of relativity is carried
out in first order logic, using the full power and machinery of this formal
framework, including advanced results (such as definability theorems).
The motivation of this logical reconstruction of the theory of relativity
is to have a clear understanding of which assumptions of the theory are
necessary to derive the well-known novel predictions of the theory (such
as the twin paradox) – the results obtained advance our understanding of
the logical structure of the theory of relativity greatly, allowing us to see
very intricate, fine logical relations.
The investigation in his paper in this volume by L. E. Szabó also
concerns the theory of relativity. Szabo focuses on some conceptual is-
sues in the theory of relativity, which he approaches in the spirit of strict
physicalism. Szabó’s physicalism, while differing from the physicalism
as interpreted by the members of the Vienna Circle as the claim that
every meaningful sentence is translatable into one in physics, is never-
theless sharing the logical empiricists’ strong empiricism and denial of
the possibility of any contentual a priori knowledge. Szabo extended the
physicalism into the philosophy of mathematics (Szabó 2003), claiming
that the axiomatic systems also can and should be interpreted as physical
claims a theoretical move that arguably might be regarded as the amal-

gamation of two of the basic tenets of logical positivism: physicalism and
heavy reliance on formal logic (syntactic systems).
The work of the ‘Budapest School’ (Butterfield 2007) on probabilis-
tic causality, in particular on the Reichenbach’s Common Cause Prin-
ciple (Hofer-Szabó/Rédei/Szabó 1999), (Hofer-Szabó/Rédei/Szabó 2000),
(Hofer-Szabó/Rédei/ Szabó 2002), (Hofer-Szabó/Rédei 2004), (Hofer-
Szabó/Rédei 2006), (Gyenis/Rédei 2004), (Gyenis/Rédei 2008), (Z. Gye-
nis/Rédei 2011) investigates a problem that goes back to the work of
Reichenbach. This Principle, which is in principle a metaphysical claim
about the causal structure of the world, is investigated by the Budapest
School in a technically explicit manner, using the concepts of measure
theoretic probability theory (both classical and quantum). The approach
adopted in these works relates a metaphysical principle to testable, em-
pirical claims, thereby specifying sharply the possible ways the Principle
can be falsified. Thus, again, not only the problem itself but the spirit
of investigating it is heavily influenced by the philosophical tradition of
logical empiricism.
IV
A remarkable follow-on story can be found in the life and work of the
Hungarian émigré and Nobel Prize-winner John Harsany (János Harsányi,
1920–2000), who in 1994 received the Nobel Prize for Economics together
with John Nash and Reinhard Selten for their work on non-cooperative
game theory: in Budapest he attended a German-speaking gymnasium
with John von Neumann and Eugene P. Wigner and, facing the threat of
forced labour in Austria as well as deportation and murder because of his
Jewish descent, he was compelled to go underground from 1944 on. In
1947 he gained a Ph.D. in Budapest before illegally escaping to Austria in
1950 because of his anti-Communist stance, in order to go on to Sydney to
complete a masters in economic science and game theory. He subsequently
emigrated again, this time to the United States (Detroit and Berkeley),
where he continued the work of Morgenstern and von Neumann and came
to appreciate the works of Rudolf Carnap. During a symposium of the
Vienna Circle Institute held in his honour he delivered a paper very much
in this spirit on ‘A Preference-Based Theory of Well-Being and a Rule-

utilitarian Theory of Morality’ (printed in: Leinfellner/Köhler 1997).
A further forum for cementing Austrian-Hungarian scientific relations
could be found in Karl Menger’s ‘Mathematical Colloquium’, which, in
parallel with the Vienna Circle, brought together inter-war Vienna’s math-
ematical and logical vanguard (Dierker/Sigmund 1998). In its programme
we find again John von Neumann (Berlin) ‘Über Maßtheorie’ (‘On mea-
sure theory’), and with a contribution to the discussion from Nöbeling
(1929/30); there is also from György von Alexits (Budapest, according
to the written announcement) ‘Über im kleinen zusammenhängende Kon-
tinua’ (‘On continua connected in smallness) and ‘Metrische Behandlung
von Torsion von Raumkurven’ (‘Metric treatment of the torsion of space
curves’ 1934/5). Alexits (1899–1978) was a member of the Alfred Rényi
Institute of Mathematics and of the Hungarian Academy of Science, with
the main focus of his work being theory of function and approximation
theory.
In the field of psychology we must mention above all Egon Brunswik
(1903–1955), who, up until his emigration to Berkeley, belonged to both
the Bühler school of Gestalt psychology and the Vienna Circle of Logi-
cal Empiricism. (Fischer/Stadler 1997). In exile he was engaged in the
Unity of Science Movement (promoted by Otto Neurath, Rudolf Carnap,
Charles Morris) along with Arne Naess and his contribution to it was The
Conceptual Framework of Psychology (1952). Brunswik was a proponent
of object-oriented psychology as a science of objective relations and is
considered the founder of the so-called ‘environmental psychology’ within
the context of the ‘probabilist revolution’. He was born in Budapest in
1903, studied in Vienna and after that maintained good contacts with
his Hungarian colleagues in the field of Gestalt and experimental psychol-
ogy, for example with his ‘Prinzipienfragen der Gestalttheorie’ (‘Ques-
tions of principle in Gestalt theory’) in the Bühler festschrift, which was
edited by Lajos Kardos, an outstanding Gestalt psychologist. (Kardos
1934). During his time in Vienna he published, among other works, “Das
Duplizitätsprinzip in der Theorie der Farbenwahrnehmung” with Kardos
(1929), along with, as single author, Wahrnehmung und Gegenstandswelt
(1934), Experimentelle Psychologie in Demonstrationen (1935). He pre-
sented a paper to the Ernst Mach Society on experimental Gestalt psy-
chology (1930), before he was given the task in 1931/32 of setting up an

institute for experimental psychology in Turkey (Ankara).
The mathematician Richard von Mises had to emigrate from Berlin to
Istanbul in 1933, where he worked until his departure to Harvard – as it
happens along with his Berlin colleague Hans Reichenbach (Stadler 2011).
It is in connection with this that his collaboration of several years with
Theodore von Kármán (1881–1963) (on, among other things, the basic
equations of mechanics and the theory of flight) is worth mentioning:
Von Kármán was born in Budapest, obtained a doctorate in Göttingen,
and went to the USA (Cal-Tech) as early as 1926. In turn, in 1941 R. von
Mises contributed his “Some Remarks on the Laws of Turbulent Motion
in Tubes” to the Kármán festschrift (Theodore von Karman Anniversary
Volume, CalTech 1941), reviewed Kármán’s The Mathematical Methods
in Engineering (Science 1940), before editing with him the Advances in
Applied Mechanics (1948-53). The Selected Papers of Richard von Mises
(1963/64) were edited by Ph. Frank, S. Goldstein, M. Kac, W. Prager, G.
Birkhoff and G. Szegő. Szegő (1895–1985, died in Palo Alto) had already
worked with R. von Mises in Berlin in 1921 and had written “Über J.von
Neumann’s Theorie der Spiele” (Mathematische Nachrichten, 1953).
Alongside Brunswik Béla Juhos (1901–1971) was the second member
of the Vienna Circle to come from Hungary: with his ‘epistemo-logical
method’ he embodied the philosophical-scientific alliance between Bu-
dapest and Vienna, although, quite unfairly, he was to remain an outsider
in academia throughout his life (cf. the essays by Wolfgang Reiter and
Gergely Ambrus in this volume).
Finally in the so-called ‘Third Vienna Circle’ at the beginning of the
Second Republic there was a clear echo of the history briefly sketched
above: among the members of the ‘Kraft Circle’, under the leadership
of Viktor Kraft, there were still Béla Juhos, Paul Feyerabend, Elisabeth
Anscombe, and Ludwig Wittgenstein. A brief visit to Vienna by Imre
Lakatos organized by Kraft represented a sort of prelude to the ‘Nordic
Connection’ and the interaction of Popper, Feyerabend and Lakatos in
England, which culminated in the LSE Conference of 1965 and the ‘his-
torical turn’ in the philosophy of science centred around Thomas Kuhn.
(Stadler/Fischer 2006; Manninen/Stadler 2010). If one attempts to con-
sider these collaborations and processes of exchange as a whole, one can
formulate the following theses, which are more or less thematized in the
individual essays in this volume:
• Before 1914 there already existed a concrete reception of philosophy

of science in Hungary, predominantly of Brentano and Mach. At the
same time there is the beginning of the reception of the new logic.
• A long neglected emigration after World War I from Hungary to the

West (incl. to and via Vienna)
• Between 1918 and 1933/34 we see mutual contacts and collabora-

tions in he context of the Vienna Circle and Logical Empiricism
• From 1933/34 both trends are already manifestations of ‘science in

exile’.
• It is notable that after 1938 there are isolated instances of continu-

ation of this work in the USA.
• After 1945, following the ‘Third Vienna Circle’, and the ‘Kraft Cir-
cle’ featuring, among others, Béla Juhos, Paul Feyerabend, Arthur
Pap, and Ludwig Wittgenstein, scientific interchanges with Imre
Lakatos and Karl Popper were continued and strengthened in Eng-
land; this too had its roots in the Austrian-Hungarian tradition in
the philosophy of science (beginning with Mach). In Hungary itself
the outsider Anton Fischer constitutes a remarkable exception as a
proponent of scientific philosophy already after World War II (cf.
the contribution by Károly Kókai in this volume).
In all, however, one must take account of the fact that, compared
to Prague or Warsaw, there was a much more moderate degree of inter-
action, despite the notable individual contacts with John von Neumann,
John Harsany, Imre Lakatos etc. We have here, surely, to distinguish two
different levels, namely the (natural) sciences and the pilosophy of science.
So, for example, Vladimir Zeman’s “The Philosophy of Science in Eastern
Europe” (1970) only refers to Kalmár’s school of logic due to the ideolog-
ical limitations after 1945. The restrictive conditions of the time are asso-
ciated with language barriers and nationality, and with the Shoah and the
exodus of Jewish scientists – originating in an intensification of monarchist
dualism. What deserves mentioning here is Horthy’s Hungary, with its
collaboration with the Nazis from 1928 and the Pfeilkreuzler in 1944, and,
beginning in 1948, the communist regime with the failed uprising of 1956,
which would entail the only political-cultural mass emigration to Austria
with positive consequences for the cultural environment of the Second
Republic, e.g., with Paul Lendvai, György Sebestyén etc. (For a general
account of bilateral relations after 1945 see Majoros/Maruzsa/Rathkolb
2010)
The Cold War formed the context for the cutting of direct relations,
which could now only be maintained by the exiled scientists of both na-
tions. The few exceptions, such as the Alpbach Forum or the Institute
for Advanced Studies (IHS) in Vienna, prove the rule that both countries
must be viewed rather as countries of emigration – we have Austria as a
place of transit until 1933/34, 1938 at the latest, and subsequently the
wave of emigration of intellectuals and researchers that ensued.
Correspondingly, we can share the conclusion (G. Palló, “Wo ist das
Zentrum?” in Habitus, Identität und die exilierten Dispositionen, eds.
Anna Wessely, Károly Kókai and Zoltán Péter 2008) that most Hungar-
ian scientists did not emigrate to Austria for good because there was no
ideally suitable milieu to be found there (in comparison to the Weimar
Republic). Science and research in the First Republic were in fact threat-
ened by marginalisation, impoverishment, and anti-Semitism, even though
a creative two-sided culture Beyond Art (Weibel 1996) came into existence
time and again. All the more, then, are the contributions in this volume to
be understood as contributions to further research which will force us to
pay attention to themes and people as yet little studied – especially from
a Hungarian perspective – and to illustrate a common scientific culture
buried by the past or forgotten. Indeed in the new Europe this oppor-
tunity to take stock of the past should be cause for new, strengthened
partnerships also within the sciences.
References
F. Altrichter, editor (1972), A Bécsi Kör filozófiája, Budapest, . Gondolat.

George Kampis, Ladislav Kvasz and Michael Stöltzner (eds.) (2002) Ap-
praising Lakatos. Mathematics, Methodology and the Man. (= Vienna
Circle Institute Library 1)
Wilhelm Baumgartner / Franz-Peter Burkhard / Franz Wiedmann / Chri-

stoph Nyı́ri (1994) (Hrsg.). Ungarn und die Brentano Schule. Würz-
burg: Verlag Josef H. Röll. (= Brentano Studien, Band 5).
J. Butterfield, “Stochastic Einstein Locality revisited.” The British Jour-
nal for the Philosophy of Science, 58: 805-867, 2007.
György Dalos (1984), A cselekvés szerelmese. Duczynska Ilona élete. Bu-
dapest: Kossuth Verlag.
Egbert Dierker/Karl Sigmund (Eds.), Ergebnisse eines Mathematischen
Kolloquiums. Wien–New York: Springer.
Ilona Duczynska (1975). Der demokratische Bolschewik. Zur Theorie und
Praxis der Gewalt. Mit einem Vorwort von Friedrich Heer. München:
List Dalos (1984).
Kurt R. Fischer/Friedrich Stadler (1997) (Hrsg.), ‘Wahrnehmung und Ge-
genstandswelt’. Zum Lebenswerk von Egon Brunswik (1903–1955).
Wien-New York: Springer.(= Veröffentlichungen des Instituts Wiener
Kreis 4)
K. Gödel. Collected Works. Correspondence H-Z, volume V. Oxford Uni-
versity Press, New York, 1986.
K. Gödel. Collected Works. Correspondence H-Z, volume V. Oxford Uni-
versity Press, New York, 2003.
B. Gyenis and M. Rédei. “When can statistical theories be causally
closed?” Foundations of Physics, 34: 1285-1303, 2004.
B. Gyenis and M. Rédei. “Causal completeness of general probability
theories.” In M. Suarez, editor, Probabilities, causes and propensities
in physics, Synthese Library, Springer, 2008. forthcoming.
Z. Gyenis and M. Rédei. “Characterizing Common Cause Closed proba-
bility spaces.” Philosophy of Science, 2011. forthcoming.
Rudolf Haller/Friedrich Stadler (1988) (Hrsg.). Ernst Mach – Werk und
Wirkung. Wien: Hölder-Pichler-Tempsky.
John Harsanyi, “A Preference-Based Theory of Well-Being and a Rule-
utilitarian Theory of Morality”, in: Leinfellner/Köhler 1997, pp. 285-
300.
Gernot Heiss/Oliver Rathkolb (1995) (Hrsg.). Asylland wider Willen.
Flüchtlinge im europäischen Kontext seit 1914. Wien: Jugend & Volk.
G. Hofer-Szabó and M. Rédei. “Reichenbachian common cause systems.”
International Journal of Theoretical Physics, 43: 1819-1826, 2004.
G. Hofer-Szabó and M. Rédei. “Reichenbachian common cause systems

of arbitrary finite size exist.” Foundations of Physics Letters, 35:
745-746, 2006.
G. Hofer-Szabó, M. Rédei, and L. E. Szabó. “On Reichenbach’s Common
Cause Principle and Reichenbach’s notion of common cause.” The
British Journal for the Philosophy of Science, 50: 377-398, 1999.
G. Hofer-Szabó, M. Rédei, and L. E. Szabó. “Common cause completabil-
ity of classical and quantum probability spaces.” International Jour-
nal of Theoretical Physics, 39: 913–919, 2000.
G. Hofer-Szabó, M. Rédei, and L. E. Szabó. “Common-causes are not
common common-causes.” Philosophy of Science, 69 (20): 623-636,
2002.
András Jakab (2010). ,,Kelsens Rezeption in Ungarn“, in: Robert Wal-
ter/Clemens Jabloner/Klaus Zeleny (Hrsg.), Hans Kelsen anderswo.
Hans Kelsen abroad. Der Einfluss der Reinen Rechtslehre auf die
Rechtstheorie in verschiedenen Ländern, Teil III. Wien: Manz, 41-72.
Georg Kampis/Ladislav Kvasz/Michael Stöltzner (2002).(Eds:) Apprais-
ing Lakatos. Mathematics, Methodology and the Man. (= Vienna
Circle Institute Library 1).
Lajos Kardos, Object and Shadow, 1934.
Károly Kókai (2002). Im Nebel. Der junge Georg Lukács und Wien.Wien–
Köln–Weimar: Böhlau.
Werner Leinfellner/Eckehart Köhler (1997) (Eds.). Game Theory, Ex-
perience, Rationality. Foundations f Social Sciences, Economics and
Ethics. In Honor f J.C. Harsanyi Dordrecht–Boston–London: Kluwer.
(= Vienna Circle Institute Yearbook 5)
Juha Manninen/Friedrich Stadler (2010) (Eds.). The Vienna Circle in
the Nordic Countries. Networks and Transformations of Logical Em-
piricism. Dordrecht: Springer.
István Majoros/Zoltán Marusza/Oliver Rathkolb (2010) (Hrsg.). Öster-
reich und Ungarn im Kalten Krieg. Wien–Budapest.
Gábor Palló (2008), “Wo ist das Zentrum?” in Habitus, Identität und die
exilierten Dispositionen. Hrsg. von Anna Wessely, Károly Kókai and
Zoltán Péter. pp.121-137.
Zoltán Péter (2010). Lajos Kassák, Wien und der Konstruktivismus 1920–
1926. Frankfurt/M.–Berlin–Bern–Bruxelles–New York–Wien: Peter
Lang.
Miklós Rédei/Michael Stöltzner (2000) (Eds.), John von Neumann and

the Foundation of Quantum Physics. Dordrecht–Boston–London (=
Vienna Circle Institute Yearbook 8).
M. Rédei, editor. John von Neumann: Selected Letters, volume 27 of
History of Mathematics, Rhode Island, 2005. American Mathematical
Society and London Mathematical Society.
Ursula Seeber (2003) (Hrsg.). Asyl wider Willen. Exil in Österreich 1933
bis 1938. Wien: Picus.
Friedrich Stadler (1982).Vom Positivismus zur “Wissenschaftlichen Welt-
auffassung”. Am Beispiel der Wirkungsgeschichte von Ernst Mach in
Österreich von 1895–1934. Wien–München: Löcker.
Friedrich Stadler (2001). The Vienna Circle. Studies in the Origins,
Development, and Influence of Logical Empiricism. Wien–New York:
Springer.
Friedrich Stadler/Kurt R. Fischer (Hrsg.). Paul Feyerabend. Ein Philo-
soph aus Wien. Wien–New York: Springer.
Friedrich Stadler (2011). “The Road to Experience and Prediction from
within: Hans Reichenbach’s scientific correspondence from Berlin to
Istanbul”, in: Synthese, forthcoming (DOI 10.1007/s11229-009-9595-
8).
L. E. Szabó. “Formal systems as physical objects: A physicalist account
of mathematical truth.” International Studies in the Philosophy of
Science, 17: 117-125, 2003.
Péter Szegedi (1997). “Der Galilei-Kreis und die Polányis“, in: Weibel,
522ff.
A. H. Taub, editor. John von Neumann: Collected Works, volume I.
Logic, Theory of Sets and Quantum Mechanics, New York and Oxford,
1961. Pergamon Press.
J. von Neumann. “Zur Hilbertschen Beweistheorie.” Mathematische
Zeitschrift, 26: 1-46, 1927. in [15] No. 12.
Peter Weibel (1997) (Hrsg.). Jenseits von Kunst. Wien: Passagen Verlag.
Vladimir Zeman (1970). “The Philosophy of Science in Eastern Europe”,
in: Zeitschrift für allgemeine Wissenschaftstheorie, Band 1, 133-141.
Károly Kókai
Anton Fischers philosophische Grundlagen der

wissenschaftlichen Erkenntnis
Es gibt zwei Arten von wissenschaftlicher Erkenntnis. Zum einen ist es

dem Wissenschaftler bewusst, dass diese Erkenntnis auf philosophischen
Grundlagen beruht, und es werden sowohl diese Grundlagen als auch das
Verhältnis zwischen wissenschaftlicher Erkenntnis und philosophischen
Grundlagen während des Erkenntnisvorganges reflektiert. Zum anderen
wird diese Einsicht ignoriert und bleiben die entscheidenden Auswirkun-
gen dieses Verhältnisses – wenn auch in der wissenschaftlichen Erkenntnis
selbst genauso wirksam – unreflektiert. Obwohl also jede wissenschaftliche
Tätigkeit methodologische und philosophische Überlegungen mit
einschließt, werden diese wesentlich seltener auch thematisiert. Und es
kommt schließlich äußerst selten vor, dass ein praktizierender Wissen-
schaftler seine Reflexionen systematisch zusammenfasst und ein Buch
zum Thema philosophische Grundlagen der wissenschaftlichen Erkennt-
”
nis“publiziert.
Anton Fischer, der Verfasser eines solchen Werkes, studierte 1919–
1926 in Wien Medizin. Bereits während seines Studiums widmete er sich
biochemischen Forschungsfragen, nach seinem Studienabschluss arbeite-
te er in einem rheumatologischen Laboratorium in Aachen.1 1933 musste
er Deutschland verlassen. Er übersiedelte zunächst nach Frankreich, im
nächsten Jahr nach Ungarn und arbeitete in einem biochemischen La-
boratorium einer Pharmafirma. Während dieser Zeit publizierte er seine
Forschungsergebnisse und auch monographische Darstellungen einzelner
Fachgebiete, insbesondere über Rheumatismus. Nach dem Zweiten Welt-
krieg arbeitete er zunächst als praktischer Arzt (1945/46), dann (1946)
1 Ich möchte János Fischer, dem Sohn von Anton Fischer, für Informationen
beim Verfassen dieses Textes danken. Für wichtige Ergänzungen und Kri-
tik bin ich außerdem Friedrich Stadler, Christoph Limbeck-Lilienau, Edwin
Glassner, Eckehart Köhler und András Máté zu Dank verpflichtet.
26 Károly Kókai
wurde er an der Universitätsklinik in Budapest zum Dozenten, später

zum Professor ernannt (bis zur Pensionierung 1971). Laut seinem Nach-
ruf veröffentlichte Fischer über einhundertfünfzig Mitteilungen, ca. 15
”
Bände, Monographien und selbstständige Abhandlungen“ 2 , er war knapp
30 Jahre lang Mitherausgeber der Ungarischen Ärztlichen Wochenzeit-
schrift und ab 1968 Herausgeber der ebenfalls in Ungarn erschienenen
Reihe Die aktuellen Probleme der Medizin. Bis 1978 sind beim Verlag
Medicina in Budapest 36 Bände der Reihe erschienen.
Anton Fischer war aber nicht nur Arzt, sondern er verfasste auch phi-
losophische Texte, so die bereits erwähnten Philosophischen Grundlagen
der wissenschaftlichen Erkenntnis.3 Das Buch wurde deutsch beim Wiener
Springer Verlag 1947 und in einer überarbeiteten zweiten Auflage 1967 pu-
bliziert. Es verdient nicht nur allein deshalb unsere Aufmerksamkeit, weil
es sich hier um die wissenschaftsphilosophische Publikation eines prak-
tizierenden Naturwissenschaftlers handelt, sondern auch deshalb, weil es
sowohl in der deutschsprachigen als auch in der ungarischen philosophi-
schen Diskussion alleine dasteht.
Grundlagen
Die durchgehende Struktur der Grundlagen ist, dass Fischer erkenntnis-
theoretische Themenbereiche systematisch abhandelt, indem er die we-
sentlichsten Probleme dieser einzelnen Themenbereiche vorstellt und die
wichtigsten Lösungsvorschläge diskutiert. Zum Abschluss jedes dieser ein-
zelnen Teile verzeichnet er die relevante Literatur. Fischer zitiert oft; in
manchen Teilen besteht sein Text aus einem Aneinanderreihen von Zita-
ten. Er gibt dazu immer die Person an, die die betreffende Ansicht vertrat,
oft auch das Werk, aus dem die zitierten Sätze stammen, nicht jedoch die
genaue Stelle. Das Buch hat dadurch einen ausgeprägten Einführungs-
2 ,,Dr. Fischer Antal (1901–1979)“, in: Orvosi Hetilap 120 Jg. Nr. 36, 1979
S. 2159, Übersetzung Károly Kókai.
3 Um die philosophische Leistung von Anton Fischer als Ganzes zu würdigen,
sollte man auch seine anderen diesbezüglichen Texte berücksichtigen, so
die 1999 publizierten ,,Fünf Dialoge über Ethik“ (1944/45), in: Existentia,
Szeged 1999, S. 136-169 und die als Manuskript im Nachlass befindliche
Sozialethik.
Anton Fischers philosophische Grundlagen der wissenschaftlichen Erkenntnis 27
charakter, auch in dem Sinne, dass es das gesamte Gebiet der Erkennt-
nistheorie abdeckt.
Das Buch besteht aus sieben Teilen: (1) Einführung über das Wesen
wissenschaftlicher Erkenntnis, (2) psychologische und (3) logische Grund-
lagen der Erkenntnis, (4) wissenschaftliche Arbeitsmethoden. Dem fol-
gen zwei Kapitel über (5) Grundprobleme der Erkenntnistheorie und (6)
Grenzfragen“ der Naturphilosophie. Den Abschluss bildet ein Kapitel
”
über die (7) Geisteswissenschaften.
Fischers Zugang führt auffallend oft über die Biologie und die Medizin,
wie das z. B. folgende Passage zeigt:
Die Welt ist ihrem Wesen nach biologistisch indifferent – sie wäre im Großen
und Ganzen ohne jedes Leben so, wie sie ist –, unsere Sinne sind biologisch
bedingt, sie haben sich an die speziellen Lebensumstände des Menschen ange-
passt, sie zeigen in erster Reihe biologisch Wichtiges an. Nicht als ob irgendein
Postulat bestünde, wonach der Mensch das Wirkliche nicht erkennen darf –
für diese Annahme fehlt jeder vernünftiger Grund –, aber die Sinnesorgane des
Menschen sind nun einmal nicht für die Erkenntnis der Wirklichkeit an sich, son-
dern nur für die relative Wirklichkeit der menschlichen Lebenssphäre geschaffen.
Von den Möglichkeiten der Erkenntniswege stehen uns direkt nur diejenigen
offen, welche für die Lebenserhaltung nützlich sind – und diese Auswahl hat
auf die Bedürfnisse der transzendenten Erkenntnis keine Rücksicht genommen.
Wir betonen nochmals: Die Beschränkung ist vom biologischen Standpunkt aus
zweckmäßig, vom Erkenntnisstandpunkt aus rein zufällig. Kein metaphysisches
Postulat hat die Sehempfindung auf einen schmalen Abschnitt des elektroma-
gnetischen Spektrums beschränkt, sondern die biologische Zweckmäßigkeit –
wiewohl es für die physikalische Erkenntnis sehr wertvoll wäre, wenn unser Au-
ge auch kurzwelligere Strahlungen empfinden könnte, da wir in diesem Fall sehr
vieles direkt sehen könnten, was wir heute durch mühevolle Versuche indirekt
erschließen müssen. Andererseits ist es zweifellos, dass unsere Sinnesempfindun-
gen die Wirklichkeit wiedergeben – wenn auch nur einen kleinen Teil der Wirk-
lichkeit und wenn das Bild auch stellenweise durch Sinnestäuschungen verzerrt
ist.4
Fischers Zugang wirkt, angesichts der Tatsache, dass Physik in der

ersten Hälfte des 20. Jahrhunderts durchgehend als Leitwissenschaft“
”
galt und etwa im Wiener Kreis die meisten Protagonisten Physiker waren,
oft erfrischend. Er eröffnet neue Perspektiven, die erkenntnistheoretische
Einsichten in einen Lebenszusammenhang bringen und neu bestärken.
4 Fischer 1947, S. 151.

28 Károly Kókai
Da die Grundlagen als Einführung verfasst sind, systematisiert Fischer

bereits Erarbeitetes und Publiziertes. Er popularisiert damit die wissen-
schaftliche Weltauffassung und steht damit in der Tradition des Wiener
Kreises, wo ein wichtiger Aspekt ebenfalls die Popularisierung der Er-
gebnisse der modernen Naturwissenschaften war. Einführung, ein Zugang
durch Beispiele aus den Bereichen der Biologie und Medizin und Popula-
risierung sind also drei charakteristische Merkmale der Grundlagen.
Das Buch ist mit einem Index ausgestattet, in dem die Namen Mo-
ritz Schlick, Ernst Mach, Immanuel Kant und Carl Stumpf am häufigsten
vorkommen. Die Gewichtung dieser Namen ist allerdings sehr unterschied-
lich. Mach gilt als Negativbeispiel. Fischer scheint mit Schlick zu sympa-
thisieren. Stumpf dürfte als gerade aktueller Kollege Geltung haben. Und
Kant ist die historische Autorität. Aufgrund der allgemeinen Thematik
des Buches, seiner Grundcharakteristik oder der Angaben im Index kann
man also Fischer nicht einer bestimmten Denkrichtung zuschlagen oder
die Frage beantworten, welche Position er sich selbst im Verhältnis zum
Wiener Kreis zuweist.
Das Verhältnis zum Wiener Kreis

Im Zuge der Vorstellung der Entwicklung und der verschiedenen Posi-
tionen der Logik kommt Fischer auf die Logistik5 und so auch auf den
Wiener Kreis:
Vom Wiener Kreis wurde die Logistik mit einem extremen Positivismus verket-
tet, der von Carnap zu einer Theorie der Relationserkenntnis ausgebaut worden
ist, die uns später beschäftigen wird. (. . . ) Bereits Wittgenstein hat die tautolo-
gische, daher unwiderlegbare Natur der logischen Sätze behauptet: Die Logik
”
besteht nur aus konventionellen Festsetzungen über Gebrauch von Zeichen und
aus Tautologien auf Grund dieser Festsetzungen.“ Diese Ansicht, welche auch
von H. Hahn vertreten wurde, betrachtet daher nicht mehr allein die Worte,
sondern auch die logische Form ihrer Verkettung als Konventionen: Der Satz
der Identität und des Widerspruches sind darnach willkürliche Annahmen, die
auch durch andere Sätze ersetzt werden könnten.6
5 Fischer 1947, S. 56ff.

6 Fischer 1947, S. 58.
An dieser Ansicht übt Fischer Kritik. Die Verkoppelung von Logistik

”
und extremem Positivismus“ kompromittiere nur die Logistik, die richtig
eingesetzt zur Klärung der wissenschaftlichen Sprache beitrage.7
Über den Wiener Kreis schreibt Fischer also auch explizit. An die-
ser hier zitierten Stelle isoliert er die beiden markantesten Elemente in
der wissenschaftstheoretischen Position des Wiener Kreises, die auch als
Logischer Positivismus respektive Logischer Empirismus8 angegeben wur-
den. Logik wertet er positiv, Positivismus lehnt er ab. Das entspricht der
zuvor bei der Aufzählung der Namen Mach, Schlick, Stumpf und Kant
festgestellten Position. Mach wird von Fischer insofern abgelehnt, als er
für einen extremen Positivismus“ steht.
”
Fischers Position dem Wiener Kreis gegenüber zeigt, dass er in den
1940er-Jahren schreibt. Dementsprechend nimmt er eine etwas distan-
ziertere Haltung ein – wie sie bald allgemein vertreten wurde.9 Mitte der
1940er-Jahre war der Wiener Kreis bereits internationalisiert worden, er
war durch die Emigration und durch die laufende Etablierung in den aka-
demischen Institutionen der Vereinigten Staaten Teil einer internationalen
Diskussion und stark umstritten. Was Fischer wiedergibt, sind also nicht
einfach die Thesen des Wiener Kreises, sondern er ortet diese in einer wis-
senschaftsgeschichtlichen Diskussion. Fischers Buch ist insofern auch ein
Beitrag zur internationalen Rezeption des Wiener Kreises.
7 Fischer 1947, S. 58.

8 Herbert Feigl veröffentlichte 1931 zusammen mit Albert Blumberg einen
Text mit dem Titel Logical Positivism“ im Journal of Philosophy 1931, S.
”
281-296. In den nächsten Jahrzehnten wurde diese Benennung verwendet.
Neuerdings scheint die zweite Benennung sich durchzusetzen. Vgl. z. B.
Alan Richardson and Thomas Uebel (Eds.) The Cambridge Companion to
Logical Empiricism, Cambridge: Cambridge University Press 2007.
9 Vgl. Alfred Jules Ayer über Verifikation 1936/1946; Hilary Putnam über
Bedeutung 1951; Willard Van Orman über Dogmen des Empirismus 1951;
Thomas Kuhn über die Entwicklung der Wissenschaft 1962. Diese vier Wer-
ke eröffneten kritische Zugänge zur Position des Logischen Empirismus, die
alle von innen“ kamen und bis in die 1970er-Jahre hinein zur allgemeinen
”
Diskreditierung der Bewegung führten. Vgl. dazu die diesbezügliche Kri-
”
tik der Kritik“ in Richardson et al. 2007. Ab den 1980er-Jahren wurde die
Bewegung vermehrt aus einer historischen Perspektive betrachtet, was zu
ihrer Neubewertung führte.
30 Károly Kókai
Insbesondere scheint ihm viel an der Positivismuskritik zu liegen. So

an der vorher zitierten Stelle10 sowie auf den Seiten 158-166, die ich im Fol-
genden etwas ausführlicher zitieren möchte, um ein Beispiel von Fischers
Argumentationsstil zu geben. Hier geht er zunächst auf Carnaps erwähnte
Theorie der Relationserkenntnis ein.
Zu einem System der Erkenntnis wurden die Relationen durch R. Carnap in

seinem Werke Der logische Aufbau der Welt ausgebaut. Es gibt nach Carnap
zwei Arten von Beschreibungen: die Eigenschaftsbeschreibungen, welche indi-
viduelle Angaben (z. B. rot, dünn) benützen, und die Beziehungsbeschreibun-
gen, welche auf relativen Angaben (z. B. Verwandtschaftsbeziehungen) beruhen.
Das Ziel jeder wissenschaftlichen Theorie ist, ihrem Inhalt nach zu einer reinen
Beziehungsbeschreibung zu werden. Jeder Gegenstandsname, der in einer wis-
senschaftlichen Angabe vorkommt, kann grundsätzlich durch eine strukturelle
Kennzeichnung des Gegenstandes ersetzt werden. Die Wissenschaft will vom
”
Objektiven sprechen, alles jedoch, was nicht zur Struktur, sondern zum Materi-
ellen gehört, ist letzten Endes subjektiv.“ Das Konstitutionssystem“, welches
”
mit Hilfe der Logistik errichtet werden soll, ergibt eine ,einheitliche Ordnung der
Begriffe, aus der heraus die Frage der einzelnen Probleme schärfer gefasst und
damit einer Lösung nähergebracht wird‘. Während jedoch Carnap die Relatio-
nen für logische Konstruktionen hält, die lediglich innerhalb der Wahrnehmungs-
welt anwendbar sind – reale“ Beziehungen gäbe es nicht –, betont Schlick (in
”
der Allgemeinen Erkenntnislehre 2. Aufl. 1925), dass die Beziehungen unse-
rem Denkakt vorausgehen; die Relationen sind also nicht Denkformen, sondern
”
müssen als Formen des Gegebenen angesehen werden.“11
Fischer referiert also nicht bloß die Position des Wiener Kreises, sondern
differenziert zwischen den einzelnen Vertretern, hier Carnap und Schlick.
Daran schließt er seine erwähnte Positivismuskritik an. Der Positi-
vismus habe seine Berechtigung als Reaktion auf die irreale Naturphi-
”
losophie des nachkantischen Idealismus“12 , er schoss aber über das Ziel
”
hinaus“ und erklärte das Wesen der Dinge für prinzipiell unerkennbar.“13
”
Wie Fischer allerdings ausführt:
Nach M. Schlick (Allg. Erkenntnislehre 2. Aufl. 1925) müsste der Positivis-

mus konsequenterweise nicht bloß die Wirklichkeit der Atome, sondern auch
die Rückseite des Mondes und das Erdinnere leugnen, die pedantisch strenge
10 Fischer 1947, S. 58.

11 Fischer 1947, S. 161f.
12 Fischer 1947, S. 164.
13 Fischer 1947, S. 165.
Durchführung seines Programmes würde einen Verzicht auf Erkenntnis über-

haupt bedeuten.“14
Zur Kritik am Positivismus fügt Fischer seine eigene Position hinzu und
zeigt damit, wie zentral die vorherige Diskussion des Positivismus für ihn
ist. Wir stünden
(. . . ) auf dem Standpunkt, dass unsere Erkenntnis das Wissen um die reale Welt
ist und dass der Mensch im Prinzip wohl fähig ist, das Innerste der Dinge zu
erkennen – allerdings unter unendlichen Schwierigkeiten, oft auf beschwerlichen
Umwegen und ständig behindert durch die biologisch bedingte Unvollkommen-
heit seiner Sinne und seines Denkapparates.15
Die vollkommene Erkenntnis des Seienden“ scheine in unerreichbarer Fer-

”
ne. Eine schrittweise Annäherung sei aber der Weg der Wissenschaften.
Die einzelnen Etappen dieses Fortschritts seien durch methodische Ent-
deckungen markiert. Das Ziel dieser Entwicklung sei deutlich nach dem
”
Wesen‘ der Dinge gerichtet (. . . ), nach der lückenlosen Erfassung des Sei-
’
enden, des Dinges an sich, im Gegensatz zu den bloßen ,Erscheinungen‘.“16
Und so gelangt er auch zu seiner Formulierung, was Wissenschaft sei:
Die Wissenschaft ist keine bloße Beschreibung, keine Erforschung von ,Erschei-
nungen‘, hinter welchen ein ewig und prinzipiell unerforschbares Ding an sich
verborgen ist – sie ist die unendlich mühevolle, schrittweise Erfassung des Sei-
enden, ein ständiges Fortschreiten in der Richtung des Dinges an sich.17
Die zweite Auflage des Buches ist als die philosophisch reifere zu be-
zeichnen – man sieht, dass der Autor nicht mehr in einer existenziellen
Notsituation schreibt wie während des Zweiten Weltkrieges und auch dass
er die Gelegenheit hatte, seine philosophischen Ansichten aufgrund der
Ergebnisse der wissenschaftlichen Entwicklung nochmals zu überprüfen –
und stellt auch in Hinsicht auf die Darstellung des Wiener Kreises einen
Fortschritt dar. Fischer 1967 berücksichtigt nicht nur den Positivismus
und den logischen Empirismus, nicht nur Schlick und Carnap, sondern
auch eine philosophische Diskussion u. a. mit Victor Kraft, Karl Popper,
Béla Juhos, Robert Reininger und Autoren der Encyclopedia of Unified
14 Fischer 1947, S. 165.

15 Fischer 1947, S. 167.
16 Fischer 1947, S. 167.
17 Fischer 1947, S. 177.
32 Károly Kókai
Science wie Philipp Frank und Charles Morris. Er ergänzt seinen Text
von 1947 mit einzelnen Argumenten – ohne jedoch an der Argumentation
seines Buches selbst etwas zu verändern. Aufmerksamkeit verdient, dass
er sich im Vorwort der Neuauflage als kritischer Realist“ deklariert.
”
So ein ergänzendes Argument ist beispielsweise das folgende in einer
Auseinandersetzung mit Béla Juhos’ Die Erkenntnis und ihre Leistung
1950:
Juhos meint, dass es auch Sätze gibt, wo ein Irrtum logisch ausgeschlossen ist
und die nicht verifizierbar, sondern entweder wahr oder Lüge sind, so ist zu-
mindest das von ihm angeführte Beispiel ( Ich fühle Schmerz“) meist leicht
”
verifizierbar. Auch ist es kein glückliches Beispiel dafür, dass solche Konstatie-
”
rungen“ keine Voraussagen zulassen: Der typische Schmerz des Herzinfarktes,
der Gallen- oder Nierenkolik oder der Schmerz einer Bandscheibenhernie las-
sen sehr wohl Voraussagen zu. Es wäre sehr riskant, die Wissenschaft auf einem
Fundament unverzichtbarer Konstatierungen“ zu errichten, umso mehr, als der
”
Übergang von der Konstatierung zu hypothetischen Sätzen, wie Juhos meint,
ein außerlogischer, irrationaler Schritt ist.18
Fischer spricht hier das in einschlägigen philosophischen Diskussionen wie-

derholt verwendete Argument an, dass die Aussage Ich fühle Schmerz“
”
nicht überprüfbar sei, weil das nur für denjenigen empirisch gegeben sei,
der diese Aussage tätige. Als Arzt weist er darauf hin, dass Schmerz alles
andere als empirisch nicht nachprüfbar“ sei, sonst wäre er ja für die Me-
”
dizin unerklärlich und auch unbehandelbar. Fischers Stellung bezüglich
der Diskussion der philosophischen Relevanz von Sätzen wie Ich fühle
”
Schmerz“ ist ein weiteres Beispiel für seinen Zugang durch Beispiele aus
der Medizin und der Biologie und zwar genau in dem Sinne, in dem vorher
dieser Standpunkt charakterisiert wurde. Der Mensch sei zwar in vielfa-
cher Hinsicht beschränkt – Fischer wies insbesondere auf die Grenzen
unserer Sinneswahrnehmungen hin –, auf indirektem Wege sei er jedoch
fähig, zur Erkenntnis zu gelangen. Die Tatsache, dass Schmerz subjektiv
sei, erklärt diese Art von Sätzen nur in einer naiv unreflektierten Argu-
mentation für eine eigene Kategorie – ein Irrtum sei logisch ausgeschlossen,
nicht verifizierbar, sondern entweder wahr oder Lüge. In einer konsequent
durchgeführten Analyse reihen sich diese Sätze genauso wie alle anderen,
denen empirische Beobachtungen zugrunde liegen, bei den überprüfbaren
ein.
18 Fischer 1967, S. 133f.

So viel zu den Philosophischen Grundlagen, Auflagen 1 und 2. Es ist

klar, dass die Publikation mit der Wirkung respektive Ausstrahlung des
Wiener Kreises zu tun hat, und es ist auch klar, dass diese Beziehung
(zwischen den Philosophischen Grundlagen und dem Wiener Kreis) keine
direkte ist, Fischer also nicht einfach die Meinung des Wiener Kreises“
”
vertritt. Wie ist aber dieses Verhältnis genau?
Um diese Frage beantworten zu können, vorweg eine Bemerkung. Der
Wiener Kreis, die vom Wiener Kreis vertretene Position, war nicht ein-
fach und homogen.19 Im Wiener Kreis gab es verschiedene Positionen
und sowohl der Wiener Kreis als auch die Position des Wiener Kreises“
”
durchliefen eine Entwicklung. Was aus den bisher zitierten Stellen von
Fischer sehr klar ersehen werden kann, indem er also die Positionen von
Carnap und Schlick einander gegenüberstellte20 und indem er den Positi-
vismus etc. historisch positionierte.21 Um die Frage nach dem Verhältnis
von Fischers Publikation zum Wiener Kreis sinnvoll stellen zu können,
möchte ich kurz auf diese Position (also die Position des Wiener Krei-
”
ses“) Mitte der 1940er- und Mitte der 1960er-Jahre (als die 1. bzw. die 2.
Auflage der Grundlagen geschrieben wurde) hinweisen.
19 Vgl. dazu die Beiträge in Richardson 2007, z. B.: “there was no homoge-
nous Vienna Circle with an undisputed program and verificationism at its
center”, Friedrich Stadler in Richardson 2007 S. 34; nachdem Thomas Mor-
mann den Standpunkt Carnaps bezüglich der Struktur von wissenschaftli-
chen Theorien vorstellt und auf jene von einigen anderen aus Platzgründen
verzichtet, schreibt er: “it should have become clear that speaking of the
logical empiricist view of theories is seriously misleading in so far as such an
expression suggests that there was a unique logical empiricist answer to the
T[eory]-question”, die lautet: “What is the structure of scientific theories?”
Und er setzt fort: “Postpositivist philosophers of science such as Feyerabend,
Putnam and others (. . . ) [reduced] the plurality of logical empiricist ans-
wers to the T[heory]-question to what the postpositivists ominously dubbed
the ‘received’ or the ‘orthodox view”’, Mormann in Richardson 2007 S. 161;
wie George Reisch in Richardson 2007 S. 58-90 zeigt, wurde die Vielfalt des
Wiener Kreises im Kalten Krieg der 1940er- und 1950er-Jahre auf apoliti-
sche Methoden (auf die Analyse) reduziert; “When taken in its extended
breadth [Uebel erwähnt insbesondere Menger, Kaufmann und Zilsel], early
logical empiricism exhibited a remarkable plurality of philosophies of social
science”, Thomas Uebel in Richardson et al. 2007 S. 252.
20 siehe Fn. 11.
21 siehe Fn. 12 und 13.
34 Károly Kókai
In den 1930er- und 1940er-Jahren spielte sich die Entwicklung des lo-
gischen Empirismus fast ausschließlich im englischsprachigen Raum ab,
vor allem in den USA. In Wien waren 1945–47 Victor Kraft und Béla
Juhos tätig. Beide gehörten zur sogenannten Peripherie des Wiener Krei-
ses. Obwohl Victor Kraft bereits in der sogenannten Vorgängergruppe“,
”
die dann zum eigentlichen Wiener Kreis“ führte, tätig war und obwohl
”
er dann auch derjenige war, der nach dem Zweiten Weltkrieg (bis 1952)
eine philosophische Professur in Wien innehatte – er also als derjenige
anzusehen ist, der die Wiederetablierung des logischen Empirismus in der
österreichischen akademischen Institutionen (wenn auch für wenige Jah-
re, aber immerhin) gewährleistete –, wird seiner Wiener Wirkung nach
dem Zweiten Weltkrieg – er starb 1975 – wenig Beachtung geschenkt.22
Kraft konnte auch einen seiner Kollegen, nämlich Juhos, an der Wiener
Universität habilitieren.
Auch konnte Kraft beim Wiener Springer Verlag in dieser Zeit einige
Bücher herausbringen, in denen er u. a. über die Geschichte des Wiener
”
Kreises“ berichtete23 sowie seine Einführung in die Philosophie 24 und sei-
ne Mathematik, Logik und Erfahrung.25 Und genau in diesem Verlag und
sozusagen in der selben Reihe erschienen auch Fischers Grundlagen.26
22 Als Ausnahmen sind anzuführen: Friedrich Stadler, Paul Feyerabend. Ein

”
Philosoph aus Wien“, in: Friedrich Stadler und Kurt Rudolf Fischer (Hrsg.),
Paul Feyerabend, Wien: Springer 2006, S. IX-XXXIV; Jan Radler, Victor
Krafts konstruktivistischer Empirismus, Berlin: Logos 2006 und Friedrich
Stadler What is the Vienna Circle“, in: Friedrich Stadler, The Vienna
”
Circle and Logical Empiricism, Wien: Springer 2002, S. XI-XXIII.
23 Victor Kraft Der Wiener Kreis. Der Ursprung des Neupositivismus. Ein
Kapitel der jüngsten Philosophiegeschichte, Wien: Springer 1950.
24 Victor Kraft Einführung in die Philosophie. Philosophie, Weltanschauung,
Wissenschaft, Wien: Springer 1950.
25 Victor Kraft Mathematik, Logik und Erfahrung, Wien: Springer 1947.
26 Sozusagen deshalb, weil es keine deklarierte Reihe gab. Vom Verlag wur-
den diese Bücher aber als thematisch verwandt angesehen, was die Or-
ganisation der Verlagsankündigungen am Deckel der Bücher beweist. So
wurden am Deckel von Krafts Der Wiener Kreis 1950 u. a. seine eigene
Einführung in die Philosophie 1950 sowie Mathematik, Logik und Erfah-
rung 1947, Fischers Grundlagen, zwei Publikationen zur Rechtsphilosophie:
A. Verdroß-Droßberg, Grundlinien der antiken Rechts- und Staatsphiloso-
phie 1948 und J. Esser, Einführung in die Grundbegriffe des Rechtes und
Staates, ein Buch zu Psychologie, P. v. Schiller, Aufgabe der Psychologie
1948, ein Buch zu Wirtschaftsphilosophie, O. Weinberger, Die Wirtschafts-
Aufgrund dieser Buchreihe“ ist anzunehmen, dass es in den 1950er- und

”
1960er-Jahren mit Victor Kraft, Ernst Topitsch und eben auch Anton
Fischer eine wissenschaftstheoretische, wenn auch eine reduzierte, Dis-
”
kussion“ gab.
Anton Fischer kannte die Arbeit des Wiener Kreises. Aus seinen
Grundlagen ist klar ersichtlich, dass er die Theorie des Wiener Kreises
nicht unkritisch übernahm, sondern sie als eine von mehreren relevanten
Positionen referierte. Aus den bisher bekannten biographischen Fakten27
ergibt sich auch eine (gespenstische) Parallele, ohne sichtbare Berührungs-
punkte. Fischer studierte in Wien exakt zu der Zeit, als sich der Wiener
Kreis um Schlick formierte. Da Fischer 1926 promoviert wurde, ist es chro-
nologisch möglich, dass er vom 1924 gegründeten (und ab Schlicks Ankunft
in Wien 1922 informell vorbereiteten) Wiener Kreis wusste und, wenn er
bei den formellen Donnerstagstreffen persönlich auch nie anwesend war,
Personen in dessen Umkreis kannte. Er vertiefte sein philosophisches Wis-
sen (wie aus den Literaturangaben in den Grundlagen ersichtlich) zu einer
Zeit, als in der freien Welt“ die Wirkung des Wiener Kreises“ dominant
” ”
war. Er veröffentlichte die erste Auflage seiner Grundlagen zu einem Zeit-
punkt, als in Wien die Wissenschaftsphilosophie, trotz propagierter po-
litischer Wende, immer noch und bezeichnenderweise verbannt war und
hätte von den damals noch lebenden Mitgliedern wahrgenommen werden
können. Als im Zuge des langsamen Auftauens“ des Kalten Krieges um
”
1960 Fischer wieder ins Ausland reisen konnte, nahm er in Salzburg an
einer Konferenz teil, als dort eine ab 1936 jahrzehntelang unterbrochene
Tradition wieder fortgesetzt wurde.28 Und 1968 plante er zur Zeit des
philosophie des Alten Testaments 1948 und zwei Bücher zu Geistesgeschich-

te: R. L. John, Dante 1948 und F. Krennbauer, Goethe und der Staat 1949
angekündigt. Der Deckel von Fischer 1947 blieb leer. 1967 waren es Victor
Kraft, Erkenntnislehre 1960, Ernst Topitsch (Hrsg.), Probleme der Wissen-
schaftstheorie. Festschrift für Victor Kraft 1960, Arthur Pap, Analytische
Erkenntnistheorie 1965, K. D. Heller, Ernst Mach 1964 und H. Stachowiak,
Denken und Erkennen im kybernetischen Modell 1965.
27 János Fischer plant nach eigenen Angaben aufgrund der in seinem Besitz
befindlichen Dokumente eine Biographie seines Vaters. Erst nach Fertigstel-
lung dieser Biographie kann man dann in dieser Hinsicht gesicherte Aussa-
gen machen.
28 (. . . ) außerdem erhielt ich von Prof. [Othmar] Anderle eine Einladung,
”
am Synopsiskongress der Internat. Gesellschaft für vergleichende Kultur-
36 Károly Kókai
Weltkongresses der Philosophie wieder eine Reise nach Wien.29 Das heißt
aber nicht bloß, dass hier eine Reihe von zufälligen Gemeinsamkeiten vor-
liegt, sondern dass Fischers Bemühungen mit denen des Wiener Kreises“
”
parallel laufen, dass er auf die Möglichkeiten der Zeit“ sensibel reagier-
”
te. Es heißt aber auch noch etwas Weiteres, was im Zusammenhang der
Fragestellung der Tagung Die Neubewertung des Einflusses des Wiener
”
Kreises in Ungarn“, auf der eine erste Version dieses Aufsatzes präsentiert
wurde, wesentlich ist: Dass er mit dem Wiener Kreis nicht zusammentraf,
ist kein bloßer Zufall. Die beiden trennte eine Wand.
Die Grundlagen in ihrem Entstehungskontext

Der Springer Verlag sammelte die Rezensionen der Philosophischen
Grundlagen und ließ sie dem Autor zukommen. Glücklicherweise hob
Anton Fischer diese Rezensionen auf und so konnte ich Einsicht in sie
nehmen. Sie sind meist kurz und erkennen die Publikation an. Eine wis-
senschaftliche Auseinandersetzung findet aber nicht statt.
Im Konvolut befinden sich zwei Rezensionen aus Österreich: Reinhard
Sexl aus Wien in den Monatsheften für Mathematik.30 In Acta Physica
Austriaca stellt Paul Urban aus Graz das Buch kurz vor31 – mit einigen
etwas kuriosen Formulierungen.
forschung in Salzburg über die Problematik der Ganzheit ein Referat zu

halten, der ich gerne Folge leiste.“ Brief Anton Fischers an Robert John
vom 5. Juni 1964; weitere Teilnehmer waren laut eines undatierten Briefes
Fischers (wahrscheinlich Sommer 1964): Schubert-Soldern, Schachermeyr,
”
Hilokinomi (Mainz) Gunther, Landtmann, Mitchells v. Freytag-Lönighoff,
Engelmann“; ebenfalls in Salzburg veranstaltete das Internationale For-
schungszentrum für Grundfragen der Wissenschaften Forschungsgespräche,
das vierte im August 1964. An diesem Forschungsgespräch über Die ana-
”
lytischen Sätze und die Grundlagen der Wissenschaften“ nahmen u. a. Her-
bert Feigl, Paul Feyerabend, Rudolf Haller, Béla Juhos, Werner Leinfellner,
Wolfgang Stegmüller und Paul Weingartner teil.
29 Ich habe einen Reisepass erhalten und gedenke am 1. September für etwa
”
10 Tage nach Wien zu fahren.“ Brief Anton Fischers an Robert John vom
13. Juni 1968. Aus den mir zur Verfügung stehenden Informationen wie der
Teilnehmerliste des Kongresses sowie den Erinnerungen von János Fischer
geht nicht hervor, ob diese Reise tatsächlich stattgefunden hat.
30 1968 (Bd. 72), H. 4.
31 1969 (Bd. 30), H. 4.
Für die Zeit bezeichnend ist eine ausführliche Rezension von Chris-
tiane Hein aus Leipzig:
Der Autor [also Fischer] lebt in einem sozialistischen Land, man hätte von ihm
erwarten können, dass er sich informiert. Dieser Mangel an Information über
den Marxismus fällt (. . . ) auf.32
Bzw.:
Die ausgewertete Literatur wird am Ende jedes Kapitels in einer Bibliographie

zusammengefasst. So gelingt es Fischer, einen Überblick über die Autoren zu
verschaffen, die zu dem behandelten Problem Stellung genommen haben. Der
Leser vermisst allerdings Hinweise auf Literatur marxistischer Autoren.33
Von der anderen Seite des Ozeans wurde das Buch ebenfalls bewertet,
Agehananda Bharati aus Syracuse stellt fest,
Fischer ist, wie man wohl erwarten könnte, dem Wiener Kreis und seinen eng-
lisch sprechenden Freunden nicht sehr gewogen. Zwar richtet er keine eigentli-
che Polemik gegen diese – oder auch gegen eine andere – Philosophierichtung,
er bedauert aber z. B. dass die Logistik‘ die Tendenz hat, sich von ontologi-
’
schen Erwägungen zu absolvieren, und zwar, wie der Autor es fasst, wegen ihrer
’
Überschätzung des formalen Denkens‘ (S. 52).
Bharati schließt seine Rezension mit den Worten:
Es ist für die westlichen Philosophen erquicklich zu wissen, dass hinter dem
Eisernen Vorhang sehr subtile, hochwissenschaftliche und vor allem hochinfor-
mierte Philosophie betrieben wird. Hoffentlich ist Anton Fischer typisch für das,
was heute dort philosophisch vorgeht.34
Fischers Arbeit ist zeitgeschichtlich bedingt im Kalten Krieg zu po-

sitionieren. Dies geht aus den zitierten Rezensionen ebenso wie aus der
bereits erwähnten35 Geschichte des Logischen Positivismus im englisch-
sprachigen Raum hervor. Fischers intellektuelle Isolation ist ebenfalls weit-
gehend diesem Umstand zuzuschreiben. Der Kalte Krieg definierte ideo-
logische Fronten und ordnete intellektuellen Positionen jeweils eindeutige
Orte zu. Damit wurde jede noch so apolitische Stellungnahme zur ideologi-
32 Hein in: Deutsche Zeitschrift für Philosophie 1969 (Jg. 17), H. 12, S. 1522.
33 Hein, S. 1523.
34 Zeitschrift für philosophische Forschung 1970 (Bd 24), H 3.
35 Vgl. Fn. 9.
38 Károly Kókai
schen Manifestation, der sich kein Autor entziehen konnte und die für das
jeweilige Werk bestimmend wurde. Fischers Insistieren auf ein fortschrei-
tendes Erkennen der Wahrheit ist dementsprechend ebenfalls in diesem
Zusammenhang zu sehen.
Das Buch wurde in Budapest geschrieben und ist in Wien erschienen;
so gesehen ist es auch ein Beispiel dafür, dass die wissenschaftliche Kom-
munikation nicht nur in eine Richtung lief, sondern dass eine Diskussion
stattfand. Die erste Auflage des Buches ist 1947 erschienen. Dieses Datum
ist in der ungarischen Geschichte äußerst signifikant, weil kurz darauf das
gesamte kulturelle, gesellschaftliche und politische Leben durch die Kom-
munistische Partei Ungarns gewaltsam gleichgeschaltet wurde. Bezeich-
nenderweise wurde die (1892 gegründete und unter der Ägide ihres ersten
Herausgebers, Imre Pauer, eine positivistische Auffassung vertretende)
Zeitschrift Athenaeum 1948 eingestellt. Diese war die bis dahin wichtigste
philosophische Zeitschrift Ungarns, in der auch einige Berichte über die
Tätigkeit des Wiener Kreises bzw. dem Wiener Kreis nahestehende Au-
toren abgedruckt wurden. Etwa über Moritz Schlicks Naturphilosophie36
und ab 1934 vermehrt37 von den Autoren József Somogyi, Ferenc Lehner,
Frigyes Pozsonyi und László Faragó. Die Periodisierungen, die wir in der
österreichischen Kulturgeschichte kennen, gelten in Ungarn nicht. In Un-
garn wurde 1948 eine Zäsur gesetzt. Und deshalb ist es interessant, die
Rezensionen in Athenaeum und Fischer zusammenzunehmen. Die Zwi-
schenkriegsperiode und die drei Jahre nach 1945 bilden eine Einheit, wie
es diese Personen (nämlich Somogyi, Lehner, Pozsonyi sowie Fischer) ver-
deutlichen. Und für diese Periodisierung erscheint es auch interessant, dass
36 József Somogyi, Moritz Schlicks Naturphilosophie“, in: Athenaeum, Buda-

”
pest 1927, S. 187f.
37 Ferencz Lehner Rudolf Carnaps Die Aufgabe der Wissenschaftslogik 1934,
”
Einheitswissenschaft Heft 3“, in: Athenaeum, Budapest 1934, S. 243; Ferenc
Lehner Popper Logik der Forschung“, in: Athenaeum, Budapest 1935, S.
”
147; Ferenc Lehner, Reichenbach Wahrscheinlichkeitslehre“, in: Athenae-
”
um, Budapest 1935, S. 146; Frigyes Pozsonyi, Rudolf Carnap The logi-
”
cal syntax of language 1937“, in: Athenaeum, Budapest 1938, S. 115-117;
Frigyes Pozsonyi, Zur Enzyklopädie der Einheitswissenschaft. Vorträge von
”
Neurath, Brunswick, Hull, Mannoury, Woodger. Einheitswissenschaft Heft 6
1938“, in: Athenaeum, Budapest 1940, S. 81f.; László Faragó, Bela von Ju-
”
hos Erkenntnisformen in Natur und Geisteswissenschaften 1940“, in: Athe-
naeum, Budapest 1940, S. 290f.
Rezensionen zum Wiener Kreis in der Zeitschrift Athenaeum ab dem Jahr

1934 mit einiger Regelmäßigkeit erscheinen, also zu der Zeit, in der Fischer
nach Ungarn zurückkehrt. Was Fischer schreibt, kann man daher auch im
Rahmen dieser (virtuellen) Diskussion in Ungarn betrachten.
Die (neben der nach dem Verhältnis Fischers zum Wiener Kreis) an-
dere interessante Frage ist, wie die Lage der Wissenschaftsphilosophie in
Ungarn 1934–1947 bzw. 1948–1967 zu charakterisieren sei, die Frage nach
dem Milieu also, in dem Fischer lebte und die erste und die zweite Ver-
sion seiner Grundlagen verfasste. Aufgrund der Philosophischen Grund-
lagen scheint Fischer sich mit der ungarischen Philosophie zwar so gut
wie überhaupt nicht auseinandergesetzt zu haben,38 diese Frage erscheint
aber trotzdem angebracht, ist es ja völlig auszuschließen, dass er sich für
die philosophische Kultur seiner Umwelt nicht interessiert hätte.
Bezogen auf die ungarische philosophische Kultur 1948–1967 ist festzu-
halten, dass es sich um eine erstens zentralisierte und zweitens sich wis-
senschaftlich nennende (der Terminus lautete wissenschaftlicher Mate-
”
rialismus“) handelte. Es musste daher von Fischer als grotesk empfun-
den worden sein, dass die offizielle Kultur und, wegen der konsequenten
Durchdringung des intellektuellen Lebens durch die Organe des Staats-
apparates, alle anderen öffentlichen Diskurse vom Gedankenkreis seines
Buches überhaupt keine Kenntnis nahmen.
Fischer ist der Beweis, dass es in Ungarn möglich war, Philosophie in
den dunkelsten Zeiten des Krieges und auch des Sozialismus auf internatio-
nalem Niveau zu betreiben. Fischer ist auch der Beweis, dass folgerichtige
und vertiefte Wissenschaftsphilosophie nicht nur dann möglich ist, wenn
man der internen Dynamik einer formellen Gruppe oder Schule (wie etwa
des Wiener Kreises) folgt. Das alles hat aber seinen Preis. Fischer konnte
veröffentlichen, und das wiederholt. Sein 1967er-Buch bleibt der ersten
Ausgabe treu und spiegelt in einer sehr dezenten Form die Entwicklun-
gen von 20 Jahren wider. Er hat aber als Philosoph nichts erreicht“. Er
”
war nicht bekannt, war nicht anerkannt, wurde institutionell nicht wahr-
genommen, kommt weder in der ungarischen noch in der internationalen
Philosophiegeschichte vor. – Andererseits ist allerdings hier zu bemerken,
dass Fischer gerade dadurch, dass er nicht Teil einer Gruppe war, in einer
polarisierten Zeit ein solides und von Extremen weitgehend freies Werk
38 Als Ausnahme kann angeführt werden, dass er sich einmal auf Ákos Pauler
bezieht: Fischer 1947, S. 52.
40 Károly Kókai
vorlegen konnte. Die in den akademischen Institutionen und in Philoso-

phieschulen sich entwickelnde Dynamik hätte eine solche Neutralität kaum
erlaubt.
Wir sehen, Anton Fischer steht sowohl in der deutschsprachigen als
auch in der ungarischen philosophischen Diskussion alleine da“. Dies
”
nicht allein deshalb, weil er durch äußere Umstände dazu gezwungen
”
war“. Vier dieser Umstände waren: (1) Er konnte nicht Philosophie stu-
dieren, sondern war von seinem Vater dazu angehalten worden, ein Me-
dizinstudium abzuschließen. (2) Während des Zweiten Weltkrieges39 zog
er sich zurück und schrieb das Buch Die philosophischen Grundlagen der
wissenschaftlichen Erkenntnis. (3) Ab 1934 lebte er in Ungarn in einem
zunehmend radikalisierten politischen Klima, in einer von der interna-
tionalen Diskussion weitgehend abgekoppelten Welt, wo außer in den
erwähnten Rezensionen im Athenaeum niemand von der gleichzeitig statt-
findenden internationalen wissenschaftstheoretischen Diskussion Kenntnis
genommen zu haben schien. (4) Nach einer kurzen Zwischenphase (1945–
1948), als es möglich schien, eine adäquate Diskussion zu führen – und
Anton Fischer nutzte die Möglichkeit, 1947 sein Buch herauszubringen –,
folgte wieder eine für die Philosophie äußerst ungünstige Zeit. Als sich
dies in den 1960er-Jahren langsam änderte, brachte Fischer die zweite
Auflage seines Buches heraus – bezeichnenderweise wieder auf Deutsch
und in Wien. Und damit schaffte er offenbar den ersehnten Durchbruch
wiederum nicht. Aus den zitierten Rezensionen kann man ersehen, dass
seine Leistung zwar anerkannt wurde, aber in eine lebendige Diskussion
fand er trotzdem keinen Eingang. Trotz dieser ungünstigen biographi-
schen und politischen Umstände hätte natürlich auch Fischer einen Weg
”
wählen können“, der ihn stärker in Kontakt mit institutionalisierten philo-
sophischen Diskussionen gebracht hätte. Fischer wollte anscheinend seine
39 Er konnte seinen Beruf nicht ausüben, wie in: Orvosi Hetilap 1979, S. 2159
formuliert wurde: Inmitten der erbarmungslosen und tragisch werdenden
”
Ereignisse des Zweiten Weltkrieges und mit den immer seltener werdenden
Möglichkeiten der experimentellen Arbeit begann auch für ihn die Epoche
des Ausgeschlossenseins, des Nachdenkens über das Schicksal des Menschen,
über die Wissenschaften. Er flüchtete aus dem Laboratorium zum Schreib-
tisch und schrieb seine Gedanken über den beschwerlichen Gang des wis-
senschaftlichen Fortschritts und der wissenschaftlichen Erkenntnis nieder.“
Übersetzung Károly Kókai.
Unabhängigkeit wahren – was, wie wir sehen, produktiv war und zu einem
singulären Ergebnis, nämlich seinen Grundlagen, führte.
Kritischer Realismus
Fischers philosophische Position zu bestimmen ist nicht schwer, gibt es
doch seine Grundlagen. Fischers Position in der Philosophie zu bestim-
men, ist desto schwerer. Er bemühte sich ja, ein objektives Bild der Lage
der Wissenschaftsphilosophie zu liefern und dabei so weit wie möglich
neutral“ zu bleiben. In diesem Fall ist es aber natürlich auch trotzdem
”
klar, dass er eine Position haben musste.
Er schreibt im Vorwort zur zweiten Auflage im Juli 1967, die Grund-
”
haltung des Buches [hat sich] nicht geändert, sie beruht nach wie vor auf
dem kritischen Realismus“.40 In einem Brief aus dem nächsten Jahr heißt
es:
Gespräche über meine Philos. Grundlagen mit kompetenten Persönlichkeiten (es

gibt nur wenige hierzulande) legten mir nahe, zur Ergänzung eine Abhandlung
über das Thema Warum kritischer Realismus? zu schreiben, da gerade diese
Frage im Buch zu kurz kommt. Wann werde ich aber Zeit dafür haben?41
Was heißt nun kritischer Realismus“?42 Dem erkenntnistheoretischen

”
kritischen Realismus“ entsprechend wird die Existenz einer realen Welt
”
angenommen, die zwar sinnlich erkennbar sei, allerdings (im Gegensatz
zum sogenannten naiven Realismus“) nicht unmittelbar, sondern in ei-
”
nem intellektuell kritischen Prozess des ständigen Erkenntnisfortschritts.
Diese Position haben u.a. Nicolai Hartmann und Alois Riehl vertreten
und sie wurde dann u.a. von Moritz Schlick in seiner Allgemeinen Erkennt-
nislehre mit Bezug auf Riehl auch wiederaufgenommen. In der letzten Zeit
hat Michael Heidelberger im Zuge einer Diskussion über den so genann-
40 Fischer 1967, S. III.

41 Brief Anton Fischer an Robert John, 13. Juni 1968.
42 Der Begriff hat einige Bedeutungen. So wurde er in katholischen Kreisen im
Nachkriegswien etwa von Alois Dempf verwendet. Georg Lukács bezeichnet
seine eigene literaturtheoretische Position ebenfalls so. Vgl. Georg Lukács,
Probleme des Realismus, Neuwied: Luchterhand 1971. Vgl. auch Durant
Drake et al. (Eds.), Essays in Critical Realism, New York: Macmillan 1920.
42 Károly Kókai
ten psychophysischen Parallelismus“ auf die Bedeutung dieser Position

”
hingewiesen.43
Die wichtigste Weiterführung des Psychophysischen Parallelismus, die uns hier

interessieren muss, war (. . . ) die des sogenannten kritischen Realismus“. Der
”
Hauptvertreter dieser Auffassung zur Zeit Fechners und Machs war der aus
Österreich stammende Philosoph Alois Riehl (. . . ). Riehl (. . . ) nahm an, dass
die den physischen und psychischen Aspekten unserer Wahrnehmung unterlie-
gende Realität mit dem Ding an sich im Sinne Kants zu identifizieren sei. (. . . )
[Er] interpretierte (. . . ) das Ding an sich als vom menschlichen Bewusstsein
unabhängige objektive und kausal wirksame Realität und verteidigte, im Ge-
gensatz zu Kant, seine teilweise Erkennbarkeit.44
Bzw.:
Moritz Schlick und Rudolf Carnap [standen] in ihrer Behandlung des Leib-
Seele-Problems ganz in der Tradition des Psychophysischen Parallelismus (. . . ).
In seiner Allgemeinen Erkenntnislehre von 1925 bezeichnete Schlick sich selbst
ausdrücklich als Anhänger dieser Doktrin (Schlick 1979, 336). Er betonte (. . . ),
sein Parallelismus [sei] nicht metaphysischer, sondern rein erkenntnistheoreti-
scher Natur (Schlick 1979, 336).45
Die von mir zu Fischers Darstellung der Position des Wiener Kreises, zur
Kritik des Positivismus und zur Wiedergabe seiner eigenen Position zitier-
ten Stellen zeigen, dass Fischer das Konzept des Dinges an sich genau in
diesem Sinne, also im Sinne des kritischen Realismus von Moritz Schlick,
auslegte. Ebenfalls Beachtung verdient, dass diese Ausführungen Heidel-
bergers in einem Text stehen, in dem er die Wurzeln von Herbert Feigls
Ende der 1950er-Jahre vertretenen Ansichten46 nachweist:
43 Michael Heidelberger, Wie das Leib-Seele Problem in den Logischen Empi-

”
rismus kam“, in: Michael Pauen und Achim Stephan (Hrsg.), Phänomenales
Bewusstsein – Rückkehr zur Identitätstheorie?, Paderborn: Mentis 2002, S.
40-72.
44 Heidelberger 2002, S. 51f.
45 Heidelberger 2002, S. 60f.
46 Herbert Feigl “The Mental and the Physical”, in Herbert Feigl et al. (Eds.),
Minnesota Studies in the Philosophy of Science Vol 2. Concepts, theories
and the Mind-Body Problem, 1958, S. 370-497 bzw. Herbert Feigl, The Men-
tal and the Physical with a Postscript after Ten Years, Minneapolis: Uni-
versity of Minnesota Press 1967.
Herbert Feigl war lange Zeit ein Anhänger von Schlicks kritischem Realismus
(. . . ). Spätestens in seinem Essay von 1958 kehrt Feigl (. . . ) zu den Ansichten
Schlicks und seinen eigenen von vor 1934 zurück. (Im Vorwort zu seinem Essay
bemerkt er übrigens, er sei mit dem philosophischen Monismus“ erstmals durch
”
Lektüre Alois Riehls bekannt geworden, dessen Position er im Wesentlichen bei
”
Moritz Schlick wiederfand“. (Feigl 1967, V; vgl. auch 79f., Fn.).47
Fischer bekannte sich also zu einer Position, die von Schlick 1918/25 ver-
treten wurde. Ob er Feigls Arbeiten vom Ende der 1950er-Jahre kannte
und somit auf ihn reagierte, geht aus den zur Verfügung stehenden Doku-
menten nicht hervor: In den Grundlagen kommt Feigls Essay 195848 nicht
vor, die in Fischers Brief von 1968 erwähnte Darstellung seines kritischen
Realismus scheint nicht ausgearbeitet worden zu sein und Feigl kannte
Fischer mit einer an Sicherheit grenzenden Wahrscheinlichkeit nicht. So
muss ich diesen Aufsatz mit einem negativen Bild schließen: mit dem Man-
gel an einer echten Auseinandersetzung, mit dem Bild von zwei Philoso-
phen, die zeitlich parallel an zwei weit entfernten Orten des Globus – Min-
neapolis und Budapest –, aus den selben Wurzeln entspringend, nämlich
aus Moritz Schlicks kritischem Realismus 1925, und zugleich voneinander
überhaupt nichts wissend an vergleichbaren Ideen arbeiteten. Was für die
Rezeption des Wiener Kreises in Ungarn ein realistisches Ergebnis sein
dürfte.
47 Heidelberger 2002, S. 66–68.

48 Auf die sich Heidelberger bezieht; Fischer gibt ebenfalls keinen Hinweis
auf Herbert Feigl, Existential Hypotheses“, in: Philosophy of Science 17
”
(1950), S. 166-169, in der es um den Kritischen Realismus geht.
Gábor Á. Zemplén1
Early 20th Century Conventionalism in Hungary.

A Case for Győző Zemplén’s Reappreciation
Introduction
At the end of the 19th century an unprecedented number of scientists
– mostly physicists but also many mathematicians – wrote essays, pam-
phlets and books on the scientific method, on the foundations of modern
science, on the status of hypotheses, and theories. This paper presents
the views of one such writer, Győző Zemplén (1879–1916). The rather
well-known physicist of the Austro-Hungarian Monarchy has almost only
been dealt with from the point of view of physics, and his essays that are
significant from a philosophical perspective are still unknown. This short
survey investigates his originality in the Hungarian context – and his con-
formity to a stronger European-based conventionalism that in the early
years of the 20th century only began to emerge in Hungary. The paper
also looks at some of the factors that explain why Zemplén’s work fell into
oblivion for a century in the Hungarian intellectual scene and contrasts it
to the flourishing philosophical endeavours of the Vienna Circle that grew
out of a similar philosophical background.
Conventionalism – the European Movement

Around the end of the 19th century, scientists advocating modern science
– and for nature, too – often stressed the importance of conventions in
the development of science. This loosely termed conventionalist attitu-
de could take many forms, at times supporting very different views on
science. Probably one of the best known sceptical use of conventionalism
1 The work was supported by the OTKA K72598 grant, and the Bólyai post-
doctoral scholarship.
46 Gábor Á. Zemplén
is the famous “ignoramus et ignorabimus” speech by the eminent phy-

siologist Emil Du Bois-Reymond in 1872. In his Über die Grenzen des
Naturerkennens 2 (“On the limits of our understanding of nature”). While
in this speech conventional elements in science are used to support scepti-
cal views, they play an important, albeit optimistic role in Ernst Mach’s
and later Otto Neurath’s writings on science. Both discussed conventions
as a necessary but welcome feature of research, the realization of which
could yield new scientific insights.3 The two aspects, one more “critical”,
often used as an argument against realist interpretations of theories, and
one more “constructive”, stressing the potentials of conventions for the
2 Emil du Bois-Reymond, Reden, 2 vols. (Leipzig: Veit 1912), pp. 441-473. For
a modern reader the 19th century’s emphasis on conventions could easily be
seen as a plea for increased attention to the history of science. Similarly, as
from a contemporary perspective, admitting the significance of conventions
automatically begs the questions of why these conventions were accepted,
whether these choices can be considered rational, etc. This does not appear
to be the case for du Bois-Reymond (unlike for Mach, or even more so
for Neurath a few decades later), yet the famous 1872 talk given on the
14th of August was preceded by much shorter and less known talk with the
title “Über Geschichte der Wissenschaft” (On the History of Science) on
the 4th of July the same year (ibid. pp. 431-440). In this paper, however,
history is mainly discussed as an educational entry-point that offers a more
interesting insight into the world of science than does a dogmatic teaching
method.
3 Examples include Ernst Mach, Die Mechanik in Ihrer Entwickelung:
Historisch-Kritisch Dargestellt (Leipzig: F.A. Brockhaus 1883). As he wri-
tes: ,,Die historische Untersuchung fördert nicht nur das Verständniss des
Vorhandenen, sondern legt auch die Möglichkeit des Neuen nahe, indem
sich das Vorhandene eben theilweise als conventionell und zufällig erweist.
Von einem höhern Standpunkt aus, zu dem man auf verschiedenen Wegen
gelangt ist, kann man mit freierm Blicke ausschauen, und noch neue Wege
erkennen“. p. 238. The early optical lectures by Neurath later discussed
in more detail also offer similar perspectives for history of science. For the
connections see Michael Stöltzner, “The Auxiliary Motive in the Forest and
in Optics”, in: Elisabeth Nemeth and Friedrich Stadler (Eds.), Encyclopedia
and Utopia. Vienna Circle Institute Yearbook 4. Dordrecht: Kluwer 1996,
pp. 113-26. , Id., “Otto Neurath 1913–1915”, in: J. Blackmore, R. Itagaki,
and S. Tanaka (Eds.), Ernst Mach’s Vienna 1895–1930. On Phenomenalism
as Philosophy of Science. Dordrecht: Kluwer 2001, pp. 105-22.
Early 20th Century Conventionalism in Hungary 47
future development of theories, have both been common topoi of the turn
of the century.4
In the philosophical discourse, however, the term “conventionalism”
is used to describe a much narrower set of ideas. As opposed to the rat-
her vague and many-faceted use of the term in the second half of the
nineteenth century, “conventionalism” for historians of philosophy is the
conventionalism attributed first and foremost to Poincaré.5 This, of cour-
se, has many reasons, like the favourable reception of Poincaré’s work
by – among others – the members of the Vienna Circle, or the depth
and detail to which Poincaré developed his ideas. As opposed to the ma-
ny passing remarks about “conventional” aspects of science in the period,
Poincaré attempted to systematically understand the consequences of this
position.6 His conventionalism was developed as primarily based on for-
mal considerations in mathematics, but later applied to physics, and it
was a theoretical attempt to give an account of the growth of scientific
knowledge in a period where abrupt changes in that knowledge emerged
and were to be accounted for.7
It is well-known that the early Vienna Circle has taken up and further
discussed many of Poincaré’s works. In a later recollection Philipp Frank
cites Poincaré as one of the major influences on the Circle, only to be
matched by that of Mach:
4 In his recent book Rheinberger surveys many of the examples, and finds the
sources of a historical epistemology in this fertile conventionalist soil: Hans-
Jörg Rheinberger, Historische Epistemologie, Zur Einführung; 336 (Ham-
burg: Junius 2007).
5 For a definition of the narrower (geometrical) conventionalism of Poincaré,
see Gerhard Heinzmann, “Convention and Observability – Poincaré Once
Again,” in: Between Leibniz, Newton, and Kant: philosophy and science in
the eighteenth century, 2001. p. 138.
6 See his books: Henri Poincaré, Science and Hypothesis, trans. William John
Greenstreet, foreword by, and Joseph Larmor (London, New York: Scott
1905), Ibid., Science and Method, trans. Francis Maitland (London, New
York [etc.]: T. Nelson and sons 1914). Poincaré’s views clearly influenced
many of the members of the later Vienna Circle, like Rudolf Carnap or Otto
Neurath.
7 For a detailed discussion from the origin to the recpetion of Poincaré’s
views see Jerzy Giedymin, Science and Convention: Essays on Henri Poin-
caré’s Philosophy of Science and the Conventionalist Tradition (Oxford;
New York: Pergamon Press 1982).
According to Mach, the general principles of science are abbreviated economical

descriptions of observed facts; according to Poincaré, they are free creations of
the human mind which do not tell anything about observed facts. The attempt
to integrate the two concepts into one coherent system was the origin of what
was later called logical empiricism.8
In a volume discussing the relationship of the Vienna Circle and Hungary

the question naturally arises: to what extent were these two strands of
thought known and discussed in Hungary? As the influence of Mach is
comparatively well studied, I will focus on the reception of conventionalism
in Hungary (Mach’s influence will cursorily be discussed in Section 5).
For this aim in the next sections I will broadly outline the Hungarian
scene, followed by a short overview of Győző Zemplén’s oeuvre, and his
conventionalist views.
The Hungarian scene
Philosophy in Hungary
Hungarian Philosophy contributed little to the development of European
Philosophy in the 19th century, probably only to be matched by the
amount of academic attention it received until the few decades. As Perecz
observed “Historians of an unfortunate philosophical culture . . . have to
be positivists and historians of ideas at the same time. In Hungarian
history of philosophy the basic sources are unstudied, and the formative
canons are undetermined, so a Historian of Philosophy has to explore pri-
mary sources as well as determine canons.”9 The present article also aims
at the following: to recover sources and place them on an intellectual map
that still has large uncharted territories.
Due to the late institutionalisation of philosophy, sources are scarce
in this first period, but after some early attempts to form a philosophical
circle (1876–78, Filozófiai társaskör), the first journal was established (Ma-
gyar Philosophiai Szemle, 1882–91). The early philosophical scene appears
to have been strongly influenced by positivist views and Kant. As in the
8 Philipp Frank, Modern Science and Its Philosophy (Cambridge, MA: Har-
vard University Press 1949). pp. 11-12.
9 László Perecz, A pozitivizmustól a szellemtörténetig. Athenaeum, 1892–
1947, Horror Metaphysicae (Budapest: Osiris 1998). pp. 15-16.
programmatic article for Hungarian Philosophy by Károly Böhm – who

was also director of the famous “Fasori” Lutheran high-school from 1873–
189610 – we read: “Our epoch in Philosophy has two outstanding works,
the authors of which do not know each other, yet these two only offer one
systematic whole when taken together: Kant’s criticism and . . . Comte’s
positivism”.11 In the later years of the short-lived journal the articles be-
come more and more dogmatically positivistic, instead of following the
more recent trends of neo-Kantianism and of Lebensphilosophien.12 The
successor journal, the longer running Athenaeum (1892–1947), the most
important philosophical journal until the Second World War, shows simi-
lar tendencies until the First World War. By the time of the launching
of the journal the first editor, Imre Pauer (1845–1930) was a staunch po-
sitivist, guarding the rather narrow ideological focus of the journal with
little originality. The scientistic, anti-metaphysical articles also exhibit
strong naturalistic tendencies: philosophy needs to become (like) a na-
tural science, and its aim is to summarize and theoretize the findings of
special sciences.13
In the first years of the 20th century, however, other approaches star-
ted to appear in the works of Ákos Pauler (1876–1933), Jenő Posch (1859
–1923), and Menyhért Palágyi (1859–1924).These already show refinement
of and at times turning away from the rather naive dogmatism charac-
teristic of 19th century works. In general, psychological and sociological
considerations play an increasing role, and more contemporary philosophi-
cal insights also surface in the articles.14 A quarter of a century later than
10 Katalin Demeter, “Természettudomány és pozitivizmus a századelő magyar

kultúrájában,” in: Magyar Filozófiai Szemle 31, no. 3, 1987.
11 Károly Böhm, “Kriticismus és Positivismus,” in: Magyar Philosophiai Szem-
le 2, no. 4, 1883. p. 245.
12 Perecz, A Pozitivizmustól a szellemtörténetig. Athenaeum, 1892–1947. pp.
26-28.
13 Ibid. p. 40.
14 For the sociological tradition of the period see the works of Kristóf Nyı́ri,
many of which are available in English and German as well. For a bibliogra-
phy and the appreciation of his oeuvre see Tamás Demeter (Ed.) Essays on
Wittgenstein and Austrian Philosophy: In Honour of J. C. Nyı́ri, Studien
Zur Österreichischen Philosophie, 38. Amsterdam: Rodopi 2004. For key
figures in the psychological tradition, see Csaba Pléh, History and Theories
of the Mind (Budapest: Akadémiai Kiadó 2008). pp. 157-176.
in the centres, Hungarian philosophy has incorporated ideas from Neo-

Kantianism, idealism, and other movements, and the broadly understood
conventionalist ideas started to surface in philosophical writings.
The best example is probably Ákos Pauler, who – starting as a Kanti-
an and becoming for a decade a devoted positivist – reflecting on method,
mathematics, and scientific understanding states in 1912 that “relativity
of our cognizing” is more and more clearly understood.15
Philosophical views of scientists – philosophies of science?
At the turn of the 20th century sciences were more institutionalized in

Hungary than philosophy. The sciences were held in high esteem, but in-
novative research was rare, and the first chairs for such activity were only
established in the first decade of the century. From the early 19th century
the teaching of science was (and still is) seen as part of a greater political
agenda, promoting the development of the nation-state (or linguistic com-
munity) and had throughout positive connotations. Most natural sciences
at the turn of the century were still struggling to establish their own Hun-
garian vocabulary.16 A main aim of the period was to raise the level of
general education, and, in connection with this, to develop elite teacher
training facilities, like the Eötvös Collegium (founded in 1895). As in most
other European countries, science was both hailed for its internationality
(something standing above the nation-state), but also for its nationality,
and opponents of science were portrayed as enemies of development and
prosperity.
15 Ákos Pauler, “Az ismerés viszonylagossága és a matematikai fogalomal-

kotás,” in: Athenaeum, no. 1-2, 1902. p. 17.
16 The late turn to the vernacular was not so uncommon, yet it is surprising
to know that in the second half of the eighteenth century the translation
of the Wolffian Krüger’s multi-volume Naturlehre into Latin by Kováts is
still hailed as a significant step in the modernization of the curriculum. The
original was Johann Gottlob Krüger, Die Ersten Gründe Der Naturlehre
(Halle 1740-49)., for the translation József Kováts, Elementa Philosophiae
Naturalis a D. Jo. Gottlob Krügero (Kolozsvár 1774). The first proper text-
book in vernacular appears in 1777, but by 1897 its vocabulary is mostly
unintelligible. See János Molnár, A Természetiekről, Newton tanitványinak
nyomdoka szerént hat könyv (Pozsony–Kassa: Landerer 1777)., and for the
dictionary needed to understand it after a hundred years “Az első magyar
fizika műszavai Molnár János 1777-es kötetében,” in: Magyar Nyelvőr, 1897.
In Hungary, University science education was still lagging behind the

centres and most talented students had to study abroad if they wanted to
learn new approaches, and to establish contacts with eminent researchers
of the field. This openness of the scientific community was facilitated by
the multi-nationality of the Monarchy, but was partly a necessity of the
underdeveloped University educational system.
The many languages spoken, the many cultural traditions accepted
also kept alive a special type of eclecticism in science that started in the
eighteenth century but survived the nineteenth and existed well into the
twentieth. The eclectic tradition developed as a response to “sectarianism”
in the eighteenth century (in science as in religion). It promoted recon-
ciliatory approaches and typically meant turning away from too strong
adherence to “isms”, as – at least so it was claimed - exclusivity leads to
dogmatism, and is thus hindering the search for truth. Eclecticism was
not seen as a sect, but rather the “judge of sects”, and as empirical as op-
posed to being dogmatic, as eclectics do not accept anything for granted,
but rather only upon investigation.17 Not adhering to clear philosophical
views was seen as freedom, and, as it is not suitable for a philosopher to
exchange the freedom afforded by nature for slavery,18 as a philosophically
superior position.
This eclecticism – common in other, mostly non-Catholic countries on
the periphery of Europe – was often not more than a rhetorical device,
but one result of this attitude was that scientists often rejected taking up
explicit philosophical positions. In the early 20th century this aphilosophi-
cal attitude still characterized popular science writing in the vernacular.
As Sándor Mikola, one of the most well-known popularizers of physics
(and also teacher at the famous “Fasori” Lutheran high school, where
John von Neumann, Eugene Wigner, and many other eminent scientists
studied) wrote already in the 20th century:
The realists are correct, as acquiring knowledge is the process whereby reality
is imprinted in our spirits. . . . The idealists are also correct, as acquiring know-
ledge is enriched by the concept- and hypothesis-creation of our spirit. . . . The
empiricists are also correct, as acquiring knowledge does stem from experience.
17 See the work of the first clear eclectic, András Ádány, Philosophiae Naturalis
Pars Prima Physica Generalis in Usum Disciplorum a R. P. Andreas Adani
E Soc. Jesu (Tyrnaviae 1755).
18 Ibid. p. 20
. . . Positivists are also correct; the theses of physics do express relationships

between sense data. . . . The conventionalists are also correct, scientific con-
cepts become precise most often due to arbitrary agreements. . . . Pragmatists
are correct to the extent that the laws of physics are truly useful. . . . The new
physico-philosophers are correct also, as we can treat the concepts of physics as
signs, symbols.19
By this time it was not a question of Cartesians versus Newtonians or

Wolffians, but of Conventionalists versus Positivists and Physico-philoso-
phers; the general attitude of scientists was to circumvent philosophical
questions and to concentrate on the empirical success of science. The pre-
viously described state of philosophy in Hungary only strengthened this
anti-theoretical attitude. In Hungary it was rare for scientists to promote
philosophical agendas and to subscribe to modern, well-developed phi-
losophical positions. This stood in opposition to international trends of
the period, with many notable scientists, to mention just Mach, Poin-
caré or Duhem, vigorously arguing for their own philosophical views. The
next section introduces a little known Hungarian figure doing just that: a
physicist purporting Conventionalist views in the first decade of the 20th
century.
Győző Zemplén, a Conventionalist in Hungary

Below I give a very cursory overview of the life of Győző Zemplén. The
extant sources are rather sparse, and the secondary literature is redundant
when it comes to discussing the life and achievements of Zemplén.20 Fol-
19 Quote from Sándor Mikola, A fizika gondolatvilága (Budapest: Szerző

kiadása 1933). pp. 382-3. Mikola, who in his youth was still a Machian,
succeeded Győző Zemplén as co-editor of the “Mathematikai és Physikai
Lapok” after Zemplén’s death.
20 For more information on the bibliography and the intellectual climate see
Iván Abonyi, “Zemplén Győző”, in: István Pénzes (Ed.) Műszaki Nagyjaink,
IV. Budapest: Gépipari Tudományos Egyesület 1981, pp. 305-23. László
Balogh, Gyula Grédics, and László Kovács, “Zemplén Győző a Tudós És
Tanár,” in: Fizikai Szemle, 1979. László Kovács, A Múlt Hagyományaival a
Jövő Fizikusaiért. Zemplén Győző és Mikola Sándor élete, munkássága (ma-
nuscript) (Zalaegerszeg: MTESZ-ELFT 1975), Ibid. (Ed.) Zemplén Győző
emlékkönyv, edited by László Kovács. Szombathely: BDTF 2004. Tibor
Nemes, “A Kollégium és az École Normale Supérieure kapcsolatai (1897–
1947)”, in: József Zsigmond Nagy and István Szı́jártó (Eds.), Tanulmányok
lowing this short introduction I will analyse a number of texts unknown

and unclassified until 2004.
Short account of Zemplén’s life and career
Zemplén Győző was born in 1879 in Nagykanizsa, but grew up in Fiu-

me from the age of four. Learning Italian, German, Hungarian, and the
classical languages, he graduated from secondary school in Fiume with
excellent grades. From 1896, he was member of the Eötvös Collegium, the
elite training college of the University of Budapest – modelled after the
École Normale Supérieure. He started his scientific career already as a
freshman, and at the age of nineteen he received the Pasquich Prize for
his paper “On the Viscosity of Gases”, and in 1901 the University awarded
him the Than Prize for his precision measurements. From the beginning
of the century he took up a position as research assistant to Baron Loránd
Eötvös (1848–1919), the most significant Hungarian physicists of the ti-
me. He started publishing in national and international journals,21 and,
based on his work on experimental measurements concerning the visco-
sity of gases, he completed his doctorate in 1901, and obtained from the
University Board his doctoral degree sub auspiciis regis in 1902.22
az Eötvös Kollégium történetéből. Budapest: Eötvös József Kollégium 1989.

Gábor Palló, “Tudományos Intézmények, Konzervativizmus, Kreativitás. A
Budapesti Műszaki Egyetem modernizáló szerepe”, in: Gábor Palló (Ed.) A
honi Kopernikusz-recepciótól a magyar Nobel-dı́jakig, Budapest: Áron Kiadó
2004, pp. 269-90. Gábor Á. Zemplén, “Zemplén Jolán”, Asszonysorsok a 20.
Században. Budapest: BME Szociológia és Kommunikáció Tanszék 2000, pp.
99-107.
21 Of these first and foremost is the Mathematikai és Physikai Lapok, where
he published several papers on algebra and number theory, and where he
was later editorial board member as well. He also wrote 5 articles to the
Annalen der Physik between 1900-1904 on mostly mechanical and kinetic
theory.
22 “Zemplén, Győző was a man of driving ambition, aspiring – in the best sen-
se of the word. At the turn of the century the title ‘royal doctor’, (today’s
title ‘doctor with a golden ring’, sub auspiciis popularis) was established
for students achieving an excellent performance both throughout their uni-
versity studies and at their doctoral examination.” László Balogh, Gyula
Grédics, and László Kovács, “Scientist and Teacher Zemplén, Győző”, in:
László Kovács (Ed.) Zemplén. The Scientist and the Teacher. Szombathely:
Berzsenyi College 2004, pp. 7-20. p. 8.
In 1902 he was appointed assistant lecturer by Loránd Eötvös at the

Institute for Natural Sciences and in 1904 it was Eötvös’ recommendation
that made Zemplén’s peregrinations to Göttingen possible. In Göttingen,
he became interested in shock waves and non-continuous phenomena23 .
After returning to Hungary and getting married, he left again, this time
for Paris, to work on the theory of shock waves.24
After his academic peregrination Zemplén continued to work as Eötvös’
assistant, to be appointed as honorary lecturer at the University of Bu-
dapest in 1905, and of the Budapest University of Technology (József
Műegyetem) in 1907. In 1908 he became a member of the Hungarian Aca-
demy of Sciences, and in 1912 became the head of the newly established
Department of Theoretical Physics at the University of Technology.
Apart from research publications, he wrote dozens of articles popula-
rizing modern physics. He received the Bugát Prize in 1905 for his book
“On the radioactive behaviour of bodies” (1905), and translated Mme
Curie’s “Investigations of radioactive substances” (1906). His most signi-
ficant, heavily illustrated and highly original textbook was on “Electricity
and its practical applications” (1910).
Taking up many societal responsibilities, memberships in several so-
cieties (including the Hungarian Philosophical Society), Zemplén’s life ab-
ruptly changed with the Great War. After joining the Army he fought at
the Serbian and later Italian front. Recovering from severe typhoid fever,
he was killed by enemy fire at the age of 37 in Italy.
23 Academic peregrination was a crucial step in a young researcher’s life in

the period – just as his mentor, Eötvös studied under Kirchoff, Bunsen
and Helmholtz, so did his student Zemplén travel to Germany to learn of
new methods and problems. See more in Jolán M. Zemplén and László
Egyed, Eötvös Loránd, A Múlt Magyar Tudósai (Budapest: Akadémiai
Kiadó 1970). In Göttingen he contributed to Klein’s Encyclopaedia of Ma-
thematical Sciences (Enzyklopädie der mathematischen Wissenschaften),
writing a chapter on the non-continuous motions of liquids. His work on
explosion-like processes and shock waves was considered important achie-
vements in the period.
24 His paper on the impossibility of rarefying shock waves in gases was pre-
sented at the University of Paris: Győző Zemplén, “Sur l’impossibilité des
ondes de choc négatives dans les gaz,” in: Comptes Rendus 141, 1905.
Zemplén’s Conventionalism
Zemplén’s papers on Conventionalism appeared in 1906, a significant year

for the Hungarian cultural history. Many tensions surfaced this year, in-
cluding a public clash between radicals and conservatives in the Society for
Social Science. Groundbreaking artistic works appeared, like Endre Ady’s
volume of poetry Új versek, or the founding of the Circle of Impressionists
and Naturalists. And, as Nyı́ri observes, this is the year when philosophi-
cal views of thinkers in Austria and Hungary start to converge.25
The emblematic journal of the period is the Nyugat (West), run and
funded by a group of liberal intellectuals. Before the Nyugat was launched,
however, more or less the same group of intellectuals created a short-lived
journal, called Szerda (Wednesday). Edited and owned by Antal Gundel
(son of the famous restaurant owner), the journal had various sections,
the literary one edited by Zoltán Ambrus, the theatrical by Ignotus (later
editor in chief of Nyugat) and the scientific one by Győző Zemplén. The
journal’s attitude was critical in all its sections,26 with most editorial
decisions made by Ambrus and Ignotus.
Zemplén published regularly in the short-lived journal – together with
Ady, Margit Kaffka, and Géza Gárdonyi. Among the other scientists
Zoltán Gombocz, a linguist, also formerly a member of the Eötvös Col-
legium, among the philosophers the young György Lukács, with an early
article on drama, also published in the journal.
In his articles for Szerda, Zemplén discussed issues for a wider public
and readership, not specifically schooled in science. The articles show his
appreciation and understanding of modern physics,27 of measurements
techniques28 and of contemporary philosophies of science. It is this latter
aspect that is our concern here, where Zemplén discusses his views on the
status of scientific knowledge.29 In the article On the value of Science (‘A
25 János Kristóf Nyı́ri, A Monarchia szellemi életéről. Filozófiatörténeti ta-

nulmányok (Budapest: Gondolat 1980).
26 Júlia Bendl, Lukács György Élete a Századfordulótól 1918-Ig (Budapest:
Scientia Humana Társulás 1994).
27 Győző Zemplén, “Boltzmann Lajos (1844–1906),” in: Szerda I, no. 2, 1906.
28 Ibid., “A földnehézség változásainak méréséről,” in: Szerda I, no. 1, 1906.
29 Ibid., “A Tudomány Értéke,” in: Szerda I, no. 1, 1906. I thank László Kovács
for drawing my attention to this article, though its bibliographical details
were lost. It also helped me uncover some other, previously unknown pa-
tudomány értéke’) he starts from a fallibilist stance, claiming that though

we often talk of certain scientific or mathematical views as “final”, in fact
“we might have to give up our strongest bastions [of science] in light of
future research.”30 His argument stresses the importance of conventions
in science from the empirical bases to the accepted styles of reasoning.
He discusses in detail perceptual issues, the individual differences when
it comes to observation and refers directly to Poincaré’s works that ap-
peared in the previous years. In the second part of the paper he extends
conventions all the way to logical axioms:
But who decides on the laws of logic, who guarantees their correctness? It might
be not a little disappointing that the answer to even this question is that nobody
does; even the laws of logic are assumptions, hypotheses, or if you like it, axioms.
The gist of the matter remains: these are laws which cannot be proven, rules,
the correctness of which we cannot be ascertained.31
When arguing for his views, Zemplén states that the correctness of
laws of logic cannot be ascertained by either reasoning, as this process
would use the same laws, or by experience, which could only result in
inductive generalizations that have predictive power, but cannot provi-
de conclusive proof: “We therefore have to admit that even logic, which
governs all the sciences and our whole life, is nothing more than a few
assumptions”32 .
Towards the end of his paper, Zemplén clearly sides with Poincaré
when discussing the non-arbitrary nature of conventions. To my knowledge
this is the first Hungarian appearance of full-blown, Poincaré-
influenced Conventionalism by an accepted scientist, and is rather un-
typical for the time. Even the role of the social factors appears in the con-
clusion: “Taking a look at the sciences, we have shown that all of them –
including logic and mathematics – are built on hypothetical foundations,
pers of Zemplén. These were cursorily discussed in: Gábor Á. Zemplén,
“A Hundred Years Make No Small Difference: Popularization of Science in
Hungary at the Turn of Two Centuries”, in: László Kovács (Ed.) Zemplén.
The Scientist and the Teacher. Szombathely: Berzsenyi College 2004, pp.
40-49.
30 Zemplén, “A tudomány értéke,” in. p. 17.
31 Ibid. p. 20.
32 Ibid. p.21.
or to formulate it differently, are the results of freely made conventions

that are significantly influenced by the human environment.”33
Conventionalism and Machianism in Hungary

The articles in Szerda appeared a year after Zemplén’s return from Paris,
where he had personal contact with Poincaré. A few years later Poincaré’s
book on “Science and Hypothesis” was translated.34 The translator was
Béla Szilárd (1884–1926 (Paris)), a chemist-pharmacist. Szilárd went to
Paris in 1907, also had contacts with Poincaré, and worked in the Cu-
rie laboratory for three years. His fascination with Conventionalist ideas
was connected to his French academic peregrination, as was Zemplén’s,
and he seems to have had contacts very similar to that of Zemplén. The
foreword of the book was written by Lajos Ilosvay (1851–1936), a well-
known chemist of the time. Ilosvay was the Head of the General Chemistry
Department and earlier rector of the University of Technology (József
Műegyetem), He studied with Robert W. Bunsen and Adolf von Baey-
er in Germany before, and also had “French connections”, as he was a
student of Marcellin Berthelot during his peregrination in Paris.
In the Hungarian edition of Poincaré’s work, the explanatory notes
were written by Győző Zemplén. He explains some of the physical aspects
of Poincaré’s views and connects these ideas to Einstein’s theory of rela-
tivity. With these marginalia, he became one of the first to reflect on and
popularize Einstein’s views in Hungary.35
One of the most influential books of French Conventionalism was
translated into Hungarian in the year of the forming of the Galilei Circle
(1908, among others influenced by Mach), a further sign of the early 20th
century revival of academic and cultural life. Ilosvay, the author of the
Foreword, apologized for the “abstract philosophical work that will not
33 Ibid. p.23.
34 Henri Poincaré, Tudomány és föltevés, trans. Béla Szilárd (Budapest:
Királyi Magyar Természettudományi Társulat 1908).
35 Ibid. pp. 251-2. See also István Gazda (Ed.) Einstein és a magyarok.
Budapest: Akadémiai Kiadó 2004. pp. 36-51. Zemplén later developed a
theoretical alternative to the solutions of Einstein and Lorentz, a third,
“(light)speed-dilatational” view that received little attention outside of
Hungary.
be a sought-after reading”. Yet the work was known and circulated – but
received little critical appraisal.
In the same years, Mach’s influence also increased among Hungarian
scientists, again, to a large extent due to peregrination. Tódor Kármán
(1881–1963), a few years after Zemplén, also visited Göttingen. As a stu-
dent he worked on a similar problem (shock waves) as Zemplén, by this
time with Klein’s student, Prandtl, and this work clearly touched on topics
also researched by Mach. Even more significant is probably Georg Heve-
sy’s admiration towards Mach’s ideas (including philosophical ones).36
There is, however, very little data about anyone being influenced by
both of these traditions. Machian economy of thought and French Conven-
tionalism seems not to have influenced the same groups and individuals
in Hungary. For Philipp Frank, cited earlier, the joining of these two ap-
proaches created the intellectual cradle of logical empiricism.37 Although
it is probably wise not to take Frank’s account at face value, it is nevert-
heless worth noting that by the first decade of the twentieth century both
of these traditions were present in Hungary. The ingredients existed, yet
nothing like the lively society of the “Verein Ernst Mach”, nor the ambi-
tious project of the “Vienna Circle” could take its roots in the intellectual
soil of Hungary.
36 Hevesy wrote a devotional letter to Mach on 9th of July, 1904. See more in
e.g. the works of John Blackmore, esp. John Blackmore, “Mach hatása hat
nagy magyar tudósra”, in: Peter Weibel (Ed.) A Művészeten Túl / Jenseits
Von Kunst. Budapest: Kortárs Művészeti Múzeum - Ludwig Múzeum Bu-
dapest; Soros Alapı́tvány C3 Kultúrális és Kommunikációs Központ n.d.,
pp. 511-20. See also Péter Szegedi, “Ernst Mach’s Influence on Hungari-
an Physicists – an Outline,” in The reception of Austrian Philosophy in
Hungary around 1900. Conference in Budapest, October 8-9, 1998.
37 See also F. Stadler’s reconstruction of this period in Friedrich Stadler, “Vi-
enna Circle: Context, Profile, and Development”, in: Thomas Ernst Uebel
and Alan W. Richardson (Eds.), The Cambridge Companion to Logical Em-
piricism. Cambridge: Cambridge University Press 2007, pp. 13-40. When
discussing the intellectual sources of the Vienna Circle, one could add Ein-
stein’s theory of relativity, but even here Hungarians (first among these
Gyula Farkas and Győző Zemplén) were quickly responding, and – at least
for Zemplén – clearly grasping the significance of conventionalist insights
when appreciating Einstein’s work.
Closing Remarks on the Asymmetrical Development of Philosophy

of Science in the Monarchy
At the beginning of the 20th century the slowly institutionalizing phi-
losophical community had close ties with natural scientists in Hungary.
Not only Győző Zemplén was a member of the Hungarian Philosophi-
cal Society (founded in 1900), but also was the eminent electrotechni-
cian Károly Zipernovszky (1853–1942), and honorary members included
Loránd Eötvös, Károly Than or Frigyes Korányi.38
Supporters of novel philosophical ideas, however, generally picked the-
se up during their academic peregrination in Germany, or, as we have seen,
in France. When returning to Hungary they often popularized these ideas,
which could therefore find their way to Hungarian readers. Usually, howe-
ver, they were neither challenged, nor forced to explicate, clarify or modify
these views.39 Hungarian culture in the first decade of the 20th century
was thus receptive to novel ideas, yet did not create the fertile ground for
the growth and dissemination of these. This is in sharp contrast to the
Viennese culture of the period, where many of these ideas found local sup-
porters and followers, and from small intellectual circles (like the Verein
Ernst Mach) significant novel schools of thought could develop.
Institutionalization also contributed to this difference. The level of
education was excellent in at least some of the high-schools, including the
already discussed Lutheran high-school (“Fasori gimnázium”), building
on Comte’s positivism and later on the philosophical eclecticism, as was
seen in Böhm’s and Mikola’s work. But even in these schools the reception
of philosophical ideas lagged behind that of the natural sciences. This was
even more true for university education.40 Furthermore very few of the
38 Perecz, A pozitivizmustól a szellemtörténetig. Athenaeum, 1892–1947, p.

100.
39 Some important ideas, however, were not received, see e.g. Gábor Palló,
“Boltzmann Magyarországon: egy meg nem történt eset,” in: Világosság
36, no. 11, 1995.
40 In Budapest, contemporary physics was dominated by the conservative re-
search agenda of Baron Eötvös, but the University where he worked (“Tu-
dományegyetem”), had an even more conservative approach in most areas.
The significant, innovative chairs of period (even in theoretical physics) we-
re established at the University of Technology (“József Műegyetem”), where
Győző Zemplén and his younger brother, Géza became professors. See mo-
philosophers (if one can legitimately use this term) were schooled in con-
temporary philosophies of science. Though some of the eminent scientists,
like Győző Zemplén dabbled in philosophy, their primary interests, in line
with the national(istic) agenda, remained the popularization of science,
and not of philosophy.
Contingent individual life-histories also clearly influenced the deve-
lopment of philosophies of science in Hungary. Some key figures of the
national scene were lost in The Great War, among them the young and
talented philosopher Béla Zalai41 (1882–1915) and the physicist Győző
Zemplén. In an intellectual community made up of relatively few active
members, these losses had great significance for the development of whole
areas. As Géza Laczkó remembered Zemplén in his obituary in the journal
Nyugat (1916/15):
The figure of Zemplén, Gőző, whose death is an irrevocable loss not only to the
Hungarian but to the European scientific community, is inseparable from the
institution where I met him for the very first time, from the Eötvös Collegium,
which has already given so many great minds to this country. . . . The Eötvös
Collegium was opened in 1896 and Zemplén was among the first generation of
its students. . . . Zemplén was a man of driving ambition bubbling over with
vitality, a genuinely talented man with all the superb qualities Jókai made up
for his heroes.42
Apart from all the admiration and praise, the ideas and movements
purported by these individuals were not taken up or followed in many
cases. To make matters worse, the tone used by contemporaries often in-
fluenced the evaluations of later generations as well. Comparing scientists
to heroes fits obituaries and the genus demonstrativum, when the aim
is to eulogize. Much of the Hungarian historiography, however, has used
the grande/sublime style typical of obituaries instead of the critical ap-
re in: Palló, “Tudományos Intézmények, Konzervativizmus, Kreativitás. A

Budapesti Műszaki Egyetem Modernizáló Szerepe”, pp. 269-290. It appears
to me that the institutionalization separated philosophy at the older Uni-
versity from modern approaches to science at the Technical University.
41 György Kampis, “Zalai Béla: egy eredeti magyar gondolkodó a Monar-
chiában”, in: Peter Weibel (Ed.) A Művészeten Túl / Jenseit Von Kunst.
Budapest: Kortárs Művészeti Múzeum – Ludwig Múzeum Budapest; Soros
Alapı́tvány C3 Kulturális és Kommunikációs Központ n.d., pp. 506.
42 Translated in Balogh, Grédics, and Kovács, “Scientist and Teacher Zemplén,
Győző”, pp. p. 14.
praisal of the historical actors as scientists or philosophers. Furthermore,

although the mapping of intellectual circles and traditions has begun, we
still fail to have a rich understanding of the early 20th century intellectu-
al movements in Hungary, and a comparative analysis still requires much
research on primary sources.
The article introduced the reception of Poincaré-influenced conventio-
nalism in Hungary, and discussed in some detail its first supporter, Győző
Zemplén, whose work deserves more attention (and a critical evaluation)
for understanding the Hungarian context than it has received so far. The
type of conventionalism visible in Zemplén’s work is commensurable with
the general attitude characterised by the examples in section 1. Neither in
detail nor in profundity does it match Poincaré’s views, though Zemplén’s
ideas were clearly influenced by them. These views did not contribute to
the development of European philosophies of science, yet they testify to
the early – and neglected – reception of these ideas in Hungary.
Bibliography
Abonyi, Iván, “Zemplén Győző.ı̈n: István Pénzes (Ed.) Műszaki Nagy-
jaink, IV. Budapest: Gépipari Tudományos Egyesület 1981, pp. 305-
23.
Ádány, András, Philosophiae Naturalis Pars Prima Physica Generalis in
Usum Disciplorum a R. P. Andreas Adani E Soc. Jesu. Tyrnaviae
1755.
“Az első magyar fizika műszavai Molnár János 1777-es kötetében” [The
Technical Terms in the First Hungarian Physics Book, Molnár’s 1777
volume], in: Magyar Nyelvőr 1897, pp. 252–55.
Balogh, László, Gyula Grédics, and László Kovács, “Scientist and Teacher
Zemplén, Győző.” in: László Kovács (Ed.) Zemplén. The Scientist and
the Teacher. Szombathely: Berzsenyi College 2004, pp. 7-20.
———, “Zemplén Győző a Tudós és Tanár”, [Scientist and Teacher Zem-
plén, Győző] in: Fizikai Szemle 1979, pp. 321.
Bendl, Júlia, Lukács György élete a századfordulótól 1918-ig. [The Life
of Georg Lukács from the Turn of the Century to 1918] Budapest:
Scientia Humana Társulás 1994.
Blackmore, John, “Mach hatása hat nagy magyar tudósra.” [Mach’s Influ-
ence on Six Hungarian Scientists] in: Peter Weibel (Ed.) A Művészeten
túl / Jenseit Von Kunst. Budapest: Kortárs Művészeti Múzeum - Lud-

wig Múzeum Budapest; Soros Alapı́tvány C3 Kultúrális és Kommu-
nikációs Központ n.d., pp. 511-20.
Böhm, Károly, “Kriticismus és Positivismus”, in: Magyar Philosophiai
Szemle 2, no. 4, 1883, pp. 245-58.
Demeter, Katalin, “Természettudomány és pozitivizmus a századelő ma-
gyar kultúrájában”, [Natural Science and Positivism in the Turn of
the Century Hungarian Culture] in: Magyar Filozófiai Szemle 31, no.
3, 1987, pp. 403-56.
Demeter, Tamás (Ed.) Essays on Wittgenstein and Austrian Philosophy:
In Honour of J. C. Nyı́ri, Studien Zur Österreichischen Philosophie,
38. Amsterdam: Rodopi 2004.
Du Bois-Reymond, Emil, Reden. 2 vols. Leipzig: Veit 1912.
Frank, Philipp, Modern Science and Its Philosophy. Cambridge, MA: Har-
vard University Press 1949.
Gazda, István (Ed.) Einstein és a magyarok. [Einstein and the Hunga-
rians] Budapest: Akadémiai Kiadó 2004.
Giedymin, Jerzy, Science and Convention: Essays on Henri Poincaré’s
Philosophy of Science and the Conventionalist Tradition. Oxford; New
York: Pergamon Press 1982.
Heinzmann, Gerhard, “Convention and Observability – Poincaré Once
Again”, in: Between Leibniz, Newton, and Kant: philosophy and science
in the eighteenth century 2001, pp. xvi, 281 p.
Kampis, György, “Zalai Béla: egy eredeti magyar gondolkodó a Mon-
archiában.” [Béla Zalai, an Original Thinker in the Monarchy] in:
Peter Weibel (Ed.) A Művészeten Túl / Jenseits Von Kunst. Buda-
pest: Kortárs Művészeti Múzeum – Ludwig Múzeum Budapest; Soros
Alapı́tvány C3 Kultúrális és Kommunikációs Központ n.d., pp. 506.
Kovács, László, A múlt hagyományaival a jövő fizikusaiért. Zemplén Győző
és Mikola Sándor élete, munkássága (Kézirat). [On the Lives and
Works of Győző Zemplén and Sándor Mikola]. Zalaegerszeg: MTESZ-
ELFT 1975.
——— (Ed.) Zemplén Győző Emlékkönyv. Edited by László Kovács. Szom-
bathely: BDTF 2004.
Kováts, József, Elementa Philosophiae Naturalis a D. Jo. Gottlob Krügero.
Kolozsvár1774.
Krüger, Johann Gottlob, Die Ersten Gründe Der Naturlehre. Halle 1740-
49.
M. Zemplén, Jolán, and László Egyed, Eötvös Loránd, A múlt magyar
tudósai. Budapest: Akadémiai Kiadó 1970.
Mach, Ernst, Die Mechanik in ihrer Entwickelung: Historisch-Kritisch
Dargestellt. Leipzig: F.A. Brockhaus 1883.
Mikola, Sándor, A fizika gondolatvilága. [The Worldview of Physics] Bu-
dapest: Szerző kiadása 1933.
Molnár, János, A természetiekről, Newton tanitványinak nyomdoka sze-
rént hat könyv. [Six Books On the Nature of Things, Following New-
ton’s Disciples] Pozsony–Kassa: Landerer 1777.
Nemes, Tibor, “A Kollégium és az École Normale Supérieure kapcsolatai
(1897–1947).” [The Connections of the Eötvös College and the École
Normale Supérieure] in: József Zsigmond Nagy and István Szı́jártó
(Eds.), Tanulmányok az Eötvös Kollégium történetéből. Budapest: Eötvös
József Kollégium 1989, pp.
Nyı́ri, János Kristóf, A Monarchia szellemi életéről. Filozófiatörténeti ta-
nulmányok. [On the Intellectual Life of the Monarchy: Studies in the
History of Philosophy] Budapest: Gondolat 1980.
Palló, Gábor, “Boltzmann Magyarországon: Egy meg nem történt eset”,
[Boltzmann in Hungary: a Case that has Never Been] in: Világosság
36, no. 11, 1995, pp. 68-80.
———, “Tudományos intézmények, konzervativizmus, kreativitás. A Bu-
dapesti Műszaki Egyetem modernizáló szerepe.” [Scientific Instituti-
ons, Conservativism, Creativity; The Modernising Role of the Tech-
nical University in Budapest] in: Gábor Palló (Ed.) A honi Koperni-
kusz-Recepciótól a magyar Nobel-dı́jakig. Budapest: Áron Kiadó 2004,
pp. 269-90.
Pauler, Ákos, “Az ismerés viszonylagossága és a matematikai fogalom-
alkotás”, [The Relativity of Our Cognizing and Mathematical Concept
Formation] in: Athenaeum no. 1-2, 1902, pp. 17-32, 129-42.
Perecz, László, A pozitivizmustól a szellemtörténetig. Athenaeum, 1892-
1947, [From Positivism to “Geisteswissenschaften” Athenaeum, 1892-
1947 ] Horror Metaphysicae. Budapest: Osiris 1998.
Pléh, Csaba, History and Theories of the Mind. Budapest: Akadémiai
Kiadó 2008.
Poincaré, Henri, Science and Hypothesis. Translated by William John

Greenstreet, foreword by Joseph Larmor. London, New York: Scott
1905.
———, Science and Method. Translated by Francis Maitland. London,
New York: T. Nelson and sons 1914.
———, Tudomány és föltevés. Translated by Béla Szilárd. Budapest:
Királyi Magyar Természettudományi Társulat 1908.
Rheinberger, Hans-Jörg, Historische Epistemologie, zur Einführung; 336.
Hamburg: Junius 2007.
Stadler, Friedrich, “Vienna Circle: Context, Profile, and Development”, in:
Thomas Ernst Uebel and Alan W. Richardson (Eds.), The Cambridge
Companion to Logical Empiricism. Cambridge: Cambridge University
Press 2007, pp. 13-40.
Stöltzner, Michael, “The Auxiliary Motive in the Forest and in Optics”,
in: Elisabeth Nemeth and Friedrich Stadler (Eds.), Encyclopedia and
Utopia-Vienna Circle Yearbook 4 . Dordrecht: Kluwer 1996, pp. 113-
26.
———, “Otto Neurath 1913–1915.” in: J. Blackmore, R. Itagaki and S.
Tanaka (Eds.), Ernst Mach’s Vienna 1895–1930. On Phenomenalism
as Philosophy of Science. Dordrecht: Kluwer 2001, pp. 105-22.
Szegedi, Péter. “Ernst Mach’s Influence on Hungarian Physicists – an
Outline.” In The reception of Austrian Philosophy in Hungary around
1900. Conference in Budapest, October 8-9, 1998.
Zemplén, Gábor Á., “A Hundred Years Make No Small Difference: Popu-
larization of Science in Hungary at the Turn of Two Centuries.” in:
László Kovács (Ed.) Zemplén. The Scientist and the Teacher. Szom-
bathely: Berzsenyi College 2004, pp. 40-49.
———, “Zemplén Jolán.” Asszonysorsok a 20. Században. Budapest: BME
Szociológia és Kommunikáció Tanszék 2000, pp. 99-107.
Zemplén, Győző, “Boltzmann Lajos (1844–1906)”, in: Szerda I, no. 2,
1906, pp. 100-03.
———, “A földnehézség változásainak méréséről”, [On measuring Chan-
ges in Gravity] in: Szerda I, no. 1, 1906, pp. 53-56.
———, “Sur l’ impossibilité des ondes de choc négatives dans les gaz”,
in: Comptes Rendus 141, 1905., pp. 710-12.
———, “A tudomány értéke”, [On the value of Science] in: Szerda I, no.
1, 1906, pp. 17-23.
Wolfgang L. Reiter
Wer war Béla Juhos?

Eine biographische Annäherung
I. Die ersten Jahrzehnte
Die Wirkungsgeschichte des Wiener Kreises“ an der Universität Wien
”
in der Zeit nach 1945 ist eng mit den Namen Victor Kraft und Béla
Juhos verbunden. Während die Person und das Wirken von Kraft in
der historiographischen Auseinandersetzung mit der Philosophie des lo-
gischen Empirismus vor allem durch die Arbeiten von F. Stadler1 gut
dokumentiert sind, blieb Juhos bisher im Schatten der Diskussion. Die
vorliegende Arbeit unternimmt den Versuch, erstmals zusammenfassend
die Biographie von Béla Juhos zu beleuchten. Dieser Versuch muss vor-
erst skizzenhaft bleiben, da für die Dokumentation seines Lebens nicht
auf einen Nachlass zurückgegriffen werden kann und die zugänglichen
Archivalien nur ein bruchstückhaftes Bild zu zeichnen erlauben. Immer-
hin ermöglichen diese Quellen die Korrektur von Unschärfen und Un-
genauigkeiten der bisher vorliegenden verstreuten Ausführungen sowie
Präzisierungen und Ergänzungen. Die vorliegende Arbeit setzt sich nicht
zum Ziel, eine Würdigung des philosophischen Werkes von Béla Juhos zu
leisten.
Béla von Juhos, aus kleinadeliger ungarischer Familie stammend, die
bis 1918 in Südungarn begütert war, wurde am 22. November 1901 in Wien
geboren (rel. Bekenntnis ev. A. C.2 ). Seine Kindheit verbrachte Juhos in
Budapest, wo eine Zweigniederlassung der in Wien II., Nordbahnstrasse
42 ansässigen Eisengrosshandlung seines Vaters bestand; dort besuchte er
auch die ungarische Volksschule. Béla sprach als Kind ungarisch und er-
lernte die deutsche Sprache erst nach der Übersiedlung nach Wien im Jah-
re 1909, wo er 1912 in das Staatsrealgymnasiums Wien XXI eintrat und
1 Friedrich Stadler, Studien zum Wiener Kreis. Ursprung, Entwicklung und

Wirkung des Logischen Empirismus im Kontext. Frankfurt am Main: Suhr-
kamp 1997.
2 Archiv der Universität Wien, Personalakt Béla Juhos.
66 Wolfgang L. Reiter
1920 mit Auszeichnung maturierte. Daran schloss sich das Studium der
Philosophie bei Moritz Schlick (1882–1936) und Robert Reininger (1869–
1955) an der Universität Wien an; nebenher belegte Juhos die Fächer
Physik und Kunstgeschichte.
Das Studium schloss Juhos 1926 mit der Dissertation zum Thema In-
”
wieweit ist Schopenhauer der Kantschen Ethik gerecht geworden?“ ab und
wurde 1927 zum Dr. phil. promoviert. Zuvor musste Juhos 1925 noch die
Ergänzungsprüfung aus Griechisch vor der Prüfungskommission für Real-
schulabsolventen ablegen.3 Die Beurteilung der Dissertation mit Datum
vom 22. März 1926 verfasste Reininger, Schlick stimmte der Approbation
der Arbeit am 9. Mai 1926 zu. Da die Beurteilung von Reininger verfasst
wurde, ist zu vermuten, dass Reininger Betreuer der Dissertation war und
Schlick lediglich als Zweitgutachter auftrat; dies legt auch das Thema der
Dissertation nahe. In einer kurzen biographischen Skizze aus dem Jahre
1966 verwendet Juhos die folgende Formulierung: Er [Juhos] promovierte
”
1926 an der Universität Wien unter Moritz Schlick und Robert Reininger
zum Doktor phil . . . “4
Reiningers Beurteilung ist durchgehend kritisch:
Der V. [Verfasser] bearbeitet sein Thema in der Weise, dass er die Äußerungen
Schopenhauers und Kants Schritt für Schritt folgt & seine Zustimmung oder
Ablehnung durch immanente Kritik zu würdigen unternimmt. Eine allgemei-
ne Bemerkung über die grundverschiedene Einstellung der beiden Denker zum
ethischen Problem wäre da vorauszuschicken gewesen. Der V. kommt zu dem
Ergebnis, dass zwischen beiden Ethiken trotz vieler Verschiedenheiten im einzel-
nen ein Verhältnis innerer Verwandtschaft besteht, welches den gleichgerichteten
metaphysischen Anschauungen beider Philosophen entspringt. Diese Ansicht ist
überhaupt nur dann erörterungsfähig, wenn man, wie es der V. tut, vor allem die
transzendentalen“ Zuspitzungen der beiden Ethiken in Betracht zieht, welche
”
eigentlich nicht das Wichtigste daran sind. [. . . ] Auch mit seiner sonstigen Auf-
fassung der Kantischen Ethik kann ich mich nicht einverstanden erklären, wie er
sich denn überhaupt durch Schopenhauers voreingenommenen Standpunkt dar-
in allzu sehr beeinflussen lässt. Immerhin zeigt der V., dass er auch selbstständig
zu denken vermag & bringt auch im einzelnen manches Richtig bei. Die Arbeit
3 Im Curriculum Vitae vom 5. März 1926 bezeichnet er sich als Adalbert von
”
Juhos, [. . . ] evang. A. C., zuständig nach Ungarn“. Archiv der Universität
Wien, Personalakt Béla Juhos.
4 Institut Wiener Kreis, Brief Juhos an Mulder vom 4. November 1966, Kopie
im Besitz des Verf.
Wer war Béla Juhos? 67
mag daher, ihrer problematischen Ergebnisse ungeachtet, als genügend befun-

den werden, um den Kandidaten zu den Rigorosen zuzulassen.“5
Angesichts der kritischen Beurteilung der Dissertation durch den Neo-

kantianer Reininger liegt es, nahe Moritz Schlick als philosophischen Men-
tor der weiteren intellektuellen Entwicklung des jungen Juhos anzuspre-
chen. Dies scheint auch insofern plausibel, als Juhos auf Empfehlung von
Schlick ab 1925 – und bis zu dessen Auflösung 1934 – an den Zusam-
menkünften des Wiener Kreises“ teilnimmt.
”
Ab den späten Zwanzigerjahren tritt Juhos als Vortragender in der
Philosophischen Gesellschaft der Universität Wien“, der Ortsgruppe
”
Wien der Kant Gesellschaft, hervor. Ein erstes Buch Über die Grund-
lagen der Gewissheit des reinen Denkens erscheint 1928 im Verlag Gerold
in Wien. 1940 folgt im Pan Verlag, Leipzig sein zweites Buch Erkennt-
nisformen in Natur- und Geisteswissenschaften, eine relativ kurze Arbeit
von 57 Seiten. Zwischen 1931 und 1940 erscheinen ingesamt 16 Arbeiten,
davon 5 in englischsprachigen Journalen (Analysis, Mind, The Journal of
Unified Science), die Problemen des logischen Empirismus gewidmet sind
und in einem Fall die Philosophie Schopenhauers im Lichte der Erkennt-
niskritik reflektiert, quasi im Nachklang zum Thema seiner Dissertation.
Eine Arbeit mit dem Titel “Principles of Logical Empiricism”, erschienen
in Mind 1937, befasst sich mit der Präzisierung der Form des logischen
Empirismus Schlick’scher Prägung und dem Physikalismus von Neurath
und Carnap im Anschluss an die Diskussionen am International Congress
for Unity of Science in Paris 1935. Mit Arbeiten aus den Jahren 1934
und 1936 beteiligt sich Juhos an der Protokollsatzdebatte des Wiener
”
Kreises“, wobei der mit Schlicks Konstatierungen gegen die konventiona-
listische Auffassung argumentiert.6 Zwischen 1940 und 1945 erscheinen
keine Arbeiten.
Soweit wir derzeit wissen, leistete Juhos in den Jahren 1942 bis 1944
Militärdienst, offen bleibt dabei, ob im deutschen oder ungarischen Heer.
Sollte Juhos (als ungarischer Staatsbürger) im ungarischen Heer gedient
haben, so könnte dies der Ursprung für das in Wien nach 1945 kursierende

6 Béla Juhos, Kritische Bemerkungen zur Wissenschaftstheorie des Physi-
”
kalismus“, in Erkenntnis 4, 397-418. Ders., Negationsformen empirischer
”
Sätze“, in Erkenntnis 6, 41-55.
hartnäckige Gerücht sein, er wäre Mitglied der ,,Pfeilkreuzler“ gewesen.

Juhos selbst wies diese Unterstellung heftig zurück.7 Bis Ende des Zweiten
Weltkriegs besaß Juhos die ungarische Staatsbürgerschaft, die Annahme
der österreichischen Staatsbürgerschaft erfolgte wahrscheinlich 1945.8
II. Wien nach 1945

Victor Kraft (1880–1975)9 und Béla Juhos sind die einzigen Mitglieder des
Wiener Kreises“, die während der Zeit des Nationalsozialismus in Wien
”
7 Interview mit Hubert Schleichert und Wolfgang Katzenberger 2008, auf-

genommen von Károly Kókai und Christoph Limbeck-Lilienau, Institut
Wiener Kreis. Schleichert erwähnt in diesem Zusammenhang den Namen
Zehetner. Katzenberger bemerkt, dass Walter Hollitscher dieses Gerücht
pflegte, was allerding angesichts der – wenn auch nur sporadischen – Ko-
operation Hollitschers mit Juhos am Institut für Wissenschaft und Kunst
unverständlich ist.
8 Victor Kraft, Nachruf auf Béla Juhos“, in Zeitschrift für allgemeine Wis-
”
senschaftstheorie Bd. 2 (1971) 163-173, auf S. 164. Archiv der Universität
Wien, Personalakt Béla Juhos. Die Angabe zu Staatsbürgerschaft ist dem
Habilitationsansuchen vom 12. April 1947 entnommen.
9 Victor Kraft, geboren in Wien, 4. Juli 1880, 1903 Dr. phil., Dissertati-
on Erkenntnis der Aussenwelt“ bei Friedrich Jodl, Habilitation bei Adolf
”
Stöhr in theoretischer Philosophie 1913, 1925 tit. a. o. Univ. Prof., 1938
Entlassung aus dem Staatsdienst und Entzug der venia, 1945 rehalibitiert,
1947 a. o. Univ. Prof., 1950 o. Univ. Prof., Prof. emeritus 1952. 1954 Er-
nennung zum Mitglied der Österreichischen Akademie der Wissenschaften.
Kraft war Mitglied des Wiener Kreises“ und später – neben Friedrich Kainz
”
– der Betreuer der Dissertation von Paul Feyerabend ( Zur Theorie der Ba-
”
sissätze“, 1951). Um Kraft bildete sich in Anlehnung an den Wiener Kreis“
”
ein studentischer Zirkel, der sogenannte Kraft-Kreis“, dem u. a. Béla Ju-
”
hos, Paul Feyerabend, Walter Hollitscher, Georg Henrik von Wright, Eli-
zabeth Anscombe und Ludwig Wittgenstein verbunden waren. Der Kreis
traf sich 1949 bis 1952/53 zweimal im Monat in den Räumlichkeiten des
Österreichischen Colleges in der Kolingasse im 9. Wiener Gemeindbezirk.
Vgl. dazu: Paul Feyerabend, Zeitverschwendung. Frankfurt am Main: Suhr-
kamp Verlag 1995, S. 107-108. Ders., Science in a Free Society. London: New
Left Books 1978, S. 109. Kraft stirbt 1975 in Wien. Eine Monographie über
Kraft und dessen moralphilosophischen Arbeiten liegt als gedruckte Disser-
tation nunmehr vor: Oliver Vollbrecht, Victor Kraft: Rationale Normenbe-
gründung und Logischer Empirismus. Eine philosophische Studie. München:
Herbert Utz Verlag 2004.
verblieben und nach 1945 auch weiter hier wirkten.10 Die Situation an
der Universität Wien nach 1945 war von einer ausgeprägten Feindseligkeit
gegenüber jeglicher Nähe zur Philosophie des Wiener Kreises“ geprägt.
”
Drastische Worte für das intellektuelle Klima eines Kulturkampfes an den
österreichischen Universitäten nach 1945 findet der Kraft-Schüler Ernst
Topitsch (1919–2003):
Nun hatte ich unter den nationalsozialistischen Terror von eher Wiederherstel-
lung der Geistesfreiheit im Zeichen eines christlichen Humanismus geträumt,
doch was dann wirklich kam, war eine erstickende provinzielle Restauration,
und ein kläglicher Klerikalismus verbreitete in den Hallen der Alma Mater eine
fast mit Händen greifbare Atmosphäre intellektueller Unredlichkeit, ohne auf
entschiedenen Widerstand zu stoßen.11
Es ist wohl vor allem der Position von Kraft als Universitätsprofessor
und dessen Förderung eines philosophischen Mitstreiters zuzuschreiben,
dass Juhos überhaupt auf akademischem Boden Fuß fassen konnte. Zur
Situation der Philosophie an der Universität Wien nach 1945 schreibt
Juhos:
Gleich nach dem Kriege kam es zu einem Wiederaufleben einer von theologisch,
klerikaler Seite gelenkten reaktionären Kulturpolitik, des traditionellen klerika-
len Geistes, der in der logischfundierten wissenschaftlichen Philosophie des ehe-
maligen Wiener Kreises für sich eine Gefahr erblickte. Sämtliche Lehrkanzeln
für Philosophie an den österreichischen Universitäten wurden mit spekulativen
Philosophen besetzt (mit Ausnahme von Viktor Kraft, dem seine frühere Stelle
an der Universität Wien zurückgegeben wurde).12
Die von Juhos hier angesprochenen spekulativen Philosophen“ waren die

”
Ordinarien Leo Gabriel (1902–1987) und Erich Heintel (1912–2000), zu
welchen sich unmittelbar nach 1945 Alois Dempf (1891–1982) gesellte,
der 1948 einen Ruf nach München erhielt, sowie der Privatdozent Ulrich
Schöndorfer. Der Sprachphilosoph und -psychologe Friedrich Kainz (1897–
1977) wird in Rudolf Hallers Beurteilung der Situation der Philosophie in
10 Als weiteres Mitglied des Wiener Kreises, der während des Zweiten Welt-
kriegs in Wien lebte, sei Heinrich Neider genannt, der jedoch keine univer-
sitäre Position bekleidete.
11 Ernst Topitsch, Naturrecht im Wandel des Jahrhunderts, S. 2, in Aufklärung
und Kritik 1/1994.
12 Bela Juhos, Formen des Positivismus“, in Zeitschrift für allgemeine Wis-
”
senschaftstheorie Bd. 2 (1971) 27-62.
Österreich nach 1945 – zusammen mit Theodor Radokovic (Graz) und

Theodor Erismann (Innsbruck) – nicht dem Lager einer weltanschaulich
gebundenen, klerikal-konservativen Haltung zugeordnet, deren Dominanz
in den beiden ersten Jahrzehnten nach 1945 mit gezielter Unterstützung
seitens des Unterrichtsministeriums jegliche davon abweichende philoso-
phische Strömung zu unterdrücken verstand.13
An dieser Stelle sei ein Einschub gestattet, der auf einen bisher nicht
bekannten Versuch Bezug nimmt, die nach Schlick verwaiste Lehrkanzel
für Philosophie der Naturwissenschaften wieder aufleben zu lassen. Kraft
wendet sich in einem Brief vom 13. September 1946 an den Dekan für
das Studienjahr 1946/47, Hans Thirring (1888–1976).14 Kraft berichtet
in seinem Brief von einer Mitteilung Berta Karliks (1904–1990), Erwin
Schrödinger (1887–1961) wäre bereit nach Wien zu kommen, jedoch nicht
an eines der physikalischen Institute, vielmehr möchte [er] bloß dozieren,
”
u. zw. – Naturphilosophie¡‘ Hierin sieht nun Kraft die Chance,
die Lehrkanzel für Philosophie der Naturwisenschaften nach Schlick wieder zu

besetzen. Die Fakultät hat zwar dafür schon einen Vorschlag gemacht, aber
einen schlechten, nämlich [Bernhard] Bavinck und Wenzel (in München). Beide
sind Metaphysiker und es ist Ihnen [sic!] besonders um die Versöhnung von
Naturwissenschaften und Religion zu tun. [. . . ] Vorläufig ist überhaupt noch
kein Ordinariat dafür vorgesehen, denn für Philosophie sind, wie mir Meister
sagte, nur 2 Ordinariate im Stellenplan, eines hat Dempf, das andere ist für den
Psychologen (Rohracher) bestimmt. Es käme also darauf an, 1) das Ordinariat
für Naturphilosophie wieder durchzusetzen und 2) den Vorschlag der Fakultät
zu revidieren. Ich darf Ihnen auch nicht verhehlen, dass schon seit längerer Zeit
Bestrebungen im Gange sind, dass die Philosophie des Wiener Kreises“, die sich
”
unter Schlick internationalen Ruf erworben hat, hier wieder offiziell vertreten
wird. Denn unter Dempf kann darüber niemand dissertieren und Prüfungen
machen.
Mit seiner letzten Bemerkung zur Situation der Philosophie des Wiener
”
Kreises“ spielt Kraft wohl auf jene Bestrebungen an, die 1947 zu sei-
13 Rudolf Haller, Die philosophische Entwicklung im Österreich der fünfziger

”
Jahre“. In: ders., Fragen zu Wittgenstein und Aufsätze zur Österreichischen
Philosophie. Amsterdam: Rodopi 1986, S. 238-239. Vgl. dazu auch: Gerhard
Benetka, Der ,Fall‘ Stegmüller“, in Friedrich Stadler (Hrsg.): Elemente
”
moderner Wissenschaftstheorie. Wien: Springer 2000, S. 123-176.
14 Österreichische Zentralbibliothek für Physik, Nachlass Hans Thirring, B35-
1622.
ner Berufung als a.o. Professor für Philosophie an der Universität Wien
führten, nachdem er in seiner Position als Bibliothekar an der Univer-
sitätsbibliothek Wien 1945 rehabilitiert und 1947 zum Generalstaatsbi-
bliothekar ernannt worden war. Durchaus überraschend ist die von Kar-
lik mitgeteilte Bereitschaft Schrödingers, nach Österreich zurückzukehren.
(Leider fehlt dazu derzeit der dokumentarische Beleg, in dem diese Absicht
Schrödingers zum Ausdruck kommt.) Wie ernst seine Absicht war, lässt
sich dem Brief Krafts an Thirring nicht entnehmen. Jedenfalls stünde eine
solche belegbare Absicht im Gegensatz zu den bisherigen biographischen
Darstellungen, die davon ausgehen, dass Schrödinger keinesfalls in ein
unter den vier Besatzungsmächten aufgeteiltes Wien zurückkehren woll-
te. Schrödinger kehrte erst 1956, ein Jahr nach Abschluß des Staatsver-
trags von Wien nach Österreich zurück. Neben der Frage einer allfälligen
Rückkehr Schrödingers nach Österreich ist für unseren Zusammenhang
hier vor allem die Bestrebung von Kraft von Interesse die Philosophie
des Wiener Kreises“ wieder zu beleben. Dies führt uns wieder zu Juhos
”
zurück.
Eine erste Betrachtung seiner Person mag nahe legen, Juhos als Pri-
vatgelehrten zu charakterisieren. Dies trifft zumindest auf seine sozia-
le Verankerung an der Universität zu. Denn seine spätere universitäre
Tätigkeit als Privatdozent und tit. a.o. Professor an der Universität Wi-
en kann nicht als Erwerbsberufs“ charakterisiert werden, da damit kei-
”
ne regelmäßigen Einkünfte verbunden waren. Aufgrund seiner finanziellen
Unabhängigkeit konnte Juhos seinen intellektuellen und akademischen In-
teressen ohne die Beschränkungen einer Berufsausübung im engeren Sinne
nachgehen. Juhos’ finanzielle Unabhängigkeit gewährleisteten Einkünfte
aus dem 1/4-Mitbesitz an der von seinem Vater begründeten Eisengroß-
handlung Jul. Juhos & Sohn, die später von einem seiner drei Brüder
geführt wurde und die unter anderem auch im Brückenbau tätig war.15
Weiters besaß die Familie eine Großtankstelle im 2. Wiener Gemeinde-
bezirk und eine Tiefgarage. Neben seiner Stadtwohnung (Wien I., Franz-
Josefs-Kai 65) bewohnte Juhos eine Villa in Döbling und war mit zwei
15 Archiv der Republik, Personalakt Béla Juhos, Curriculum Vitae vom 12.
VI. 1947. Juhos führt hier u. a. aus: Ich war weder Pg [Parteigenosse,
”
Anm. d. Verf.], noch Pa [Parteianwärter, Anm. d. Verf.], habe keinem der
der Partei angeschlossenen Verbände angehört und mich auch sonst jeder
politischen Betätigung enthalten.“
Autos und zwei Jagdhunden mobilisiert.16 Juhos war ein passionierter

Jäger mit einem eigenen Revier in den Donauauen, und seine Vorlesung
bestritt er – sich souverän über den akademischen dress code hinwegset-
zend – gelegentlich auch in Jagdkleidung.17 Man liegt wohl nicht ganz
falsch, Juhos’ äußere Erscheinung als jene eines wohlbestallten, von fei-
ner Zurückhaltung geprägten aristokratischen Herrn zu zeichnen, der von
seinem privaten Leben nur sehr spärlich Mitteilung machte, jedoch umso
offener für die wissenschaftliche Diskussion war und dabei auch die soziale
Rollenzuweisung von Professor und Student zu durchbrechen verstand.18
Am 12. Juni 1947 ersucht Juhos um Erteilung der venia legendi auf
”
dem Gebiet der reinen Philosophie für das Sonderfach ,Erkenntnistheo-
rie‘“.19 Als Adresse gibt er Franz-Josefs-Kai 65 an. Wie im Zuge der
Entnazifizierungsmaßnahmen an der Universität vorgeschrieben, legt Ju-
16 Interview mit Hubert Schleichert und Wolfgang Katzenberger 2008, aufge-

nommen von Károly Kókai und Christoph Limbeck-Lilienau, Institut Wie-
ner Kreis.
17 Werner Leinfellner, persönliche Mitteilung an den Verfasser, Januar 2009.
18 Interview mit Wolfgang Katzenberger 2008, aufgenommen von Károly
Kókai und Christoph Limbeck-Lilienau, Institut Wiener Kreis.
19 Im Oktober 1945 unternahm Juhos, der sich noch aus Anlass der Eva-
kuierung Wiens in Ehrwald in Tirol aufhielt, einen Versuch, sich an der
Universität Innsbruck für das Fach Logik und Erkenntnistheorie“ zu ha-
”
bilitieren, der jedoch vom Dekan der philosophischen Fakultät, Theodor
Erismann, am 8. November 1945 mit dem Hinweis, es gäbe bereits aus-
reichend viele Dozenten der Philosophie in Innsbruck, mit dem Ersuchen
um Rücknahme des Habilitationsansuchens und mit dem Verweis auf Ha-
bilitationsmöglichkeiten in Wien und Graz abschlägig beantwortet wurde.
Offensichtlich war für Erismann das Vorhaben von Juhos sich mit Fragen
”
der neueren Logik (Logistik), der neueren Erkenntnistheorie (Sprach- und
Erkenntnislogik), sowie mit Grundlagenfragen der Natur- und Geisteswis-
senschften“ zu beschäftigen, kein hinreichendes Motiv, diese Richtung der
Philosophie in Innsbruck zu fördern, denn der Verweis Erismanns in seinem
Antwortschreiben an Juhos vom November 1945, ein weiterer Zuwachs
”
von Lehrkräften [bedeute] doch ein Missverhältnis für unsere normalerwei-
se nicht grosse Universität“, erschöpfte sich in einer bürokratisch formalen
Formulierung. Ich danke Michael Schorner, Innsbruck, für diese Informati-
on.
hos dem Ansuchen eine Bescheinigung der politischen Unbedenklichkeit“

”
bei.20
Als Habilitationsschrift reichte Juhos sein Manuskript Die Erkennt-
”
nis und ihre Leistung“(440 Seiten) ein. Den Kommissionsbericht über das
Habilitationsgesuch verfasst Kraft.21 Kraft verweist darin auf die bisher
vorliegenden 17 Veröffentlichungen und bezeichnet sie als ,,auf dem Bo-
den des logischen Neopositivismus“ stehend, innerhalb dessen sich Juhos
kritisch gegen einen radikalen Flügel, die Hypothetisten“, wendet. Die
”
Habilitationsschrift fasst frühere Arbeiten zusammen und führt diese fort.
Die im Folgenden kurze Darstellung der Habilitationsschrift von Ju-
hos stützt sich auf den Kommissionsbericht von Kraft. Einleitend formu-
liert Juhos die Theorie der empirischen Sätze, die Unterscheidung von
empirisch-hypothetischen und empirisch-nichthypothetischen Sätzen (bei
Schlick als Konstatierungen“ bezeichnet, deren Gegner innerhalb des
”
Neopositivismus die Hypothetisten“ sind); die hypothetischen Sätze sind
”
grundsätzlich dem Irrtum unterworfen.22 Im Anschluss an seine Arbeit
aus 194023 referiert Juhos nun die empirischen Gesetze der 1. Stufe“,
”
das sind gesetzmäßige Wiederholungen beobachtbarer Ereignisfolgen, wie
sie in singulären Sätzen bei der Wiederkehr der gleichen Umstände ent-
halten sind; in mathematischer Formulierung treten sie als Konstanten
auf. Darauf bauen Gesetze der 2. Stufe“ auf, welche Gesetze der 1. Stu-
” ”
fe so miteinander verknüpfen, dass die einander folgenden Ereignisse sich
in ihren stetigen Änderungen gegenseitig entsprechen“.24 In mathemati-
scher Formulierung sind dies Funktionen ohne explizite Konstanten. Die
Unterscheidung von Gesetzen 1. und 2. Stufe dient Juhos der erkenntnis-
logischen und methodischen Analyse der physikalischen Gesetze. Physi-
kalische Gesetze sind überwiegend Mischformen aus Gesetzen 1. und 2.
Stufe.
20 Archiv der Universität Wien, Personalakt Béla Juhos. Lebenslauf und Un-
bedenklichkeitsbescheinigung (polizeiliche Bescheinigung) fehlen im Konvo-
lut.
21 Archiv der Universität Wien, Personalakt Béla Juhos. Kommissionsbericht
über das Habilitationsgesuch des Dr. Béla Juhos, 5 Seiten.
22 Béla Juhos, Die Erkenntnis und ihre Leistungen. Wien: Springer Verlag
1950, S. 7ff.
23 Béla Juhos, Erkenntnisformen in Natur- und Geisteswissenschaften. Leip-
zig: Pan Verlag 1940, Kap. IV.
24 Ibid., Kap. V.
Juhos folgt hier der Verifikationstheorie des Neopositivismus, die er

eingehend analysiert. Die Wahrheit von nicht-hypothetischen Sätzen (Kon-
statierungen) ist nicht durch andere Sätze bestimmt, sondern durch Be-
obachtungen als etwas Außersprachliches, die man als eine der Sprache
gegenüberstehende Wirklichkeit bezeichnen kann, d. h. die empirischen
”
Sätze geben die Ordnung und Formen der Wirklichkeit“. Entsprechend
den Gesetzen 1. und 2. Stufe unterscheidet Juhos auch solche der Er-
kenntnis und auch analoge Stufen der Kausalität, des Determinismus und
des Indeterminismus. Wenn etwas durch Gesetze 2. Stufe beschreibbar
ist, dann ist der Zufall ausgeschlossen (Determinismus 2. Stufe). Wenn
das Geschehen durch Gesetze 1. Stufe beschreibbar ist, dann ist Zufall als
Anomalie möglich (Determinismus/Indeterminismus 1. Stufe). Die Fra-
ge nach Determinismus/Indeterminismus kann nur empirisch entschieden
werden. In der modernen Physik sei dies zum Zeitpunkt der Abfassung der
Habilitationsschrift (1947) nicht entscheidbar: im Mikrobereich herrscht
nur Determinismus 1. Stufe, vielleicht sogar vollständiger Indeterminis-
mus. Im Makrobereich kann Determinismus 2. Stufe angenommen werden.
Aufbauend auf seine Unterscheidung der empirischen Sätze gibt Juhos ne-
ben einer sprachlogischen Definition des Physischen und des Psychischen
eine klare Kritik des sogenannten Physikalismus und der Einheitssprache
”
der Wissenschaft (Übersetzbarkeit von Erlebnisaussagen in solche über
körperliche Vorgänge).“
Schließlich merkt Kraft im Kommissionsbericht über das Habilitati-
onsgesuch an: Weniger gelungen ist dagegen seine Begründung der in-
”
tersubjektiven Verständlichkeit von Konstatierungen, bei der er [Juhos]
sich ganz auf eine Abhandlung von Schlick stützt, wieder ohne ihn zu
nennen“.25 Durchaus kritisch fasst Kraft seine Beurteilung der Habilita-
tionsschrift zusammen:
Wenn auch an der Arbeit nicht die Originalität hervorgehoben werden kann,
so hat sie doch den großen Vorzug der Solidität und Gründlichkeit und Reife.
Sie zeigt ein hohes Niveau und eingehende Vertrautheit mit der Physik in er-

über das Habilitationsgesuch des Dr. Bela Juhos, 5 Seiten. Vgl. dazu auch
Artur Koterski, Béla von Juhos and the concept of ‘Konstatierungen’“,
”
in Friedrich Stadler (ed.), The Vienna Circle and Logical Epiricism. Vien-
na Circle Institute Yearbook 10 [2002]. Dordrecht–Boston–London: Kluwer
Academic Publishers 2003, S. 163-169.
kenntnistheoretischer Hinsicht, was in einer ausführlichen Diskussion der beiden

Konstanten c [Lichtgeschwindigkeit] und h [Planck’sches Wirkungsquantum] zur
Geltung kommt. Durch diese Arbeit wird nicht nur der Neopositivismus teils
ergänzend, teils kritisch in beachtenswerter Weise weitergeführt, sondern auch
die allgemeine Erkenntnistheorie bereichert.26
Die bereits genannte Arbeit aus dem Jahr 1940 Erkenntnisformen

”
in Natur- und Geisteswissenschaften“ scheint in mehrfacher Weise eine
Vorstufe zu den dann in der Habilitationsschrift breiteren thematischen
Ausführungen zu bilden, da schon in dieser Arbeit zwei Kapitel (Kap. VI.
und Kap. VII) den oben genannten Konstanten c und h gewidmet sind.
Die Abstimmung des Professorenkollegiums der philosophischen Fa-
kultät über die Zulassung zu den weiteren Schritten der Habilitation ergab
folgendes Ergebnis: 28 JA, 2 NEIN, 3 Enthaltungen; die Schlussabstim-
mung ergab 35 JA, 1 NEIN, 1 Enthaltung. Die Probevorlesung zum The-
ma Das Fundament der Verifikation“ fand sodann am 19. Jänner 1948
”
statt, die Juhos erfolgreich bestand.27 Damit lag der Ernennung zum Do-
zenten an der Universität Wien für theoretische Philosophie nichts mehr
im Wege.28 Die als Habilitationsschrift eingereichte Arbeit erschien 1950
bei Springer, Wien unter gleichem Titel.
Juhos nahm regen Anteil am internationalen wissenschaftlichen Le-
ben seines Fachgebiets durch Vorträge auf Kongressen, so u. a. am Deut-
schen Kongress für Philosophie (Bremen 1950) mit einem Referat über
logisch-semantische Fragen, am XI. Internationalen Kongress für Philo-
sophie (Bruxelles 1953) mit einem Referat zum Thema Wahrscheinlich-
”
keitsschlüsse als syntaktische Schlußformen“ und am Kongress für Phi-
losophie der Wissenschaften (Zürich 1954) mit dem Referat Erkennt-
”
nisanalytische Untersuchung physikalischer Gesetzesformen“. In der Zeit
zwischen 1945 und 1955 entstehen weitere elf Arbeiten, von denen eine
größere Anzahl in der Zeitschrift Methodos publiziert wird.

über das Habilitationsgesuch des Dr. Bela Juhos, 5 Seiten.
27 Der Habilitationskommission gehörten die Professoren Dempf, Meister,
Rohracher, Kainz, Ebert, Storch und Kraft unter dem Vorsitz von Dekan
Duda an.
28 Schreiben von Dekan Duda an das Bundesministerium für Unterricht vom
30. Jänner 1948. Archiv der Universität Wien, Personalakt Béla Juhos.
Bestätigung durch Bundesminister für Unterricht, Felix Hurdes am 24. Fe-
bruar 1948. Archiv der Universität Wien, Personalakt Béla Juhos.
Wie lässt sich die Position von Juhos in Wien in der Tradition des
Wiener Kreises“ stehend und insbesondere nach der Emeritierung von
”
Kraft im Jahre 1952 charakterisieren? Nun, zum einen bestand Juhos’
Rolle an der Universität Wien darin, die Ideen des Wiener Kreises“ wei-
”
ter zu tragen und weiter zu entwickeln. Neben der profunden Kenntnis
der Schlick’schen Philosophie befähigte ihn dazu vor allem seine intensive
Beschäftigung mit den Problemen der modernen Logik und seine auf ei-
nem technisch ausreichend differenzierten Niveau stehende Kenntnis der
modernen Physik, der speziellen und allgemeinen Relativitätstheorie und
der Quantentheorie, um die Diskussion auf einem dem geforderten tech-
nischen Anspruch entsprechenden Niveau aufzunehmen und fortzuführen.
Die Tragödie“ von Juhos bestand darin, dass er auf dem ihm entspre-
”
chenden Stand der akademischen Diskussion in Wien als eine singuläre
Erscheinung zu sehen ist, eine Position, die der Sichtbarkeit und Durchset-
zungsfähigkeit seiner Ideen nur eingeschränkte positive Impulse zu geben
vermochte und die sich weitgehend auf den unmittelbaren Kreis seiner
Schüler beschränkte. In diesem Kreis allerdings fand Juhos intellektu-
elle Resonanz, ja menschliche Bewunderung. In diesem Zusammenhang
der Wirksamkeit und Sichtbarkeit seiner Person und seines Werkes ist es
doch bemerkenswert, dass Juhos am Institut für Wissenschaft und Kunst
mit der damaligen Abteilung für Wissenschaftstheorie und Psychologie“
”
unter der Leitung von Walter Hollitscher oder der Abteilung für Mathe-
”
matik und Logik“, an der u. a. auch Kraft und der Mathematiker Edmund
Hlawka mitwirkten, nur einmal als Referent hervortrat, obschon dies seine
Wirkungsmöglichkeit über die universitäre Sphäre hinaus hätte erweitern
können. Im Januar 1950 hielt Juhos eine Vortrag zum Thema Neuere
”
logisch-semantische Untersuchungen“, zu dem er möglicherweise von Hol-
litscher eingeladen wurde, in dessen Abteilung unter anderen auch Kraft,
Paul Feyerabend, Franz Austeda Hans Novotny und Friedrich Waismann
referierten.29 Für eine weitere Mitwirkung Juhos’ am Institut für Wissen-
schaft und Kunst gibt es keine Hinweise. Zu ergänzen ist hier, dass Feyer-
abend über eine Einladung von Juhos in den Kraft-Kreis“, einen um 1950
”
29 Michael Schorner, Comeback auf Umwegen – die Rückkehr der Wissen-

”
schaftstheorie nach Österreich.“ In: Friedrich Stadler (Hg.) Vertreibung und
Rückkehr der Wissenschaftstheorie vor und nach dem Zweiten Weltkrieg –
Am Beispiel von Rudolf Carnap und Wolfgang Stegmüller. Wien: LIT Ver-
lag 2010.
aktiven studentischen Diskussionszirkel, berichtet, wo dieser über die In-

”
terpretation mathematischer Aussagen“ sprach. Die Frage, ob im Umfeld
des Kraft-Kreises“ Juhos in einen näheren wissenschaftlichen Kontakt
”
mit Elizabeth Anscombe ( [. . . ] versuchte Wittgenstein zu erklären, aber
”
mit wenig Erfolg“30 , so Feyerabend) oder Georg Henrik von Wright trat,
bleibt hier offen.
III. Wirken in der Nische

Mit der Ernennung von Leo Gabriel im Jahr 1951 zum ordentlichen Pro-
fessor für Philosophie war die von ihm innegehabte außerordentliche Pro-
fessur zur Wiederbesetzung frei geworden. Die a.o. Lehrkanzel für Phi-
”
losophie (naturwissenschaftliche Richtung)“ war für die Vertretung der
Naturphilosophie, Erkenntnistheorie und Logik bestimmt.31 In der Sit-
zung vom 30. Juni 1951 erstellte das Professorenkollegium der philosophi-
schen Fakultät dazu den folgenden Dreiervorschlag: Primo et aequo loco:
Friedrich Waismann (Oxford) und Karl Friedrich Freiherr von Weizsäcker
(Göttingen), secundo loco: Pd. Dr. Bela Juhos (Wien), tertio loco: Pd. Dr.
Erich Heintel (Wien).32
Das Protokoll der VII. Sitzung (Zl 7 aus 1950/51) der Sitzung des Pro-
fessorenkollegiums der philosophischen Fakultät unter Vorsitz von Dekan
Erich Schenk hält unter Punkt 3 zum Kommissionsbericht: Besetzung
”
der Lehrkanzel für Philosophie nach Prof. Kraft“ folgendes fest: Referent
Prof. Kraft; En bloc-Abstimmung; I. Weismann [sic!] – Weizsäcker, II. Ju-
hosz [sic!]; III. Heintel. Empfehlung: Ordinariat für Gabriel. Der Antrag
wurde mit 35 JA, 7 NEIN, 1 Stimmenthaltung angenommen.33 Weitere
Informationen, insbesondere über das Zustandekommen des Dreiervor-
schlags, sind diesem Protokoll nicht zu entnehmen. Ob die Reihung des
Besetzungsvorschlags als ein (letzter) Versuch Krafts zu sehen ist, diese
Lehrkanzel in die von ihm in der Nachfolge Schlicks vertretene Richtung
30 Ibid.
31 Archiv der Republik, Personalakt Erich Heintel, BMU, Gz. 34246/I-2/52.
33 Archiv der Universität Wien, Gz. Ph 31.35, Phil. Sitzungsprotokolle
1950/51, Prot. Professorenkollegium der Philosophischen Fakultät, 30. Juni
1951.
zu besetzen, erscheint möglich; da jedoch Belege dazu fehlen, wird im

folgenden der weitere formale Entscheidungsweg nachgezeichnet.
Friedrich Waismann (1896–1959), enger Mitarbeiter von Moritz Schlick,
Teilnehmer und Mitorganisator des Wiener Kreises“, emigrierte 1937
”
nach Großbritannien, wo er von 1937 bis 1939 an der Cambridge Univer-
sity als Lecturer wirkte; ab 1939 lehrte Waismann in Oxford, anfänglich
als Reader in the Philosophy of Mathematics, dann als Reader in the Phi-
losophy of Science. Karl Friedrich von Weizsäcker (1912-2007) war zum
Zeitpunkt der Erstellung des Dreiervorschlags Leiter der Abteilung für
Physik des Max-Planck-Instituts in Göttingen und Honorarprofessor an
der dortigen Georg-August-Universität. 1957 erhielt von Weizsäcker einen
Lehrstuhl für Philosophie an der Universität Hamburg.
Zu diesem Dreiervorschlag stellte das Bundesministeriums für Unter-
richt fest:
Was die primo loco genannten Kandidaten betrifft, so ist, abgesehen von den be-
kannten Schwierigkeiten, Gelehrte aus dem Auslande nach Österreich zu berufen
zu bemerken, daß sich Prof. Dr. Waismann in erster Linie mit mathematischen
Problemen befasst und lediglich Grenzgebiete der Philosophie vertritt, während
Prof. Dr. Weizsäcker sich mit allgemeinen Fragen der Philosophie beschäftigt
und es zweifelhaft ist, ob er einer Beschränkung seiner Lehrtätigkeit auf die
Erkenntnistheorie und Logik zustimmen würde.34
Im Abfertigungsentwurf des Bundesministeriums für Unterricht an das

Bundeskanzleramt betreffend die Zustimmung zur Ernennung von Hein-
tel zum ao. Prof. für Philosophie wurde die zitierte abwägend fachliche
Würdigung der primo loco genannten Kandidaten zu Gunsten einer kla-
ren Linie der Ablehnung fallen gelassen. In dem im Entwurf gestrichenen
Passus heißt es:
Die an erster Stelle genannten ausländischen Kapazitäten wurden teils ihrer

Ehrung wegen, teils für den Fall vorgeschlagen, dass es möglich wäre im Rah-
men des Dienstpostenplans 1952 für die Berufung eines der Genannten durch
Umwandlung der freien außerordentlichen Lehrkanzel in eine ordentliche Pro-
fessur die notwendige Voraussetzung dafür zu schaffen, einen der beiden die
Berufung auf die gegenständliche Lehrkanzel als ord. Professor anzubieten. Da
diese Möglichkeit im Dienstpostenplan nicht gegeben ist, kommen sie für die
Berufung nicht in Betracht.

So einfach ist das: Die Professoren des Kollegiums der philosophischen

Fakultät waren – trotz bester Beziehungen zum Bundesministerium für
Unterricht – bei der Erstellung des Dreiervorschlags Ende Juni 1951 in
völliger Unkenntnis des Dienstpostenplans 1952! So war man die bei-
den Erstgereihten dienstpostenplanmäßig“ losgeworden, und Waismann
”
hätte sich eine Ehrung“ dieser Art wohl verbeten, die von Weizsäcker
”
wohl nicht nötig hatte. Warum mit den beiden Erstgereihten nicht ver-
handelt wurde, bleibt aktenmäßig“ unerörtert. Resultat: Es galt nun zwi-
”
schen secundo und tertio loco zu entscheiden.
Der an zweiter Stelle gereihte Pd. für theoretische Philosophie Dr. Bela Ju-
hos besitzt wohl eingehende Kenntnisse der Physik in erkenntnistheoretischer
Hinsicht und eine genaue Kenntnis der mathematischen Logik. Er hat bisher
2 größere Abhandlungen und etwa 20 Zeitschriftenaufsätze veröffentlicht, sei-
ne Vorlesungen über Grundlagenprobleme wissenschaftlicher Methoden sind je-
doch im Gegensatz zu denen des Priv. Doz. Dr. Heintel schwach besucht. Es
wäre sohin im Hinblick auf vorstehende Ausführungen die Ernennung des Pd.
Dr. Heintel zum ao. Prof. für Philosophie zu erwirken. Er kommt vor allem in
Betracht, daß [?] die mathematische Logistik ein Teilgebiet der Philos. (natur-
wissenschaftl. Richtung] ist, während Heintels biologische Richtung die gesamte
Naturwissenschaft umfasst. [Der letzte Satz ist handschriftlich eingefügt. Anm
d. Verf.]35
Im Antrag auf Ernennung von Heintel zum außerordentlichen Professor

für Philosophie, unterfertigt von Bundesminister Ernst Kolb (1912–1978)
am 7. März 1952, wird sodann die größere Anzahl der Publikationen Hein-
tels (30 Abhandlungen gegenüber 20 Arbeiten von Juhos) hervorgehoben
und unter Verschweigen der von Juhos’ publizierten Bücher jenen Hein-
tels ( Nietzsche’s System in seinen Grundbegriffen“ und Metabiologie
” ”
und Wirklichkeitsphilosophie“) das Zeugnis ausgestellt, sie hätten infol-
”
ge ihrer zahlreichen neuen Aspekte die Forschung maßgeblich erweitert.
Sie waren bald nach ihrem Erscheinen vergriffen.“
Heintel abschließend als populären Vortragenden herauszustreichen,
verschaffte dem Antrag ein weiteres Atout gegenüber Juhos.
Während Juhos im vorliegenden Besetzungsantrage erstmalig von der Wiener

Philosophischen Fakultät für das Extraordinariat für Philosophie vorgeschlagen
wird, ist Heintel bereits 1950 an erster Stelle unico loco mit Dr. Gabriel für die
zu besetzende Lehrkanzel vorgeschlagen worden. Heintel hat im vergangenen
35 Ibid.
Jahrzehnt über sämtliche Hauptdisziplinen der Philosophie, insbesondere über

Erkenntnistheorie, Logik und Naturphilosophie, Teilgebiete, deren Vortrag dem
Inhaber der zu besetzenden Lehrkanzel zukommen, Vorlesungen abgehalten die,
obwohl er als Privatdozent weder prüfte noch Pflichtvorlesungen abhielt, infol-
ge seines glänzenden Lehrtalentes in jedem Semester von 100 bis 200 Hörern
besucht waren, während die Vorlesungen des Privatdozenten Juhos knapp 10
Hörer aufzuweisen pflegte. Alle diese Umstände rechtfertigen es, bei der Wie-
derbesetzung der gegenständlichen Lehrkanzel Heintel zu bevorzugen.36
Aus formalen Rücksichten musste auf Heintels NS-Mitgliedschaft Be-

zug genommen werden. Heintel ist seinerzeit aus beruflichen Gründen
”
der NSDAP ohne gesinnungsmäßige Bindung beigetreten und war als min-
derbelastet [gem. §17, Abs. 3, VG 1947, Anm. d. Verf.] registriert. Mit-
gliedsnummer 9,018.395.“37 Das Bundeskanzleramt hatte bereits mit Note
vom 8. Februar 1952 keinen Einwand gegen die beabsichtigte Ernennung
erhoben.38
Mit Entschließung vom 25. März 1952 wurde Heintel von Bundespräsi-
dent Körner zum außerordentlichen Professor für Philosophie ernannt. In
allen diesen akademischen Schachzügen blieb Juhos wenig mehr als eine
passive Rolle, während seine akademischen Kollegen mit Nachdruck ihren
Interessen folgten. Für Juhos hatte das Professorenkollegium nun eine
Kompensation parat.
Die nach der Habilitation nächste und damit zugleich auch letzte Stufe
seiner Karriere als universitärer Lehrer erreichte Juhos mit der Verleihung
36 Ibid.
37 Ibid.
38 Ibid.; BKA Zl. 508-PrS/52. In der Note des BKA wird ausgeführt: Er
”
hat sich um die Parteimitgliedschaft im Sommer 1938 aus Rücksicht um
die Existenz seiner Familie und im Hinblick auf seine bevorstehende Ha-
bilitierung an der Universität beworben.“ Im Zusammenhang mit Heintels
Habilitation 1940 gibt dieser im Fragebogen zu Gz. 6305/40 (Archiv der
Republik, Personalakt Erich Heintel) an: VF [Vaterländische Front] Mit-
glied seit 5. 12. 1934; weiters: Ich war SA-Anwärter.“ Heintel war vom
”
Juli 1938 bis zu seinem Austritt 1944 im Nationalsozialistischen Kraftfahr-
korps (Kraftbootstaffel Wien) organisiert. Der NS-Dozentenbundführer Ar-
thur Marchet führt mit Datum vom 14. 12. 1939 im Zusammenhang mit
dem Habilitationsansuchen aus, Erich Heintel bietet die Gewähr, dass er je-
”
derzeit rückhaltlos für den nationalsozialistischen Staat eintritt.“ E. Heintel
1959: Ich war nicht beim NS-Dozentenbund.“ (AdR, Personalakt Heintel,
”
Gz. 38.260-4/59)
des Titels eines a.o. Professors“. Der Kommission zur Verleihung dieses
”
Titels gehörten Gabriel, Heintel, Hopmann, Leitmaier, Meister, Rohra-
cher und Kainz (als Berichterstatter) an. Die Verleihung war von den
Vorständen des Philosophischen Instituts, Friedrich Kainz und Leo Ga-
briel, im Mai 1952 angeregt worden. In der Sitzung der Kommission vom
17. Juni 1952, bestehend aus Kainz, Gabriel, Heintel, Kraft, Hlawka, Hop-
mann, Thirring, Leitmeier, Meister und Gottschalk, wurde der Antrag ein-
stimmig angenommen, worauf das Professorenkollegium in seiner Sitzung
vom 21. Juni 1952 den Antrag zur Abstimmung brachte, mit folgendem
Ergebnis: 46 JA, 0 NEIN, 2 Enthaltungen. Hiervon wurde das Bundesmi-
nisterium für Unterricht mit Schreiben des Dekans Johann Radon vom 6.
August 1952 in Kenntnis gesetzt. Dieser Antrag wird jedoch vom Ministe-
rium nicht weiter behandelt, eine Tatsache, die zu Interpretationen Anlass
gibt. Hat das Ministerium aus eigenem Gutdünken gehandelt? Wurde von
außen her Einfluss ausgeübt Juhos nicht zu ernennen? Die vorliegenden
Akten geben darüber keine Auskunft und eröffnen so zugleich Raum für
Spekulationen, welchen strategischen Stellenwert im Kontext personal-
politischer und forschungspolitischer“ (i. e. ideologischer) Überlegungen
”
eine Ernennung von Juhos auf den für die philosophische Fakultät unbe-
deutenden Posten eines tit. a.o. Professors hatte. Hinweisen möchte ich
an dieser Stelle lediglich auf die schon oben referierte Tatsache, dass die
Initiative zur Verleihung des Titels formal von Kainz und Gabriel aus-
ging, wobei einer – Gabriel – in dezidierter Ablehnung zu der von Juhos
vertretenen philosophischen Richtung stand. Hatte man sich mit Juhos
auf eine Art von Minimallösung verständigt, die als Argument dafür die-
nen konnte, die verfemte Philosophie des Wiener Kreises sei mit Juhos
ausreichend repräsentiert, um jeden weiteren Vorstoß in Richtung einer
erkenntiskritischen Philosophie“ zu verhindern?
”
Kainz und Gabriel werden in dieser Angelegenheit im Dezember 1954
neuerlich aktiv und beantragen die Erstellung einer Kommission zur Be-
”
ratung der Verleihung des Titels eines a.o. Univ.-Professors an den Privat-
dozenten Dr. Bela Juhos.“39 Nunmehr wurde in der Sitzung des Profes-
sorenkollegiums vom 11. Dezember 1954 eine Kommission, bestehend aus
den Professoren Kainz, Gabriel, Heintel, Kraft, Hlawka, Hopmann, Thir-
ring, Leitmeier, Meister, Gottschalk und Rohracher zur weiteren Beratung

eingesetz.40 Der Antrag, zu dem sich Gabriel und Meister ausdrücklich zu-
stimmend äußerten wurde in der Sitzung der Kommission vom 14. Jänner
1955 einstimmig angenommen.
Ein zweiseitiges Gutachten legte dazu Kainz vor, in welchem er sich
vorbehaltlos positiv zu den bisherigen akademischen Leistungen Juhos’
äußert. Kainz hebt drei Hauptrichtungen der Tätigkeiten hervor: (1) die
akademische Lehre (Juhos hält jedes Semester mehrstündige Vorlesungen,
die bei den Studierenden viel Anklang und Beifall finden“), (2) seine An-
”
teilnahme am internationalen wissenschaftlichen Leben“ durch Referate
”
bei Kongressen und (3) seine sehr intensive Forschungsarbeit vorab auf
”
erkenntnislogischem Gebiet“. Kainz hebt insbesondere das 1954 im Hum-
boldt Verlag, Berlin erschienene Werk Elemente der neuen Logik hervor.
Es gibt – ohne die Brücke zur logischen Tradition völlig abzubrechen –
”
eine vollständige und gründliche Darstellung der modernen Logik, wofür
man bislang auf amerikanische Darstellungen angewiesen war. Dadurch
unterscheidet sich das Werk von den rein logistisch orientierten Darstel-
lungen Carnaps, dem Juhos im übrigen nicht fern steht.“ Kainz betont
weiters, dass sich Herr Juhos kritische Zurückhaltung auch mit bezug auf
”
Schulanschlüsse auferlegt und sich keinem ,Kreis‘ mit Haut und Haaren
verschrieben hat“.41 Diese Feststellung Kainz’ ist wohl als Entschärfung“
”
der Nähe von Juhos zum Wiener Kreis“ zu lesen, so, als ob er damit
”
seine Professorenkollegen zu beruhigen versuchte. Der zentrale Stellen-
wert der neuen Logik (Frege, Russell/Whitehead) für die philosophischen
Untersuchungen des Wiener Kreises“ ist in Juhos’ Terminologie der er-
”
kenntniskritischen (nicht erkenntnistheoretischen!) Untersuchungen deut-
lich spürbar. Von ähnlicher Funktion bezüglich der damaligen Rezepti-
on der analytischen Philosophie in Wien lassen sich die abschließenden
Ausführungen im Gutachten von Kainz verstehen:
Höchst anerkennenswert ist es, dass er ständig den Zusammenhang mit der
Erkenntnistheorie wahrt, also Erkenntnislogik betreibt und seiner Logik nach
wie vor die Aufgabe eines Organons zur Bearbeitung philosophischer Probleme
zuweist. Dadurch entfernt er sich zu seinem Vorteil von den Autonomie- und
Autarkieansprüchen der modernen Logistiker und analytischen Philosophen, für
welche die Bearbeitung logischer Darstellungsprobleme reiner Selbstzweck ge-
40 Archiv der Republik, Personalakt Béla Juhos.

41 Archiv der Universität Wien, Personalakt Béla Juhos. Antrag auf Verlei-
hung des Titels eines a.o. Professors, Gutachten F. Kainz.
worden ist und in zunehmendem Maß und völlig bewusst jede Verbindung mit
der logischen Tradition einer-, der gesamtphilosophischen Problematik anderer-
seits negiert.
Im zweiten Anlauf ging alles glatt. In der Fakultätssitzung vom 29.

Jänner 1955 kommt der positive Antrag der Kommission vom 14. Jänner
1955 zur Abstimmung und wird mit 42 JA, 3 NEIN, bei 0 Enthaltungen
angenommen. Die Verleihung des Titels eines außerordentlichen Univer-
sitätsprofessors durch Bundespräsident Theodor Körner erfolgt mit Ent-
schließung vom 17. Mai 1955.42 Ist mit dieser Ernennung möglicherweise
ein Schachzug – glatt – aufgegangen, der die Repräsentanz in der Nach-
folge der Philosophie des Wiener Kreises auf dem Niveau der univer-
sitätspolitisch unbedeutenden Position mit der Verleihung eines Titels
einfrieren ließ? War die Titelverleihung an Juhos jenes minimale Op-
”
fer“ an die nicht klerikal-konservativ gebundenen Kollegen, das univer-
sitätspolitisch als Zugeständnis fungieren konnte, um die nahezu zeitglei-
che Verhinderung einer Berufung von Wolfgang Stegmüller zuerst auf ein
Extraordinariat für Geschichte der Philosophie und systematische Philo-
sophie an der Universität Innsbruck 1956 zu kaschieren und sodann die Be-
strebungen von Hubert Rohracher, Stegmüller auf einen Wiener Lehrstuhl
zu berufen, zu unterlaufen? Diente Juhos als Schachfigur in einem Spiel,
das einerseits zur Kalmierung jener Kräfte in der Fakultät beitrug, die wie
Rohracher für analytische, antimetaphysische Philosophie offen waren, es
andererseits jedoch erlaubte, die Berufungspolitik im Sinne einer christlich
formierten Philosophie zu stabilisieren? Ein, wenn auch schwaches Indiz
in diese Richtung ist die Teilnahme von Rohracher an der Kommission
des Jahres 1954, deren Mitglieder ansonsten identisch mit jener des Jah-
res 1952 waren. Im Sinne dieser – durchaus spekulativen – Überlegungen
wäre dann Gabriel jene Schlüsselrolle im universitären Ränkespiel zuge-
fallen, wie dies von Benetka im Fall“ Stegmüller verdeutlicht wurde.43
”
Von einem Abstellgeleise“ für Juhos zu sprechen mag allzu hart klingen,
”
doch diese Qualifizierung seiner Position an der Universität Wien ent-
spricht den tatsächlichen Gegebenheiten seiner institutionellen Rolle, von
der seine Wirkung als universitärer Lehrer und produktiver Forscher zu
trennen ist.
42 Archiv der Republik, Personalakt Béla Juhos.

43 Gerhard Benetka, op. cit., S. 123-176.
1955 wurde Juhos zum Mitglied des Institut International de Philo-

”
sophie“, Paris als Vertreter Österreichs gewählt, was über die persönliche
Anerkennung für Juhos hinaus auch als Stellungnahme zu der in Wien ge-
pflogenen Philosophie zu deuten ist. Im Jahr 1956 Heirat mit Pia Juhos,
Dolmetscherin; die Ehe bleibt kinderlos.
Zurückhaltend, doch deutlich genug, formuliert Kraft im Nachruf auf
Juhos das am Institut für Philosophie der Universität Wien herrschende
Klima einer Feindseligkeit gegenüber der von Juhos (und Kraft) vertre-
tenen philosophischen Richtung, wenn er schreibt: Der philosophische
”
Gegensatz zu den Ordinarien hat es verhindert, dass ihm mehr als der Ti-
tel eines Extraordinarius zuteil wurde.“44 Aufgrund seiner akademischen
Position konnte Juhos lediglich Dissertationen betreuen, die offiziell“ von
”
einem der Ordinarien vergeben wurden und bei Rigorosen auch nicht of-
”
fiziell“ als Prüfer auftreten.
Im Jahre 1961 kommt es zwischen Juhos und Heintel zu einer Kon-
troverse, die in der Österreichischen Hochschulzeitung ausgetragen wird.
Juhos’ Artikel trägt den provokanten Titel Nichtmaterielle Gründe der
”
Abwanderung einheimischer Wissenschaftler“, auf den Heintel polemisch
antwortet.45
IV. Juhos als Lehrer

Wenn auch die Anzahl seiner Hörerinnen und Hörer gering war – sie be-
wegte sich zwischen 15 und 20, so vermochte er die Wenigen durch seine
offene Art des Vortrags, der – damals durchaus ungewöhnlich – durch Fra-
gen der Studierenden unterbrochen werden konnte und keineswegs dem
üblichen autoritären Stil von Professoren entsprach, zu faszinieren. Eine
Dame stenographierte in einer der letzten Reihen des Hörsaals während
seiner Vorlesungen eifrig mit; es war die Sekretärin der Firma Juhos. Seine
Vorlesungen fanden regelmäßig eine Fortsetzung zur weiteren Diskussion
im Café Landtmann, wozu Juhos einlud – um sich selbst mit einem Glas
Wein zu begnügen, wie Katzenberger von dem fast familiären Kreis sei-
44 Victor Kraft, Nachruf auf Béla Juhos“, op. cit., S. 163-173, auf S. 164.
”
45 Erich Heintel, Leserbrief“ in: Österreichische Hochschulzeitung XIII/12,
”
Wien, 15. Juni 1961 (Erwiderung auf den Leserbrief von Béla Juhos in der
Österreichischen Hochschulzeitung XIII/10, Wien, 15. Mai 1961, S. 2 - 3.).
4 Sp.
ner Hörer berichtet.46 Wenn Juhos auch engen Kontakt zu seinen Hörern
pflegte, so war er doch insgesamt in einer sehr isolierten Position, insbe-
sondere nach der Emeritierung von Kraft 1952 und auch zu den Physikern
– so naheliegend dies aufgrund von Juhos’ fachlichen Interessen scheinen
mag – bestand kein nachweisbarer Kontakt.
Anzumerken ist das in seinen späteren Jahren erneut aufkeimende In-
teresse an ethischen und moralischen Fragen, die an das Thema seiner
Dissertation anknüpfen, jedoch von Juhos in seinen Vorlesungen nicht be-
handelt wurden (siehe Anhang). Dies ist insofern bemerkenswert, da der
Wiener Kreis und der logische Empirismus (abgesehen von Otto Neurath
und vor allem Victor Kraft47 ) diesen Fragestellungen einer praktischen
Philosophie kein systematisches Interesse entgegenbrachte. Andererseits
lag es ganz außerhalb der Aufmerksamkeit von Juhos zu sehen, dass die
radikal aufklärerische Philosophie des Wiener Kreises diesen politisch ten-
denziell auf der Seite der Linken“ ansiedelte, wie Schleichert betont und
”
in diesem Zusammenhang Juhos als ganz apolitisch“ charakterisiert.48
”
Auch in dieser Hinsicht steht Juhos seinem Mentor Schlick nahe.
Als Schüler von Béla Juhos seien hier genannt: Werner Leinfellner,
Hubert Schleichert, Franz Oppacher, Wolfgang Katzenberger, Torsten
Kronfellner49 und Veit Pittioni.50 Schleichert nennt weiters den späteren
Schulpsychologen Gregor Pecko als Schüler von Juhos. Leinfellner und
Schleichert konnten sich an der Universität Wien nicht habilitieren, ein
deutlicher Hinweis auf die feindselige Haltung der Ordinarien für Phi-
losophie gegenüber der von Juhos vertretenen Richtung einer antimeta-
46 Interview mit Wolfgang Katzenberger 2008, aufgenommen von Károly

Kókai und Christoph Limbeck-Lilienau, Institut Wiener Kreis.
47 Victor Kraft, Die Grundlagen einer wissenschaftlichen Wertlehre. Wien
1937, 19512 .
48 Interview mit Hubert Schleichert 2008, aufgenommen von Károly Kókai und
Christoph Limbeck-Lilienau, Institut Wiener Kreis.
49 Katzenberger plante zusammen mit Kronfeller ein Buch über induktive Lo-
gik zu verfassen, das jedoch durch den Selbstmord Kronfellners nicht zu-
stand kam. Die geplante Arbeit erschien dann mit Juhos als Ko-Autor:
Béla Juhos, Wolfgang Katzenberger, Wahrscheinlichkeit als Erkenntnis-
form, Berlin: Duncker & Humblot 1970.
50 Der Ethnologe Hans Peter Duerr studierte in den 1960er Jahren bei Juhos
Wissenschaftstheorie. Hans Peter Duerr (Hg.), Paul Feyerabend. Briefe an
einen Freund. Frankfurt am Main: Suhrkamp Verlag 1995, S. 7.
physischen analytischen Philosophie. Schleichert stellt fest, das er für die

Drucklegung des mit Juhos gemeinsam erarbeiteten Textes für den Band
Die erkenntnislogischen Grundlagen der modernen Physik“ auf die Mit-
”
autorschaft verzichtete, um seine Chancen nicht zu schmälern, sich an
der Universität Wien zu habilitieren, wozu er 1967 noch Absicht hatte.51
Der Band Die erkenntnislogischen Grundlagen der klassischen Physik, pu-
bliziert 1963, hingegen nennt Schleichert neben Juhos als Autor.52 Sein
Habilitationsgesuch zog Schleichert zurück und verließ wenige Wochen
später Österreich.53 Leinfellners Habilitation wurde abgelehnt, diese Ab-
lehnung jedoch durch ein Gerichtsurteil als unzulässig aufgehoben.54 Alle
Juhos-Schüler, die akademische Karrieren anstrebten, waren gezwungen
aus Österreich wegzugehen: Leinfellner ging – nach einigen Jahren bei
Stegmüller in München – in die USA (University of Nebraska, Lincoln),
Schleichert nach Deutschland (Universität Konstanz), Oppacher nach Ka-
nada (Carleton University, Ottawa).
Béla Juhos verstarb im 70. Lebensjahr völlig unerwartet am 27. Mai
1971 in seinem Heim in Wien an einem Herzanfall, nachdem er am Vortag
noch seine Vorlesung abgehalten hatte.
Die wenigen persönlichen Zeugnisse,55 die uns vorliegen, lassen uns
Juhos als einen introvertierten, sehr distanzierten, durchaus professoralen
Menschen sehen; auch in der Stadt im Hubertusmantel mit schweren Schu-
hen unterwegs, was zum Jäger Juhos passt, ein durch und durch urbaner
Mensch, der seiner kleinaristokratischen Herkunft dezent Ausdruck ver-
leiht, autonom in intellektueller wie sozialer und ökonomischer Hinsicht,
51 Interview mit Hubert Schleichert 2008, aufgenommen von Károly Kókai und
Christoph Limbeck-Lilienau, Institut Wiener Kreis.
52 Béla Juhos und Hubert Schleichert, Die erkenntnislogische Grundlagen der
klassischen Physik, Berlin: Duncker & Humblot 1963. Schleichert promo-
vierte 1957 an der Universität Wien bei Juhos und habilitierte sich 1968 an
der Universität Konstanz.
53 Persönliche Mitteilung von Hubert Schleichert an den Verfasser, email, 25.
Mai 2008.
54 Heintel zu Leinfellner: Solange ich hier bin, werden Sie nie hier Professor.“
”
Leinfellner zu Heintel: Wenn nicht hier, dann sicher woanders“. Werner
”
Leinfellner, persönliche Mitteilung an den Verfasser, Januar 2009.
55 Interview mit Hubert Schleichert und Wolfgang Katzenberger 2008, aufge-
nommen von Károly Kókai und Christoph Limbeck-Lilienau, Institut Wie-
ner Kreis.
der Missachtung seiner akademischen Kollegen ausgesetzt und darunter

leidend, beliebt bei seinen Studenten, die in ihm einen kommunikativen
Gegenpol zum autoritären Gehabe der Professoren seiner Zeit erleben.
Die aufgezwungene Isolation vermochte er jedoch nicht aufzubrechen.
V. Bemerkungen zu Juhos’ erkenntnislogischen Grundlagen“ der

”
klassischen und modernen Physik
Im Folgenden wird eine kurze Darstellung der wissenschaftlichen Arbeiten
von Béla Juhos geboten, die sich an die Ausführungen im Nachruf von
Victor Kraft anschließt und insbesondere auf seine Auseinandersetzung
mit der klassischen und modernen Physik eingeht.56
Ausgehend von der im Wiener Kreis“ begründeten Philosophie des
”
Neopositivismus sowie der Logik nach Russell/Whitehead hat Juhos seine
philosophischen Untersuchungen als erkenntnislogische“ (nicht erkennt-
”
nistheoretische) bezeichnet und sich mit den neueren Entwicklungen in
den USA und Großbritannien (Carnap, Feigl, Quine; Popper, Lakatos)
auseinandergesetzt. Mit der logischen Analyse der Erkenntnis verbindet
Juhos die Forderung nach der Überprüfbarkeit der Erfahrung durch die
Wirklichkeit.
In Differenz zur Orthodoxie des Wiener Kreises“ (Schlick, Carnap)
”
und deren Begründung der Erkenntnis im sinnlich Gegebenen weist Ju-
hos darauf hin, dass das Gegebene“ nicht alleine mit den Sinnesdaten
”
identifiziert werden könne, da mit der sprachlichen Darstellung begriff-
liche Inhalte hinzukommen, d. h. die Analyse der Sprache eine wesent-
liche Komponente der philosophischen Analyse ausmache. Gleichrangig
dazu ist die Nachprüfbarkeit von Aussagen über die Realität durch die
Erfahrung Aufgabe der erkenntnislogischen“ Arbeit im Übergang vom
”
Neopositivismus zum logischen Empirismus.
Das Arbeitsgebiet von Juhos ist die logische Analyse der Erkenntnis
mit einem ausgeprägten Schwerpunkt in der Physik, wobei die Analy-
se der Sprache der (Natur-) Wissenschaft (hier wieder insbesondere der
klassischen und modernen Physik) besonderes Gewicht zukommt.
56 Victor Kraft, Nachruf auf Béla Juhos“. Zeitschrift für allgemeine Wissen-
”
schaftstheorie Bd. 2 (1971), S. 163-173.
Die Analyse der Sprache richtet sich (1) auf die logischen Elemente in
der sprachlichen Darstellung, (2) auf die Bedeutungen (Beziehungen der
sprachlichen Zeichen zu Inhalten). Juhos unterscheidet hier zwei Untersu-
chungsverfahren als formale und inhaltliche Analyse.
Die formale Analyse untersucht . . . die sprachlichen Ausdrücke . . . unabhängig

von ihrer Bedeutung (also unabhängig von den Objekten, die sie bezeichnen).
[. . . ] Demgegenüber untersucht die inhaltliche Analyse die Beziehungen der
sprachlichen Ausdrücke [. . . ] in ihren Beziehungen zu den Objekten, die sie be-
zeichnen, vor allem durch welche Zuordnungsregeln die Bedeutung der Sprach-
zeichen und Zeichenkomplexe festgelegt wird.57
Das Hauptziel der/seiner erkenntnisanalytischen Methode besteht in der

Ermittlung des eindeutigen Sinnes der sprachlichen Ausdrücke (Zeichen,
Begriffe, Sätze, Satzsysteme). Mit diesem Zugang lässt sich zeigen, dass
Grundbegriffe in verschiedenen Wissenschaften in verschiedenem Sinn ge-
braucht werden (z. B. Gesetzesbegriff in der klassischen Physik als Kau-
salgesetz und der Quantenphysik als Wahrscheinlichkeitsgesetz).
Als triadische Methode“ bezeichnet Juhos die Unterscheidung einer
”
Zuordnung wissenschaftlicher Begriffe zu drei Bereichen: (1) Bewusstseins-
daten, (2) logisch-mathematische Konstruktionen, (3) vom Bewusstsein
unabhängige Gegenstände. (Juhos wendet diese Zuordnung in seiner Ana-
lyse auf Kategorien wie Raum und Zeit an.) Unter diesem Gesichtspunkt
betrachtet er auch dreierlei Begriffe von Wahrheit: (1) Erlebniswahrheit
(das Für-wahr-Halten), (2) formale“ Wahrheit, (3) inhaltliche Wahrheit.
”
Demnach ist nach verschiedenen Kriterien zu prüfen, ob eine Satz wahr ist.
Weiters unterscheidet Juhos dreierlei Begriffe von Wahrscheinlichkeit: (1)
Erlebniswahrscheinlichkeit (das Für-wahrscheinlich-Halten als subjektiver
Glaube), (2) logisch-mathematische Wahrscheinlichkeit (logisch mögliche
Relation zwischen Sätzen), (3) physikalische Wahrscheinlichkeit (die for-
malen wahrscheinlichkeits-metrischen Systeme erhalten hier eine inhaltli-
che Bedeutung als empirisch reale Phänomenordnungen bestimmter Form).
Diese Unterscheidungen werden in der Analyse der klassischen und re-
lativistischen Physik (Kausalgesetze) und der Quantenphysik (objekti-
ve Wahrscheinlichkeitsaussagen und statistische Gesetze, die empirisch
prüfbar sind) zur Geltung gebracht. Kraft bemerkt dazu: Der dreifa-
”
57 Béla Juhos, Die erkenntnisanalytische Methode“. Zeitschrift für philoso-

”
phische Forschung, 6, 1951, S. 43f.
che Gesichtspunkt für die Analyse der Bedeutung verhilft somit zu einer
wesentlichen Klärung von Grundbegriffen der Wissenschaft.“58 Juhos kri-
tisiert mit seinem Ansatz Carnaps induktive Logik“ und deren Wahr-
”
scheinlichkeitsmetrik, stellt Kraft fest.59
Mit seinen logischen Untersuchungen des Aufbaus der Erkenntnis hat
Juhos sich Fragen der Induktion gewidmet und hier eine induktive Lo-
”
gik“ auf deduktiver Grundlage entwickelt, wobei er ,,induktive Schlüsse“
als Schlüsse aus disjunktiven Prämissen p1 oder p2 oder ...pn darstellt,
d. h. als nicht-tautologische Folgebeziehungen zwischen Prämissen und
Schlusssätzen. Da die Ableitung von Schlusssätzen aus disjunktiven Prä-
missen nicht eindeutig ist, d. h. auf verschiedene Weise möglich ist, muss
eine Festsetzung über die Auswahl getroffen werden, wodurch induktive
Schlüsse Wahrscheinlichkeiten verschiedener Art erhalten, die rein logi-
scher Natur (also nicht empirisch) sind. Diese induktiven Schlüsse lassen
sich zur Beschreibung der Wirklichkeit verwenden, wenn zwei Bedingun-
gen erfüllt sind: (1) diskrete unterscheidbare Fälle in einem empirischen
Bereich und (2) dürfen diese nicht in einer eindeutigen Ordnung abfol-
gen. So werden die logisch-mathematischen Wahrscheinlichkeitsformen zu
nachprüfbaren empirischen Wahrscheinlichkeitsaussagen, beschreiben rea-
le Zustände und deren Abfolge.
Damit ist der Schritt von der formalen Analyse der Erkenntnis zur
Analyse der empirischen Erkenntnis vollzogen. Durch die Anwendung von
Wahrscheinlichkeitsaussagen erhält man nun zweierlei Arten von Beschrei-
bungen von Vorgängen in der Natur: (1) Kausalgesetze und (2) Wahr-
scheinlichkeitsgesetze, die beide selbständig und gleichberechtigt sind,
nicht aus einander abgeleitet werden können. Diese Analyse dient nun
dem Verstehen der Nichtreduzierbarkeit von Wahrscheinlichkeitsaussa-
gen in der Quantentheorie. Diese Wahrscheinlichkeit [von quantenme-
”
chanischen Zuständen, Anm. d. Verf.] ist eine neue Erkenntnisform, die
nicht subjektive Ungewissheit bedeutet, sondern objektive Verhältnisse
darstellt. Juhos hat ihr mehrfache Abhandlungen gewidmet, die er in ei-
nem Buch Wahrscheinlichkeit als Erkenntnisform 1970 zusammengefasst
und ergänzt hat.“60
58 Victor Kraft, Nachruf auf Béla Juhos“, op. cit., S. 166.

”
59 Ibid., S. 166.
60 Ibid., S. 168.
In zwei Büchern, die der Untersuchung der klassischen und der moder-
nen Physik gewidmet sind, hat sich Juhos ausführlich mit der erkenntnis-
logischen Analyse der physikalischen Wissenschaft auseinandergesetzt.61
Juhos untersucht die Grundlagen der klassischen Physik, die Begrif-
fe von Raum und Zeit, sowie Fernwirkungen und Nahwirkungen. Me-
chanik und Elektrodynamik lassen sich in der klassischen Physik nicht
vereinen (Fernwirkung/Nahwirkung). Die Geometrie des (empirischen)
Raums muss aufgrund der Erfahrung ermittelt werden. Die spezielle Rela-
tivitätstheorie hat die Unhaltbarkeit der Voraussetzungen der klassischen
Physik erwiesen.
Allgemeine Relativitätstheorie und Quantenmechanik werden als mo-
derne Physik verstanden. Die allgemeine Relativitätstheorie basiert auf
der Invarianz der grundlegenden Gleichungen (Gesetze) gegenüber Trans-
formationen. Zur Begründung der Grundsätze der allgemeine Relativitäts-
theorie wird festgestellt, dass die relativistische Raum-Zeit-Metrik das
Prinzip der konstanten Lichtgeschwindigkeit als gültig voraussetzt und
darum die empirische Nachprüfung dieses Prinzips innerhalb der Rela-
tivitätstheorie nicht möglich ist. Die Relativität der Bewegung ist eine
”
Annahme in Bezug auf das gesamte System aller Maße und deshalb einer
Verifikation unzugänglich.“62
Die historische Entwicklung der Quantentheorie wird von Juhos aus-
führlich erläutert und auf die Energiequantelung und den Welle-Teilchen-
Dualismus eingegangen, der als ungeklärt beschrieben wird. Im Gegensatz
zur älteren (Bohr’schen) Quantentheorie
benützt die Quantenphysik eine Anzahl physikalisch zunächst nichtgedeuteter

Formeln und Ausdrücke, auf die je nach den empirischen Bedingungen entspre-
chend ausgesuchte Deutungsregeln angewendet werden. [. . . ] Je nach den em-
pirischen Bedingungen können dann die formalen Ausdrücke als Phänomene“
”
dieser oder jener Art, einmal etwa als Zustände, ein andermal etwas als Ver-
teilungen von Zuständen, gedeutet werden. Dadurch erst erhalten formale Aus-
drücke jeweils einen physikalischen Sinngehalt.63
61 Béla Juhos und Hubert Schleichert, Die erkenntnislogischen Grundlagen der

klassischen Physik, op. cit. Béla Juhos, Die erkenntnislogischen Grundlagen
der modernen Physik. Berlin: Duncker & Humblot 1967.
62 Béla Juhos, Die erkenntnislogischen Grundlagen der modernen Physik, op.
cit., S. 88. Zit. nach V. Kraft, Nachruf auf Béla Juhos“, op. cit., S. 169.
”
63 Ibid., S. 161.
Juhos gibt eine Interpretation der Unschärferelationen und bezeichnet den

Begriff der Unschärfe, der sich nun nicht mehr auf klassische Vorstellungen
bezieht, als einen nicht weiter zurückführbaren Grundbegriff einer neuen
Beschreibungsform des Geschehens auf der mikroskopischen Ebene.
Wenn man einen Korpuskularzustand nie durch Messung so bestimmen kann,

daß man die Folgezustände eindeutig voraussagen kann, dann erscheint die Be-
hauptung der Realität der Korpuskeln genauso mangelhaft begründet wie die
Realität der Materiewellen auf Grund der Interferenzmuster.64
Erst die kritischen Fälle, die in neuester Zeit zur Aufstellung der Quanten- und
Relativitätstheorie Anlaß gaben, erforderten zu ihrer Bewältigung die Aufstel-
lung von Naturgesetzen von so extremer Entartung, daß zu ihrem Verständnis
die Besinnung auf die gesonderten Regeln des physikalischen Sprachgebrauchs
und damit auf die einzelnen Erkenntnismethoden der Physik nötig wurde.65
Bei der Analyse der Grundlagen der klassischen und modernen Physik
bezeichnet Juhos als Grundbegriffe“ alle jene, welche die inhaltliche Deu-
”
tung der formalen Ausdrücke festlegen (absolute/r Raum, Zeit, Bewegung
in der klassischen Physik; Fernwirkung). Die Fernwirkung wird so zum
Schlüsselbegriff“ der klassischen Physik, da mit der Problematisierung
”
der Fernwirkung ein Aufgeben der Grundbegriffe“ verbunden ist. Für die
”
Theorie des Elektromagnetismus ist die Nahwirkung der Schlüsselbegriff“.
”
Dies gilt auch für die Relativitätstheorie und ihren Nahewirkungsbegriff,
der die ,,Schlüsselbegriffe“ der klassischen Physik relativiert, allerdings er-
folgen Zustandsänderungen stetig und Stetigkeit wird hier zum Schlüssel-
”
begriff“ der allgemeinen Relativitätstheorie.
Stetigkeit gilt allerdings nicht mehr in der Mikrophysik, da eine Mes-
sung von stetigen Zustandänderungen nicht möglich ist. Hier wird eine
neue Form der Beschreibung eingeführt, nämlich Wahrscheinlichkeitsge-
setze. Diese absoluten“ (nicht subjektiven) Wahrscheinlichkeitsgesetze
”
sind objektive Beziehungen zwischen Anfangs- und Endzuständen, die von
der subjektiven Kenntnis unabhängig sind. Diese Wahrscheinlichkeitsge-
setze sind nun die Schlüsselbegriffe“ der atomaren Physik. Juhos gelangt
”
hier zu einer Dualität von Schlüsselbegriffen in der modernen Physik: Ma-
64 Béla Juhos, Die erkenntnislogischen Grundlagen der modernen Physik, op.

cit., S. 227.
65 B. von Juhos, Erkenntnisformen in Natur- und Gesisteswissenschaften.
Leipzig: Rudolf Birnbach Abteilung Pan-Verlag 1940, S. 4.
krophysik charakterisiert durch Stetigkeit und Nahewirkung; Mikrophysik

charakterisiert durch Unstetigkeit und Wahrscheinlichkeit. Dieser Dua-
”
lismus von kontinuierlich veränderlichen Zuständen und einer Verteilung
von diskreten körperlichen Elementen“66 ist aber nicht in einer einheitli-
chen Sicht aufgehoben.
Die beiden verschiedenen Bereiche der Makro- und der Mikrophänomene wer-
den dadurch miteinander verknüpft, dass ein Kollektiv diskreter Mikrovorgänge,
wenn entsprechend empirische Bedingungen vorliegen, ein beobachtbares, even-
tuell messbares Makrophänomen zugeordnet wird. Damit wird eine Vielzahl von
diskreten Mikrovorgängen mit einem relativ stetigen Makrophänomen gleichge-
setzt. Diese Gleichsetzung kann nicht logisch abgeleitet werden, sondern erfolgt
durch Zuordnung.67
Anknüpfend an Juhos’ Darstellung der klassischen und modernen Physik

sei abschließend auf seine Unterscheidung zweier Arten von Naturgeset-
zen, die er in seiner Habilitationsschrift ausführt (s. o.), nochmals kurz
eingegangen. Als Gesetze 1. Stufe bezeichnet er solche, die nur beob-
”
achtete Einzelereignisse enthalten und deshalb nur eine Voraussage der
Wiederholung der gleichen Ereignisse erlauben.“68 ; als Gesetze 2. Stufe
gelten nun solche, die verallgemeinerte Ereigniszusammenhänge enthalten
und Voraussagen ermöglichen, die über solche hinausgehen, die zur Ge-
winnung des Gesetzes nötig waren. Dieser hier implizierten zweistufigen
Kausalität entspricht weiters ein zweistufiger/doppelter Determinismus,
der jeweils durch Gesetze der 1. bzw. 2. Stufe beschrieben werden kann.
Der Determinismus der 1. Art lässt Raum für einen Indeterminismus,
weil es Ereignisse geben kann, die nicht Ereignisse gemäß einem Gesetz
der 1. Stufe sind. In diesem Indeterminismus sieht Juhos die Begründung
des freien Handelns, die ihn zu seinen Reflexionen über Fragen der Ethik
führt, welchen jedoch hier nicht weiter nachgegangen wird.

”
67 Ibid., S. 172 und Béla Juhos, Makrophänomene und ihre Zusammensetzung
”
aus Mikrophänomenen“. Philosophia naturalis, Bd. 12 (1970) S. 413.
”
VI. Denker ohne Wirkung

Juhos hinterließ ein wissenschaftliches Lebenswerk von acht Büchern69
und ca. 110 Aufsätzen.70 Die Wirksamkeit und Rezeption seiner Arbeiten
und deren Sichtbarkeit im philosophischen Diskurs ist bis heute kaum
über den engeren Kreis seiner Schüler und die einschlägige Diskussion
hinaus gediehen. In einer neueren Arbeit zur Geschichte des Positivismus
und Neopositivismus im Österreich des 19. und 20. Jahrhunderts widmet
Gerhard Schurz neben einer Würdigung des gleichfalls kaum rezipierten
Victor Kraft eine längere kompakte Zusammenfassung der Philosophie
von Béla Juhos.71
Unter den Titel Denker ohne Wirkung. Béla Juhos – ein typisches
”
Schicksal“ schreibt Hubert Schleichert in einem Nachruf über seinen Leh-
rer:
Gelegentlich diente er als Aushängeschild, um Toleranz und Pluralismus der

österreichischen Philosophie zu demonstrieren – aber alles, was er geschaffen
hat, ist ohne jede Unterstützung durch Österreichs Hochschulen geschehen. Nur
die Zufälligkeit eines hinreichenden Besitzes erlaubte ihm die wissenschaftliche
Arbeit. Die Schande, dass ein Gelehrter vom Range Béla Juhos sein Leben in der
akademischen Verbannung führen musste, fällt auf seine Feinde zurück. Es soll
hier über diese Leute kein Wort verloren werden. [. . . ] Die Wirkungsarmut des
Verstorbenen wird durch lokale Wiener Intrigen nur teilweise erklärt. Sie liegt
auch in seinem Werk begründet. Es gibt eine Geisteshaltung und eine Art, zu
69 Über die Grundlagen der Gewissheit des reinen Denkens, Wien: Gerold
1928; Erkenntnisformen in Natur- und Gesisteswissenschaften, Leipzig: Pan
Verlag 1940; Die Erkenntnis und ihre Leistung, Wien: Springer Verlag 1950;
Elemente der neuen Logik, Frankfurt und Wien: Humboldt Verlag 1954; Das
Wertgeschehen und seine Erfassung, Meisenheim am Glan: Verlag Anton
Hain 1956; Die erkenntnislogische Grundlagen der klassischen Physik (zu-
sammen mit Hubert Schleichert), Berlin: Duncker & Humblot 1963; Die er-
kenntnislogische Grundlagen der modernen Physik, Berlin: Duncker & Hum-
blot 1967; Wahrscheinlichkeit als Erkenntnisform (zusammen mit Wolfgang
Katzenberger), Berlin: Duncker & Humblot 1970.
70 Eine Bibliographie der Arbeiten von Juhos enthält der Band Béla
Juhos, Selected Papers on Epistemology and Physics. Gerhard Frey (ed.),
Dordrecht/Boston: D. Reidel Publishing Company 1976.
71 Gerhard Schurz, Wissenschafts- und Erkenntnistheorie, Logik und Spra-
”
che: Positivismus und Neopositivismus und das Umfeld“, in K. Acham
(Hg.), Cognitio Humana. Der Beitrag Österreichs zu den Wissenschaften
vom Menschen im 19. Und 20. Jahrhundert. Akademie-Verlag 1998.
schreiben, mit der man mit großer Sicherheit in akademischen oder überhaupt
in gebildeten“ Kreisen ignoriert wird.72
”
Vielleicht kann die hier vorgetragene Skizze des Lebens und Wirkens
eines unbeirrt geradlinigen und allen akademischen Eitelkeiten abholden
Forschers einen Beitrag dazu leisten, den engen Zirkel der Rezeption von
Juhos’ Arbeiten zu erweitern.
Danksagung
Eckehart Köhler, Wien, Werner Leinfellner, Wien, Franz Oppacher, Mont-
real, Veit Pittioni, Innsbruck, Hubert Schleichert, Konstanz uhnd Micha-
el Schorner, Innsbruck haben mir wertvolle Hinweise und Anregungen
für diese Arbeit gegeben. Károly Kókai und Christoph Limbeck-Lilienau
danke ich für die Zurverfügungstellung von Interviews mit Wolfgang Kat-
zenberger und Hubert Schleichert. Friedrich Stadler gilt mein Dank für
die Anregung dieser Arbeit.
72 Hubert Schleichert, Denker ohne Wirkung. Béla Juhos – ein typisches

”
Schicksal“. Conceptus, 5 (1971), S. 5-12.
Anhang
Béla Juhos
Verzeichnis der Vorlesungen und Seminare,
gehalten an der Universität Wien 1948 bis 1971
WS 1948/49 Neuere Erkenntnislehre, 2st., Do. 11-13; Hs. 37

SS 1949 Neuere Erkenntnislehre II (Forts.), 2st., Do. 11-13; Hs. 37
WS 1949/50 Raum, Zeit, Kausalität, 2st., Do. 11-13; Hs. 37
SS 1950 Raum, Zeit, Kausalität (Forts.), 2st., Do. 11-13; Hs. 37
WS 1950/51 Erkenntnisanalytische Behandlung philosophischer Pro-
bleme, 2st., Do. 11-13; Hs. 37
SS 1951 Erkenntnisanalytische Behandlung philosophischer Pro-
bleme (Forts.), 2st., Do. 11-13; Hs. 37
WS 1951/52 Grundlagenprobleme wissenschaftlicher Methoden, 2st.,
Do. 11-13; Hs. 37
SS 1952 Elemente der neuen Logik I, 2st., Do. 11-13; Hs. 37
WS 1952/53 Elemente der neuen Logik II, 2st., Do. 11-13; Hs. 37
SS 1953 Elemente der neuen Logik III, 2st., Do. 11-13; Hs. 37
WS 1953/54 Neuere Erkenntnislehre I, 2st., Do. 11-13; Hs. 37
Elemente der neuen Logik IV, 2st., Fr. 11-13; Hs. 35
SS 1954 Neuere Erkenntnislehre II, 2st., Do. 11-13; Hs. 37
Elemente der neuen Logik V, 2st., Fr. 11-13; Hs. 35
WS 1954/55 Naturphilosophie I (Begriffs- und Theoriebildung in der
klassischen Physik), Do. 11-13; Hs. 37
Neuere Erkenntnislehre III, 2st., Fr. 11-13; Hs. 37
SS 1955 Elemente der neuen Logik I, 2st., Do. 11-13; Hs. 37
Naturphilosophie II (Erkenntnislogische Grundlagen der
klassischen Mechanik), 2st., Fr. 11-13; Lektoren-Hs.
WS 1955/56 Elemente der neuen Logik II, 2st., Do. 11-13; Hs. 37
Naturphilosophie III (Erkenntnislogische Grundlagen der
Elektromagnetik), 2st., Fr. 11-13; Hs. 37
SS 1956 Elemente der neuen Logik III (Der Prädikatenkalkül), 2st.,
Do. 11-13; Hs. 37
Naturphilosophie IV (Erkenntnislogische Grundlagen der
Elektromagnetik), 2st., Fr. 11-13; Hs. 37
WS 1956/57 Philosophische Grundlagen der relativistischen Physik,

2st., Do. 11-13; Hs. 37
Seminar: Besprechung neuerer philosophisch-
wissenschaftslogischer Werke, 2st., Fr. 11-13; Hs. 37
SS 1957 Raum, Zeit, Masse, 2st., Do. 15-17, Hs. 35a
Übungen in der Interpretation erkenntnislogischer Arbei-
ten, 2st., n. Ü.
WS 1957/58 Geometrie und empirische Erkenntnis, 2st., Do. 15-17, Hs.
35b
SS 1958 Begriffsformen der Allgemeinen Relativitätstheorie, 2st.,
Do. 15-17; Hs. 37
WS 1958/59 Die philosophischen Grundlagen der anschaulichen Quan-
tenphysik, 2st., Do. 15-17; Hs. 35 b
SS 1959 Die Wahrscheinlichkeit“ als empirische Beschreibungs-
”
form, 2st., Do. 15-17; Hs. 35 b
WS 1959/60 Die Wahrscheinlichkeitsbeschreibung in der Quantenphy-
sik, 2st., Do. 15-17; Hs. 35 b
SS 1960 Die Quantisierung als Operation und ihre Anwendung in
der Biologie, 2st., Do. 15-17; Hs. 35 a
WS 1960/61 Neuere Erkenntnislehre, 2st., Do. 15-17; Hs. 35 b
SS 1961 Neuere Erkenntnislehre II, 2st., Do. 15-17; Hs. 35 a
WS 1961/62 Die neue Logik als Voraussetzung der wissenschaftlichen
Erkenntnis. Der Aussagenkalkül, 2st., Do. 15-17; Hs. 35 b
SS 1962 Elemente der neuen Logik II: Entscheidbarkeitsfragen im
Aussagenkalkül, 2st., Do. 15-17; Hs. 35 a
WS 1962/63 Die philosophischen Grundlagen der Mathematik, 2st., Do.
15-17; Hs. 35 b
SS 1963 Die logizistische und formalistische Begründung der Ma-
thematik, 2st., Do. 15-17; kl. Hs. d. Mathem. Inst., IX.,
Strudelhofgasse 4
WS 1963/64 Die formalistische Nichtentscheidbarkeit und die institu-
tionistische Begründung der Mathematik, 2st., Do. 15-17;
kl. Hs. d. Mathem. Inst., IX., Strudelhofgasse 4
SS 1964 Die finitistisch-intuitionistischen Grundlagen der Mathe-
matik, 2st., Do. 15-17; kl. Hs. d. Mathem. Inst., IX., Stru-
delhofgasse 4
WS 1964/65 Die erkenntnislogischen Grundlagen der exakten Natur-

wissenschaften, 2st., Do. 14-16; Neues Institutsgebäude,
Hs. III
Seminar (Probleme der erkenntnislogischen Grundlagen-
forschung), Mi. 18-20; I. Philos. Inst.
SS 1965 Die erkenntnislogischen Grundlagen de klassischen Physik,
2st., Do. 14-16; Neues Institutsgebäude, Hs. III
Seminar (Probleme der Anwendung formaler Wissenschaf-
ten), 2st., Mi. 18-20; Hs. 39
WS 1965/66 Die philosophisch-erkenntnislogischen Grundlagen der
klassischen Elektromagnetik, 2st., Do. 14-16; Neues Insti-
tutsgebäude, Hs. III
Seminar (Besprechung erkenntnislogischer Grundlagen-
probleme), 2st., Mi. 18-20; Hs. 39
SS 1966 Die Grundlagenkrise der klassischen Physik und der
Übergang zur Relativitätstheorie, 2st., Do. 14-16; Neues
Institutsgebäude, Hs. III
Seminar (Besprechung neuerer erkenntnislogischer Unter-
suchungen), 2st., Mi. 18-20; Hs. 39
WS 1966/67 Die erkenntnislogischen Grundlagen der Relati-
vitätstheorie, 2st., Do. 14-16; Neues Institutsgebäude, Hs.
III
Seminar (Besprechung erkenntnislogischer Probleme),
2st., Mi. 18-20; Hs. 39
SS 1967 Grundlagenfragen der Relativitätstheorien, 2st., Do. 14-
16; Neues Institutsgebäude, Hs. III
Seminar (Erkenntnislogische Grundlagenprobleme in
neueren Darstellungen) 2st., Mi. 18-20; Hs. 39
WS 1967/68 Erkenntnislogische Analyse der Grundlagen der modernen
Physik. Der Übergang von der klassisch-relativistischen
Physik zur anschaulichen Quantentheorie, 2st., Do. 14-16;
Neues Institutsgebäude, Hs. III
SS 1968 Grundlagenfragen der Quantenphysik. Logische und em-
pirische Wahrscheinlichkeit, 2st., Do. 14-16; Neues Insti-
Seminar: Besprechung neuerer Arbeiten zu Problemen

der Erkenntniskritik (Wissenschaftstheorie, Methodologie,
Ideologiekritik), 2st., Mi. 18-20; Hs. 39
WS 1968/69 Die erkenntnislogischen Grundlagen der nichtanschauli-
chen Quantenphysik. Die absoluten Wahrscheinlichkeits-
gesetze, 2st., Do. 14-16; Neues Institutsgebäude, Hs. III
Seminar: Besprechung neuerer Arbeiten auf dem Gebie-
te der Erkenntniskritik, Wissenschaftstheorie, Ideologie-
kritik, 2st., Mi. 18-20; Hs. 39
SS 1969 Grenzformen der Erkenntnis in der Quantenphysik. Ele-
mente der Quantenbiologie, 2st., Do. 14-16; Neues Insti-
Seminar: Besprechung neuerer Arbeiten über Probleme
der Wissenschaftstheorie, Methodologie, Ideologiekritik,
2st., Mi. 18-20; Hs. 39
WS 1969/70 Neuere Erkenntnislehre, Do. 14-16; Neues Instituts-
gebäude, Hs. III
Seminar: Besprechung neuerer Arbeiten auf dem Gebie-
te der Erkenntniskritik, Wissenschaftstheorie, Ideologie-
kritik, 2st., Mi. 18-20; Hs. 39
SS 1970 Neuere Erkenntnislehre II, 2st., Do. 14-16; Neues Instituts-
gebäude, Hs. III
Seminar: Besprechung neuerer Arbeiten über Grundlagen-
fragen der Wissenschaftstheorie, Methodologie und Er-
kenntnisanalyse, 2st., Mi. 18-20; Hs. 39
WS 1970/71 Elemente der neuen Logik, 2st., Do. 14-16; Neues Insti-
Seminar: Besprechung neuerer Arbeiten auf dem Gebiete
der Erkenntniskritik, Wissenschaftstheorie, Methodologie,
Ideologiekritik, 2st., Mi. 18-20; Hs. 39
SS 1971 Das Aussagenkalkül II, 2st., Do. 14-16; Neues Instituts-
gebäude, Hs. III
Seminar: Referate über Grundlagenfragen der Wissen-
schaftstheorie, Methodologie und Erkenntnisanalyse, 2st.,
Mi. 18-20; Hs. 39
Gergely Ambrus
Juhos’ Antiphysicalism and his Views on the

Psychophysical Problem
Introduction
In this paper my primary aim is to discuss Bela Juhos’ views on the mind-
body problem (or the psychophysical problem, as the logical positivists
preferred to call it). In order to achieve this, I find it necessary to provide
some background against which his ideas can be located: I shall outline
Juhos’ metaphilosophical views on the nature and goal of philosophical
inquiry, and the diverse accounts of the psychophysical problem in and
around the Vienna Circle.
Juhos, the only member of the Vienna Circle of Hungarian origin,1 stu-
died mathematics, physics, and philosophy at the University of Vienna.
In philosophy, he was a student of Schlick, and his views on most issues
followed his teacher’s rather closely. Juhos’ philosophical method was a
certain sort of linguistic analysis (“die sprachlogische Methode”), accor-
ding to which one has to deal with philosophical problems the following
way. First, one has to clarify the meaning of the concepts used in for-
mulating the philosophical problem and provide them with a new, scien-
tifically acceptable (i.e., verifiable) meaning. Then the original problem
was to be reformulated with these new concepts. As a result, the old pro-
blem (often) dissolves, but new empirical and logical-linguistic problems
1 He came from a (possibly Transsylvanian) gentry family, who had a steel
business at the turn of the 20th century both in Budapest and Vienna.
Bela Juhos himself was born in Budapest and moved to Vienna in 1909,
at the age of 8. Besides Juhos, the psychologist Egon Brunswik, who was
associated with the Vienna Circle, also came from a (historical) Hungarian
family. However, the branch of Brunswiks he belonged to, had lived in
Vienna for long, so his relations to Hungary may have been rather weak.
(This information is credited to Csaba Pléh.)
100 Gergely Ambrus
also arise.2 This approach fits well with the general logical positivist me-
thodology of the 1930s, represented especially by Carnap. According to
his dictum, “philosophy is the logic of science”, i.e., the task of philoso-
phy is the linguistic-logical analysis of those philosophical concepts, which
are important for the sciences (e.g., reality, mind, body, causation, laws,
etc.), and also the unfolding of the logical connections between scientific
statements.
As for Juhos’ main philosophical interests: he took part in the protocol
sentence debate in the first part of the 1930s, siding with Schlick and
attacking the physicalist wing. He published several papers on this topic3
in the 1930s; his views from this period are summarized in his book, Die
Erkenntnis und ihre Leistung, which appeared much later.4 Later in his
career, from the end of the 1940s until 1970, his interests lay mainly in
issues of philosophy of science and epistemology, and in philosophy of
physics in particular. He discussed the nature of laws and causation;5 he
wrote many papers and two books on the philosophical interpretation of
physics;6 several papers on general epistemological issues. He criticized
Popper, claiming that verification and falsification are symmetrical;7 he
2 Cf. e.g. Bela Juhos, Die Erkenntnis und ihre Leistung. Wien: Springer 1950,
pp. 1-2.
3 “Kritische Bemerkungen zur Wissenschaftstheorie des Physikalismus”, in:
Erkenntnis 4, 1934, pp. 397-418., “Empiricism and Physicalism”, in: Analy-
sis 2/6, 1935, pp. 81-92., “Some Modes of Speech of Empirical Science”, in
Analysis 3/5, 1936, pp. 41-55. All reprinted in Bela Juhos, Selected Papers
on Epistemology and Physics. Dordrecht–Boston: D. Reidel 1976.
4 This book was Juhos’ Habilitationsschrift, which he submitted to the Uni-
versity of Vienna in 1948.
5 Already early in his career, as e.g. in “Stufen der Kausalität”, in: Jahresbe-
richt der Philosophischen Gesellschaft zu Wien 1931/32, pp. 1-19., and also
later, e.g. in “Wie gewinnen wir Naturgesetze?” in: Zeitschrift für philoso-
phische Forschung 22/4, 1968, pp. 534-548.
6 (together with Hubert Schleichert) Die erkenntnislogischen Grundlagen
der Klassischen Physik. Berlin: Duncker und Humblot 1963, Die erkennt-
nislogischen Grundlagen der Modernen Physik. Berlin: Duncker und Hum-
blot 1967.
7 “Die methodologische Symmetrie von Verifikation und Falsifikation” in:
Journal for General Theory of Science 1/2, 1970, pp. 41-70.
Juhos’ Antiphysicalism 101
also argued for an operational conception of the relativistic notion of space

and time.8 Besides, he also published some works in ethical theory.9
*
Juhos’ views on the psychophysical problem may be exposed aptly against
the background of the physicalism debate in the first half of the 1930s, so I
will locate them in the dialectic of this controversy. First, I shall clarify his
general philosophical method somewhat further, and explain his notions
of the “physical” and the “mental”. Then, I shall shortly introduce the
main views on the psychophysical relation in the Vienna Circle, in order
to define Juhos’ position more precisely.
As already mentioned, Juhos’ method starts with an analysis of the
concepts involved in the formulation of a philosophical thesis, including
the determination of their scientific, i.e., verifiable meaning. As a result
of this analysis, the problems involved in the original formulation may
bifurcate into empirical and analytical questions, i.e., synthetic questions
that can be settled a posteriori, and analytic questions which concern only
language use, the meaning of terms and sentences, and require no empiri-
cal inquiry. In Juhos’ opinion, as we shall see shortly, the psychophysical
problem has both of these aspects: it involves issues of a purely lingui-
stic/analytic character, and issues which are to be decided empirically.
In order to explain the competing alternative positions Juhos out-
lined concerning the psychophysical problem, it is necessary to clarify
Juhos’ notions of the “physical” and the “mental”. The meaning of these
terms are based on a distinction Juhos drew between two types of empiri-
cal sentences, i.e., empirical-hypothetical and empirical-non-hypothetical
sentences. The empirical-nonhypothetical sentences are the so-called Kon-
statierung- or affirmation-sentences. Such sentences allow for absolute cer-
tainty, for it is impossible to be mistaken about their truth values. For such
sentences there is no process of verification: for, as we may put it, their as-
sertion conditions are their truth conditions. The empirical-hypothetical
8 See e.g. “The Influence of Epistemological Analysis on Scientific Research:
Length and Time in the Special Theory of Relativity”. In: Imre Lakatos and
Alan Musgrave (eds.), Problems in the Philosophy of Science. Amsterdam:
North-Holland 1968.
9 “Über juristische und etische Freiheit”, in: Archiv für Rechts- und Sozial-
philosophie 29/3-4, 1937, pp. 406-431., Das Wertgeschehen und seine Er-
fassung. Meisenheim an Glan: A. Hain 1956.
102 Gergely Ambrus
sentences are never certain. The verification of such sentences proceeds

basically the following way. One finds logical consequences of the sentence
which are empirical-nonhypothetical, i.e., Konstatierung-sentences, and
then check whether these are true. Juhos accepted that such a verificati-
on process can never be final, hence the “hypothetic” character of these
sentences.
Juhos, following Schlick, strictly upheld both the analytic/synthetic
distinction, and the idea of there being incorrigible Konstatierung- or
affirmation-statements. He also followed his teacher in interpreting Kon-
statierung-sentences in a way analoguous with analytic sentences. Accor-
dingly, Konstatierung-sentences are “quasi-analytic”: if they can be for-
mulated meaningfully, they cannot be false. Consequently, he was against
the view that all empirical sentences – protocol sentences included – are
hypothetical (he dubbed this view as “hypothetism”).10
Juhos defined the terms “physical” and “mental” relying on this con-
ceptual frameswork. Accordingly, a term is “mental” if it can also appear
in a Konstatierung-sentence: names of psychophysical objects like “pain”,
or verbs referring to experiencing, as “feel”, “see”, “hear”, etc., are such,
for example. Such terms may also appear in empirical-hypothetical sen-
tences, as e.g., “While burning one feels pain”. In contrast, a term is
“physical”, if it can appear only in empirical-hypothetical sentences, i.e.
not in Konstatierungen-sentences.
Now let’s have a look at the main views about the psycho-physical pro-
blem in the Vienna Circle. In modern philosophy, the problem of how the
mind relates to the body originates from Descartes. In the original Car-
tesian discourse the most perplexing question seemed to be the issue of
interaction: how a causal connection is possible between two such radically
different substances as the body and the soul?
Later, however, the importance of this problem diminished, at least
for those with empiricist leanings; for according to the Humean-empiricist
approach to causation, the nature of the relata of causal relations is irre-
levant.
10 Neurath, Carnap, Popper and Morris advocated “hypothetism” from the

early 1930s.
In the second half of the 19th century, the advocates of “scientific

philosophy”11 embraced psychophysical parallelism as the dominant con-
ception concerning the mind-body relation. This broad view, according
to which every mental phenomenon is conjoined with some bodily (i.e.,
brain or neural) phenomenon, was both a scientific-empirical hypothesis
and a philosophical (metaphysical) theory. Among the proponents of par-
allelism, two main strands may be distinguished. One group advocated
a realist version of neo-Kantianism, according to which the mental and
the physical are two aspects of the one underlying reality (Riehl’s views
may be an example). The alternative was some sort of phenomenalism-
cum-double language view (as e.g. of Mach and Avenarius). According to
the latter, there is only one type of phenomena, which we may describe
(or rather, construct) in two ways: either as mental (phenomenal) or as
physical. These two ways of description, however, do not signify an onto-
logical difference: there are only one sort of – neutral – elements (like, for
example, Mach’s Empfindungen).12
In the “classical period” of the Vienna Circle (i.e., from the mid-
20s until the mid-30s) these two views, or rather, their logical positivist
offsprings, were the major competing conceptions about the psychophysi-
cal relation.13 Here I follow Juhos’ classification of the different versions
of parallelism. According to him, parallelism comes in three main varie-
ties: (a) the traditional metaphysical parallelist views, (b) empirical-cum-
logico-linguistic views, (c) purely logico-linguistic (or syntactic) views. (A)
may be divided further into two types. Either (1) the simultaneously in-
stantiated mental and physical processes are different sorts of realities,
which are harmonized by a transcendent cause, God (Leibniz’s view, for
11 This label denotes some scientifically oriented philosophers or philosopher-

scientists, as e.g. Fechner, Mach, Avenarius and Riehl, whose views were an
important source of the logical positivists’ ideas.
12 There were, of course, adversaries of parallelism as well. Critics from the tra-
dition of Lebensphilosophie included Dilthey and Husserl, but James was al-
so a firm opponent. See Michael Heidelberger, “The Mind-Body Problem in
the Origin of Logical Empiricism: Herbert Feigl and Psychophysical Paral-
lelism”, in: Paolo Parrini, Wesley C. Salmon, and Merrilee H. Salmon (eds.),
Logical Empiricism: Historical and Contemporary Perspectives. Pittsburgh,
PA: University of Pittsburgh Press 2003, pp. 258-9.
13 There were some others as well, see e.g. Kaufmann’s view in Felix Kauf-
mann, Methodenlehre der Sozialwissenschaften. Wien: Springer 1936.
104 Gergely Ambrus
example). Or (2) the parallel mental and the physical phenomena are dif-
ferent types of appearances of one and the same underlying reality (the
Kantian and neo-Kantian approaches).
The traditional metaphysical views can be discarded, according to
Juhos, for their main motivation was to by-pass the question concerning
the nature of the causal nexus between mind and body, which, for them,
seemed utterly problematic. For Juhos and for all logical positivists, ho-
wever, this was hardly a problem, for they adopted a broadly Humean
conception of causation14 – hence, for them, the metaphysical parallelist
solutions to the problem of psychophysical interaction were otiose.
The empirical-cum-logical approach holds that whenever there is a
“mental” event, there is also an adjoining “physical” event, and these two
are independent from each other, in the following sense. It is an empirical
fact, that certain types of “mental” events are co-instantiated with certain
types of “physical” events: it could be otherwise, and it is the task of
empirical science to discover the correlations between the “mental” and
the “physical” events.
According to the purely logico-linguistic view, when it seems that
there are two independent “mental” and “physical” processes paralleling
each other, there is, in fact, only one process, which may be described in
two languages, either in “physical”, space-time language, or in “mental”,
phenomenal language. The sentences describing “mental” and “physical”
phenomena may be translatable into each other; if a mental sentence is
translatable into a physical sentence (or a set of physical sentences), then
the content of the mental and the physical sentence(s) is identical.
According to the empirical-cum-logico-linguistic view, the issue whe-
ther certain mental phenomena are related to certain physical (i.e., brain
or behavioral) phenomena has both empirical and logico-linguistic aspects.
The empirical task is to find out whether for an observed series of mental
events (i.e., a series of Konstatierungen), there exist another parallel se-
ries of events – which may turn out to be “physical”. The logico-linguistic
14 Gustav Bergmann writes: “as long as one sticks to cautious generalities

all Logical Positivists could still agree that they (a) hold Humean views
on causality and induction [. . . ]”. (in: “Logical Positivism”, in: Vergilius
Ferm (ed.), A History of Philosophical Systems. New York: Philosophical
Library 1951, pp. 471-82. Reprinted in: Gustav Bergmann, The Metaphysics
of Logical Positivism. Westport, Conn.: Greenwood Publishers 1954, p. 2.
issue is to find out whether the series parallel with the observed “mental”
series is “physical”, or not.
According to the purely logico-linguistic view, there is no empirical
issue concerning the relation between the mental and the physical, as
the meaning of mental sentences is identical with the meaning of certain
physical sentences. For example, according to the logical behaviorist ver-
sion of the purely logico-linguistic view, the content of “Mr. N. has a red
sensation” is that Mr. N.’s body is in a red sensing state. Hence no empi-
rical issue could arise whether N’s having red sensations, i.e., a “mental”
state, is conjoined with a “physical” state, namely N’s body being in a
red-sensing state. For N.’s having a red sensation means that N’s body is
in a red-sensing state; thus there is no logical possibility such that N has
a red sensation but N’s body is not in a red-sensing state.
The main difference between the empirical-cum-logico-linguistic and
the purely logico-linguistic view is that the former accounts for parallelism
as there being two independent series of observed events between which
certain empirical relations may hold (or may not), while according to the
latter, the issue whether such an empirical relation exists or not, cannot
arise. This question cannot even be formulated meaningfully, since the
meaning15 of a mental sentence is identical with the meaning of the phy-
sical sentences into which it can be translated; hence one cannot formulate
the question of how the mental and the physical series are related.
In the first part of the 1930s, there was an ongoing debate about a series
of interrelated issues in the Vienna Circle, including the correct account
of protocol sentences, whether they should be physical or phenomenal,
whether they are incorrigible, etc.; the related issues concerning the cor-
rect notion of truth; the criteria of empirical meaningfulness; and so on.
For my purposes, i.e., to illuminate Juhos’ view on the psychophysical
problem, it is revealing to treat the protocol sentence debate as a part, or
an aspect of the physicalism debate. It is clear why: the doctrine of physi-
calism is fundamental for the interpretation of the protocol sentences and
the interpretation of the psychophysical relation as well; they both turn
15 The cognitive or empirical meaning, the emotive and the pictorial meaning
is another issue.
106 Gergely Ambrus
on whether physical language should and could be the universal language

of sciences.
The Physicalism Debate: Its Origin and the Main Issues

The doctrine of physicalism in the logical positivist movement may be
characterized by two tenets: (i) physical language is the universal langua-
ge of science, (ii) all scientific laws and explanations can be reduced to
physical laws and explanations. The former was accepted by all physica-
lists, the latter was much more controversial.16 Here I only deal with (i),
the language thesis.
Physicalism, as it is well-known, was put forward first by Neurath, but
quickly adopted by Carnap as well.17 The doctrine had several formulati-
ons and aspects, and grew out, partly at least, from the opposition to Car-
nap’s Aufbau idea, according to which the constructional system of scien-
tific concepts should be based on private experience (or more precisely, on
the recollection of similarity of experiences (Ähnlichkeitserinnerung)).18
There are, of course, several interpretations of the aims of the Aufbau pro-
gramme, as well as of its failures,19 but these need not concern us here.
16 For example, Carnap, still in 1963, considered it only as a hypothesis, not

sufficiently grounded as yet (see his “Intellectual Autobiography”, in: Paul
A. Schilpp (ed.), The Philosophy of Rudolf Carnap. La Salle: Open Court
1963, pp. 1-84., and “Reply to Feigl”, ibid. pp. 859-60.); Neurath definitely
rejected it.
17 For a detailed history, see Thomas E. Uebel, Overcoming Logical Positi-
vism from Within. Neurath in the Vienna Circle’s Protocol Sentence Deba-
te. Amsterdam–Atlanta: Rodopi 1992. Focusing on Wittgenstein’s claims to
priority and the “plagiarism issue” with Carnap, see Jaakko Hintikka, “Lud-
wig’s Apple Tree: On the Philosophical Relations between Wittgenstein and
the Vienna Circle”, In: Friedrich Stadler (ed.), Scientific Philosophy: Orig-
ins and Developments. Dordrecht: Kluwer 1993, pp. 27-46.
18 Neurath’s motivation for physicalism were manifold, and the issue of pro-
tocol sentences was just one among them. However, for our topic, i.e., the
debate between a Schlickian/Juhosian “phenomenalist/parallelist” and the
Neurathian/Carnapian physicalist/logical behaviorist approach to the psy-
chophysical relation, the fundamental issue concerning physicalism is the
meaning of mental sentences, on which the phenomenalists (“empiricists”)
and physicalists radically disagreed.
19 As e.g. the debate between the empiricist and the Kantian interpretations.
What is of interest for us now, and that is fairly uncontroversial, that one
of the major objections against the Aufbau was that its constructional
system cannot provide for the intersubjectivity of the meaning of scien-
tific concepts.20 It is also clear that choosing a physicalist base eo ipso
solves this problem, physical concepts being intersubjective by definition.
To see the connections between Carnap’s and Neurath’s physicalism
and the debate on the psychophysical problem, I shall characterize phy-
sicalism somewhat more extensively. It is well-known that the physicalist
doctrine, from its inception in 1929 or 193021 underwent several chan-
ges. Here I present what may be seen as the full-blown radical physica-
list/logical behaviorist view of Carnap.22
(UnivPhys) Physical language is the universal language of science.

(UnivPhys1) The meaning of phenomenal sentences is identical with the
meaning of some physical sentences. (In the material mode: phenomenal
sentences refer to physical (bodily/behavioral) states.)
(UnivPhys2) The observational base sentences (protocol sentences) are
also physical sentences.
(AntiFund) There are no incorrigible (non-hypothetic) observational sen-
tences. Protocol sentences are also corrigible, they can be rejected.
(Intersubjectivity) Physical language is the only intersubjective language.
(Syntacticism) For a statement to be scientific, it must be possible to for-
mulate it in the formal mode.
Carnap’s early view was different, as presented in his “Physikalische Spra-

che”,23 where he still held that protocol sentences are incorrigible and to
be formulated in phenomenal language. This early view was already criti-
cized by the phenomenalists side, namely by Duncker and Zilsel.24 But it
20 Put forward by Neurath and also by Neider. Cf. e.g., Uebel, op. cit. pp.
93-96.
21 See e.g., Uebel, op. cit.
22 See Rudolf Carnap, “Psychologie in physikalischer Sprache”, in: Erkenntnis
3, 1932/33, pp. 107-142.
23 Rudolf Carnap, “Die physikalische Sprache als Universalsprache der Wis-
senschaft”, in: Erkenntnis 2, 1932, pp. 432-465.
24 See Karl Duncker, “Behaviorismus und Gestaltpsychologie. Kritische Be-
merkungen zu Carnap’s ‘Psychologie in physikalischer Sprache’”, in:
108 Gergely Ambrus
was also criticized “from the left”, by Neurath. Carnap accepted Neurath’s
criticism, and radicalized his physicalism: he claimed that protocol sen-
tences are also physical sentences, and that there were no incorrigible
observational sentences at all.25
In reply to the radical physicalist doctrine, Schlick published his “Über
das Fundament der Erkenntnis” (in 1934), in which he attacked prima-
rily the coherentist theory of truth suggested by Neurath. In the same
issue of the Erkenntis appeared Juhos’ first contribution to the debate,
his “Kritische Bemerkungen zur Wissenschaftstheorie des Physikalismus”.
The next important move was the publication of Neurath’s “Radikaler
Physikalismus und ‘wirkliche Welt’” in the end of 1934, in which he at-
tacked Schlick’s ideas put forward in the “Fundament” article. Schlick
found Neurath’s paper absurd, and did not intend to reply directly to
Neurath at all. In early 1935, however, also Hempel published a paper, “On
the Logical Positivists Theory of Truth”, in which he supported Carnap’s
and Neurath’s physicalism, and criticized Schlick’s “Fundament” article,
as well as Juhos’ “Kritische Bemerkungen”. To this paper both Schlick
and Juhos answered, with “Facts and Propositions” and ,,Empiricism and
Physicalism” respectively (both in 1935). Hempel replied to Schlick with
“Some Remarks on ‘Facts and Propositions’” (June, 1935), and to Juhos
with “Some Remarks on Empiricism” (January, 1936). Schlick’s counter-
reply was “Sur les ‘Constatations’” (at the end of 1935), Juhos’ “Some
Modes of Speech of Empirical Science” (June, 1936).26
Erkenntnis 3, 1932/33, pp. 162-176., and Edgar Zilsel, “Bemerkungen zur

Wissenschaftslogik”, in: Erkenntnis 3, 1932/33, pp. 143-161.
25 See his “Psychologie in physikalischer Sprache”.
26 Otto Neurath, “Radikaler Physikalismus und ‘wirkliche Welt’”, in: Erkennt-
nis 4, 1934, pp. 346-362. Moritz Schlick, “Über das Fundament der Erkennt-
nis”, in: Erkenntnis 4, 1934, pp. 79-99.; “Facts and Propositions” in: Ana-
lysis 2, 1935a, pp. 65-70.; “Sur les ‘Constatations’”, in: Sur le fondement de
la connaissance (Actualités Scientifiques et Industrielles, No. 289.). Paris:
Herman et Cie 1935b. English transl.: “On Affirmations”, in: Henk L. Mul-
der, Barbara F. B. van de Velde-Schlick (eds.), Philosophical Papers Vol.
II, 1925–1936, Dordrecht and Boston: Reidel 1979, pp. 407-413. Carl G.
Hempel, “On the Logical Positivists’ Theory of Truth”, Analysis 2, 1935,
pp. 49-59.; “Some Remarks on ‘Facts and Propositions’”, Analysis 2, 1935,
pp. 93-96.; “Some Remarks on Empiricism”, Analysis 3, 1936, pp. 33-40.
For the publication data of Juhos’ papers, see note 3.
Juhos’ Account: Description and Evaluation

As already mentioned, the issue connecting physicalism and the psycho-
physical problem is whether physical language could and should be the
universal language of science, i.e., whether the mental language is part of
the physical language, or not. If the physicalist language thesis should be
rejected, as Juhos argued, then the physicalist approaches to the psycho-
physical relation have serious problems and Carnap’s logical behaviorism
of the early 1930s ought to be discarded.27 In the following, I will sketch
Juhos’ own view on the meaning of mental (phenomenal) terms and his
arguments against the Carnapian view that mental sentences are content-
identical with physical sentences.
Juhos’ account of the psychophysical relation and of the meaning of
mental sentences (both protocol and system sentences) are rather close to
Schlick’s, which he proposed in the 1930s,28 and also to Carnap’s earlier
views put forward in the Aufbau.
Carnap’s Aufbau view may be summarized as follows. (i) As for the ba-
se language: the constructional base is phenomenal, i.e., the content of the
basic observation sentences are the subject’s experiences. (ii) Concerning
the psychophysical relation, a constructional and a metaphysical problem
ought to be distinguished. As for the former: according to the paralle-
list assumption, when there is a series of observations (Konstatierungen)
of mental events (e.g. representations of sounds, imagining a melody),
a parallel series of observations (Konstatierungen) may always be made
(namely, visual experiences of certain brain events). The constructional
problem is to show that the elements of the parallel series are such that
they allow for the construction of “physical” events. The metaphysical
problem concerns the nature of the relation between the two parallelly
observed series: whether they are representations of one and the same
27 Along with the other physicalist views some Vienna Circle members put
forward, e.g., the consciousness-brain identity theory of Feigl. I shall discuss
this later.
28 Cf. Schlick 1934, Schlick 1935a, Schlick 1935b, and “De la relation entre les
notions psychologiques et les notions physiques”, in: Revue de la Synthese
10, 1935, pp. 5-26. English transl.: “On the Relation between Psychological
and Physical Concepts”, in: Herbert Feigl and Wilfrid Sellars (eds.), Rea-
dings in the Philosophical Analysis. New York: Appleton-Century-Crofts
1949.
110 Gergely Ambrus
process (as according to neo-Kantianism), or they are two different sort

of processes (dualism). Now, this question is meaningless according to the
Aufbau, for it cannot be formulated meaningfully: one cannot construct
the notion of the relation between the two series.
Juhos presents a view rather close to this Carnapian approach. In fact,
Juhos’ view is in certain respects closer to the Aufbau conception than to
Schlick’s position in the 1930s.29 Schlick then held that the parallel mental
and physical observations are the observations of the same process. This
is a contingent fact, however, which could be otherwise. Now, on the one
hand, this goes clearly against Carnap’s logical behaviorist physicalism,
according to which it is of logical (or linguistic) necessity that the obser-
vations of mental events are ipso facto observations of physical events,
since the content of a mental report is the same as that of some physical
reports: mental reports are about physical (brain or behavioral) events.
Hence, according to logical behaviorism, it is not a contingent fact that
the observational content of physical and mental series is the same. Ju-
hos accepted the contingency claim, so in this respect he followed Schlick.
But on the other hand, Schlick’s account of the 1930s was not so entirely
ametaphysical, as the Aufbau, and also Juhos, strived to be. It echoed
Schlick’s earlier pre-positivist (critical-realist) view on the psychophysical
problem, which, according to the Aufbau, was a scientifically meaningless
metaphysical view as well. On this latter issue Juhos seems to be closer
to Carnap’s Aufbau view than to Schlick.
Now, Carnap’s radical physicalist/logical behaviorist view in the early
30s, which Juhos criticized, may be summarized as follows.30 (i) As for the
base language: the protocol sentences are also physical sentences. Hence
protocol sentences are also corrigible and not immune to error. (ii) As
for the psychophysical relation: there is only one type of events that may
occur in observation. The sentences describing observations are physical
sentences, the meaning of a mental report is identical with the meaning
of physical reports: for example, “Mr. N. senses red” means that Mr. N.’s
body is in a red-sensing state. The question concerning the nature of the
psychophysical relation is senseless, of course, but for a different reason,
as in the Aufbau. It is senseless to ask what is the relation between the
29 Ibid.
30 See “Psychologie in physikalischer Sprache”.
reference of certain mental terms and physical terms, if they are by stipu-
lation the same. (And, it is also senseless to claim that they are identical
in the metaphysical sense, i.e., that a metaphysical identity relation holds
between them.)
Juhos argued both against (i) and (ii). Let’s examine them in turn.
Juhos’ Arguments Against the Physicalist Account

of Protocol Sentences
Juhos took over some of Schlick’s argument, for example the following.
(1) Knowledge must have an indubitable foundation, the coherentist
account of knowledge (proposed e.g. by Neurath) is untenable.31 Hence
the sentences describing the basic observations, which ought to provide
the foundations of empirical knowledge, cannot be physical sentences, for
physical sentences are corrigible, one may err about the content of a sen-
tence describing some physical state of affairs. Hence, they cannot serve
as the indubitable foundation of knowledge.
(2) The physicalist wing of the Circle (Neurath, Carnap) and also
Popper put forward arguments against the Konstatierung account of pro-
tocol sentences (or base sentences, in Popper’s terminology), presented
by Schlick. Juhos provided some counterarguments and tried to develop
Schlick’s account somewhat further.
(2.1) One of Neurath’s major objections against the idea, that empiri-
cal knowledge should be based on Konstatierungen, was, in contemporary
terminology, that propositions may only be justified by propositions; they
cannot be justified by experiences. Schlick sketched an answer in “Facts
and Proposition”, which were modified and developed further in “Sur les
‘Constatations’”.
(2.2) Another important objection, which can be directed both against
the phenomenal account of protocol sentences and mental sentences in
general, is that on such an account they have no intersubjective meaning
– and this is a fatal weakness, since the meaning of scientific concepts
ought to be intersubjective.32
31 See Schlick’s “Über das Fundament der Erkenntnis”.

32 The objection was first proposed by Neurath against the base-choice of the
Aufbau, i.e., that the constructional system starts from experiences. Though
112 Gergely Ambrus
Juhos’ answer drew on Schlick’s. The upshot of these arguments is

that the content of experiential reports may be intersubjective, even if we
understand them as referring to experiences, not to bodily states. (I shall
come back to this later).
Juhos’ Arguments Against the Physicalist Account of

Mental Sentences in General
I shall consider three objections of Juhos against the physicalist account
of mental reports, which may be dubbed as the argument from standard
use, the argument from ignorance, and the argument from corrigibility.33
The Argument from Standard Use
According to this argument, the physicalist/behaviorist interpretation of

sensation terms is contrary both to everyday and to scientific usage. We
do not just use sensation terms as if they referred to behavioral or brain
states, neither in everyday parlance, nor in scientific psychology. Now,
this objection is easy to meet on Carnapian or Neurathian grounds. Pre-
sumably, their answer would have been that ’folk’ and actual scientific
use of mental terms is irrelevant, for the criterion of cognitive significance
is normative, not descriptive. If one is to make real scientific psychology,
one ought to use an intersubjective nomenclature; psychological theories
the content of which cannot be formulated with intersubjectively under-
standable notions, ought not to be considered as scientific theories at all.
The Argument from Ignorance
The argument from ignorance is harder to disarm. According to this ob-

jection, if the physicalists are right about the meaning of mental reports,
then in most (or all) cases we do not know what mental states we are in.
For example, in most cases people do not know whether they are in pain
or not, because in most cases they do not know the physical (behavioral
or neural) states in which they are when in pain. Hence they do not know
Carnap was at pains to provide for the intersubjectity of the concept-system

of the Aufbau, he was not successful. Cf. note 20.
33 See “Physicalism and Empiricism”.
whether the state they are in is a pain state, or not.34 Now, this sounds
rather counterintuitive.
There may be several counter-strategies. First, one might bite the
bullet, and say that it is true that we never know what mental state
we are in, it is only an illusion that we do. For the real psychological
state types are behavioral or brain state types, which are not accessable
to introspection. (This does not mean to deny that there are sensations,
only that they are not types which can be identified solely by having them,
or introspecting them.) There are many ways to spell out such a view, but
Schlick and Juhos surely did not accept it anyway: their account of the
Konstatierungen embraces the idea that we can identify our sensations
(or at least the simple ones) just by having them, (we cannot even be
mistaken about them, when having them). So here we are at a dialectic
stalemate.
It might seem that there is a plausible strategy to counter the argu-
ment from ignorance, which can be reconciled with physicalism - though
it fits the mind-brain identity theory version of physicalism better than
logical behaviorism. Among the logical positivists, Schlick advocated an
identity-thesis e.g. in his “De la relation entre les notions psychologi-
ques et les notions physiques”, and also Feigl proposed a similar view
in his “The ‘Mental’ and the ‘Physical’” (1958), and in “Mind-Body not
a Pseudo-Problem”(1960).35 Smart in his “Sensations and Brain Proces-
ses” (1959),36 the manifesto of consciousness-brain identity theory (aka
central state materialism), replied to the argument from ignorance re-
lying on Frege’s sense/reference distinction. According to this proposal,
certain mental and physical terms have different senses, but nevertheless
refer to the same entity, as e.g., “pain” and “c-fiber firing” (similarly as
34 Ibid.
35 An important difference being that Feigl’s view was emphatically a meta-
physical view. See Schlick, “De la relation entre les notions psychologiques
et les notions physiques”; Herbert Feigl, “The ’Mental’ and the ’Physical’”,
in: Herbert Feigl, Michael Scriven and Grover Maxwell (eds.), Minneso-
ta Studies in the Philosophy of Science, Vol. II. Minneapolis: University
of Minnesota Press 1958, pp. 370-497., and “Mind-Body, not a Pseudo-
Problem”, in: Sydney Hook (ed.), Dimensions of Mind. New York: New
York University Press 1960, pp. 24-36.
36 Jack J. C. Smart, “Sensations and Brain Processes”, in: The Philosophical
Review, 1959/2, pp. 141-156.
114 Gergely Ambrus
“Morningstar” and “Eveningstar”). If this is so, then we may counter

the argument from ignorance by saying that we know in what state we
are in under one description (i.e., under an experiential description), but
we do not know this under another (i.e., under a neural description). (It
should be mentioned that the sense/reference distinction in defence of the
consciousness-brain identity theory was also used by Feigl (and others)
much earlier: already in Feigl 195037 he refers to Frege’s example of the
identity of the Morning Star with the Evening Star as analoguous with
the identity of sensations with brain states.)38
I think, however, that such a strategy to establish the identity of cons-
ciousness with brain states is unlikely to succeed, for the following reason.
If we hold, as the “right wing” of the Circle certainly did, that sense-
qualities or qualia are constitutive of sensations (as e.g., the painfulness
of pain), then this at least a dualism of properties follows. For clearly, the
qualia of pain, the property of painfulness, cannot be identical with the
property of being a c-fiber firing. (Unless we adopt a structuralist view
of physical terms of some sort, according to which physical terms refer
to dispositional, not to categorical properties. In this case, it is prima
facie conceivable that the categorical base of certain physical – disposi-
tional – properties are certain qualia; hence certain physical terms (e.g.,
“c-fiber firing”) select the same categorical property to which sensation
terms (“pain”) refer. Similar views were also present in logical positivism:
Schlick’s early account of physical theories was a version of structuralism.
Russell’s view in his Analysis of Matter (1928) and also in Human Know-
ledge (1948) was also structuralist; in the latter, he explicitly combined it
with a consciousness-brain identity theory.)39
37 Herbert Feigl, “The Mind-Body Problem in the Development of Logical

Empiricism”, in: Revue Internationale de Philosophie 11, 1950, pp. 64-83.
38 It should also be emphasized that such an answer is clearly not in the vein
of Carnap’s and Neurath’s view of the early 1930s. At that time Carnap
was still before his “semantic turn”, he formulated his intension/extension
distinction later. Thus, he would not have accepted a view which relies on
something like the concept of the Fregean sense.
39 Cf. Moritz Schlick, Allgemeine Erkenntnislehre. Berlin: Springer, 1918.
Bertrand Russell, The Analysis of Matter. London: Kegan Paul, Trench,
Trubner 1927, and Human Knowledge. Its Scope and Limits. London: Al-
len and Unwin 1948. Feigl also drew attention to the similarities between
these Russellian and Schlickian views; see his “Russell and Schlick: A Re-
It should also be noted, however, that Smart’s conception was diffe-

rent. He analysed the meaning of “pain” by reference to the characteristic
stimuli which cause pain (and later, following Armstrong, the causal role
of pain). Such a view does not have to face directly the problem for Feigl
(namely that it leads to property-dualism), but only on pain of leaving
out the experiential aspect of pain from the meaning of “pain” altogether,
which is rather counter-intuitive.
So, I think, Juhos’ argument from ignorance does not allow for such
an easy answer, as it might first seem.
The Argument from Corrigibility

The argument from corrigibility deserves deeper scrutiny, so I shall discuss
it at some length. First I shall present Juhos’ argument from corrigibility
against the physicalist interpretation of sensation reports, Hempel’s rep-
lies to Juhos, and Juhos’ counter-replies. Then I shall discuss a close kin
of Juhos’ objection, put forward by Kurt Baier against the consciousness-
brain identity theory, some 30 years later. I shall go into some detail and
show that the Juhos-Baier argument is directly relevant to some contem-
porary views and debates on the consciousness-brain relation, and also
that, the many diverse strategies to disable it notwithstanding, a strong
case can be made for this argument.
Juhos’ argument from corrigibility against Neurath’s and Carnap’s
physicalistic interpretation of sensation reports, may be summarized as
follows. The physicalistic interpretation of sensation reports (e.g., that
the meaning of “Mr. N. has pains” is that Mr. N.’s body is in a pain
state) implies that honest reports of someone’s own sensations can be
corrected by external observers. This is absurd, according to Juhos. For
think about the absurdity of a situation, when I reassure someone who
honestly complains that he is in pain, by saying: “Cheer up, you are
mistaken, you have no pains. I checked your nervous system!”40
Hempel, in his “Some Remarks on Empiricism” came up with two
answers to Juhos’ argument. He claimed that (i) someone may lie about
his sensations, and this report can be corrected by observing the person’s
markable Agreement on a Monistic Solution of the Mind-Body Problem”,
in: Erkenntnis 9, 1975, pp. 11-34. However, there are other grave problems
for such views, which I cannot discuss here.
40 See Juhos, “Empiricism and Physicalism” p. 85.
116 Gergely Ambrus
bodily states and behavior. Even if someone claims that he has no pains
when putting his hand into the flames, the observation of his bodily states
and behavior may speak against him, and such evidence is surely legiti-
mate grounds to correct someone’s reports. (ii) He also put forward cases
of corrigible judgements about one’s own mental states. For example: I
may honestly believe that “I change my political party out of conviction”,
or that “I have not got the least desire for material advantages”. But, in
fact, I am mistaken: unbeknownst to me, I had some other motivations
for changing my party, e.g., to please a person who is important for me,
or something else; or I do desire material advantages, I just suppress or
conceal this desire so successfully that I do not realize it.41
Juhos’ answers were the following.42 As for (i): lying is irrelevant. Of
course, dishonest reports can be falsified by contrary behavioral evidence.
But the relevant issue is whether someone could err about his own mental
states in his honest reports. The cases mentioned in (ii) are honest re-
ports. However, as Juhos claims, the logic of simple sensation statements
differs from the logic of the statements Hempel cites. The latter are about
unconscious motives, which are not directly given. First person sensation
reports, as e.g., “I feel pain”, are such that understanding them by the
subject simply allows him to know whether they are true or not. This is
not the case, however, in Hempel’s examples: the fact that someone un-
derstands what it means that one has a certain unconscious motive, e.g.,
understands the statement “I desire material advantages unconsciously”,
does not imply that he knows whether the statement is true or not.
Now, we shall elaborate the argument somewhat further. It is note-
worthy that Kurt Baier put forward basically the same argument (as his
original) against Smart’s consciousness-brain identity theory in 1962 (in
his “Pains” and “Smart on Sensations”).43 I shall present Baier’s argu-
ment, and examine Smart’s answers to it as well. This is interesting for us,
because these answers can be targeted against Juhos’ original argument
as well; hence seeing how they fare is important for adjudicating Juhos’
position.
41 See Hempel, “Some Remarks on Empiricism”, p. 36.

42 Cf. Juhos, “Some Modes of Speech of Empirical Science”, p. 68.
43 Kurt Baier, “Pains”, in: Australasian Journal of Philosophy 40, 1962,
pp. 1-23, and “Smart on Sensations” in: Clive Vernon Borst (ed.), The
Mind/Brain Identity Theory. London: Macmillan 1970, pp. 95-106.
Baier’s argument may be summarized as follows.

(1) A honestly judges that he is in pain (he feels pain).
(2) B observes that A is not in the bodily/behavioral state which is cha-
racteristic of being in pain.
Hence
(C) B legitimately concludes, that A has no pains, i.e., B can legitimately
correct A’s own judgement about his being in pain.
Smart had two answers to this argument.44 According to the first, the “it
will never happen” answer,45 (1) and (2) are never actually true together,
hence the conclusion does not follow. Smart first granted that if such a
case were actually true, this would refute his consciousness-brain identity
theory.46
Now, my objection to Smart is that it is irrelevant whether (1) and
(2) ever actually happens, i.e., whether it is ever actually true that so-
meone honestly judges that he feels pain, but his body is not in the bodi-
ly/behavioral state which is characteristic of being in pain. For the possi-
bility of such a case ((1) and (2) being true together) is sufficient for the
argument to go through. And Smart (and the logical positivists as well)47
allowed for such a possibility. To see this, we may alter Baier’s argument
as follows.
(1 and 2) It is possible that A honestly judges that he is in pain (feels
pain) and B observes that A is not in the bodily/behavioral state which
is characteristic of being in pain.
(3) If A is incorrigible about his being in pain, then if he judges honestly
that he is in pain, then he is necessarily in pain. So it is impossible that
A judges honestly that he is in pain, but he is not in pain.
(4) If A is incorrigible about his being in pain, and if his pain state is
44 Cf. Smart: “Brain Processes and Incorrigibility”, in: Clive Vernon Borst
(ed.) The Mind/Brain Identity Theory. London: Macmillan 1970, pp. 107-
109.
45 Ibid.
46 Ibid.
47 To see this, it is enough to turn to Schlick’s and Juhos’ argument for the
intersubjectivity of the meaning of sensation terms. They argue that it is
(logically) possible that pains are not correlated with the bodily states they
are actually correlated with: they may correlate with some states of another
body, or with no bodily state whatsoever. I shall discuss this later.
118 Gergely Ambrus
identical with a bodily state of his, then he is also incorrigible about his
being in that bodily state. Hence it is impossible that A judges honestly
that he is in pain, but A is not in the bodily state characteristic of being
in pain.
(5) So if it is possible that (1) and (2) are true together, then
either
(C1) A’s bodily state characteristic of being in pain is not identical with
pain,
or
(C2) A is not incorrigible about his being in pain.
Hence, it is irrelevant whether (1) and (2) are ever actually true together,
for (C1) or (C2) follows solely from the possibility of (1) and (2) being
true together.
Smart’s other reply was that we are corrigible about our own sensa-
tions. His rather sketchy argument was the following. It is arbitrary or
artificial to make the sincerity of someone’s reports of his own states the
criterion of its correctness. For such a move, Smart claims, “would ensure
that a man who sincerely reported a bruised toe nail had a bruised toe
nail” (and, of course, this is false). However, it is clear what Juhos could
have answered: “I have a bruised toe nail” is a hypothetic-empirical state-
ment, not a non-hypothetic-empirical one. Hence the sincere assertion of
the statement “I have a bruised toe nail” need not and does not imply
that the statement is true – in line with Juhos’ theory.
Against Incorrigibility
Now, the question of the corrigibility or incorrigibility of first-person sen-
sation reports is a many-facetted issue, with many arguments and aspects
Juhos was in no position to consider in the mid-1930s. As an illustration,
here are some argument-types or argumentative directions indicating that
we have no incorrigible knowledge about any sort of sensations of ours.
(a) An experience cannot justify a proposition or a belief, propositions or
beliefs can only be justified by other propositions or beliefs (Neurath, Sel-
lars, Bonjour, Davidson, McDowell). Hence the idea of having incorrigible
knowledge about our sensations is untenable, it rests on conceptual con-
fusion. As we may put it: there cannot be incorrigible beliefs about our
sensations, which would be justified just by having the sensations, for no

belief whatsoever could be justified just by having experiences.
(b) Wittgenstein’s challenge. Incorrigible knowledge is a contradiction in
terms: one cannot know what one is incorrigible about.
(c) The identification of any sensation relies on external factors, external
objects (e.g., the characteristic stimuli of the sensations), or the linguistic
communities’ use of the sensation term, to which we may have no access.
Hence we are not incorrigible about our own sensations.
Now, I shall sketch some answers to these arguments which may be related
to Juhos’ position.
1. Against (a): Experience can justify beliefs/propositions
The inconsistent triad (Sellars)

Sellars put forward the so-called inconsistent triad argument, aiming to
show that experience cannot justify beliefs.48 Accordingly
A. x senses red sense content s entails x non-inferentially knows that s is
red.
B. The ability to sense sense contents is unacquired.
C. The ability to know facts of the form x is F is acquired.
A and B together entail not-C; B and C entail not-A; A and C entail
not-B. (Sellars, op.cit., section 6).49
Chalmers50 answers the following: we can deny A. Having an experience
of red sense content by x does not entail that x non-inferentially knows
that s is red. For
(i) Knowledge is to be analysed in terms of belief.
(ii) But having an experience does not imply having any beliefs.
48 See Wilfrid Sellars, “Empiricism and the Philosophy of Mind”, in: Herbert
Feigl and Michael Scriven, (eds.), Minnesota Studies in the Philosophy of
Science, Volume I: The Foundations of Science and the Concepts of Psycho-
logy and Psychoanalysis. Minneapolis: University of Minnesota Press 1956,
pp. 253-329.
49 Here I follow Chalmers’ brief reconstruction of Sellars’ argument, see Da-
vid Chalmers, “The Content and Epistemology of Phenomenal Belief”, in:
Quentin Smith and Aleksandar Jokić (eds.), Consciousness: New Philoso-
phical Perspectives. Oxford: Oxford University Press 2003, pp. 220-272.
50 Chalmers ibid.
120 Gergely Ambrus
(iii) However, if a (first person present tense phenomenal) belief accom-

panies the having of the relevant experience, this belief is justified just by
having this belief.
The justificatory dilemma (Bonjour, Davidson, McDowell) 51

We may reconstruct the justificatory dilemma as follows:
(1) There can be no inferential relation between a non-conceptual experi-
ence and a belief, as inference requires connections within the conceptual
domain.
(2) But a mere causal relation between an experience and a belief cannot
justify the belief.
hence
(3) Non-conceptual experiences cannot justify beliefs.
According to Chalmers,52 we may accept (1) and (2). But we needn’t
accept (3), for there is an alternative form of justification: not justifica-
tion by inference and not justification by causation, but justification by
(partial) constitution. According to this, the content of the belief I sense
red now 53 (or There is redness now ) involves/is constituted (partly) by
the presence of an instance of phenomenal redness. So if I am able to
form a genuine belief I sense red now, this implies that an instance of
phenomenal redness is present. (This is a special feature of phenomenal
concepts/belief. Cf. e.g. the belief I sense water. Arguably, its content is I
sense H 2 O, hence it is possible to have the honest belief I sense water, and
be mistaken about it; namely, in cases when I have a watery experience,
which was caused not by H2 O, but by something else, e.g. XYZ.)
Now, the Schlick-Juhos view is analoguous with Chalmers’ in import-
ant respects. Juhos’ formulation is the following: if I make an honest Kon-
statierung-statement, e.g. “There is pain”, correctly, (i.e. in accordance
with Schlick’s and Juhos’ meaning theory, using the word “pain” as re-
51 See Lawrence BonJour, “Can Empirical Knowledge Have a Foundation?”,

in: American Philosophical Quarterly 15, 1978, pp. 1-13.; Donald Davidson,
“A Coherence Theory of Truth and Knowledge”, in: Ernest Lepore (ed.),
Truth and Interpretation: Perspectives on the Philosophy of Donald David-
son. Oxford: Blackwell 1986, pp. 307-19.; John McDowell, Mind and World.
Cambridge Mass.: Harvard University Press 1994.
52 Chalmers op. cit.
53 Belief contents are marked by italics.
ferring to a painful state), then the statement cannot be false. Hence the
correct making of an honest Konstatierung-statement justifies its content.
In somewhat more detail: the meaning of empirical statements are their
truth conditions. In the case of empirical-hypothetical statements (“The-
re’s a table in the next room”), the truth conditions coincide with the
verification conditions (If you go to the next room you will have a table-
formed such and such coloured visual experience.). In contrast, empirical-
nonhypothetical statements, i.e., Konstatierung-statements (“There is
pain”) refer directly to their truth condititions. There is no verificati-
on procedure. Having the experience directly justifies the Konstatierung-
statement. In other words: asserting sincerely and correctly a Konstatie-
rung-statement implies that it is true.
2. Against (b): Incorrigible knowledge is not contradiction in terms
1. Against Wittgenstein’s challenge, we may argue in a way which seems

to me consistent with Juhos’ and Schlick’s position. We can accept the
Wittgensteinian point that an utterance of a sentence has no meaning if
there is no criterion which determines whether it is true or false (i.e., if
there is no way to distinguish a correct application of a sentence from an
incorrect one). But we may hold that the incorrigibility of simple sensation
reports (e.g., “I feel pain”), does not imply that there is no criterion which
tells whether such reports are true or not.
Remember how Juhos and Schlick accounted for such Konstatierung-
statements. I can make such a Konstatierung-statement correctly viz.
using the word “pain” in the Konstatierung-sentence with its proper mea-
ning), only if I have the sensation to which the word “pain” refers to, i.e.,
pain. Otherwise the content of my statement is different. Remember: this
is not the case usually. I can make an utterance “I sense water” using
word “water” with its proper meaning, without there being water in my
neighborhood, (there may be watery stuff around instead). So there is a
criterion of the truthfulness of “I feel pain”. It is true only if I formulate
the statement using the correct concept of pain. However, if I formulate
the statement correctly, then the statement is incorrigibly true. (But it
is possible that I do not formulate it correctly: then the meaning of that
statement differs.)
So, if we accept that there are sensations with a directly given phe-
nomenal nature, and sensation words refer directly to these, then we may
122 Gergely Ambrus
have a set of examples of incorrigible but nevertheless meaningful judge-

ments, namely certain sensation reports. Of course, there are many argu-
ments against such an account of simple sensations and sensation words,
including some of Wittgenstein, but that is a separate issue: Juhos and
Schlick clearly adhered to such an account.
2. A further argument may be built on Wittgenstein’s challenge, na-
mely the following. The possibility of making an error by misnaming (ap-
plying a wrong word to) a sensation is necessary for the meaningfulness
of “I know I am in pain”. For if one cannot make a mistake by apply-
ing the word “pain” wrongly, then there would be no difference between
the correct and incorrect use of “pain”, which implies that “pain” has no
meaning.
Some Wittgenstein interpretators, e.g. Rorty,54 carry on as follows.
One cannot differentiate between false judgements originating from mis-
naming and originating from being mistaken about the nature of our sen-
sation. Hence the distinction that one cannot err about the nature of his
sensations, but can err by misnaming it, is senseless. However, the distinc-
tion is required for “pain” to have a meaning. Hence the word “pain” (as
referring to some sensation) has no meaning, and the sentence “I know I
am in pain” neither.
We may answer in the vein of Schlick and Juhos: it is possible to dis-
tinguish between error by misnaming and error by being mistaken about
the nature of a sensation. If we accept that sensations have a directly
given phenomenal nature, then the distinction is tenable. Wittgenstein’s
arguments (about need for a criterion) are neutral on this issue.
3. Against (c): Externalism about phenomenal content is an open issue
The argument from externalism against being incorrigible about our own
sensations runs as follows. The identification of any sensations relies on
external factors, either external objects, or the linguistic communities’
use of the sensation term. Since we may have no access to these external
factors from our first-person perspective, we are not incorrigible about
our own sensations.
54 Richard Rorty, “Mind-Body Identity, Privacy and Categories”, in: The Re-
view of Metaphysics 19, 1965, pp. 24-54.
Now, we may counter this objection by noting that Juhos was clearly
not an externalist in the sense given here, and also that externalism about
phenomenal content is an open issue even today. And, if externalism is
not true, i.e., the content of an experience is exhausted by what is given
to the experiencer in the first-person perspective, then one could only
err about one’s own sensations either by referring with a wrong name to
the sensation, or by remembering falsely that the actual sensation is the
same type as another one experienced earlier: one could not err about
what sensation he has (i.e., what his sensation is like) now. Hence the
objection from externalism may be put aside.
The Arguments for the Intersubjectivity of the

Meaning of Sensation Terms
Besides attacking the physicalist language thesis directly, Juhos also att-
empted to defend his position against the objection that the phenomena-
list interpretation of mental sentences cannot provide for the intersubjec-
tivity of mental language. He presented an argument to the point that the
phenomenalist (“empiricist”) account of the meaning of sensation terms is
indeed intersubjective.55 This claim sounds rather odd, but the proposed
argument is interesting and worth discussing.
In Juhos’ reconstruction, the thesis to be proved is that the statement
“I feel a pain (sensation)” is intersubjectively verifiable, which, according
to him, amounts to saying that it is logically possible to verify it.56 We may
55 It must be noted, however, that this argument was principally taken over
from Schlick. See Schlick, “De la relation entre les notions psychologiques
et les notions physiques”, and “Meaning and Verification”, in: The Philo-
sophical Review 44, 1936, pp. 339-369. Kraft, in his official evaluation of
Juhos’ Habilitationsschrift (which appeared later as Die Erkenntnis und ih-
re Leistung) even criticized Juhos for not mentioning that this argument
is credited to Schlick. In the book version, however, Juhos does refer to
Schlick, saying that he provided a very similar argument.
56 According to Schlick’s and Juhos’ usage, a sentence expresses a logically
possible case, if it is consistent with the meaning of the terms forming the
sentence, i.e., if it is not self-contradictory. On Schlick’s and Juhos’ phe-
nomenalist reading, sensation terms refer solely to the phenomenal states,
i.e., their meaning does not involve reference to any bodily state. Hence it is
logically possible, i.e., it is not ruled out by the meaning of sensation terms,
124 Gergely Ambrus
query right at the outset why it ought to be logically possible to verify it.
Why not empirically possible? Schlick’s answer is briefly the following.57
We do not and will not ever have indubitable knowledge of what is em-
pirically possible, since empirical law-statements which constrain what is
empirically possible are only hypothetic. If the meaningfulness of a term
depended on the empirical possibility of a certain state of affairs, we would
never know for sure whether the term is meaningful or not – and this
would be unacceptable. Hence we have to rely on the logical-conceptual
possibility in demarcating the meaningful from the meaningless.
The intersubjectivity argument
We may accept this answer for the sake of argument, and turn to the
master argument of Juhos and Schlick. The master argument starts with
a discussion of the possible interpretations of “my pain” in the sentence
“I can only feel my pain”. Accordingly, we have two options, which we
may call the solipsistic (S) and the non-solipsistic (NonS) interpretation.
(S) Any experienced pain-sensation is necessarily my pain: pains are logi-
cally private.
or
(NonS) An experience of a pain-sensation is my pain, if it is co-instantiated
with (the experience of) certain states of my body M.
According to (S), necessarily, I can only feel my pain. If I feel pain co-
instantiated with certain characteristic (pain-)states of another person’s,
O’s body, then I feel my pain in O’s body, not O’s pain. (NonS) al-
lows for the logical possibility that an experience of pain-sensation is co-
instantiated with (the experience of) certain states of O’s body. This is
to be understood as “I feel O’s pain”. (Similarly, it is logically possible
that O feels my pain.) Hence, according to (NonS), “I feel my pain” is
intersubjectively verifiable, since it is logically possible, that someone else
also feels my pain.
Schlick and Juhos argue for the non-solipsistic reading, claiming that
the solipistic reading is meaningless. Here is the core of their argument.
that sensations occur without a body, or that they are correlated not with
the body of the subject of sensations, but with someone else’s body.
57 See section III of Schlick’s “Meaning and Verification”.
(1) According to (S), there is no difference between the meaning of “I feel

my pain”, “I feel pain” and “There is pain”, because there is no logically
possible situation in which they would describe different states of affairs.
(2) If there is no difference between the meaning of “I feel my pain”, “I
feel pain” and “There is pain”, then “I” and “my” in “I feel my pain” are
redundant, since it it logically impossible that it is not me who feels pain
or that it is not my pain that I feel.
Hence
“My” and “I” do not have a meaning, they fall out.
thus
The solipistic thesis “I can only feel my pain” is meaningless.
Therefore, we must accept interpretation (NonS). On the (NonS) inter-
pretation of “my pain”, “I feel my pain” is intersubjectively verifiable,
since it is logically possible that someone else feels my pain. Hence “I
feel my pain” has an intersubjective meaning, contrary to Carnap’s and
Neurath’s claim.
So far, so good. However, even if we accept this argument against (S),
we may wonder whether (NonS) is acceptable in itself. That is, what justi-
fies positively the (NonS) interpretation? Schlick and Juhos held that it is
the meaning, the “grammar” of “pain” that grounds (NonS). Accordingly,
“pain” refers to pain sensation exclusively, the concept of pain is indepen-
dent of the concept of bodily states. But we may ask: what justifies this
understanding of the meaning of “pain”?
My understanding of the Schlick-Juhos position is the following. In
identifying a sensation as pain, we need not refer to any bodily state. The
qualia of pain can be identified directly, by acquaintance, i.e., “inwardly
pointing” at it. There are several well-known objections to such a view, put
forward by many, including Wittgenstein. For example, there is one family
of objections to the point that it is impossible to define the meaning of
any word based solely on acquaintance. Another line of argument claims
that the meaning of sensation words cannot be defined ostensively, for
sensations cannot be intersubjectively “pointed at”, as they are private
to their beholder. Prima facie, these objections seem to be begging the
question against Schlick and Juhos, for according to them the content of
Konstatierungen can be determined just by having the experience, i.e.
ostensively, and sensations are not (logically) private. However, I will not
126 Gergely Ambrus
address these issues here. Instead, I shall present an argument of mine,

according to which the Schlick-Juhos interpretation of sensation terms is
inconsistent.
My objection
For Juhos’ and Schlick’s argument for the intersubjectivity of sensation

terms to go through, one has to make sense of the meaning of “my” (or
“I”), i.e., the terms referring to the subject of pain in (NonS), which claim
that “An experienced pain-sensation is my pain, if it is co-instantiated
with (the experience of) certain states of my body M.” Now, it seems,
that a person may be identified either
(a) by the pain (or any other) sensations to which she has access (which
she can feel), or
(b) by her body.
However, a particular pain (or other sensation) of yours can only be felt
by me, if you and I are distinguishable. Distinguishing between you and I
may be done only by way of (a) or (b). If (a), then I cannot feel your pain,
for any pain I could feel is mine, by definition. Hence (a) is out. If (b),
however, then for the pain felt by me to be yours, it must be necessarily
related to your body. But if (b), then in order to be able to feel your
pain, I have to experience your and my body. Hence, it is a precondition
of being able to feel other’s pain that we have bodies and that we have
separate bodies.
So, it seems, there is a contradiction in the Schlick-Juhos view. On
the one hand, according to their account of the meaning of “pain”, so-
meone’s pain can be identified without referring to a particular body or
bodily state. (Schlick claimed that the concept of a bodiless individual
consciousness is coherent, so it must be possible to identify someone’s
pain without refering to any body or bodily state).58 On the other hand,
to make sense of “It is logically possible that I feel your pain”, we have to
presuppose that when there are particular pains, (1) there are also indi-
vidual bodies, and that (2) every pain is correlated with just one body in
a special way (so that the changes in the pain sensations are coordinated
58 Cf. Schlick “De la relation entre les notions psychologiques et les notions
physiques” and also Virgil C. Aldrich, “Messrs. Schlick and Ayer on Im-
mortality”, in: The Philosophical Review 47, 1938, pp. 209-213.
with the changes of that body’s states exclusively). Now, if my argument

is correct, then Schlick’s and Juhos’ attempt to show that the phenome-
nalist interpretation of the meaning of sensation terms is intersubjective,
is defective. Hence, it seems, physicalism fares better on this count.
Summary and conclusion

My aim in this paper was to show Juhos’ views on the psychophysical
problem from the early 1930s onwards. These are intimately connected
with his views on physicalism, hence I also tried to locate his role in
the physicalism debate, and discussed his main arguments against the
physicalist language thesis, according to which mental reports are content-
identical with physical reports.
Now, as for the general results of my inquiry: it is not a very bold
assumption that Juhos’ positive views on the psychophysical problem,
the “empirical-cum-logico-linguistic” phenomenalist approach, is hardly
shared by anyone today. This fact may be explained by several factors
(besides the obvious one, namely the almost complete ignorance of Ju-
hos’ work). One major factor is, I think, that his approach comprises
elements that have been rather unpopular ever since the demise of logical
positivism. For example, he denied the meaningfulness of the diverse me-
taphysical views about the consciousness-body relation; maintained the
possibility of logically constructing the concepts of physical objects out of
sense experiences; held onto a strict analytic/synthetic distinction . . . In
other words, he did not seem to react to changes even within the logical
positivist movement, and stuck to a rigid (and, in retrospect, rather sim-
plicistic) linguistic-ametaphysical framework, characteristic of the early
days of logical positivism; he maintained this approach still in his book,
Erkenntnis und ihre Leistung, published in 1950, and did not seem to al-
ter it substantially even later. For this reason, one may hold that Juhos’
views are deservedly neglected and are interesting only from an historical
point of view.
Be it as it may, I believe that Juhos’ arguments against the physicalist
language thesis do deserve more attention. I find Juhos’ objections against
physicalism important on their own, and it is also interesting, that they
turned up again later, in a somewhat different context, namely in the
debate about central-state materialism in the 1950s and 1960s.
128 Gergely Ambrus
I argued that both the argument from ignorance and the argument
from corrigibility is much harder to tackle than it might first seem. The
argument from ignorance cannot be answered simply by appealing to a
Fregean sense/reference distinction, as most famously Smart proposed
(among others, e.g. Feigl). The argument from corrigibility cannot be
countered by Smart’s replies, and I tried to show, that other strategies to
dispel it, as e.g. Sellars’, Bonjour’s, Davidson’s, or McDowell’s arguments,
may be rejected as well, relying on a conception of sensation terms and
phenomenal judgements similar to Juhos’ and Schlick’s theory of Konsta-
tierungen.
I think, both arguments had an effect in rejecting central state mate-
rialism, which stated the identity of conscious states with brain states: the
inability to answer them persuasively led to the proposal of other sorts of
physicalist views, as e.g., eliminative materialism, and later non-reductive
materialism. Thus Juhos’ views on the psychophysical problem, at least
his criticism of the physicalist language thesis, did effectively contribute
to the debate on the mind-body problem.
Ladislav Kvasz
Classical Mechanics between

History and Philosophy
Abstract
In a series of papers (Kvasz 1998, 2005 and 2006) I proposed an interpre-
tation of the development of scientific theories as changes of the pictorial
form in the sense of the Tractatus. In the development of geometry and
of algebra it was possible to identify six different forms, each of which
determines the way how linguistic representations are coordinated with
each other as well as with the particular subject matter, represented by
the language of the theory. The aim of the present paper is to use this
approach in the epistemological interpretation of the development of clas-
sical mechanics.
It is open to question which technical tools are appropriate in the interpre-

tation of scientific texts of the past. Are we supposed to use only means
which were available also to the authors of the texts we interpret, or is
it possible to enlist also modern tools, which were developed much later?
I believe that here the answers of the historian and of the philosopher
will be radically different. While a historian will probably opt for the
first possibility, a philosopher, at least one working in the tradition of the
Vienna Circle, will opt for the second. As in the heydays of the Vienna
Circle it was common to use modern logic in order to see more clearly the
empirical content of scientific theories, I believe that today we can simi-
larly employ the tools of modern semantics in order to understand better
the development of scientific theories. Therefore in the epistemological
interpretation of the historical texts of classical mechanics I will use the
terms of modern mathematics like manifold or phase flow, which make
it possible to grasp the meaning of the analyzed theories more clearly.
This use of anachronistic tools illustrates clearly the difference between a
130 Ladislav Kvasz
purely historical description and an epistemological reconstruction of his-

tory. Thus for instance in the reconstruction of Lagrange’s views I will use
the notion of a differentiable manifold. To a historian this could appear
problematic, because Lagrange did not use this term. When Lagrange
writes: “Let us assume, that taking into account the formal properties of
the equation, we express the coordinates x, y, z of each body as functions
of other variables ξ, ψ, ϕ, ..., which are mutually totally independent and
they allow us to determine the position of the system in every moment.”
(Lagrange 1788, p. 413) then it could appear as a usual change of co-
ordinates. Nevertheless, from the epistemological point of view there is
an important innovation: the three points after ξ, ψ, ϕ together with
the requirement that these coordinates should be “mutually totally inde-
pendent”. Lagrange leaves here the three-dimensional Euclidean space as
the framework of representation of a mechanical system and turns to an
n-dimensional space. Of course, this transition is only implicit, but for its
implicit nature it is by no means less important.
In geometry the idea of an n-dimensional space appears more than
50 years later in the paper Chapters in the analytical geometry of (n)
dimensions of Arthur Cayley (1821–1895) published in the Philosophical
Magazine in 1843; in the book Die lineale Ausdehnungslehre, ein neuer
Zweig der Mathematik of Hermann Grassmann (1807–1877) published in
1844; and in the Über die Hypothesen, welche der Geometrie zu Grunde
liegen, a lecture read by Bernhard Riemann (1826–1866) in 1854. While
these mathematicians introduced the idea of an n-dimensional space in
an explicit way, Lagrange himself never mentioned any such space. Nev-
ertheless, even though he did not mention it, he used it, and therefore it
is an inseparable component of the epistemological structure of his work.
Thus modern mathematics by making the structure of the funda-
mental notions of classical mechanics fully explicit reveals the implicit
epistemological innovations contained in the classical works of this dis-
cipline. Similarly as formal logic makes it possible to reconstruct the
logical structure of scientific theories, so in a same way the tools of mod-
ern mathematics enable us to reconstruct the epistemological structure
of these theories. In the following text I try to reconstruct the devel-
opment of classical mechanics, leading from Newton and Euler, through
d’Alembert to Lagrange. I will argue that in a similar way as in geometry
and algebra also in mechanics the development was founded on a series of
Classical Mechanics Between History and Philosophy 131
changes of the (pictorial) form of language and led from the perspectivis-
tic form through the projective, coordinative, compositive, interpretative
right to the integrative and constitutive forms.
1. Perspectivistic form of language of classical mechanics:

Newton 1687
I will start the reconstruction of the development of the form of lan-
guage of classical mechanics with the work of Isaac Newton (1643–1727)
Philosophiae naturalis principia mathematica (1687). Here Newton for-
mulated the laws of motion of material bodies using the so-called perspec-
tivistic form of language.1 In order to show this we have to show first that
besides terms which have clear denotation Newton used in his mechanics
also expressions without any denotation. These esxpressions constitute
the pictorial form of his language. (Wittgenstein introduced the notion
of the pictorial form to mean all that cannot be expressed in a language
but which is only shown by its use.) And secondly we have to show that
this form is the perspectivistic one, which means that it is the form which
represents the world from a unique point of view, while the viewpoint itself
is not represented. With respect to Newton we are in a lucky situation,
because this has already been shown by Ernst Mach. In his famous book
Die Mechanik in ihrer Entwicklung (1893) Mach presented a criticism of
Newtonian mechanics. By unveiling the metaphysical notions (the notion
of absolute space and time), tautologies (the first law), and apparent def-
initions (the definition of mass) of the Newtonian system, this criticism
became the paradigm for the philosophy of the Vienna circle that started
originally as Verein Ernst Mach.
In his analysis Mach first quotes the Principia:
Absolute, true and mathematical time, of itself, and from its own nature, flows
equably without relation to anything external, and by another name is called
1 An interpretation of the equations of mechanics as pictures that unequivo-

cally represent reality is so natural that for a long time no other alternatives
were considered. Thus even though mechanics on the technical level long
time ago abandoned the perspectivistic form of language (i.e. the form
which understands a theory as a picture of reality) and passed to the inte-
grative form, in the philosophical reflection mechanics was still thought of
as a picture of reality.
132 Ladislav Kvasz
duration; relative, apparent, and common time, is some sensible and external
. . . measure of duration by the means of motion, which is commonly used instead
of true time; such as an hour, a day, a month, a year. . . . the natural days are
truly unequal, though they are commonly considered as equal and used for a
measure of time; astronomers correct this inequality for their more accurate
deducing of the celestial motions. It may be that there is no such thing as an
equable motion, whereby time may be accurately measured. All motions may
be accelerated and retarded, but the true, or equable, progress of absolute time
is liable to no change. The duration or perseverance of the existence of things
remains the same, whether the motions are swift or slow, or none at all; and
therefore, it ought to be distinguished from what are only sensible measures
thereof. (Newton 1687)
Then Mach presents his criticism of these views: “This absolute time can
be measured by comparison with no motion; it has therefore neither a
practical nor a scientific value; and no one is justified in saying that he
knows aught about it. It is an idle metaphysical conception.” (Mach 1893)
Mach subjects to a similar criticism also the notions of absolute space
and absolute motion, the notion of mass and Newton’s laws. According
to Mach’s criticism some definitions of Newton’s physics are illusory defi-
nitions (Scheindefinition) because they are circular. For instance Newton
defines mass as the product of volume and density. But according to Mach
density is nothing else than mass divided by volume and so Newton’s def-
inition of mass is circular. Similarly according to Mach Newton’s first law
is a tautology because force is the measure of acceleration and therefore
to a zero force corresponds zero acceleration, i.e. uniform motion in a
straight line. This criticism was the most original analysis to which a sci-
entific theory was ever subjected. Mach demonstrated a new method of
analyzing theories which makes it possible to discover metaphysical con-
cepts, illusory definitions, and tautologies. The aim of such analysis was
to discard these metaphysical concepts, illusory definitions, and tautolo-
gies and replace them by empirical concepts, non-circular definitions, and
non-tautological propositions. In this respect Mach was a forerunner of
the theory of relativity, which achieved what Mach wanted. The meta-
physical notion of absolute time was replaced by the empirical notion of
local time; Newton’s notion of mass was split into gravitational and in-
ertial masses; and Newton’s first law was turned into the definition of an
inertial frame.
Nevertheless, another reading of Mach’s criticism is possible, not a

physical one but rather a philosophical one. Its aim is not to discard
the metaphysical notions from science but rather to understand their role
in the building of scientific theories. Thus we will consider Mach as the
thinker who discovered the existence of the pictorial form of language of
mechanics. Mach believed that it is possible to discard the metaphys-
ical concepts and fully replace them by empirical ones. I do not share
Mach’s optimism in this respect. I believe that by such a replacement
we simply move to some other language, the form of which will again be
non-empirical. Of course, each transition to a new language brings about
undeniable progress from the physical point of view. It is sufficient to
mention the theory of relativity which led to the discovery of new phe-
nomena (contraction of length, dilatation of time) as well as new laws
(the equivalence of mass and energy). Nevertheless, the purpose of epis-
temology is not the understanding of how nature operates, but rather to
get insight into the processes of functioning of science. Thus, even though
I am far from belittling the importance of Mach’s effort to replace the
metaphysical concepts by empirical ones, I am convinced that for episte-
mology it is similarly important to understand their role in the structure
of the language of scientific theories.
It seems that the presence of non-empirical (or “metaphysical”) no-
tions and principles is a systematic feature of every formal language. On
the one hand this means that in the case of Newton this was not the result
of some kind of inconsequence or naivety. The non-empirical notions con-
stitute the form of language and therefore there is no escape from them.
It is, of course, possible to move to some other language in which the
form of the previous language becomes “empirical”, but the new language
has its own form, which will be again “metaphysical”. Mach was able to
discover the form of the language of Newtonian mechanics precisely be-
cause he himself analyzed this language from the viewpoint of a new lan-
guage (energetism), in the framework of which the metaphysical elements
of Newtonian mechanics are visible. On the other hand the systematic
nature of the presence of non-empirical notions and principles in formal
languages shows that a language is not a collection of some isolated, ran-
domly introduced independent elements. A language is a coherent unity,
and it is its form which unites the expressions of the language in such a
134 Ladislav Kvasz
way that the language can function in very complex ways (it can perform
complex calculations and proofs, the results of which fit together).
The form of language of Newtonian mechanics was discovered by
Mach. It is sufficient to reinterpret his criticism of Newtonian mechanics
and read this criticism as a description of the form of Newtonian language.
This form is constituted by elements such as absolute space and absolute
time which correspond to the fixed viewpoint of a unique external epis-
temic subject. This subject is not represented in the language and thus
it is analogous to the subject of perspective representation in geometry.
Therefore we will describe this form as the perspectivistic form of lan-
guage which represents the universe of discourse from a unique external
viewpoint.
2. Projective form of language of classical mechanics: Euler 1736

Newton designed his Principia using Euclid’s Elements as a model. In
contrast to the contemporary way of formulating scientific theories New-
ton gave his propositions the form of theorems and proved them as it
was usual in geometry. Such approach is logically unassailable but on
the other hand it is cumbersome to such a degree that Leonhard Euler
(1707–1783) felt the need to remark that:
Although the reader has no doubts about the correctness of the presented propo-
sitions, he does not understand them clearly and precisely enough and if the
same problems would be even slightly changed, he would be hardly able to solve
them independently. (Euler 1736)
The discontent with Newton’s purely geometrical formulation of his me-

chanics led the 26 years old Euler to rewriting it using the language of
the differential calculus in his Mechanica sive motus scientia analytice
exposita (Euler 1736). Thus Euler was the first to formulate Newton’s
second law as a differential equation
dt
dc = np , (1)
A
where c is the velocity, p is the force, A is the mass of the body, t is
time and n is a coefficient that is determined by the choice of units (Euler
1736, p. 124). This change was of fundamental importance for the whole
further development of mechanics because the calculus offered a great

variety of methods for the solution of mechanical problems. The advan-
tage of analytical methods is that “if the same problems would be even
slightly changed” they can be solved by almost the same methods (using
for instance perturbation theory developed in order to solve such slightly
changed mechanical problems). In this way Euler removed the main defi-
ciency of Newton’s approach to mechanics – the lack of universal methods.
Among the technical tricks which were made possible by the new formu-
lation of mechanics an important role was played by substitutions.
In the analysis of the development of algebra I argued that the emer-
gence of substitutions in a language is the characteristic feature of the
transition to the projective form. A substitution is a transition from one
coordinate system to another. From an epistemological point of view it
is analogous to a central projection enabling us to move from one view
(or perspective) of some object to another. Therefore Euler did not bring
about a simple reformulation of Newton’s mechanics; he did not simply
replace Newton’s cumbersome geometrical formulation of mechanics with
a much smoother analytic one. Euler did substantially more; he created a
new form of language, he turned mechanics into an analytical discipline.
He replaced Newton’s synthetic arguments by analytic calculations.2
2 For the sake of completeness let me quote here Newton’s original formula-
tion of his second law: “The change of momentum of a body is proportional
to the impulse impressed on the body, and happens along the straight line
on which the impulse is impressed.” (Newton 1687) In 1765 Euler gave
Newton’s law a new form:
d2 x λp d2 y λq d2 z λr
= , = , = ,
dt2 A dt2 A dt2 A
here p, q, r are projections of the resulting force into the directions of the
coordinate axes and λ is a coefficient of proportionality by which Euler
replaced the coefficient n from the equation (1) (Euler 1765, p. 328). I
mention these formulas only in a footnote, because they belong already to
the coordinative form of language. Comparing Newton’s original formula-
tion from 1687 with Euler’s formulations from 1736 and 1765 we can get a
feeling of the differences which exist in the formulation of the same propo-
sition in three different forms of language: perspectivistic, projective, and
coordinative.
136 Ladislav Kvasz
3. Coordinative form of language of classical mechanics:

Maclaurin 1742
In his Mechanica Euler used systematically the decomposition of forces
into tangential and normal components. The coordinate system in which
he analyzed the motion of a particular body was thus chosen specially
from the point of view of the described motion. This is advantageous as
long as the system of moving bodies is sufficiently simple. The tangen-
tial component of force causes an increase of the bodies’ velocity while
the normal component only changes the curvature of the trajectory of
the motion without changing its velocity. This way of description is not
suitable for the description of systems formed by several bodies because
the normal and the tangential directions are for each of the bodies dif-
ferent and so a messy situation occurs. Colin Maclaurin (1698–1746) in
his book A Complete System of Fluxions (1742) introduced a systematic
decomposition of all movements into three fixed directions. This change
may appear small but its consequences are far-reaching. It was such a de-
composition of forces into three independent directions that enabled the
discovery of the equations of the rotation of a solid body. These equations
relate the angular momentum to the moment of force.
The moment of force was known already to Archimedes who used
it in his description of the equilibrium of a lever. Similarly the angular
momentum was already (at least implicitly) present in Kepler’s second
law. But to relate these quantities, to bring into mutual relation the
moment of force acting on a rotating body to the change of the angular
momentum of its rotation is by no means easy. The problem is that both
these quantities are defined as vector products (of the radius vector with
the force or velocity respectively). This makes working with them rather
complicated and if we have no fixed referential framework at our disposal
to which we could relate the directions of all vectors that occur in the
problem (of the radius vector, of the vector of the angular momentum, of
the vector of force, and of the vector of the moment of force), it is prac-
tically impossible to find our bearings in this system. Therefore progress
in the theory of motion of rigid bodies could appear only after Maclaurin
introduced the idea of a fixed coordinate system. And indeed a few years
later Euler published his Découvert d’un noveau principe de la mécanique
(1750) where he formulated the law of the momentum of forces. Euler’s
first formulation of this law was still rather cumbersome and so he simpli-
fied his equations step by step and in 1758 he discovered the idea to use
the main axis of the body as a referential frame. This move simplified the
equations to a great extent:
P = A.dp/dt + (C - B ).q.r,
Q = B.dq/dt + (A - C ).r.p,
R = C.dr /dt + (B - A).p.q.
This system of equations bears Euler’s name. Here P , Q, R are the com-
ponents of the vector of momentum of force; A, B, C are the components
of the vector of momentum of inertia of the rotating body, and p, q, r are
the components of the vector of angular velocity. We see that the first
equation contains the x-component of the vector of momentum of force
(represented by P ), the y-component and the z-component of the vector of
angular velocity (represented by q and r), the x-component of the vector
of angular acceleration (represented by dp/dt), and all three components
of the vector of momentum of inertia (represented by A, B, C). These
three equations show clearly the importance of Maclaurin’s innovation
which I interpret as the transition to the coordinative form of language.
Similarly as in algebra Michael Stifel united (coordinated) into a single
polynomial all the different kinds of equations of the same degree (which
until then were studied separately), Maclaurin unified (coordinated) all
elements of a mechanical system into a scheme which makes it possible
to represent rather complex dependencies among quantities, as illustrated
by the above equations.
4. Compositive form of language of classical mechanics:

d’Alembert 1743
On the example of the compositive form of language of classical mechanics
it is possible to illustrate the process of successive emergence of the new
form of language, which can be seen also in algebra. The three stages –
fragmentary, implicit, and explicit – are here relatively clearly distinguish-
able and their formulations are separated by sufficiently great intervals of
time.
138 Ladislav Kvasz
4.1 Fragmentary variant of the compositive form of language:

Johann Bernoulli 16863
The description of mechanical systems which are formed by several bodies

tied together with bonds of different kinds represents a particular problem
as the forces of bonds are usually unknown. The first problem of this kind
was posed in 1646 by Marin Mersenne (1588–1648). It can be formulated
as the question: with which period will move a pendulum composed of a
weightless rod on which at distances l1 < l2 from the point of suspension
there are two balls with masses m1 and m2 . Each ball, if it would be
alone,
would swing with a period given by the known law of Galileo T =
2π gl , where l is the distance from the point of suspension and g is the
acceleration of free fall. The period of the pendulum is thus proportional
to the square root of the distance l, and as l1 < l2 , the ball m1 would
swing with a smaller period than the ball m2 . The problem is that both
balls are fastened to the same rod and therefore they must swing with a
common period. We have to find the period of swings of this composed
system. Intuitively it is clear that the period ofthe composed system
must be somewhere between the values T1 = 2π lg1 and T2 = 2π lg2 ,
because the first ball will accelerate the motion of the second while the
second ball will decelerates the motion of the first. The two balls act on
each other by means of the bond. The following illustration shows two
positions of the system.
In 1686 Johann Bernoulli (1667–1748) presented a solution of this
problem based on an idea which contains the germ of d’Alembert’s prin-
ciple. Bernoulli reasoned in the following way: imagine that the rod is
in the horizontal position. If there were no bonds (i.e. no rod), the two
3 The fact that the fragmentary variant of the compositive form of language
appeared almost sixty years earlier than the previous (coordinative) form
reached its climax is a proof of a peculiar feature of the interpretation of the
history as the development of the form of language. The consecutiveness
of the particular stages appears only at the fully developed variants of the
particular forms. Thus for instance the explicit variant of the compositive
form is linked to the explicit variant of the coordinative form. On the other
hand, in the fragmentary variant most forms of language have a rather
long prehistory. Various fragments that later give rise to a united form of
language can have a long history. It is even possible that in the same text
several fragments that will later belong to different forms coexist.
balls would fall by free fall. But since they are joined by the weightless
rod, this kind of motion is impossible. The rod will decelerate the fall of
the ball m1 and accelerate the fall of the ball m2 . In this way the ball
m1 will lose and the ball m2 will gain acceleration. Bernoulli’s main idea
was that this loss and gain must be in balance. In their motion from the
horizontal position the balls m1 and m2 fall with accelerations
l1 l2
b1 = g b2 = g
l l
where l is an unknown value laying somewhere between l1 and l2 . It is
the value which should be inserted into Galileo’s formula for the period of
the swings. As l1 < l < l2 , it is obvious that b1 < g < b2 . This is what we
would expect, the first ball loses, the second wins in comparison with the
free fall. During a short period of time dt the first ball loses the momentum
m1 (g − b1 ).dt, which corresponds to the “lost force” m1 (g − b1 ). On the
other hand the second ball moves with an acceleration that is greater
than its acceleration in the free fall. Its “gained force” is thus m2 (b2 − g).
These two forces must be in equilibrium, which Bernoulli described using
the law of Archimedes for the equilibrium of the lever
m1. (g − b1 ).l1 = m2 .(b2 − g).l2 .

m l2 +m l2
By means of simple transformations we obtain l = m11 l11 +m22 l22 . This length
determines the period of the described motion computed according to the
law of Galileo

m1 l12 + m2 l22
T = 2π .
(m1 l1 + m2 l2 ).g
The fundamental idea that emerged here only in the form of a trick in the
process of solution of a particular problem is that the forces of bonds are
140 Ladislav Kvasz
in equilibrium. The description of the equilibrium of the forces of bonds

by means of the law of the lever exploits the specific setting of the problem
and is therefore restricted to a small set of similar problems. This is why
I suggest calling it the fragmentary variant of the compositive form of
language of classical mechanics.
4.2 Implicit variant of the compositive form of language: D’Alembert 1743
The universal principle that makes it possible to solve every problem in

which the motion of a mechanical system is determined by bonds was
formulated by Jean le Rond d’Alembert (1717–1783) in his Traité de Dy-
namique in 1743 as:
The universal task: Let us have a system of bodies which are positioned in an
arbitrary way; suppose that each body possesses a particular motion, which it
nevertheless cannot realize due to the action of the other bodies: we have to
determine the motion that each body has to take. Solution: Let A, B, C etc.
be bodies forming a system and let us suppose that they have natural motions
a, b, c which they are, nevertheless, forced to change to motions a, b, c due to
their mutual interactions. It is obvious that the motion a which is natural to the
body A can be composed from the motion a which the body really performs and
some other motion α; and that similarly can be represented also the motions b,
c as represented from the motions b, β; c, γ etc; . . . From this follows that the
motions of the bodies A, B, C would be the same if they would be bestowed
instead of the motions a, b, c etc. simultaneously the pairs of motions a, α; b, β;
c, γ etc. But now, according the assumption the bodies A, B, C etc. would by
themselves take the motions a, b, c etc. Therefore the motions α, β, γ etc. must
be such that they would change nothing on the motions a, b, c etc., that means
that if the bodies would take only the motions α, β, γ etc. they would cancel
each other and the system would remain in the state of a rest. (d’Alembert
1743, p. 108).
D’Alembert formulated here in its full generality the condition which

Bernoulli bound only to the context of the lever. This condition in the
contemporary formulation says that the forces of the bounds are in equi-
librium. D’Alembert formulated here his principle in an explicit form.
The term implicit in the title of this chapter refers not to the principle
but to the form of language in which the principle is expressed. The lan-
guage in which d’Alembert formulates his principle does not express the
forces of bonds automatically. It is still necessary to add these forces and
to handle them with due care (to represent the motions as composed from
two components). In contrast to Bernoulli we have here a universal prin-

ciple which is applicable to any mechanical problem thus being common
to all fragments of the language of mechanics. But on the other hand in
each case we have to take care and decide how to apply it. In this sense
the principle is still a synthetic one. It refers to an implicit structure (the
equilibrium of the forces of bonds) which we have to introduce into the
description of the system (by decomposing the motions into their compo-
nents in our imagination). The principle leads us towards the equilibrium,
but we must in each case determine it by looking at the system.
4.3 Explicit variant of the compositive form of language: Lagrange 1788
In the reconstruction of the development of geometry we discovered a

surprising phenomenon which we described as the incorporation of the
form of language into the language. The form of language which was at a
particular stage of evolution bound to synthetic intuition was at the next
stage turned into a succession of analytic operations, subjected to fully
explicit rules. For instance the construction of the perspective which at
the beginning was a fully intuitive and thus a synthetic task was during
the 15th century turned into a fully explicit technique based on the theory
of central projection. Language thus took over the role which formerly
belonged to intuition. So the relation of the analytic and the synthetic
is in a sense the relation between the implicit and the explicit variants
of the form of a language. As long as the form of language is implicit,
the transformations of different proposition of the language can be based
only on synthetic intuition. Only when the form of language becomes
fully explicit can these transformations be turned into analytic operations
governed by explicit rules.
In the case of the compositive form of language of classical mechan-
ics its fully explicit variant was introduced by Lagrange in his Mécanique
analytique (Lagrange 1788). In the Lagrangian formulation d’Alembert’s
principle does not consist in any appeal to decompose the motions (in
our imagination) into a really occurring and an imaginary component.
Lagrange incorporated this principle into the very way of formal repre-
sentation of a mechanical system. Let us shortly mention the steps of
Lagrange’s derivation of his equations in which the explicit variant of the
compositive form of language is clearly recognizable. So let us take a sys-
tem of material particles with masses m1 , m2 , . . . mn . Let the coordinates
142 Ladislav Kvasz
of the k-th particle be (xk , yk , zk ) and the components of the external

force acting on this particle (Xk , Yk , Zk ). Let us suppose that the system
fulfills r bonds, i.e. the positions of the individual particles are not fully
independent, but they fulfill the equations
F1 (x1 , y1 , z1 , . . . , xn , yn , zn ), ..., Fr (x1 , y1 , z1 , . . . , xn , yn , zn ) (2)
As a consequence of these bonds on the particle mk act, besides the exter-

nal forces (Xk , Yk , Zk ), also additional forces, the components of which
let be (Pk , Rk , Sk ). From Newton’s second law we obtain:4
mk ẍk = Xk + Pk , mk ÿk = Yk + Rk , mk z̈k = Zk + Sk (3)
The acceleration of a particular body is thus the result of the action of

the external forces and the forces of the bonds. So far we have moved
fully in the realm of the coordinative form of language. We are describing
the complex system of bodies using a fixed system of coordinates.
D’Alembert’s principle says that the forces of bounds are passive, and
so the total work done by them is zero. Therefore if we imagine that we
give the k-th body a small variation (δxk , δyk , δzk ), the work in the course
of this variation done by the forces Pk , Rk and Sk will be zero. Thus

n
n
n
Pk .δxk + Rk .δyk + Sk .δz = 0 (4)
k=1 k=1 k=1
This is one single equation that contains the contributions from all bonds.
The only problem is that we do not know the forces Pk , Rk and Sk . These
forces occur as a result of the bonds and are analogous to the forces by
which the two balls acted on the weightless rod in the problem solved by
Bernoulli.
The first step of Lagrange was to express the forces of the bonds from
the equations (3) as
mk ẍk − Xk = Pk mk ÿk − Yk = Rk mk z̈k − Zk = Sk
4 A point above a particular letter indicates the temporal derivative of the

particular quantity, two points indicate the second derivative, etc. This
convention comes from Newton.
Then he substituted these expressions into the equation (4). This step
can be seen as an illustration of Wittgenstein’s proposition 7 from the
Tractatus: “Whereof one cannot speak, thereof one must be silent.”5 As
we do not know the forces of the bonds, they should not appear in the
description of the system. The equation (4) thus obtains the form:

n
n
n
(mk ẍk − Xk ).δxk + (mk ÿk − Yk ).δyk + (mk z̈k − Zk ).δzk = 0 (5)
k=1 k=1 k=1
This formula represents the incorporation of d’Alembert’s principle into

the language of mechanics. The forces of the bonds are taken into account
automatically without being mentioned. They became incorporated into
the form. The formula (5) expresses the fact that the work done by the
bonds is zero, and it does so without letting the forces of the bonds appear
in the formula.
The second step was motivated by the intention to decompose the
equation (5) into a system of independent equations. Thus Lagrange
wanted to represent the dynamics of the system as a composition of in-
dependent motions. The problem consisted in the circumstance that the
variations δxk , δyk , and δzk are not independent. On the contrary, they
must obey the conditions (2) of the bonds. From the equations (2) we
can get for the variations δxk , δyk , and δzk the following conditions
n

∂Fi ∂Fi ∂Fi
δF1 = δxk + δyk + δzk =0 i = 1, 2, . . . , r
∂xk ∂yk ∂zk
k=1
Each of these conditions Lagrange multiplied by the so called Lagrange’s

coefficients λi , added them together and their sum he added to the equa-
5 To mention Wittgenstein in connection with Lagrange is not as anachro-

nistic as it might seem. The influence was probably from Lagrange to
Wittgenstein. The influence of Hertz’s Mechanik on Wittgenstein is known,
and Hertz’s program can be seen as a radicalization of the Lagrangean ap-
proach. Our hint wanted only to indicate that some aspects of Wittgen-
stein’s Tractatus are closely related to Lagrange’s approach in mechanics.
144 Ladislav Kvasz
tion (5). He obtained a somewhat complicated formula:

n r
∂Fi
mk ẍk − Xk + λi . δxk +
i=1
∂xk
k=1

n r
∂F
+ mk ÿk − Yk + λi . δyk + (6)
i=1
∂yk
k=1

n r
∂Fi
+ mk z̈k − Zk + λi . δzk = 0
∂zk
k=1 i=l
The meaning of this equation is not difficult to grasp. In the equation

(5) the variations δxk , δyk , and δzk had to fulfill a system of r additional
bonds. Therefore it was not possible to pass from the formula (5) to a
system of independent equations (i.e. the motion of the system could
not be decomposed into independent components). On the other hand in
the equation (6) such decomposition becomes possible thanks to the fact
that Lagrange added to the system r new parameters λi . Thus Lagrange
succeeded in decomposing the motion of a system with bonds into 3n
independent components.
Thus we see the aim of the compositive form of language: to represent
a complex problem as a composition of simpler problems. It is fascinating
to realize that Lagrange attempted to do the same thing also in algebra
when he tried to decompose the equation of the fifth degree into a series
of equations of lesser degree, the so called resolvents. Unfortunately, in
algebra such a decomposition of the posed problem is not possible and
therefore this attempts remained unsuccessful. On the other hand, in me-
chanics Lagrange managed to decompose the motion of a complex system
of bodies and the equations describing the motion of a mechanical system
with bonds are today called Lagrange’s equations of the first kind. Their
form is:

r
∂Fi
r
∂Fi
mk ẍk = Xk + λi , mk ÿk = Yk + λi ,
i=1
∂xk i=1
∂yk

r
∂F1
mk z̈k = Zk + λi
i=1
∂zk
From an epistemological point of view it is interesting to note that these

equations contain, besides quantities which have direct physical interpre-
tation, also the parameters λi introduced purely formally as parameters
enabling the decomposition of the system into independent components.
It is possible to show that these parameters are closely connected to the
forces of the bonds. Nevertheless, it is important to realize that they
are quantities obtained not by measurement as is usual in physics. Their
source is not experience but the formalism. They are not empirical quan-
tities but constituents of the compositive form of language.
5. Interpretative form of language of classical mechanics:

Lagrange 1788
The interpretative form of the language of classical mechanics appeared
for the first time in the Mécanique Analytique (Lagrange 1788). This cir-
cumstance underlines a peculiar feature of the symbolic languages which
we have found also in algebra: fragments belonging to different forms of
language can appear in the same text. In the first part of his Mécanique
Analytique Lagrange introduced the above described method of decom-
posing the motion of a complex mechanical system with bonds into a set
of independent motions.
This method is from a technical point of view impressive but in a sense
it is superfluous. The need to introduce additional parameters was caused
by the fact that at the beginning we have chosen too many coordinates
for the description of the system. We have chosen 3n parameters, one
for the x, y, and z-coordinates of each of the n bodies. In fact the sys-
tem had only 3n − r degrees of freedom, where r is the number of bonds.
We introduced too many coordinates, they lost their independence due
to the bonds and then we had to introduce the additional parameters
λi to restore the lost independence. This shows a basic weakness of the
compositive form of language: the representation of the system is chosen
independently of the problem. The three coordinative axes (x, y, z the
introduction of which was the main achievement of the coordinative form
of language) are fixed in space and every body automatically obtains its
three coordinates whether we need them for the description of the system
or not. Lagrange realized that the technique of indeterminate coefficients
λi is a means for the compensation of the original redundancy in choice of
146 Ladislav Kvasz
coordinates for the description of the system. Had we chosen at the very
beginning only as many coordinates for the description of the mechanical
system as many degrees of freedom the system has, then we would not
have needed to introduce any additional parameters to restore indepen-
dence. All coordinates would automatically have been independent and
the decomposition of the problem into independent components would
have been automatically granted.
The main idea of the change to the interpretative form of language
consisted in the transition from the representational to the configuration
space, i.e. the transition from the 3n-dimensional space in which three
coordinates corresponded to every body of the system representing its
position in space to a (3n − r)-dimensional space, every point of which
represents a unique configuration of the system as a whole, without any
special reference to the particular bodies forming the system. The transi-
tion from the representational to the configuration space can be illustrated
on the pendulum analyzed by Bernoulli, consisting of two material balls
fastened to a weightless rod. The position of the pendulum is uniquely
determined by the angle ϕ that indicates the deflection of the weightless
rod from the vertical direction. The positions of the particular balls are
then given by the formulas
x1 = l1 . sin(ϕ), y1 = l1 .(1 − cos(ϕ)), x2 = l2 . sin(ϕ), y2 = l2 .(1 − cos(ϕ)).
In the framework of the compositive form of language we use, instead of

the unique coordinate ϕ, four parameters (x1 , y1 , x2 , y2 ) characterizing
the positions of the two balls. Nevertheless, these four parameters are not
independent. They are tied together by three bonds:
x21 + y12 = l12 , x22 + y22 = l22 , x1 : y1 = x2 : y2

We see that after we unnecessarily complicated the problem by introduc-
ing too many parameters, we must deal with the three bonds. The main
idea of the interpretative form of language is a transition to the configu-
ration space which has precisely as many dimensions as many parameters
we need to determine the configuration of the system. Thus in the case
of the above discussed pendulum the configuration space would have only
one dimension corresponding to the angular parameter φ.
The approaches to mechanics based on the compositive form of lan-

guage had, nevertheless, one undeniable advantage. They described the
motion in Euclidean space and therefore the equations were simple. The
acceleration of a particle was proportional to the acting force. If we de-
termine the force by means of potential energy V , Newton’s second law
will have the form:
d2 xi ∂V (x1 , . . . , xn )
m = (7)
dt2 ∂xi
Lagrange’s decision to describe mechanical problems by means of the con-
figuration space leads to a difficulty. A motion that corresponds to uni-
form increase of the coordinates is not necessarily an inertial one. It is
sufficient to look at the pendulum. To a uniform increase of the angle φ
corresponds in reality the circular motion of the two balls. Nevertheless,
circular motion is an accelerated one and thus it is the result of the action
of a force. Thus a force is needed not only for accelerating a motion, but
it is required already for a uniform motion in the direction of ϕ. So a
simplification of the representation of a mechanical system (the omission
of the superfluous parameters) leads to a complication of the equations
describing its motion. If we let the coordinates be curvilinear, we must
insert into the equations some additional terms compensating for the cur-
vature of the coordinate axes. Therefore the inertial nature of a motion is
no more equivalent to the uniform increase of coordinates. In curvilinear
coordinates we must change Newton’s first law. Lagrange succeeded in
finding the equations describing motion in curvilinear coordinates.
This situation resembles in many respects the idea of Gauss to in-
troduce internal coordinates on a surface. Before Gauss, it was usual to
describe surfaces as submerged in three-dimensional space. Therefore in
the calculation of the length of a curve situated on a surface, mathemati-
cians used all three spatial coordinates (x, y, z). However, one of them
was superfluous. Gauss introduced onto the surface the so-called induced
metric that uses only two coordinates (u, v). In this way he reduced the
number of parameters but on the other hand he had to use more compli-
cated formulas for the metric. In the three-dimensional Euclidean space
the square of the distance ds is given by the formula:
ds2 = dx2 + dy 2 + dz 2 , (8)

148 Ladislav Kvasz
which is the three-dimensional analogue of the theorem of Pythagoras. In

curvilinear coordinates the distance cannot be expressed so simply. It has
a more complicated form
ds2 = Edu2 + 2F dudv + Gdv 2 ,

where E, F , and G are the so-called Gauss’ coefficients. Thus the Eu-
clidean case can be characterized as having E, and G equal 1 and F equal
zero. In the general case Gauss’ coefficients can take different values from
point to point. Similarly as the reduction of the number of parameters
in a geometrical problem forced Gauss to introduce a more complicated
formula representing distance, the analogous reduction of the number of
parameters in mechanics forced Lagrange to introduce more complicated
equations describing motion. We could say, that Lagrange introduces an
“induced dynamics” just like Gauss introduced an induced metric.
We come thus to the second central idea of the interpretative form of
language which is the translation of the equations describing motion into
the coordinates of the configuration space. From an epistemological view-
point the translation between the external and the internal description
places Lagrangean mechanics alongside Beltrami’s model of non-Euclidean
geometry, or Gauss’ theory of the internal geometry of curved surfaces.
We will illustrate this translation on the case of polar coordinates which
is perhaps the simplest example of curvilinear coordinates. In polar coor-
dinates the position of a point is given by its distance r from the origin
(the pole) and the angle ϕ, between the radius vector of the point with
a given direction. If we introduce in the given point with coordinates (r,
ϕ) local axes in the directions of the growth of the coordinates r and ϕ,
then we can mark the unit vectors of these axes as er and eϕ . The radius
vector of the given point will then be expressed as:
−
→
r = r.er ,
which means that the point is at a distance r in the direction er . If we
want to determine the velocity of this point, we have to differentiate this
expression using the rule for the differentiation of a product:
−
→
v = ṙ.er + rϕ̇.eϕ .
We see that the point’s velocity has two components. The first is called
radial velocity. It represents the escape of the point from the origin in
the direction of the radius vector er , and it consists in the growth of
the parameter r. The second is called tangential velocity. It represents
the rotation of the point around the origin and consists in the growth
of the parameter ϕ. The total velocity is the vector sum of these two
components. By another differentiation we obtain the acceleration:
−
→
a = (r̈ − rϕ̇2 ).er + (2ṙ ϕ̇ + rϕ̈).eϕ (9)
The meaning of the particular terms is the following:
r̈.er – acceleration of the radial motion, consisting in the increase of the
radial velocity
rϕ̈.eϕ – acceleration of the rotation, consisting in the increase of the an-
gular velocity
−rϕ̇2 .er – centripetal acceleration of the circular motion with a constant
angular velocity
2ṙ ϕ̇.eϕ – Coriolis acceleration of the motion in a spiral with constant
radial and angular velocities
The interpretation of the terms was made from the point of view of the
external language, i.e. from the point of view of the rectilinear coordinates.
From the internal point of view of the curvilinear coordinates the third
term does not correspond to circular motion – it is the motion along the
coordinate axes ϕ. Similarly, the fourth term is by no means a motion
along a spiral but along a straight line that lies across the axes r and ϕ.
These two motions are curved only from the point of view of the rectilinear
coordinates. From the internal point of view they are rectilinear because
they are described by linear equations of the coordinates. Thus we have
to interpret the last two terms from the point of view of the curvilinear
coordinates. Let us therefore write down Newton’s second law in the
external language:
∂V
mr̈ − mrϕ̇2 = − (10)
∂r
∂V 1
2mṙϕ̇ + mrϕ̈ = − . . (11)
∂ϕ r
150 Ladislav Kvasz
The first equation says that if we want to impart to a body some radial
acceleration (the first term in (10)) or if we want to keep it on a circular
trajectory (the second term in (10)), we need a radial force. Similarly the
second equation says that if we want to increase the tangential velocity
of the body or we want to keep it on a spiral trajectory, we need a tan-
gential force. This is not yet a translation but simply a transcription of
the formula of the external language (9) by means of the parameters of
the internal language r and ϕ. The kinematical aspect of the motion is
thus expressed by means of the new parameters but the dynamical one,
i.e. the connection of the forces and accelerations, is made in the external
language. That is for instance clear from the term mrϕ̇2 being on the left
side of the equation, i.e. it is viewed as an acceleration. From the internal
point of view the term mrϕ̇2 corresponds to the uniform rectilinear mo-
tion along the ϕ axes and therefore it has no place at the left hand side of
the equation. Here we have a similar possibility of double interpretation
as we encounter in Beltrami’s model of the non-Euclidean geometry. In
that model, a particular line is from the external (Euclidean) viewpoint
a chord of a Euclidean circle, while from the internal (non-Euclidean)
viewpoint the same line is an infinite straight line of the non-Euclidean
plane. Similarly to a line in geometry, also a motion in mechanics can be
from the external viewpoint a uniform circular motion (i.e. a curvilinear
motion in the plane x, y) and as such it is an accelerated motion which
must be a result of the action of a force. But the same motion is from the
internal point of view a uniform motion along the coordinate axes ϕ, and
as such its acceleration is zero. Therefore if we want to rewrite the equa-
tions describing motion from the point of view of the internal language,
we must first of all transfer the term with the centripetal acceleration to
the right hand side of the first equation:
∂V
mr̈ = mrϕ̇2 − . (12)
∂r
This means that from the point of view of the internal language a new
force appears which can be called centrifugal force. A body that moves in
a “straight line parallel with the ϕ-axes” (from the external perspective
we know that it in reality moved in a circle), “feels” the action of a force.
But in this way an asymmetry occurs. A “real” Newtonian force is
given through the potential energy V , while the centrifugal force is given
just as an ad hoc term. Nevertheless, also the centrifugal force can be

introduced by means of a potential. Its potential is the kinetic energy.
Indeed:
∂ 1 2 2 ∂ 1 2 2 1 2 ∂T
mrϕ̇2 = ( mr ϕ̇ ) = ( mr ϕ̇ + mṙ ) = . (13)
∂r 2 ∂r 2 2 ∂r
If we label kinetic energy 12 mr2 ϕ̇2 + 12 mṙ2 with letter T , then we can write
the equation (12) as:
∂V ∂ 1 2 2 1 2 ∂V ∂
mr̈ = mrϕ̇2 − = ( mr ϕ̇ + mṙ ) − = (T − V ) , (14)
∂r ∂r 2 2 ∂r ∂r
which means that the acceleration of the motion (viewed from the internal
point of view) is the result of the action of a force the potential of which
is the difference (T − V ) of the kinetic energy T and the potential V .
Now we could try to introduce a similar correction also for the Coriolis
acceleration. We could try to transfer this acceleration existing from the
point of view of the external language of the rectilinear coordinates into
the internal language as a Coriolis force and then look for the potential
of this new force. But there is no need for this, because we obtain the
Coriolis force automatically from the kinetic energy T . To see this we
must first express the radial component of the acceleration by means of
the kinetic energy. We have

d d ∂ 1 2
mr̈ = (mṙ) = mṙ =
dt dt ∂ ṙ 2
(15)
d ∂ 1 2 1 2 2 d ∂T
= mṙ + mr ϕ̇ = .
dt ∂ ṙ 2 2 dt ∂ ṙ
Thus the term mr̈ that corresponds to the radial component of accelera-
tion can be obtained from the kinetic energy. When we now apply similar
steps, but instead of ṙ take ϕ̇, we will get

d ∂T d ∂ 1 2 1 2 2
= mṙ + mr ϕ̇ =
dt dϕ̇ dt∂ ϕ̇ 2 2
(16)
d
= (mr2 ϕ̇) = 2mrṙ ϕ̇ + mr2 ϕ̈ .
dt
152 Ladislav Kvasz
We see that we obtained besides the term describing the acceleration

of the rotation automatically also the term expressing the Coriolis force
(multiplied by r ). Thus if we substitute for the left hand side of the
equation (11) the expression from (16), we see that the factor r will be
cancelled and for the component of the motion in the direction of the ϕ
axes we get:

d ∂T ∂V
=− (17)
dt ∂ ϕ̇ ∂ϕ
An analogical substitution from the equation (15) instead of the left hand
side of (14) leads to:

d ∂L ∂
= (T − V ) (18)
dt ∂ ṙ ∂r
Equations (17) and (18) are the translation of the equations (10) and
(11) into the internal language. In them there is no sign of centripetal
accelerations and similar strange things because in the internal language
there is no center. All terms in these two equations are fully comprehen-
sible for an inhabitant of the world with curvilinear coordinates. There
is still a small esthetical imperfection: these two equations have different
right hand sides, thus the radial and the angular coordinates are not fully
equivalent. Nevertheless, if we realize that the potential energy V does
not depend on the velocities ϕ̇ and ṙ, we can in the equations (17) and
(18) just formally replace on the left hand side the kinetic energy T by
the difference (T − V ). Similarly, as the kinetic energy does not depend
on the angle ϕ, we can replace on the right hand side of the equation (17)
−V by T − V . Therefore if we label L = T − V = 12 m(ṙ2 + r2 ϕ̇2 )− V (r, ϕ),
we can give to both equations (17) and (18) the same form

d ∂L ∂L d ∂L ∂L
= , = .
dt ∂ ṙ ∂r dt ∂ ϕ̇ ∂ϕ
The common form of the two equations shows that from the internal per-
spective there is no difference between the radial and the angular coor-
dinates (which of course from the external perspective are fundamentally
different). Therefore we can introduce the so-called generalized coordi-
nates qi . Thus the introduction of the generalized coordinates is based
on the point of view of the internal language, from which all kinds of
coordinates are equivalent. By means of these coordinates the Lagrange’s

equations of a system with l coordinates q1 , q2 , . . . , ql receive the form

d ∂L ∂L
= , i = 1, 2, . . . , l . (19)
dt ∂ q˙i ∂qi
I illustrated the transition to Lagrange’s equations on the simplest system
of curvilinear coordinates – the polar coordinates in a plane. Nevertheless,
the general scheme of this transition is the same also in the general case.
Newton’s equations are a special case of the Lagrange’s equations for the
case of rectilinear coordinates when the kinetic energy does not depend
on the position but it depends only on the velocities. In that case we
get from Lagrange’s equations directly the Newton’s ones. Helmholtz
called the function L kinetic potential but usually it is called Lagrange’s
function.
6. Conclusions
We could follow the development of classical mechanics further and inter-
pret in a similar way the works of Hamilton and Poincaré, the representa-
tives of the two further developmental stages. Nevertheless, it seems to be
better to stop here and look shortly back on what we have achieved. First
of all the above reconstruction indicates that the method of reconstruction
developed originally in geometry and algebra can be used also for the re-
construction of the development of classical mechanics. The other point,
which is perhaps even more important, is the close connection between
our method of reconstruction and the reconstruction of the Newtonian
mechanics by Ernst Mach.
References
Vladimir Arnold, Mathematical Methods of Classical Mechanics. New

York: Springer 1978.
Jean D’Alembert, Traité de Dynamique. Paris 1743. Russian translation:
Dinamika. Moskva: GITTL 1950.
Rene Dugas, A History of Mechanics. New York: Dover 1988.
154 Ladislav Kvasz
Leonard Euler, Mechanica sive motus scientia analytice exposita. Petropoli

1736. Russian translation: Osnovy dinamiki točky. Leningrad: GRTTL
1938, pp. 29-262.
Leonard Euler, Theoria motus corporum solidorum seu rigidorum ex primis
nostrae cognitionis principiis stabilita et ad omnes motus, qui in huius-
modi corpora cadere possunt, accomodata, Rostochii 1765. Russian
translation: Osnovy dinamiki točky, Leningrad: GRTTL 1938, pp.
263-467.
Ladislav Kvasz, “History of Geometry and the Development of the Form
of its Language”, in: Synthese 116, 1998, pp. 141-186.
Ladislav Kvasz, “Similarities and Differences Between the Development
of Geometry and of Algebra”, in: Carlo Cellucci and Donald Gillies
(Eds.), Mathematical Reasoning and Heuristics. London: King’s Col-
lege Publications 2005, pp. 25-47.
Ladislav Kvasz, “History of Algebra and the Development of the Form of
its Language”, in: Philosophia Mathematica 14, 2006, pp. 287-317.
Ladislav Kvasz, Pattern of Change, Linguistic Innovations in the Devel-
opment of Classical Mechanics. Basel: Birkhäuser 2008.
Joseph Lagrange, Mécanique Analytique, Paris 1788. English translation:
Analytical Mechanics. Dordrecht: Kluwer 1998.
Ernst Mach, Die Mechanik in ihrer Entwicklung. Leipzig: Brockhaus
1893. English translation: The Science of Mechanics. Chicago: The
Open Court 1902.
Isaac Newton, The Principia, A New Translation by I. B. Cohen and A.
Whitman, Berkeley: University of California Press, 1999.
Istvan Szabó, Geschichte der mechanischen Prinzipien. Baseel: Birkhäuser,
1996.
Ludwig Wittgenstein, Tractatus Logico-philosophicus. Frankfurt am Main:
Suhrkamp, 1989.
Judit X. Madarász and Gergely Székely
Comparing Relativistic and Newtonian

Dynamics in First-Order Logic
Introduction
In this paper we introduce and compare Newtonian and relativistic dy-
namics as two theories of first-order logic (FOL). To illustrate the similar-
ities between Newtonian and relativistic dynamics, we axiomatize them
such that they differ in one axiom only. This one axiom difference, how-
ever, leads to radical differences in the predictions of the two theories.
One of their major differences manifests itself in the relation between
relativistic and rest masses, see Thms. 5 and 6.
The statement that the center-lines of a system of point masses viewed
from two different reference frames are related exactly by the coordinate
transformation between them seems to be a natural and harmless assump-
tion; and it is natural and harmless in Newtonian dynamics, see Cor. 11.
However, in relativistic dynamics it leads to a contradiction, see Thm.
4. Showing this surprising fact, which also illustrates the great difference
between the two theories, is the main result of this paper.
Our work is directly related to Hilbert’s 6th problem on axiomatiza-
tion of physics. Moreover, it goes beyond this program since our general
aim is not only to axiomatize physical theories but to investigate the re-
lationship between the basic assumptions (axioms) and the predictions
(theorems) of the theories and to compare the axiom systems of related
theories. Our another general aim is to provide a foundation of physics
similar to that of mathematics.
For good reasons, the foundation of mathematics was performed strictly
within FOL. One of these reasons is that staying within FOL helps to avoid
tacit assumptions. Another reason is that FOL has a complete inference
system while second-order logic (and thus any higher-order logic) cannot
have one, see, e.g., (11, §IX. 1.6). For further reasons for staying within
FOL, see, e.g., (5, §Why FOL?), (1), (35, §11), (19), (20).
156 Judit X. Madarász and Gergely Székely
There are many FOL axiomatizations of relativistic kinematics both

special and general, see, e.g., (1), (8), (9), (16), (30). However, as far as we
know, our co-authored paper (6) is the only one which deals with the FOL
axiomatization of relativistic dynamics, too. Newtonian and relativistic
kinematics are compared in the level of axioms in (5, §4.1). The main aim
of this paper is to compare the key axioms and theorems of Newtonian
and relativistic dynamics, too.
A first-order logic frame for dynamics

Our choice of vocabulary (basic concepts) is explained as follows. We
represent motion as the changing of spatial location of bodies in time. To
do so, we have reference-frames for coordinatizing events (sets of bodies)
and, for simplicity, we associate reference-frames with observers. There
are special kind of bodies which we call photons. For coordinatizing events,
we use an ordered field in place of the field of real numbers.1 Thus the
elements of this field are the quantities which we use for marking time and
space. In our axioms of dynamics we use relativistic masses of bodies as
a basic concept.
Motivated by the above, we now turn to fixing the FOL language of
our axiom systems. First we fix a natural number d ≥ 2 for the dimension
of spacetime. Our language contains the following non-logical symbols:
• unary relation symbols IOb (inertial observers), B (bodies), Ph

(photons) and Q (quantities),
• binary function symbols +, · and a binary relation symbol < (the

field operations and the ordering on Q),
• a 2 + d-ary relation symbol W (world-view relation), and
• a binary function symbol M (mass function).
We translate IOb(x), B(x), Ph(x) and Q(x) into natural language as

“x is an (inertial) observer,” “x is a body,” “x is a photon,” and “x is a
1 Replacing the field of real numbers by an ordered field not just increases the
flexibility of our theories but makes it possible to keep them within FOL,
which is crucial in axiomatic foundations, see, e.g., (5, Appendix: Why
first-order logic?).
Relativistic and Newtonian Dynamics in FOL 157
quantity.” (A more careful wording would be “x is a possible observer,”

“x is a possible body,” etc.) The bodies play the role of the “main char-
acters” of our spacetime models and they are “observed” (coordinatized
using the quantities) by the observers. This observation is coded by the
world-view relation by translating W(x, y, z1 , . . . , zd ) as “observer x coor-
dinatizes body y at spacetime location z1 , . . . , zd ,” (i.e., at space location
z2 , . . . , zd at instant z1 ). Finally we use the mass function to speak about
the relativistic masses of bodies according to observers, i.e., “M(x, y) is
the relativistic mass of body y according to observer x.”
IOb(x), B(x), Ph(x), Q(x), W(x, y, z1 , . . . , zd ), x = y and x < y are
the atomic formulas, where x, y, z1 , . . . , zd can be arbitrary variables or
terms built up from variables by using the field-operations and the mass
function M. The formulas are built up from these atomic formulas by
using the logical connectives not (¬), and (∧), or (∨), implies (→), if-and-
only-if (↔) and the quantifiers exists x (∃x) and for all x (∀x) for every
variable x.
The models of this language are of the form
U ; IOb, B, Ph, Q, +, ·, <, W, M,
where U is a nonempty set and IOb, B, Ph and Q are unary relations on

U , etc. For simplicity we write k ∈ IOb in place of IOb(k), etc.
We use the notation Qn := Q × . . . × Q (n-times) for the set of all n-
tuples of elements of Q. If p ∈ Qn , then we assume that p = p1 , . . . , pn ,
i.e., pi ∈ Q denotes the i-th component of the n-tuple p. We write
W(k, b, p) in place of W(k, b, p1 , . . . , pd ), and we write ∀p in place of
∀p1 , . . . , pd , etc.
We present each axiom at two levels. First we give an intuitive for-
mulation, then we give a precise formalization using our logical notation
(which can easily be translated into FOL formulas by inserting the defini-
tions into the formalizations). We seek to formulate easily understandable
axioms in FOL.
Our first axiom expresses very basic assumptions, such as: photons
are bodies, etc.
AxFrame Ph ⊆ B, the quantity part Q; +, ·, < is a Euclidean2 ordered

field, and the masses are positive elements of the quantity part, i.e.,

Q M(k, b) ∧ M(k, b) > 0.
For the FOL definition of linearly √ ordered field, see, e.g., (9). We use
the usual field operations 0, 1, −, /, definable within FOL. We also use
the vector-space structure of Qn , i.e., if p, q ∈ Qn and λ ∈ Q, then
−p, λp ∈ Qn . The Euclidean length of p ∈ Qn is defined as
p + q,
|p| := p21 + . . . + p2n , for any n ≥ 1. The Euclidean distance of p, q ∈
Qn is defined as |pq| := |p − q|. As usual,
is called a line iff there are
p, q ∈ Qd such that q = 0, . . . , 0 and
= {p + λq : λ ∈ Q}. And
Q+ := { λ ∈ Q : 0 < λ} denotes the set of positive elements of Q.
Set Qd is called coordinate system and its elements are referred to as
coordinate points. We use the notations
pσ := p2 , . . . , pd and pτ := p1
for the space component and the time component of p ∈ Qd , respec-

tively. The event evk (p) is the set of bodies observed by observer k at
coordinate point p, i.e.,
evk (p) := { b ∈ B : W(k, b, p) } .
The world-line of body b according to observer k is defined as the set of

coordinate points where b was observed by k, i.e.,

wlk (b) := p ∈ Qd : W (k, b, p) .
Kinematics
In this section we formulate our axioms on kinematics. Our first axiom
on observers states that they see the same events.
AxEv All observers coordinatize the very same events:
∀k, h ∈ IOb ∀p ∈ Qd ∃q ∈ Qd evk (p) = evh (q).
2 That is, a linearly ordered field in which positive elements have square roots.
To introduce our next axiom, we need a concept of inertial bodies. A

body is called inertial iff its world-line is a line for every observer. The
set of inertial bodies is denoted by IB, i.e.,
IB := {b ∈ B : ∀k ∈ IOb wlk (b) is a line}.
AxThEx below states that each observer can make thought experi-
ments in which it assumes the existence of “slowly moving” inertial bodies
(see, e.g., (4, p. 622)):
AxThEx For each observer there is a positive speed limit such that in each
spacetime location, in each direction, with any speed less than this
limit it is possible to “send out” an inertial body:
∀k ∈ IOb ∃λ ∈ Q+ ∀p, q ∈ Qd ∃b ∈ IB

|(p − q)σ | < λ(p − q)τ → p, q ∈ wlk (b) .
The following axiom system will be the common core of our axiom
systems for relativistic and Newtonian kinematics:
Kin := {AxEv, AxThEx, AxFrame}.
The world-view transformation between the coordinate systems

of observers k and h is the set of pairs of coordinate points p, q such
that k and h observe the same event in p and q, respectively, i.e.,
wkh := {p, q ∈ Qd × Qd : evk (p) = evh (q)}.
If R is a binary relation and X is a set, R[X] denotes the R-image

of X, i.e., R[X] := {b : ∃a ∈ X a, b ∈ R}.
Proposition 1. Assuming Kin, the world-view transformations are bi-
jections and take lines to lines, i.e., wkh [
] is a line for every line
and
k, h ∈ IOb.
A proof can be obtained from that of Thm. 3.1.1 in (5, pp. 160-170).
We extend Kin to an axiom system for special relativity by assuming

that the speed of light is 1 according to any observer.
AxPh The world-lines of photons are of slope 1, and for every observer,
there is a photon through two coordinate points if their slope is 1:

∀k ∈ IOb ∀p, q ∈ Qd |pσ − qσ | = |pτ − qτ | ↔

∃ph ∈ Ph p, q ∈ wlk (ph) .
We axiomatize special relativistic kinematics as follows:
SpecRelKin := Kin ∪ {AxPh}.
Convention 1. Whenever we write “wkh (p),” we mean that there is a

unique q ∈ Qd such that p, q ∈ wkh , and wkh (p) denotes this q.
Convention 2. We use the equation sign “=” in the sense of existential
equality (of partial algebra theory (1)), i.e., α = β abbreviates that both
α and β are defined and they are equal. See (22, Conv. 2.3.10, p. 31)
and (5, Conv. 2.3.10, p. 61). Similar convention applies for the binary
relations “<” and “=.”
To get an axiom system for Newtonian kinematics, we extend Kin
by an axiom saying that the simultaneity of events is independent from
observers.
AxAbsSim Simultaneity is absolute, i.e.,

∀k, h ∈ IOb ∀p, q ∈ Qd pτ = qτ → wkh (p)τ = wkh (q)τ .
We axiomatize Newtonian kinematics as follows:
NewtKin := Kin ∪ {AxAbsSim}.
Let us note that SpecRelKin and NewtKin differ in one axiom only.
But we will see in Prop. below that these two axiom systems are very
different, e.g., they are inconsistent together if we assume that there are
observers moving relative to each other. To formulate this statement we
need the following definition.
The speed vk (b) of body b according to observer k is defined as:
|pσ − qσ |
vk (b) := , for p, q ∈ wlk (b) with pτ = qτ
|pτ − qτ |
if wlk (b) is a subset of a line and contains coordinate points p and q with
pτ = qτ , otherwise vk (b) is undefined.
Ax∃IOb There are observers moving relative to each other.
∃k, h ∈ IOb ∃b ∈ IB vh (b) = vk (b) = 0.
Proposition 2. SpecRelKin ∪ NewtKin ∪ {Ax∃IOb} is inconsistent.

This proposition is a corollary of Thm. 2 below.
While in Newtonian kinematics there is no speed limit for observers
SpecRelKin implies that no observer can move faster than light if d ≥ 3
by the following theorem.
Theorem 1. Assume d ≥ 3 and SpecRelKin. Then there are no faster
than light observers, i.e.,

∀k, h ∈ IOb ∀b ∈ B vk (b) = 0 → vh (b) < 1 .

Moreover, ∀k, h ∈ IOb ∀b ∈ B vk (b) < 1 → vh (b) < 1 .

In the first formula of the theorem, the speeds of observers are captured
by speaking about resting bodies. For proof , see, e.g., (2, Prop. 1, Thm.
3), (22, 2.3.5, 2.8.25, 3.2.13), (16, Thm. 3, Thm. 5). We note that the
theorem remains true if we omit AxThEx from SpecRelKin.
Lines
and
are said to be orthogonal in the Euclidean sense iff
there are p, p ∈
and q, q ∈
such that p = p , q = q , and
(p1 − p1 )(q1 − q1 ) + (p2 − p2 )(q2 − q2 ) + . . . + (pd − pd )(qd − qd ) = 0.
If p, q ∈ Qd and p = q, then pq denotes the line passing through coordinate

points p and q.
By Thm. 2, two clocks separated in direction not orthogonal to the
direction of movement get out of synchronism.
Theorem 2. Assume SpecRelKin. Then two clocks remain in synchro-
nism iff they are separated in direction orthogonal to the direction of
movement. Formally: Let k, h ∈ IOb, b ∈ IB and p, q ∈ Qd be such
that vk (b) = 0, p = q and pτ = qτ . Then whk (p)τ = whk (q)τ iff pq is
orthogonal to wlh (b) in the Euclidean sense.
For proof , see, e.g., (4, Thm. 11.4, p. 626).
To formulate one more theorem on SpecRelKin, we need the following

definitions: Let p, q, r, s ∈ Qd . The Minkowski length of p is
⎧
⎨ p2τ − |pσ |2 if p2τ − |pσ |2 ≥ 0
μ(p) :=
⎩ − p2 − |pσ |2 otherwise
τ
and the Minkowski distance of p and q is μ(p, q) := μ(p − q). Segments

[pq] and [rs] are called Minkowski equidistant iff μ(p, q) = μ(r, s).
Theorem 3. Assume SpecRelKin. Then the world-view transformations

preserve the Minkowski equidistance, i.e.,
∀k, h ∈ IOb ∀p, q, r, s ∈ Qd

μ(p, q) = μ(r, s) → μ(wkh (p), wkh (q)) = μ(wkh (r), wkh (s)) .
Idea of proof of Thm. 3 is in the section after the next one.
Dynamics
In this section we formulate our axioms on dynamics. For convenience
we use the notation mk (b) := M (k, b) for the relativistic mass of body b
according to observer k.
The spacetime location lock (b, t) of body b at time instance t ∈ Q
according to observer k is defined to be the coordinate point p for which
p ∈ wlk (b) and pτ = t if there is such a unique p, otherwise lock (b, t) is
undefined, see Fig. 1.
The center of the masses cenk (b, c, t) of bodies b and c at time
instance t according to observer k is defined to be the coordinate point q
such that qτ = t and q is the point on the line-segment between lock (b, t)
and lock (c, t) whose distances from these two end-points have the same
proportion as that of the relativistic masses of b and c; and it is closer to
the “more massive” body, i.e.:

mk (b) lock (b, t) − cenk (b, c, t) = mk (c) cenk (b, c, t) − lock (c, t)
k
cenk (b, c, t)
c
b
lock (b, t) lock (c, t)

t
mk (c) mk (b)
cenk (b, c)
Figure 1: Illustration of lock (b, t), cenk (b, c, t) and cenk (b, c).
if lock (b, t) and lock (c, t) are defined, and otherwise cenk (b, c, t) is unde-
fined, see Fig. 1. We note that an explicit definition for cenk (b, c, t) is the
following:
mk (b) mk (c)
cenk (b, c, t) = lock (b, t) + lock (c, t),
mk (b) + mk (c) mk (b) + mk (c)
if lock (b, t) and lock (c, t) are defined, otherwise cenk (b, c, t) is undefined.
The center-line of the masses of bodies b and c according to observer
k is defined as
cenk (b, c) := {cenk (b, c, t) : t ∈ Q and cenk (b, c, t) is defined}.
In Newtonian dynamics two bodies can be substituted by one body

living on the center-line of the two bodies and having mass equal to the
sum of the masses of the two bodies. The crucial point in this statement
is that different observers agree as for the center-line of inertial bodies (up
to world-view transformations), which can be formalized as follows.
AxCen The world-view transformations take the center-line of two inertial

bodies to the center-line of the two bodies.
∀k, h ∈ IOb ∀b, c ∈ IB wkh [cenk (b, c)] = cenh (b, c).
However intuitive and natural AxCen is, it does not hold in the “standard
model” of special relativity. Moreover, it is inconsistent with SpecRelKin
if we assume that there are observers moving relative to each other.
Theorem 4. Assume d ≥ 3. Then SpecRelKin ∪ {Ax∃IOb, AxCen} is

inconsistent.
The proof of Thm. 4 is in the next section.
Thus, by Thm. 4, two bodies cannot be replaced by one in relativistic

dynamics. Therefore, if we want to build a consistent relativistic dynamics
based on this assumption, we have to weaken AxCen. The solution is to
assume it only for meeting bodies.
AxCen− The world-view transformations take the center-line of two meet-

ing inertial bodies to the center-line of the two bodies.
∀k, h ∈ IOb ∀b, c ∈ IB

wlk (b) ∩ wlk (c) = ∅ → wkh [cenk (b, c)] = cenh (b, c) .
The main axiom of dynamics is AxCen− . The remaining axioms of our

axiom system are only simplifying axioms.
The rest mass m0 (b) of body b is defined as m0 (b) = λ if (1) there is
an observer according to which b is at rest and the relativistic mass of b is
λ, and (2) for every observer according to which b is at rest the relativistic
mass of b is λ, i.e., m0 (b) = λ iff
∃k ∈ IOb (vk (b) = 0 ∧ mk (b) = λ) ∧ ∀k ∈ IOb (vk (b) = 0 → mk (b) = λ).
AxCen− (together with SpecRelKin) implies that the mass of a body

necessarily depends on the observer. The reason for this fact is that the
simultaneities of observers moving relative to each other in SpecRelKin
differ from each other, and this implies that the proportions involved in
AxCen− change, too. See Prop. 4.1 and Fig. 3 in (6). The next axiom
states that the relativistic mass of a body depends at most on its rest
mass and its speed.
AxSpeed The relativistic masses of two inertial bodies are the same if both
of their rest masses and speeds are equal:
∀k ∈ IOb ∀b, c ∈ IB

m0 (b) = m0 (c) ∧ vk (b) = vk (c) → mk (b) = mk (c) .
Our next axiom on dynamics states that each observer can make ex-
periments by putting stationary inertial bodies with arbitrary rest mass
to any coordinate point.
AxRest In the coordinate system of any observer there is a resting inertial

body with arbitrary rest mass at any coordinate point.

∀k ∈ IOb ∀λ ∈ Q+ ∀p ∈ Qd ∃b ∈ IB m0 (b) = λ ∧ p ∈ wlk (b) .
Let IB0 denote the set of inertial bodies having rest mass.
AxMedian For every two inertial bodies having rest mass, there is an ob-
server for which they have the same speed:
∀b, c ∈ IB0 ∃k ∈ IOb vk (b) = vk (c).
Let us collect the axioms for dynamics together.
Dyn := {AxCen− , AxSpeed, AxRest, AxMedian}.

By adding Dyn to our kinematical axiom systems we get the respective
dynamical ones.
NewtDyn := Dyn ∪ Kin ∪ {AxAbsTime} = Dyn ∪ NewtKin,
SpecRelDyn := Dyn ∪ Kin ∪ {AxPh} = Dyn ∪ SpecRelKin.
Let us note that Dyn is the common dynamical core of the two axiom
systems, which also differ in one axiom only.
Thms. 5 and 6 below give the connection between the rest mass and
the relativistic mass of an inertial body. Their conclusions are well known
results of relativistic and Newtonian dynamics. However, in the usual
literature, the assumptions are stronger and not stated explicitly.
Theorem 5. Assume SpecRelDyn. Let k be an observer and b be an

inertial body having rest mass. Then vk (b) < 1 and

m0 (b) = mk (b) 1 − vk (b)2 .
Theorem 6. Assume NewtDyn. Let k be an observer and b be an inertial

body having rest mass. Then
m0 (b) = mk (b).
The proofs of Thms. 5 and 6 are in the next section.

Let us note that, by Thms. 5 and 6, axiom systems SpecRelDyn and
NewtDyn differing in one axiom have radically different consequences.
Coordinate points p, q and r are called collinear iff there is a line
such that p, q, r ∈
. A map f : Qd −→ Qd is called an affine trans-
formation, if it preserves collinearity and the ratios of distances, i.e., for
every distinct and collinear coordinate points p, q and r coordinate points
f (p), f (q) and f (r) are collinear and |pq| · |f (q)f (r)| = |f (p)f (q)| · |qr|.
By Prop. , it follows that the world-view transformations are affine
transformations composed by filed-automorphism induced mappings in
models of Kin. It can be proved that, in models of SpecRelKin and
NewtKin, the world-view transformations are not necessarily affine trans-
formations, i.e., there are models in which field-automorphism induced
mappings occur in the world-view transformations. By Thm. 7, this is
not the case in models of NewtDyn and SpecRelDyn.
Theorem 7. The world-view transformations are bijective affine trans-
formations in models of NewtDyn and SpecRelDyn.
By Thm. 5, SpecRelDyn implies that for every inertial body b the
quantity mk (b) 1 − vk (b)2 is independent of observer k if b has rest mass.
By Thm. 8, the same holds for every inertial body b moving slower than
light.
Theorem 8. Assume SpecRelDyn. Let b be an inertial body such that
vk (b) < 1 for an observer k. Then vk (b) < 1 for every observer k and

∀k, h ∈ IOb mk (b) 1 − vk (b)2 = mh (b) 1 − vh (b)2 .
By our definition, the rest mass of an inertial body can be undefined

even if there is an observer according to which the body is at rest. By
the following immediate corollary of Thm. 8, the rest mass of a body is
defined whenever there is an observer according to which the body is at
rest if we assume SpecRelDyn.
Corollary 9. Assume SpecRelDyn. Let b be an inertial body such that

vk (b) = 0 for some k ∈ IOb. Then b has a rest mass.
By Thm. 6, axiom system NewtDyn implies that relativistic mass of

an inertial body is observer independent if the body has rest mass. By
Thm. 10, the same holds for inertial bodies with “finite” speeds.
Theorem 10. Assume NewtDyn. Let b be an inertial body such that

vk (b) is defined for some k ∈ IOb. Then
∀k, h ∈ IOb mk (b) = mh (b).

By the following corollary, theory NewtDyn implies AxCen, the axiom
which is inconsistent with SpecRelDyn, see Thm. 4. This fact also shows
great difference between the two theories of dynamics.
Corollary 11. NewtDyn |= AxCen.
The proof of Cor. 11 is in the next section.
Proofs
Idea of proof of Thm. 3. Assume first that d > 2. One can prove,
by Alexandrov-Zeeman theorem, that every world-view transformation
is a composition of a Poincaré transformation, a dilation and a field-
automorphism-induced mapping, cf. (3, Thm. 1.2). All of these mappings
preserve the Minkowski equidistance. Thus the world-view transforma-
tions also preserve the Minkowski equidistance.
We note that a similar proof can be obtained for d = 2, cf. (3, Thm.
1.4).
Another proof can be obtained as follows. Coordinate points p and

q are called lightlike separated, in symbols pλq iff |pτ − qτ | = |pσ − qσ |.
Furthermore, p and q are timelike separated iff |pτ − qτ | > |pσ − qσ |.
Assume first that d = 2. Let p, q and q be distinct coordinate points. It
can be seen that
μ(p, q) = μ(p, q ) ↔

∃s, s s = s ∧ qλs ∧ q λs ∧ qλs ∧ q λs ∧ coll(p, s, s ) . (1)
Then timelike separatedness is FOL definable from lightlike separat-

edness by Alexandrov-Zeeman theorem. Let p, q and q be distinct points
such that p and q are timelike separated and the same holds for p and q .
Then one can see that (1) above holds for p, q and q .
By the above one can prove that Minkowski equidistance for timelike
separated pairs of points is FOL definable from lightlike separatedness
and collinearity. World-view transformations preserve lightlike separated-
ness and collinearity by AxPh and Prop. . Thus they preserve Minkowski
equidistance for timelike separated pairs of points. The general case can
be reduced to the timelike and lightlike cases.
Proof of Thm. 4. The proof goes by contradiction. Let M be a model

of SpecRelKin ∪ {Ax∃IOb, AxCen}. Let k, h ∈ IOb and b ∈ IB be such
that vh (b) = vk (b) = 0, see Fig. 2. Let c ∈ IB be such that vh (c) = 0
and wlh (b) and wlh (c) do not meet, i.e., wlh (b) and wlh (c) are skew lines.
Such c exists by AxThEx. Furthermore, vh (b) < 1 and vk (c) < 1 by Thm.
1. Thus wlh (b) and wlk (c) are not “horizontal” lines, i.e., for every t ∈ Q
there are p ∈ wlh (b) and q ∈ wlk (c) such that t = pτ = qτ . Let p ∈ wlh (c)
and q ∈ wlh (b) be such that pτ = qτ and pq is not orthogonal to wlh (b) in
the Euclidean sense. It is easy to see that there are such p and q. Then by
Thm. 2, whk (p)τ = whk (q)τ . Let r ∈ wlh (c) and s ∈ wlh (b) be such that
whk (q)τ = whk (r)τ and rτ = sτ . Then pq and rs are skew lines since the
world-lines of b and c are skew lines. Center-line cenh (b, c) intersects lines
pq and rs and it does not go through points p, q, r and s. Thus it does
not intersect line qr since the world-lines of b and c are skew lines. On
the other hand center-line cenk (b, c) intersects line whk (q)whk (r). That is
a contradiction since the world-view transformations are bijections taking
lines to lines and center-lines to center-lines by Thm. and AxCen.
wlh (c) wlh (b)
r τ = sτ
s r whk (r)
whk (q)
whk (q)τ = whk (r)τ
p b cenk (b, c) c
q pτ = qτ
b cenh (b, c) c
coordinate system of h coordinate system of k
Figure 2: Illustration for the proof of Thm. 4.
Proof of Thm. 5. Let k be an observer and b be an inertial body having

rest mass. Let v := vk (b).
First we prove from SpecRelDyn that v < 1. For d ≥ 3, v < 1 already
follows from SpecRelKin by Thm. 1. To see that v < 1 for d = 2, let b be
a body having rest mass such that vk (b ) = 0. Such a b exists by AxRest.
By AxMedian there is an observer according to which b and b have the
same speeds. But then, by SpecRelKin, it can be proved that v < 1, cf.
(5, Thm. 2.7.2, p. 110).
If v = 0, the conclusion of the theorem holds. Thus we can assume
that v = 0. Let c be an inertial body such that vk (c) = 0, m0 (c) = m0 (b)
and b and c meet. See Fig. 3. Such a c exists by AxRest. Let A be
the coordinate point where the world-lines of b and c meet. Let B and
C be the coordinate points on the world-lines of b and c, respectively,
such that Aτ −√ Cτ = 1 and Bτ = Cτ . Then |AC| = 1, |BC| = v
and μ(A, B) = 1 − v 2 . Let D be the center of masses of b and c at
Bτ = Cτ , i.e., D := cenk (b, c, Bτ ). By definition of center of masses,
mk (b)|BD| = m0 (b)|DC|. Thus
|BD|
m0 (b) = mk (b). (2)
|DC|
wkh
A
A whk
√
√ 1 − v2
μ(A, B) = 1 − v 2
G 1 B E
E
F
b c
F
B
D C μ(A , B ) = μ(A , E )
b
v c
coordinate system of k coordinate system of h
Let h be an observer according to which the speeds of b and c coincide.

Such an h exists by AxMedian. Let us turn our attention to the coordinate
system of observer h illustrated by the right hand side of Fig. 3. Let
A and B be the wkh -images of A and B, respectively. Let E be a
coordinate point on the world-line of c such that Eτ = Bτ . Let F be the
center of masses of b and c at Eτ = Bτ , i.e., F := cenh (b, c, Eτ ). Since the
rest masses and the speeds of b and c coincide, their relativistic masses
coincide by AxSpeed. Thus |B F | = |F E | by the definition of center
of masses. Consequently, μ(A , B ) = μ(A , E ), i.e., segments [A B ] and
[A E ] are Minkowski equidistant.
Now we turn our attention to the coordinate system of observer k
illustrated by the left hand side of Fig. 3. Let F and E be the whk
images of F and E . Then, by AxCen− , F ∈ AD = cenk (b, c) since
F ∈ cenh (b, c). Furthermore, E ∈ BF since whk takes lines to lines
by Prop. . The world-view transformation whk preserve the Minkowski
equidistance by Thm. 3. Consequently, segments [AB] and [AE] as well
as [BF ] and √ [F E] are Minkowski equidistant. Thus |AE| = μ(A, E) =
μ(A, B) = 1 − v 2 and |BF | = |F E|.
Triangles BDF and EGF are congruent and triangles AGE and ADC
are similar. Thus
|BD| |GE| |AE|
= = = 1 − v2 .
|DC| |DC| |AC|
√
By that and (2), we get m0 (b) = mk (b) 1 − v 2 . That completes the
proof.
Proof of Thm. 6. Let k be an observer and b be an inertial body having

rest mass. Then there is an observer k such that vk (b) = 0. Thus, by
AxAbsSim, vk (b) is not “infinite,” i.e., vk (b) is defined. If vk (b) = 0, the
conclusion of the theorem holds. Assume that vk (b) = 0. Let c be an
inertial body such that vk (c) = 0, m0 (c) = m0 (b) and b and c meet, see
the left-hand side of Fig. 4. Such a c exists by AxRest. Let h be an
wkh
w := whk
B C
B C D

D
b b c
c
coordinate system of k coordinate system of h
observer according to which the speeds of b and c coincide. Such an h

exists by AxMedian. Let us turn our attention to the coordinate system
of observer h illustrated by the right-hand side of Fig. 4. Let B ∈ wlh (b)
and C ∈ wlh (c) be distinct points such that Bτ = Cτ . Let D be the center
of masses of b and c at time instance Bτ = Cτ according to observer h,
i.e., D := cenh (b, c, Bτ ). Since the speeds and the rest masses of b and
c coincide, their relativistic masses coincide by AxSpeed. But then, by
definition of center of masses, D is the midpoint of segment [BC], i.e.,

|BD| = |DC|.
w := whk is a bijection taking lines to lines by Prop. . Furthermore,
it takes the midpoint of a segment to the midpoint of the w-image of the

segment, i.e., w (p + q)/2 = w(p) + w(q) /2 for every p, q ∈ Qd . The

proof of this statement is illustrated in Fig. 5.
q w
w(q)
w
p+q
2 w(p)+w(q)
2
w
w(p)
Figure 5: w takes midpoint of a segment to the midpoint of the w-image

of the segment. The proof of that is based on the fact that w takes par-
allelograms to parallelograms and the diagonals of parallelograms bisect
each other.
Now let us turn our attention to the coordinate system of observer k

illustrated by the left-hand side of Fig. 4. Let B , C and D be the w-
images of B, C and D. Then Bτ = Cτ = Dτ by AxAbsSim. Furthermore,
D is the midpoint of segment [B C ], i.e., |B D | = |D C | since D is the
midpoint of segment [BC] and w takes midpoints to midpoints. Further-
more, D ∈ cenk (b, c), by AxCen− , since D ∈ cenh (b, c). Thus D is the
center of masses of b and c at time instance Bτ = Cτ = Dτ according to
observer k, i.e., D = cenk (b, c, Bτ ). But then, by definition of center of
masses, mk (b) = mk (c) = m0 (c) = m0 (b) since |B D | = |D C |. That
completes the proof.
Proof of Thm. 7. Assume first Kin ∪ {AxRest}. Let k, h ∈ IOb. Let

w := wkh . Then w is a bijection taking lines to lines by Prop. . Thus
it takes parallel lines to parallel ones. Furthermore, it takes the mid-
point of a segment to the midpoint of the w-image of the segment, i.e.,

w (p + q)/2 = w(p) + w(q) /2 for every p, q ∈ Qd . The proof of this

statement is illustrated in Fig. 5.
To prove that the world-view transformations are affine transforma-
tions in models of NewtDyn and SpecRelDyn, it is enough to prove that
|AC|/|CB| = |w(A)w(C)|/|w(C)w(B)| for every distinct A, B, C ∈ Qd
with Aτ = Bτ = Cτ . To prove that, let A, B and C be distinct coordinate
points such that Aτ = Bτ = Cτ . We can assume that C is between A and
B. See the left-hand side of Fig. 6. Let a ∈ IB be such that A ∈ wlk (a)
D w D
D
w
a bx v = vh (bx )
a bx a bx
Fx Gx

A E
A Cx C B A
Cx B Cx
coordinate system of h coordinate system of h

B
coordinate system of k
NewtDyn SpecRelDyn
and vk (a) = 0. Let D ∈ wlk (a) be such that Dσ = Bσ . For every x ∈ Q+ ,

fix an inertial body bx such that wlk (bx ) = BD and the rest mass of b
is x, i.e., m0 (bx ) = x. Such bodies exist by AxRest. For every x ∈ Q+ ,
let Cx be the center of masses of a and bx at time-instance Aτ = Bτ
according to k, i.e., Cx := cenk (a, bx , Aτ ). By definition of the center of
masses, |ACx |/|Cx B| = mk (bx )/mk (a) = m0 (bx )/mk (a) = x/mk (a). Let
1
λ := . (3)
mk (a)
Then
|ACx |
∀x ∈ Q+ = λx. (4)
|Cx B|
Let A , B , D and Cx be the w-images of A, B, D and Cx , respectively.
Now assume NewtDyn. Then mh (bx ) = m0 (bx ) = x by Thm. 6.
Furthermore, Aτ = Bτ = (Cx )τ by AxAbsSim, see the middle of Fig. 6.
Then D Cx = cenh (a, bx ) by AxCen− . Thus Cx is the center of masses of
a and bx at time instance Aτ according to h, i.e., Cx = cenh (a, bx , Aτ ).
Thus |A Cx |/|Cx B | = mh (bx )/mh (a) = x/mh (a). Let
1
λ := . (5)
mh (a)
Then
|A Cx |
∀x ∈ Q+ = λ x. (6)
|Cx B |
Let us now consider the case x = mk (a). Then Cmk (a) is the midpoint
of the segment [AB], i.e., |ACmk (a) | = |Cmk (a) B| by (3) and (4). Then

Cm k (a)
is the midpoint of the segment [A B ], since w takes the midpoint
of a segment to the midpoint of the w-image of the segment. But then,
by (4) and (6), λmk (a) = λ mk (a) = 1. Hence
λ = λ . (7)
Thus, by (4) and (6), ∀x ∈ Q+ |ACx |/|Cx B| = |A Cx |/|Cx B |. Clearly,
there is an x such that Cx = C. Then |AC|/|CB| = |A C |/|C B |, which
completes the proof for the case of NewtDyn.
Now assume SpecRelDyn. See the right-hand side of Fig. 6. Let
us note, that the speed vh (bx ) is independent√of the choice√of x. Let
v denote this speed. Then mh (bx ) = m0 (bx )/ 1 − v 2 = x/ 1 − v 2 by
Thm. 5. Clearly, D Cx = cenh (a, bx ) by AxCen− . Thus Aτ = Dτ = Bτ .
Let E ∈ D B be such that Eτ = Aτ . Let Fx be the intersection of D Cx
and A E . Clearly, Fx = cenh (a, bx , Aτ ). Thus
|A Fx | mh (bx ) x

= = √ .
|Fx E | mh (a) mh (a) 1 − v 2
Let Gx ∈ A E be such that B Gx is parallel to D Cx . Now,
|A Cx | |A Fx | |A Fx | |Fx E | x |D E |

= = = √ .
|Cx B | |Fx Gx | |Fx E | |Fx Gx | mh (a) 1 − v 2 |D B |
Let
|D E |
λ := √ . (8)
mh (a) 1 − v 2 |D B |
Then
|A Cx |
∀x ∈ Q+ = λ x. (9)
|Cx B |
Now,
λ = λ (10)
can be proved by (3), (4) and (9) exactly the same way as λ = λ was
proved for the case of NewtDyn. The rest of the proof is analogous to the
proof for the case of NewtDyn.
Proof of Thm. 8. Let a be an inertial body and k, h ∈ IOb be such

that vk (a) < 1. Then, by Thm. 1, vh (a) < 1 if d ≥ 3. To prove that
vh (a) < 1 for arbitrary d, let bk and bh be inertial bodies having rest
masses such that vk (bk ) = 0 and vh (bh ) = 0. Such bodies exist by AxRest.
By AxMedian, there is an observer according to which bk and bh have the
same speeds. But then it can be proved that vh (bk ) < 1, cf. (5, Thm.
2.7.2, p. 110). By that, it is easy to prove that vh (a) < 1, too.
Assume first, that vk (a) = 0. We will use the proof of Thm. 7 for
the case of SpecRelDyn. So we can assume that k, h and a are as in the
second paragraph of that proof, and let A, B, bx etc. be as in that proof,
see the left-hand and right-hand sides of Fig. 6. It can be proved that
μ(D , A ) μ(D, A)

= = 1 − vk (a)2
μ(D , B ) μ(D, B)
by Thm. 3 and Thm. 7. Then

√
|D E | μ(D , E ) μ(D , E ) μ(D , A ) 1 − v 2 1 − vk (a)2
= = = .
|D B | μ(D , B ) μ(D , A ) μ(D , B ) 1 − vh (a)2
By thisequation, (3), (8) and

(10), we conclude that
mk (a) 1 − vk (a)2 = mh (a) 1 − vh (a)2 .
Now assume that vk (a) = 0. Let a∗ be an inertial body such that
wlk (a∗ ) = wlk (a) and m0 (a∗ ) = mk (a). Such an a∗ exists by AxRest.
Then
mk (a) = m0 (a∗ ) = mh (a∗ ) 1 − vh (a)2 (11)
by Thm. 5. Let b be an inertial body such that vk (b) = 0 and wlk (a) ∩
wlk (b) = ∅. Clearly, cenk (a, b) = cenk (a∗ , b). Then cenh (a, b) = cenh (a∗ , b)
by AxCen− . Thus mh (a) = mh (a∗ ). This equation together with (11)

Proof of Thm. 10. The proof is analogous to that of Thm. 8. Let a be

an inertial body and k, h ∈ IOb be such that vk (a) is defined.
Assume first, that vk (a) = 0. We will use the proof of Thm. 7 for the
case of NewtDyn. We can assume that k, h and a are as in that proof. By
(3), (5) and (7) we get that mk (a) = mh (a).
Now assume that vk (a) = 0. Let a∗ be an inertial body such that
wlk (a∗ ) = wlk (a) and m0 (a∗ ) = mk (a). Such an a∗ exists by AxRest.
Then
mk (a) = m0 (a∗ ) = mh (a∗ ) (12)
by Thm. 6. Let b be an inertial body such that vk (b) = 0 and wlk (a) ∩
wlk (b) = ∅. Clearly, cenk (a, b) = cenk (a∗ , b). Then cenh (a, b) = cenh (a∗ , b)
by AxCen− . Thus mh (a) = mh (a∗ ). This equation together with (12)
Proof of Cor. 11. Assume NewtDyn. Let k, h ∈ IOb and a, b ∈ IB. We

would like to prove that
wkh [cenk (a, b)] = cenh (a, b). (13)
First assume that vk (a) is “infinite,” i.e., undefined. Then vh (a) is

also “infinite,” by AxAbsSim. Hence, cenk (a, b) = cenh (a, b) = ∅. Then
(13) holds. The same holds if vk (b) is “infinite”.
If wlk (a) ∩ wlk (b) = ∅, (13) holds by AxCen− . Thus we can assume
that wlk (a) ∩ wlk (b) = ∅. Now assume that vk (a) and vk (b) are not
“infinite,” i.e., both are defined. Then vh (a) and vh (b) are also defined
by AxAbsSim. First we prove that wkh [cenk (a, b)] ⊆ cenh (a, b). To prove
this statement, let C ∈ cenk (a, b). Let A ∈ wlk (a) and B ∈ wlk (b)
be such that Aτ = Bτ = Cτ . Then, by definition of center of masses,
|AC|/|CB| = mk (b)/mk (a). Let A , B and C be the wkh -images of
A, B and C. Clearly, Aτ = Bτ = Cτ by AxAbsSim. By Thm. 10,
mk (b) = mh (b) and mk (a) = mh (a). By Thm. 7, wkh is a bijective affine
transformation. Consequently,
|A C |/|C B | = |AC|/|CB| = mk (b)/mk (a) = mh (b)/mh (a).

Hence C ∈ cenh (a, b). Thus wkh [cenk (a, b)] ⊆ cenh (a, b). Analogously,
whk [cenh (a, b)] ⊆ cenk (a, b) holds. Since wkh and whk are bijections and
inverses of each other, we conclude that (13) holds.
Concluding remarks
We have shown that Newtonian and relativistic dynamics can be axiom-
atized (within FOL) such that they differ in one axiom only. However,
in the level of consequences, they have radical differences. The most sur-
prising difference is that AxCen, an apparently harmless consequence of
Newtonian dynamics, is inconsistent with relativistic dynamics.
Acknowledgments: Research supported by Hungarian National Foun-
dation for Scientific Research grant No T73601 as well as by Bolyai Grant
for Judit X. Madarász.
Bibliography
[1] H. Andréka, P. Burmeister, and I. Németi. Quasivarieties of partial

algebras – a unifying approach towards a two-valued model theory for
partial algebras. Studia Sci. Math. Hungar. 16, 1981, pp. 325–372.
[2] H. Andréka, J. X. Madarász, and I. Németi. “Logical analysis of

special relativity theory,” in: J. Gerbrandy, M. Marx, M. de Ri-
jke, and Y. Venema (Eds.), Essays dedicated to Johan van Ben-
them on the occasion of his 50th birthday. Vossiuspers, Ams- ter-
dam University Press, 1999. CD-ROM, ISBN: 90 5629 104 1,
http://www.illc.uva.nl/j50.
[3] H. Andréka, J. X. Madarász, and I. Németi. “Logical axiomatiza-

tions of space-time. Samples from the literature,” in: Non-Euclidean
geometries, volume 581 of Mathematics and Its Applications. New
York: Springer, 2006, pp. 155-185.
[4] H. Andréka, J. X. Madarász, and I. Németi, “Logic of space-time and

relativity theory,” in: M. Aiello, I. Pratt-Hartmann and J. van Ben-
them (Eds.), Handbook of Spatial Logics. Dordrecht: Springer, 2007,

pp. 607-711.
[5] H. Andréka, J. X. Madarász, and I. Németi, “On the logical struc-

ture of relativity theories.” Research report, Alfréd Rényi Insti-
tute of Mathematics, Budapest, 2002. With contributions from
A. Andai, G. Sági, I. Sain and Cs. Tőke. http://www.math-
inst.hu/pub/algebraic-logic/Contents.html. 1312 pp.
[6] H. Andréka, J. X. Madarász, I. Németi, and G. Székely, “Axiomatiz-

ing relativistic dynamics without conservation postulates,” in: Studia
Logica 89, 2, 2008, pp. 163-186.
[7] J. Ax, “The elementary foundations of spacetime,” in: Foundations

of Physics 8, 7-8, 1978, pp. 507-546.
[8] S. A. Basri, A Deductive Theory of Space and Time. Amsterdam:

North-Holland 1966.
[9] T. Benda, “A Formal Construction of the Spacetime Manifold,” in:

Journal of Philosophical Logic 37, 5, 2008, pp. 441-478.
[10] C. C. Chang and H. J. Keisler, Model theory. Amsterdam: North-

Holland 1973, 1977, 1990.
[11] H. D. Ebbinghaus, J. Flum and W. Thomas, Mathematical logic. New

York: Springer-Verlag 1994.
[12] R. Goldblatt, Orthogonality and spacetime geometry. New York:

Springer-Verlag 1987.
[13] J. X. Madarász. Logic and Relativity (in the light of definability the-
ory). PhD thesis, Eötvös Loránd Univ., Budapest, 2002.
[14] J. X. Madarász, I. Németi, and G. Székely, “Twin paradox and the

logical foundation of relativity theory,” in: Foundations of Physics
36, 5, 2006, pp. 681-714.
[15] J. X. Madarász, I. Németi, and G. Székely, “First-order logic foun-

dation of relativity theories,” D. Gabbay, S. Goncharov and M. Za-
kharyaschev (Eds.), in: Mathematical problems from applied logic II.
New York: Springer, 2007, pp. 217-252.
[16] J. X. Madarász, I. Németi, and Cs. Tőke. On generalizing the

logic-approach to space-time towards general relativity: first steps.
in: V. F. Hendricks, F. Neuhaus, S. A Pedersen, U. Scheffler and
H. Wansing (Eds.), First-Order Logic Revisited, Logos Verlag, Berlin,
2004, pp. 225–268.
[17] V. Pambuccian, “Alexandrov-Zeeman type theorems expressed in

terms of definability,” in: Aequationes Mathematicae 74, 3, 2007,
pp. 249-261.
[18] G. Székely, First-Order Logic Investigation of Relativity Theory with

an Emphasis on Accelerated Observers. PhD thesis, Eötvös Loránd
Univ., Budapest, 2009.
[19] J. Väänänen, “Second-order logic and foundations of mathematics,”

in: Bulletin of Symbolic Logic 7, 4, 2001, pp. 504-520.
[20] J. Woleński “First-order logic: (philosophical) pro and contra,” in:

V. F. Hendricks, F. Neuhaus, S. A. Pedersen, U. Scheffler and
H. Wansing (Eds.), First-Order Logic Revisited. Berlin: Logos Verlag
2004, pp. 369-398.
Gergely Székely
On Why-Questions in Physics
Introduction
In the natural sciences, the most interesting and relevant questions are
the so-called why-questions. What is a why-question? A why-question
is nothing else than a question in the form “Why P ?” (or “Why is P
true?”) where P is an arbitrary statement.
There are several different approaches to why-questions and explana-
tions in the literature; see, e.g., (13), (7), (8), (18). However, most of the
literature deals with why-questions about particular events, such as “Why
did Adam eat the apple?”. Even the best known theory of explanation,
Hempel’s covering law model, is designed for explaining particular events.
Here we will only deal with purely theoretical why-questions about gen-
eral phenomena of physics, for instance “Why can no observer move faster
than light?” or “Why are Kepler’s laws valid?”.
Here we are not going to develop a whole new theory of why-questions
in physics. We will just touch upon some ideas and examples relevant to
our subject.
Answering Why-Questions
How to answer a why-question? For example, let NoFTL be the statement
“No observer can move faster than light,” which is one of the several
astonishing predictions of relativity. As it is a statement that is hard to
believe, it is natural to ask why we think it is true. The standard answers
to the question “Why NoFTL?” are
1. “NoFTL is true because the 4-dimensional Minkowski spacetime

over R (the field of real numbers) is a good model of our physical
world; and in this model NoFTL is valid.”
2. “NoFTL is an axiom (of Special Relativity).”

182 Gergely Székely
Neither of these answers is satisfactory to a logician. The problem with

the first is that it refers to one particular model and not to a list of
axioms. The second does not really answer the question, it is a kind of
“just because” answer.
How to give satisfactory answers to why-questions in physics? First of
all to answer the question “Why P ?”, we need a formal language in which
we can formulate P and the possible answers to it. Let us fix one such
language.1 The possible answers to the question “Why P ?” are consistent
theories which do not contain P as an axiom.2 Here we do not require
theories to be deductively closed systems, as we would like to separate
the assumptions of the theory from its consequences. Hence we use the
term ‘theory’ as a synonym for axiom system. Let us call a possible answer
an acceptable answer if it implies P .
To present some examples of acceptable answers to the question “Why
NoFTL?”, let us consider the following two-sorted first-order language:
{ Q , +, ·, <; B , IOb, Ph; W },
where Q is the sort of quantities and B is the sort of bodies; · and +

are binary function symbols and < is a binary relation symbol of sort
Q ; IOb (inertial observers) and Ph (light signals or photons) are unary
relation symbols of sort B ; and W (the world-view relation) is a 6-ary
relation symbol of the type B × B × Q × Q × Q × Q . Relations IOb(o) and
Ph(b) are translated as “o is an inertial observer,” and “b is a photon,”
respectively. We use the world-view relation W to speak about coordina-
tization by translating W(o, b, x, y, z, t) as “observer o coordinatizes body
b at spacetime location x, y, z, t,” (that is, at space location x, y, z and
at instant t).
In this language we can define the term speedo (b) as the speed of a
body b according to observer o. Hence we can formulate that no observer
1 Every concept we introduce here is relative to a fixed language in which P

can be formulated.
2 It is important that we restrict our definition to possible answers to why-
questions in physics because, for example, in mathematics the possible an-
swers to why-questions are usually proofs and not theories. See, e.g., (15).
On Why-Questions in Physics 183
can move faster than light. Moreover, we can prove the following:
SpecRel |= ∀o, o , p IOb(o) ∧ IOb(o ) ∧ Ph(p) =⇒ speedo (o ) < speedo (p),
where SpecRel is the consistent axiom system of the following axioms:
AxField: The quantity part Q ; +, ·, < is an ordered field.
AxSelf: Every observer coordinatizes itself at a coordinate point if and

only if its space component is the origin, that is, space location
0, 0, 0.
AxPh: The speed of light signals is 1 according to every inertial observer.
AxEv: Every inertial observer coordinatizes the very same set of events.
AxSymd: Inertial observers agree as for the spatial distance between events
if these events are simultaneous for both of them.
For the formulation of these concepts and axioms, see, e.g., (5). So SpecRel
is an acceptable answer to the question “Why NoFTL?”, see (5, Thm.1).
How to Compare the Different Answers

How to get better and better answers to a certain why-question? The
basic idea is that “the less it assumes, the better an answer is.” Let us
try to make this idea more precise by introducing the following concepts:
1. T h2 is nonworse than T h1 as an answer to the question “Why P ?”

if all the formulas of T h2 are consequences of T h1 , that is, T h1
implies ϕ2 for any formula ϕ2 ∈ T h2 .
2. T h2 is piecewise nonworse than T h1 as an answer to the question

“Why P ?” if for any formula ϕ2 ∈ T h2 there is ϕ1 ∈ T h1 such that
ϕ1 implies ϕ2 .
It is easy to see the following:
1. Both concepts above are preorders on sets of formulas, that is, they
are transitive and reflexive relations.
2. T h2 is a piecewise nonworse answer than T h1 if T h2 ⊆ T h1 .

3. T h2 is a nonworse answer than T h1 if it is piecewise nonworse.

Let us say that T h2 is (piecewise) better than T h1 if it is (piecewise)
nonworse and nonequivalent according to the equivalence relation defined
by the corresponding preorder.
These definitions fit in with the following idea of Michael Friedman
(13): “Science increases our understanding of the world by reducing the
total number of independent phenomena that we have to accept as ulti-
mate or given.”
To present another example, let us note that the following theorem
can also be proved:
SpecRel0 |= ∀o, o , p IOb(o) ∧ IOb(o ) ∧ Ph(p) =⇒ speedo (o ) < speedo (p),
where SpecRel0 consists of the following four axioms only: AxField, AxSelf,
AxPh and AxEv. Hence SpecRel0 is a piecewise better answer to the ques-
tion “Why NoFTL?” than SpecRel. Further answers to this question,
which are not comparable to SpecRel0 according to the preorders above,
can be found in (13, Thms. 3 and 5).
Let us note that, according to the definition above, a theory consisting
of two (nonequivalent) axioms is a piecewise better answer than a theory
consisting of their conjunction. That is a nice property as it makes the
introduction of the following concept possible. Let us say that a possible
answer to the question “Why P ?” is pointless if there is a piecewise better
answer which contains P as an axiom. Accordingly, the conjunction of
Kepler’s laws and Boyle’s law is a pointless answer to the question “Why
Kepler’s laws are valid?”. So it eliminates a problem that motivated
Hempel to introduce his covering law model only for the explanation of
particular events, see (7).
General Answers
In the previous section, we dealt with answers to one particular question.
However, in physics it is a general desire to search for unified theories,
that is, theories answering more questions. So a good answer in physics
is characterized by assuming little, but implying a lot.3
3 Of course there are other desired requirements of a physical theory, such as
simplicity and comprehensibility of the axioms or experimental testability.
The unification of theories is the point where we have to leave the

convenience of fixed languages and search for a suitable unification of the
languages of the theories in question, too. For example, the juxtaposition
of the axioms of relativity and quantum theories in their combined lan-
guage will result in one theory but it will not solve the problem of their
reconciliation. Though it will be consistent (as its parts are consistent)
and will imply all the prediction of relativity and quantum theories, it will
not be what we mean by a unified theory. To achieve a truly unified the-
ory, we need a richer language in which we can formulate the interrelations
between the concepts of these theories. However, unifying the languages
in an appropriate way and formulating new axioms in this unified lan-
guage to establish the interrelations between their concepts such that the
unified theory has new experimentally testable predictions is a very dif-
ficult task. Unifying theories is the subject of amalgamating theories in
algebraic logic, see, e.g., (10).
The Truth of Axioms of Physics

Most theories of why-questions require the explanations to be true; see,
for instance, (7). However, it is a fundamental requirement of a physical
theory to be experimentally testable, hence refutable. So in physics we do
not know whether an axiom is true or not, we just presume so. Therefore,
if we require the explanations to be true, we will never ascertain whether
our theory is really an explanation or not. Hence in the case of why-
questions of physics, it is better not to require the axioms to be true.
For example, if we required the axioms of physics to be true, we could
not say that Newton’s theory is an explanation of Kepler’s laws because
Newton’s theory is refuted by some experimental facts.4 That would be
inconvenient as Newton’s theory is the standard example in the literature
for explanation of Kepler’s laws. Hence it is better to treat the axioms
as possible truths and treat the matter only in conditional, that is, an
answer to the question “Why P ?” means that “If the theory is true, then
so is P .” In this sense it is still meaningful to say that Newton’s theory
However, it is not easy to define these concepts precisely if it is possible at
all.
4 For example, the perihelion advance of Mercury is a well known experimen-
tal fact that disproves Newton’s theory.
explains Kepler’s laws. And according to the above definition Newton’s

theory is an acceptable answer to the question “Why are Kepler’s laws
valid?”.
Some further Examples

We can take the Twin Paradox (TwP) and ask “Why TwP?”. In the
language of SpecRel, we cannot formulate the full version of TwP only
its inertial approximation called the Clock Paradox (ClP). Here we only
concentrate on TwP but for a similar investigation of ClP see (35), (17).
To formulate the full version of TwP we have to extend the language
above for accelerated (that is, non-inertial) observers, which is done by
adding one more unary relation for accelerated observers on the sort of
bodies. In this language we can formulate TwP, see (23), (35). Let us
denote the formulated version of TwP as TwP.
The most natural axiom to assume about accelerated observers is the
following:
AxCmv: At each moment of its world-line, every accelerated observer co-

ordinatizes the nearby world for a short while in the same way as
an inertial observer does.
For precise formulation of this axiom, see (23), (35). Let AccRel be the
axiom system consisting of AxCmv and all the five axioms of SpecRel.
Surprisingly, AccRel does not answer the question “Why TwP?”. More-
over, the following can be proved:
AccRel ∪ Th(R) |= TwP,
where Th(R) is the whole first-order theory of real numbers, see (23). At
first sight this result suggests that the question “Why TwP?” cannot be
answered within first-order logic, which would be depressing as there are
weighty methodological reasons for staying within first-order logic, see,
e.g., (5), (35). However, there is a first-order axiom scheme (IND) in the
above language such that AccRel together with this scheme answers the
question “Why TwP?”, see (23), (35).
The why-question “Why does gravity slow time down?”, can be an-
swered by the theory AccRel ∪ IND and Einstein’s equivalence principle,
see (24), (35).
For further examples, let us consider the question of “Why does rela-
tivistic mass increase?” or “Why is the sum of the rest masses of inelas-
tically colliding inertial bodies smaller than the rest mass of the original
inertial body?”. Theories SpecRelDyn and SpecRelDyn+ are possible an-
swers to these questions, respectively; see (6), (14), (35).
Concluding Remarks
In the spirit of reverse mathematics, we have introduced a precise defi-
nition of acceptable answers to why-questions in physics and some ideas
about how to compare these answers. We also presented several exam-
ples taken mainly from axiomatic relativity. Finally let us note that the
work done by the research group of Logic and Relativity led by Hajnal
Andréka and István Németi can be considered as providing explanations
to why-questions of relativity, see the bibliography.
This research is supported by the Hungarian National Foundation for

Scientific Research grant No T73601.
Bibliography
[1] Hajnal Andréka, Judit X. Madarász, and István Németi. “Logical ax-
iomatizations of space-time. Samples from the literature”, in: Non-
Euclidean geometries, volume 581 of Mathematics and Its Applica-
tions. New York: Springer, 2006, pp.155-185.
[2] Hajnal Andréka, Judit X. Madarász, and István Németi, “Logic

of space-time and relativity theory”, in: Marco Aiello, Ian Pratt-
Hartmann and Johan van Benthem (Eds.), Handbook of Spatial Log-
ics. Dordrecht: Springer, 2007, pp.607-711.
[3] Hajnal Andréka, Judit X. Madarász, and István Németi, On

the logical structure of relativity theories. Research report, Alfréd
Rényi Institute of Mathematics, Budapest, 2002. With contri-
butions from Attila Andai, Gábor Sági, Ildikó Sain and Csaba
Tőke. http://www.math-inst.hu/pub/algebraic-logic/Contents.html.
1312 pp.
[4] Hajnal Andréka, Judit X. Madarász, István Németi, and Gergely

Székely, “Axiomatizing relativistic dynamics without conservation
postulates”, in: Studia Logica 89, 2, 2008, pp.163-186.
[5] Hajnal Andréka, Judit X. Madarász, István Németi, and Gergely

Székely, Vienna Circle and Logical Analysis of Relativity Theory, in:
this volume.
[6] Michael Friedman, “Explanation and Scientific Understanding”, in:

The Journal of Philosophy 71, 1, 1974, pp.5-19.
[7] Carl G. Hempel and Paul Oppenheim, “Studies in the Logic of Ex-
planation”, in: Philosophy of Science 15, 2, 1948, pp.135-175.
[8] Jaakko Hintikka and Ilpo Halonen, “Semantics and Pragmatics for
Why-Questions”, in: The Journal of Philosophy 92, 12, 1995, pp.636-
657.
[9] Judit X. Madarász. Logic and Relativity (in the light of definability
theory). PhD thesis, Eötvös Loránd Univ., Budapest, 2002.
[10] Judit X. Madarász. “Amalgamation and Interpolation; Pushing the

limits.” Parts I-II, in: Studia Logica 61, 3, 1998, pp.311-345 and 62,
1, 1999, pp.1-19.
[11] Judit X. Madarász, István Németi, and Gergely Székely, “Twin para-
dox and the logical foundation of relativity theory”, in: Foundations
of Physics 36, 5, 2006, pp.681-714.
[12] Judit X. Madarász, István Németi, and Gergely Székely, “First-order

logic foundation of relativity theories”, Dov Gabbay, Sergei Gon-
charov and Michael Zakharyaschev (Eds.), in: Mathematical prob-
lems from applied logic II. New York: Springer, 2007, pp.217-252.
[13] Judit X. Madarász, István Németi, and Csaba Tőke, “On generaliz-
ing the logic-approach to space-time towards general relativity: first
steps”, in: Vincent F. Hendricks, Fabian Neuhaus, Stig Andur Ped-
ersen, Uwe Scheffler and Heinrich Wansing (Eds.), First-Order Logic
Revisited. Berlin: Logos Verlag 2004, pp.225-268.
[14] Judit X. Madarász, and Gergely Székely, Comparing Relativistic and

Newtonian Dynamics in First-Order Logic, in: this volume.
[15] David Sandborg, “Mathematical Explanation and the Theory of

Why-Questions”, in: The British Journal for the Philosophy of Sci-
ence, 49, 4, 1998, pp.603-624.
[16] Gergely Székely, First-Order Logic Investigation of Relativity The-

ory with an Emphasis on Accelerated Observers. PhD thesis, Eötvös
Loránd Univ., Budapest, 2009.
[17] Gergely Székely, “Geometrical characterization of the twin paradox

and its variants”, submitted.
[18] Dennis Temple, “The Contrast Theory of Why-Questions”, in: Phi-

losophy of Science 55, 1, 1988, pp.141-151.
Lorentzian Theories vs. Einsteinian Special

Relativity – A Logico-empiricist Reconstruction
Introduction
1. It is widely believed that the principal difference between Einstein’s spe-
cial relativity and its contemporary rival Lorentz-type theories was that
while the Lorentz-type theories were also capable of “explaining away” the
null result of the Michelson–Morley experiment and other experimental
findings by means of the distortions of moving measuring-rods and mov-
ing clocks, special relativity revealed more fundamental new facts about
the geometry of space-time behind these phenomena. For the sake of
brevity, I shall use the term “Lorentz theory” as classification to refer to
the similar approaches of Lorentz, FitzGerald, and Poincaré, that save the
classical Galilei covariant conceptions of space and time by explaining the
experimental findings through the physical distortions of moving objects
– of moving measuring equipments included – no matter whether these
physical distortions are simply hypothesized in the theory, or prescribed
by some “principle” like Lorentz’s principle, or they are constructively
derived from the behavior of the molecular forces. From the point of view
of my recent concerns what is important is the logical possibility of such
an alternative theory. Although, Lorentz’s 1904 paper or Chapter V. of
his The theory of electrons (1909) are good historic examples.
According to this widespread view, special relativity was, first of all, a
radically new theory about space and time. A theory about space and time
describes a certain group of objective features of physical reality, which
we call (the structure of) space-time. Consider the claims like these:
– According to classical physics, the geometry of space-time is E3 ×E1 ,

where E3 is a three-dimensional Euclidean space for space and E1 is
a one-dimensional Euclidean space for time, with two independent
invariant metrics corresponding to the space and time intervals.
192 László E. Szabó
– In contrast, special relativity claims that the geometry of space-time

is different: it is a Minkowski geometry M4 .
The two statements are usually understood as telling different things about
the same objective features of physical reality. One can express this rev-
olutionary
change by the following logical schema: Earlier we believed in
G1 M̂ , where M̂ stands for (the objective features of physical reality
(like “of type-E ×E ”).
3 1
called) space-time and G1 denotes
some
predicate
Then we discovered that ¬G1 M̂ but G2 M̂ , where G2 denotes a
predicate different from G1 (something like “of type-M4 ”).
This is however not the case. Our logico-empiricist style analysis
willshow
that the correct logical schema is this: Earlier we believed in
G1 M̂ . Then we discovered for some other features of physical reality

= M̂ that ¬G1 M
M but G2 M . Consequently, it still may (and

actually does) hold that G1 M̂ . In other words, in comparison with
the pre-relativistic Galileo-invariant conceptions, special relativity tells us
nothing new about the geometry of space-time. It simply calls something
else “space-time”, and this something else has different properties. We
will also show that all statements of special relativity about those features
of reality that correspond to the original meaning of the terms “space”
and “time” are identical with the corresponding traditional pre-relativistic
statements. Thus the only new factor in the special relativistic account
of space-time is the terminological decision to designate something else
“space-time”.
2. So the real novelty in special relativity is some G2 M . It will be also

argued, however, that G2 M does not contradict to what the Lorentz
theory
claims. Both,
the Lorentz theory and special relativity claim that
G1 M̂ & G2 M . In other words: Special relativity and the Lorentz
theory are identical theories about space and time in all sense of the words.
Moreover, I shall show that the two theories provide identical descrip-
tion of the behavior of moving physical objects. Thus, we will arrive at
the following final conclusion: Special relativity and the Lorentz theory are
completely identical in both senses, as theories about space and time and
as theories about the behavior of moving physical objects.
Lorentzian Theories vs. Einsteinian Special Relativity 193
Empirical definitions of space and time tags

3. Physics describes objective features of reality by means of physical
quantities. Our scrutiny will therefore start by clarifying how classical
physics and relativity theory define the space and time tags assigned to
an arbitrary event. When I say “definition”, I mean empirical definition,
somewhat similar to Reichenbach’s “coordinative definitions”, Carnap’s
“rules of correspondence”, or Bridgman’s “operational definitions”; which
give an empirical interpretation of the theory. Einstein too, at least in his
early writings, strongly emphasizes that all spatiotemporal terms he uses
are based on operations applying measuring-rods, clocks and light signals.
The empirical definition of a physical quantity requires an etalon mea-
suring equipment and a precise description of the operation how the quan-
tity to be defined is measured. For example, assume we choose, as the
etalon measuring-rod, the meter stick that is lying in the International
Bureau of Weights and Measures (BIPM) in Paris. Also assume – this is
another convention – that “time” is defined as a physical quantity mea-
sured by the standard clock, say, also sitting in the BIPM. When I use
the word “convention” here, I mean the semantical freedom we have in
the use of the uncommitted signs “distance” and “time” – a freedom what
Grünbaum (1974, p. 27) calls “trivial semantical conventionalism”.
4. Consider two frames of reference: K, in which the etalons are at rest,
and K which is moving at constant velocity relative to K. As usual, we
assume that the coordinate axes are parallel, and K moves relative to K
at constant velocity v in the positive x-direction.
We are going to reconstruct the empirical definitions of the space and
time tags of an arbitrary event A. I call them space and time “tags”
rather than space and time “coordinates”. By this terminology I would
like to distinguish a particular kind of space and time coordinates which
are provided with direct empirical meaning from space and time coordi-
nates in general, the empirical meaning of which can be deduced from
the empirical meaning of the space and time tags. Once we have space
and time tags defined, we can introduce arbitrary other coordinates on
the manifold of space-time tags. The physical/empirical meaning of a
point of the manifold is however determined by the space-time tags of
physical/empirical meaning. Only in this way we can confirm or falsify,
empirically, a spatiotemporal physical statement.
Preliminary facts
5. In order to explain the intuitions behind the definitions below, let us
recall a few well known facts, to the detailed discussion of which we will
return later.
(a) Both the Lorentz theory and Einstein’s special relativity “know”
about the distortions of measuring-rods and clocks when they are
in motion relative to the “stationary” reference frame (of the BIPM,
say). Einstein writes:
A rigid body which, measured in a state of rest, has the form of a sphere,
therefore has in a state of motion – viewed from the stationary system –
the form of an ellipsoid of revolution with the axes.

v2
R 1 − 2 , R, R.
c
Thus, whereas the Y and Z dimensions of the sphere (and therefore of

every rigid body of no matter what form) do not appear modified
by the
v2
motion, the X dimension appears shortened in the ratio 1 : 1− c2
[. . .]
Further, we imagine one of the clocks which are qualified to mark the time
t when at rest relatively to the stationary system, and the time τ when
at rest relatively to the moving system, to be located at the origin of the
co-ordinates of k, and so adjusted that it marks the time τ . What is the
rate of this clock, when viewed from the stationary system?
Between the quantities x, t, and τ , which refer to the position of the clock,
we have, evidently, x = vt and
1 vx
τ = t− 2 .
2 c
1 − vc2
Therefore,
v2
τ =t 1−
c2
[. . .] whence it follows that the time marked by the clock (viewed in the
stationary system) is slow [. . .]
From this there ensues the following peculiar consequence. If at the points
A and B of K there are stationary clocks which, viewed in the stationary
system, are synchronous; and if the clock at A is moved with the velocity
v along the line AB to B, then on its arrival at B the two clocks no longer
synchronize, but the clock moved from A to B lags behind the other which
2
tv
has remained at B by 2c 2 (up to magnitudes of fourth and higher order),
t being the time occupied in the journey from A to B. (Einstein 1905)
Probably, this last phenomenon is why Einstein suggested the light

signal method of synchronization, instead of the even more obvious
transportation of clocks from one place to the other.
(b) However, the loss of phase accumulated by a clock during its trans-
portation tends to zero if the transportation is very slow relative to
the stationary reference frame:
tv 2 xv
lim 2
= lim =0
v→0 2c v→0 2c2
where x is the distance of the transportation. Therefore, in the
stationary frame of reference, the synchronization by means of slow
transportation of clocks is equivalent with Einstein’s synchronization
by means of light signals.
(c) Moreover, this equivalence is true in an arbitrary reference frame

moving relative to the stationary frame. Imagine a clock at rest
relative to the moving frame of reference K at a point A . Then, we
slowly (relative to K ) transport the clock from A to another point
B of the moving frame. As it will be shown in details (Point 14), the
loss of phase accumulated by the clock during its slow transportation
does not tend to zero but to a finite value, which is exactly equal to
the one corresponding to Einstein’s light signal synchronization.
6. Thus, when we will define distance by means of measuring-rod, we will

follow Einstein’s empirical definitions in his 1905 paper; we will do the
same when we will define time at the origin of a reference frame by means
of a co-moving copy of the standard clock; and when we will define the
time tags of events distant from the origin of the frame by means of the
readings of slowly transported copies of the standard clock, we will use
an operation which is equivalent to Einstein’s light signal definition of
simultaneity.
The reader may wonder, why do we choose this old-fashioned way
of defining time and space tags, by means of measuring-rods and slowly
transported clocks; instead of the more modern “one clock + light signals”
method offered in many textbooks – not to mention the obvious opera-

tional circularities related with the conceptions like “rigid body”, “refer-
ence frame”? We have three main reasons for this choice: First, our aim in
this paper is to reconstruct the definitions of the spatiotemporal terms as
they are understood in Einstein’s special relativity and in the Lorentz-type
theories (that is, in classical physics), without questioning whether these
definitions are tenable or not. Second, by applying measuring-rods and
slowly transported clocks, it becomes more explicit that the Lorentz trans-
formations express the physical behavior of measuring-rods and clocks. In
Einstein’s words:
A Priori it is quite clear that we must be able to learn something about the
physical behavior of measuring-rods and clocks from the equations of transfor-
mation, for the magnitudes z, y, x, t are nothing more nor less than the results
of measurements obtainable by means of measuring-rods and clocks. (Einstein
1920, p. 35)
Third, the “one clock + light signals” method is actually much more
complex than suggested in the textbook literature (see Szabó 2009).
Definitions
7. For the sake of simplicity consider only one space dimension and assume
that the origin of both K and K is at the BIPM at the initial moment
of time.
There will be no difference between the Einsteinian and Lorentzian
definitions in case of space and time tags in the stationary frame of refer-
ence, that is, in the frame of the etalons.
(D1) Time tag in K according to special relativity
Take a synchronized copy of the standard clock at rest in the

BIPM, and slowly1 move it to the locus of event A. The
time tag
tK (A) is the reading of the transferred clock when A
occurs.
1 By “slowly” is usually meant that we move the clock from one place to the
other over a long period of time, according to the reading of the clock itself.
The reason is to avoid the loss of phase accumulated by the clock during
its journey.
(D2) Space tag in K according to special relativity
The space tag xK (A) of event A is the distance from the origin
of K of the locus of A along the x-axis2 measured by superpos-
ing the standard measuring-rod, being always at rest relative
to K.
(D3) Time tag in K according to classical physics
The same as (D1): t̂(A) = t̃(A).
(D4) Space tag in K according to classical physics
The same as (D2): x̂(A) = x̃(A).
8. The difference between the Einsteinian and Lorentzian understanding

starts with the definition of the space and time tags in the moving refer-
ence frame K . The two approaches agree that measuring-rods and clocks
suffer distortions when transferred from the BIPM to the moving (relative
to the BIPM) reference frame K . In the special relativistic definitions,
however, we will ignore this fact and define the space and time tags just
as they are measured by means of the distorted equipments:
(D5) Time tag in K according to special relativity

BIPM, gently accelerate it from K to K and set it to show 0
when the origins of K and K coincide. Then slowly (relative

to K ) move it to the locus of event A. The time tag
tK (A)
is the reading of the transferred clock when A occurs.
(D6) Space tag in K according to special relativity

The space tag x K (A) of event A is the distance from the

origin of K of the locus of A along the x-axis measured by
superposing the standard measuring-rod, being always at rest
relative to K , in just the same way as if all were at rest.
2 Straight line is usually defined by a light beam.

In contrast, in the Lorentzian definitions, as it follows from the whole

3
we will take into account that the standard
tradition of classical physics,
2
clock slows down by factor 1 − vc2 and that a rigid rod suffers a contrac-

2
tion by factor 1 − vc2 when they are gently accelerated from K to K .
Therefore, if we measure
a distance and the result is X, then the “real
2
distance” is only X 1 − vc2 ; similarly, we know (see Point 14) that if the
reading of the clock is T then the “real time” is
T + X cv2

2
1 − vc2
Accordingly,
(D7) Time tag of an event in K according to classical physics

BIPM, gently accelerate it from K to K and set it to show 0
when the origins of K and K coincide. Then slowly (relative
to K ) move it to the locus of event A. Let T be the reading

of the transferred clock when A occurs. The time tag t̂K (A)
is
T + X cv2
t̂K (A) := (1)
2
1 − vc2
(D8) Space tag of an event in K according to classical physics

Let X be the “distance” from the origin of K of the locus
of A along the x-axis measured by superposing the standard
measuring-rod, being always at rest relative to K , in just the

same way as if all were at rest. The space tag x̂K (A) of event
A is
K v2
x̂ (A) := X 1 − 2 (2)
c
3 I am referring to the tradition, according to which it must be taken into ac-
count that, for example, the pendulum clock slows down as the temperature
rises.
9. Yet one might raise the historical question whether our reconstruction
is correct. We will, however, not go into these philological details. Con-
cerning definitions (D1)–(D2) and (D5)–(D6) I only refer to the following
passage from Einstein’s 1905 paper:
Let there be given a stationary rigid rod; and let its length be l as measured
by a measuring-rod which is also stationary. We now imagine the axis of the
rod lying along the axis of x of the stationary system of co-ordinates, and that
a uniform motion of parallel translation with velocity v along the axis of x in
the direction of increasing x is then imparted to the rod. We now inquire as to
the length of the moving rod, and imagine its length to be ascertained by the
following two operations:
(a) The observer moves together with the given measuring-rod and the rod to
be measured, and measures the length of the rod directly by superposing
the measuring-rod, in just the same way as if all three were at rest.
(b) By means of stationary clocks set up in the stationary system and synchro-
nizing in accordance with [the light-signal synchronization], the observer
ascertains at what points of the stationary system the two ends of the rod
to be measured are located at a definite time. The distance between these
two points, measured by the measuring-rod already employed, which in
this case is at rest, is also a length which may be designated “the length
of the rod.”
In accordance with the principle of relativity the length to be discovered by the

operation (a) – we will call it “the length of the rod in the moving system” –
must be equal to the length l of the stationary rod.
The length to be discovered by the operation (b) we will call “the length
of the (moving) rod in the stationary system.” This we shall determine on the
basis of our two principles, and we shall find that it differs from l.
Concerning (D3)–(D4) and (D7)–(D8) I only refer to Einstein’s own read-

ing of the Lorentzian approach, quoted in Point 24 below. In addition, I
refer to the calculation we will make in Point 14, proving that the slow-
transport-of-clock and light-signal synchronizations are equivalent.
10. With empirical definitions (D1)–(D8), in every inertial frame we de-
fined four different quantities for each event; such that:
x̂K (A) ≡ K (A)

x (3)
K
t̂ (A) ≡
tK (A) (4)
for the reference frame of the etalons, and

x̂K (A) ≡ K (A)
x (5)
K

t̂ (A) ≡ tK (A) (6)
for any other inertial frame of reference. (≡ denotes the identical empirical
definition.)
In spite of the different empirical definitions, it could be a contingent
K (A) and/or t̂K (A) =

fact of nature that x̂K (A) = x tK (A) for every
event A. Let me illustrate this with an example. Inertial mass mi and
gravitational mass mg are two quantities having different experimental
definitions. But, it is a contingent fact of nature (experimentally proved
by Eötvös around 1900) that, for any object, the two masses are equal,
mi = mg . However, that is obviously not the case here.
Thus, our first partial conclusion is that different physical quanti-
ties are called “space” tag, and similarly, different physical quantities are
called “time” tag in special relativity and in the Lorentz theory.4 In order
to avoid further confusion, from now on s tags will refer to
pace and time
the physical quantities defined in (D3), (D4), (D7), and (D8) – according
to the usage of the terms in classical physics; and “space” and “time” in
the sense of the relativistic definitions (D1), (D2), (D5) and (D6) will be
called
space and time.
Special relativity theory makes different assertions about somethings
which are different from In our symbolic notations of
space and time.

Point 1, classical physics claims G1 M̂ about M̂ and relativity theory

claims G2 M about some other features of reality, M . As we will see,
both the Lorentz theory and Einstein’s special relativity are sufficiently
complete accounts of physical reality to describe M̂ as well as M̃ . The
question is: What do the two theories say when they are making assertions
about the same things?
4 This was first recognized by Bridgman (1927, p. 12), although he did not
investigate the further consequences of this fact.
Special relativity does not tell us anything new about space and
time
11. Classical physics – from this point of view, Lorentz-type theories in-
cluded – calls “space” and “time” what we denoted by
space and time.

Let A be an arbitrary event and let K be an arbitrary inertial frame of

reference. Denote x̂K (A) relativity and t̂K (A) relativity the s pace and
K

time tags of A predicted by relativity theory and denote x̂ (A) classical
K
and t̂ (A) classical the similar tags predicted by classical physics. Rel-
ativity theory would tell us something new if it accounted for physical
quantities x̂ and t̂ differently. If there were any event A and any inertial

frame of reference K such that x̂K (A) relativity and t̂K (A) relativity

were different from x̂K (A) classical and t̂K (A) classical . If, for exam-

ple, there were any two events simultaneous in relativity theory which

were not simultaneous according to classical physics, or vice versa – to
touch on a sore point. But a little reflection shows that this is not the
case. Taking into account empirical identities (3)–(4), one can calculate
the relativity theoretic prediction for the outcomes of the measurements
described in (D3), (D4), (D7), and (D8). The relativity theoretic predic-

tion for x̂K (A):

x̂K (A) =x K (A) − vK (K )
tK (A) (7)
relativity
the value of which is equal to

x̂K (A) − v̂ K (K )t̂K (A) = x̂K (A) (8)
classical
Similarly,

t̂K (A) =
tK (A) = t̂K (A) = t̂K (A) (9)
relativity classical
This completes the proof of our first thesis that special relativity does not
tell us anything new about those features of reality that correspond to the
original meaning of the terms “space” and “time”.
The Lorentz theory and special relativity are identical theories of

space and time
12. It is now in order to specify in a more formal way what I exactly mean
by the Lorentz theory. I shall give the final and most general formulation
in Point 17. At this stage, we will use the following definition:
(L1) in terms of the classical concep-

The classical Galilean kinematics
tions of
space and time.
(L2) The assumption that
(L2a) the dimension

parallel to v of a solid suffers contraction
v2
by factor 1 − c2 and
(L2b) a physical
process in a physical system slows down by
2
factor 1 − vc2
when they are gently accelerated from the reference frame of the
etalons, K, to the frame K moving at velocity v relative to K.
This amounts to what Grünbaum (1974, p. 723) calls the “double amended
variant of the aether theory”, with the only difference, of course, that we
have not even mentioned the aether yet. Note that it is sometimes mistak-
enly claimed that we cannot derive the Lorentz transformation equations
from the doubly amended theory alone. As we will see it below, the truth
is what Grünbaum correctly asserts that the “theory permits the deduc-
tion of the Lorentz transformation equations no less than does Einstein’s
special theory of relativity” (p. 723).
13. Since the Lorentz theory adopts the classical conceptions of space
it does not differ from special relativity in its assertions about
and time,
s What about the other claim – G2 M̃ – about
pace and time. space
In order to prove that the Lorentz theory and special relativity
and time?
are identical theories about space and time in all sense of the words, we

also have to show that the two theories have identical assertions about x
and t, that is,

K (A)
x = x K (A)
relativity LT

tK (A) =
tK (A)
relativity LT
According to relativity theory, the tags in K and in K

space and time
are related through the Lorentz transformations. From (3)–(4) we have
K
t̂K (A) − v x̂ c2(A)

tK (A) = (10)
relativity 2
1 − vc2
x̂K (A) − v t̂K (A)
K (A)
x = (11)
relativity 2
1 − vc2
14. On the other hand, taking the immediate consequences of (L2a) and
2
(L2b) that the measuring-rod suffers a contraction by factor 1 − vc2 and

the standard clock slows
2
down by factor 1 − vc2 when they are gently

accelerated from K to K , one can directly calculate the K (A)
space tag x
K
and the time tag t (A), following the descriptions of operations in (D5)
and (D6).
First, let us calculate the reading of the clock slowly transported in

K from the origin to the locus of an event A. For the sake of simplic-
ity we continue to restrict our calculation to one space dimension. (For
the general calculation of the phase shift suffered by moving clocks, see

Jánossy 1971, pp. 142–147.) The clock is moving with a varying velocity
K K
v̂C (t̂ ) = v + ŵK (t̂K )
of the clock relative to K , that is, ŵK (0) = 0

where ŵK (t̂K ) is the velocity
when it starts at x̂C (0) = 0 (as we assumed, t̂K = 0 and the transported
K
clock shows 0 when the origins of K and K coincide) and ŵK (t̂K 1 ) = 0
when the clock arrives at the place of A. The reading of the clock at the
time t̂K
1 will be
ˆ t̂K
2
1 v + ŵK (t̂)
T = 1− dt̂ (12)
0 c2
Since ŵK is small we may develop in powers of ŵK , and we find from (12)
when neglecting terms of second and higher order

´ t̂K
t̂K
1 v+ 0
1 ŵ K (t̂) dt̂ v
K
1 −
t̂K 2 t̂K (A) − x̂ c(A)v
2
T = c
= (13)
2 v2
1 − vc2 1 − c2
(where, without loss of generality, we take t̂K K

1 = t̂ (A)). Thus, according
to the definition of
t, we have
t̂K (A) − v x̂ c2(A)

K

tK (A) = (14)
LT 2
1 − vc2

which is equal to
tK (A) in (10).
relativity
of the co-moving meter stick
Now,taking into account that the length
v2 of event A from the origin of K is the
is only 1 − c2 , the distance
following:
K v2
K K
x̂ (A) = t̂ (A)v + x (A) 1 − 2 (15)
c
and thus
x̂K (A) − v t̂K (A) K
K (A)
x = (A)
= x
LT 2 relativity
1 − vc2
This completes the proof. The two theories make completely identical
assertions not only about tags x̂, t̂ but also about
space and time space

and time tags x̃, t̃. and widehattextrmtime tags@
space and time tags and
widetildetextrmtime tags@ tags
space and time
15. Consequently, there is full agreement between the Lorentz theory and
special relativity theory in the following statements:
(a) – which is called “velocity” by relativity theory – is not an

Velocity
additive quantity,

vK (K ) + vK (K )
vK (K ) =
K (K )
v v K (K )
1+ c2
– that is, what we traditionally call “velocity” – is an

while velocity
additive quantity,

v̂ K (K ) = v̂ K (K ) + v̂ K (K )
where K , K , K are arbitrary three frames. For example,

v̂ K (light signal) = v̂ K (K ) + v̂ K (light signal)
(b) The x 3 ,
2 , x
1 , x t -map of the world can be conveniently described
through a Minkowski geometry, such that the t-simultaneity can be
described through the orthogonality with respect to the 4-metric of
the Minkowski space, etc.

(c) The x̂1 , x̂2 , x̂3 , t̂ -map of the world, can be conveniently described
through a traditional “space-time geometry” like E3 × E1 .
(d) of a light signal is not the same in all inertial frames

The velocity
of reference.
(e) of a light signal is the same in all inertial frames of

The velocity
reference.
(f) Time are invariant, the reference frame independent

and distance
and distance
concepts, time are not.
(g) t̂-simultaneity is an invariant, frame-independent concept, while

t-
simultaneity is not.

(h) For arbitrary K and K , x̂K (A), t̂K (A) can be expressed by

x̂K (A), t̂K (A) through a suitable Galilean transformation.

(i) K (A),
For arbitrary K and K , x tK (A) can be expressed by

K (A),
x tK (A) through a suitable Lorentz transformation.
..
.
To sum up symbolically, the Lorentz theory and special relativity theory

have identical
assertions
about both M̂ and M : they unanimously claim
that G1 M̂ & G2 M .
16. Finally, note that in an arbitrary inertial frame K for every event

K K K
A the tags x̂K 1 (A), x̂2 (A), x̂3 (A), t̂ (A) can be expressed in terms of

K
x 1 (A), x K 2 (A), x K K
3 (A), t (A) and vice versa. Consequently, we can
express the laws of physics – as is done in special relativity – equally well
in terms of the variables x 2 , x
1 , x 3 ,
t instead of the s pace and time tags
x̂1 , x̂2 , x̂3 , t̂. On the other hand, we should emphasize that the one-to-
one correspondence between x 2 , x
1 , x 3 ,
t and x̂1 , x̂2 , x̂3 , t̂ also entails that
the laws of physics (so called “relativistic” laws included) can be equally
well expressed in terms of the traditional space and time tags x̂1 , x̂2 , x̂3 , t̂
instead of the variables x 2 , x
1 , x
3 , t. In brief, physics could manage equally
well with the classical Galileo-invariant conceptions of
space and time.
The Lorentz theory and special relativity are completely identical

theories
17. Although special relativity does not tell us anything new about space
and time, both special relativity and the Lorentz theory enrich our knowl-
edge of the physical world with the physics of objects moving at constant
velocities – in accordance with the title of Einstein’s original 1905 paper.
The essential physical content of their discoveries is that physical objects
suffer distortions when they are accelerated from one inertial frame to the
other, and that these distortions satisfy some uniform laws.
In the Lorentz theory, the laws of deformation (L2a) and (L2b) ap-
ply to both the measuring-equipment and the object to be measured.
Therefore, it is no surprise that the “length” of a moving, consequently
distorted, rod measured by co-moving, consequently distorted, measuring-
of the rod, is the same as the length
rod and clock, that is the length of
the corresponding stationary rod measured with stationary measuring-
of a slowed down process in a moving object
rod and clock. The duration
measured with a co-moving, consequently slowed down, clock will be the
of the same process in a similar object at rest, mea-
same as the duration
sured with the original distortion free clock at rest. These regularities are
nothing but particular instances of the relativity principle. Similar obser-
vations lead Lorentz and Poincaré to conclude with the general validity
of the relativity principle:
(RP) For any two inertial frames of reference K and K , the laws
of physics in K are such5 that the laws of physics empiri-
cally ascertained by an observer in K , describing the behav-
ior of physical objects co-moving with K , expressed in vari-

ables xK
1 ,x K2 ,xK3 ,t
K , have the same forms as the similar
empirically ascertained laws of physics in K , describing the sim-
ilar physical objects co-moving with K , expressed in variables

K
x 1 ,xK
2 ,x K K
3 , t , if the observer in K performs the same mea-
surement operations as the observer in K with the same measur-
ing equipments transferred from K to K , ignoring the fact that
the equipments undergo deformations during the transmission.
Therefore, in its most general form the Lorentz theory consists of

(L1)+(L2)+(RP).
On the other hand, (RP) is the basic premise of special relativ-
ity theory.6 And, as is well known from Einstein’s 1905 paper, (RP)

implies the Lorentz transformation equations for x K K
1 ,x2 ,xK
3 ,t
K and

K
x 1 ,x K2 ,x K3 ,t
K between arbitrary two inertial frames. Applying these
transformations between the reference frame of the etalons, K, and an
arbitrary inertial frame K we can derive (see Point 11) the rules of clas-

sical kinematics, that is, (L1). It also follows from (RP) that the char-

acteristic dimensions and the characteristic periods in a physical system
co-moving with the moving reference frame K , expressed in terms of
K
x

1 ,x K

2 ,x K

3 ,t

K are the same as the similar dimensions of
and periods
the same system when it is at rest relative to K, expressed in terms of
K
x 1 ,x K
2 ,x K K
3 , t . Combining this fact with the Lorentz transformations,
we have (L2a) and (L2b). That is, (RP) ⇒ (L1) + (L2). In other words,
5 It must be emphasized that the relativity principle characterizes the laws of

physics in one single frame of reference; in other words, the laws of physics
in one single inertial frame of reference pre-determine whether relativity
principle holds or not. This is what Bell calls “Lorentzian pedagogy” (Bell
1987, p. 77).
6 Applying (RP) to electrodynamics, one finds that the velocity of a light
signal is the same in all inertial reference frames. This is certainly true if,
in the application of the principle, the meaning of the phrase “the Maxwell
equations have the same form” includes that constant of nature c has the
same value. This is sometimes regarded as the “second basic principle” of
special relativity.
Lorentz theory, too, reduces to (RP). The two theories are identical in all
sense. For, (RP) not only implies the classical rules of kinematics and the

relativistic rules of kinematics, but also implies some uniform laws gov-
erning the behavior of physical objects when accelerated from one inertial
frame to the other.
18. Let E be a set of differential equations describing the behavior of the
system in question in an arbitrary reference frame K . Let ψ0 denote a
set of (initial) conditions, such that the solution [ψ0 ] determined by ψ0
describes the behavior of the system when it is, as a whole, at rest relative
to K . Let ψv be a set of conditions which corresponds to the solution
describing the same system in uniform motion at velocity v relative to K .
To be more exact, [ψv ] is a solution of E that describes the same behavior
of the system as [ψ0 ] but in superposition with a collective translation at
velocity v. Denote E and ψ0 the equations and conditions obtained from

E and ψ0 by substituting the variables x K , tK , . . . with the variables

K ,
x tK , . . . of K . Denote Λv (E ) , Λv (ψv ) the set of equations E and
conditions ψv expressed in terms of the double-primed variables, applying
the Lorentz transformations. Now, the following two conditions guarantee
the satisfaction of the relativity principle:
Λv (E ) = E (16)

Λv ([ψv ]) = [ψ0 ] (17)
To make more explicit how this provides a useful method in the de-
scription of the deformations of physical systems when they are acceler-
ated from one inertial frame K into some other K , consider the following
situation: Assume we know the relevant physical equations and know the
solution of the equations describing the physical properties of the object in
question when it is at rest in K : E , [ψ0 ]. We now inquire as to the same
description of the object when it is moving at a given constant velocity

relative to K . If (16)–(17) is true, then we can solve the problem in the
following way. Simply take E , [ψ0 ] – by putting one more prime on each
variable – and express [ψv ] from (17) by means of the inverse Lorentz
transformation.
Now, according to the standard views, the solution belonging to con-
dition ψv describes the same object when it is moving at a given constant
relative to K . The situation is, in fact, much more complex.
velocity
Whether or not the solution thus obtained is correct, that is whether or

not relativity principle holds, depends on the details of the relaxation
process after the acceleration of the system (Szabó 2004). What must be
emphasized is, however, that within the scope of validity of the relativ-
ity principle, the Lorentz theory and special relativity provide the same
powerful problem solving method in the physics of moving objects – no
and
matter if variables x t are called “space” and “time” tags or not.
This completes the proof: special relativity and the Lorentz theory
are completely identical theories. They are identical theories about
space,

time,
space, and time; and they provide identical descriptions of the be-
havior of moving physical objects.
This is, of course, an astounding result, in contrast to how people think

about the “revolution” brought about by Einstein’s special relativity. In
order to help the reader form his or her own opinion, the rest part of this
paper is devoted to the discussion of the possible objections.
Are relativistic deformations real physical changes?

19. Many believe that it is an essential difference between the two theories
that relativistic deformations like the Lorentz–FitzGerald contraction and
the time dilatation are real physical changes in the Lorentz theory, but
there are no similar physical effects in special relativity. Let us examine
two typical argumentations.
According to the first argument the Lorentz contraction/dilatation of
a rod cannot be an objective physical deformation in relativity theory,
because it is a frame-dependent fact whether the rod is shrinking or ex-
panding. Consider a rod accelerated from the state of rest in reference
frame K to the state of rest in reference frame K . According to relativ-
ity theory, “the rod shrinks in frame K and, at the same time, expands
in frame K . But this is a contradiction, the argument says, if the de-
formation was a real physical change. In contrast, the argument says,
the Lorentz theory claims that the length of a rod is a frame-independent
concept. Consequently, in the Lorentz theory, “the contraction/dilatation
of a rod” can indeed be an objective physical change.
However, we have already clarified, that the terms “distance” and
“time” have different meanings in relativity theory and the Lorentz the-
from dilatation,
ory. We must differentiate dilatation
contraction from

contraction, and so on. For example, consider the reference frame of the
etalons, K, and another frame K moving relative to K, and a rod accel-
erated from the sate of rest in reference frame K – state1 – to the state
of
of rest in reference frame K – state2 . Denote l̂K (state1 ) the length
˜K of the rod in state1
the rod in state1 relative to K, l (state1 ) the length
relative to K, etc. Now, the following statements are true about the rod :
l̂K (state1 ) > l̂K (state2 )

contraction in K (18)
K K
l̂ (state1 ) > l̂ (state2 ) contraction in K (19)
l̃K (state1 ) > l̃K (state2 )
contraction in K (20)

l̃ K
(state1 ) < l̃ K
(state2 ) in K
dilatation
(21)
And there is no difference between relativity theory and the Lorentz the-
ory: all of the four statements (18)–(21) are true in both theories. If,
in the Lorentz theory, facts (18)–(19) provide enough reason to say that
there is a real physical change, then the same facts provide enough reason
to say the same thing in relativity theory. And vice versa, if (20)–(21)
contradicted to the existence of real physical change of the rod in relativity
theory, then the same holds for the Lorentz theory.
20. It should be mentioned, however, that there is no contradiction be-
tween (20)–(21) and the existence of real physical change of the rod. Rel-
ativity theory and the Lorentz theory unanimously claim that length is a
relative physical quantity. It is entirely possible that one and the same
objective physical change is traced in the increase of the value of a relative
quantity relative to one reference frame, while it is traced in the decrease
of the same quantity relative to another reference frame (see the example
in Fig. 7). (What is more, both, the value relative to one frame and the
value relative to the other frame, reflect objective features of the objective
physical process in question.)
21. According to the other widespread argument, relativistic deformations
cannot be real physical effects since they can be observed by an observer
also if the object is at rest but the observer is in motion at constant
velocity. And these relativistic deformations cannot be explained as real
physical deformations of the object being continuously at rest.
However, there is a triple misunderstanding behind such an argument:
K K

K K
Ekin = 12 mV 2 K
Ekin is decreasing Ekin =0

K K K
Ekin =0 Ekin is increasing Ekin = 12 mV 2
Figure 7: One and the same objective physical process is traced in the
increase of kinetic energy of the spaceship relative to frame K , while it is
traced in the decrease of kinetic energy relative to frame K
x 1
x 3

K
B moving
in K

K K

E at rest E moving
in K in K
Electromagnetic field Electromagnetic field of a point

of a point charge at rest charge moving in x 3 -direction
Figure 8: The change of the electromagnetic field of a point charge

1. Of course, no real distortion is suffered by an object which is contin-

uously at rest relative to a reference frame K , and, consequently,
which is continuously in motion at a constant velocity relative to
another frame K . Contrary to the argument, none of the inertial
observers can observe such a distortion. For example,

distortion free distortion free
=

lK lK
rod at
t1 rod at
t2

distortion free distortion free
=

lK lK
rod at
t1 rod at
t2
2. It is surely true for any

t that

distortion distortion
=

lK lK (22)
free rod at t free rod at t

This fact, however, does not express a contraction of the rod – nei-

ther a real nor an apparent contraction.
3. On the other hand, inequality (22) is a consequence of the real phys-

ical distortions suffered by the measuring equipments – with which
the tags are empirically defined – when they are
space and time
transferred from the BIPM to the other reference frames in ques-
tion. (For further details of what a moving observer can observe by
means of his or her distorted measuring equipments, see Bell 1983,
pp. 75–76.)
22. Finally, let me give an example for a well-known physical phenomenon

which is of exactly the same kind as the relativistic deformations, but
nobody would question that it is a real physical change. Consider the
electromagnetic field of a point charge q. One can easily solve the Maxwell
equations when the particle is at rest in a given K . The result is the
familiar spherically symmetric Coulomb field (Fig. 8):
xK

1K
E =
q1
(23)
at rest
K
2 K
2 32
2
1
xK 2
+ x 3
+ x
in K
xK

K
E =
q 2
(24)
2 at rest

K
2 32
2 K 2 + x
K
x +
x 3
in K 1 2
xK

K
E =
q 3
(25)
3 at rest

2 K
2 32
2
K
x + xK 3
+ x
in K 1 2

1K
B = 0 (26)
at rest
in K

2K
B = 0 (27)
at rest
in K

K
B = 0 (28)
3 at rest
in K
How does this field change if we set the charge in motion with con-
v along the x
stant velocity 3 -axis? Maxwell’s equations can also answer
this question. First we solve the Maxwell equations for arbitrary time-
depending sources. Then, from the retarded potentials thus obtained, we
derive the Lienart-Wiechert potentials, from which we can determine the
field. (See, for example, Feynman, Leighton and Sands 1963, Vol. 2.)
Here is the result (long after the acceleration; see Szabó 2004):
− 12

2
xK
q 1− v
K c2
1
E = 32 (29)
1 moving
2

2
K
x + K
x + B2
in K 1 2
− 12
2
xK
q 1− v
K c2
2
E = 32 (30)
2 moving
2

2
K
x + K
x + B2
in K 1 2

3K
E =
qB
(31)
moving

3
K
x
2
+
xK 2 + B2 2
in K 1 2
v K
1K
B = − E (32)
moving c 2
in K
v K
2K
B = E (33)
moving c 1
in K

K
B = 0 (34)
3 moving
in K
where
K

K
x3 − X3 t
B=
2
1 − c2
v
and X K (t̃) is the position

of the charge at time t. The electromagnetic
3
field of the charge changed : earlier it was like (23)–(28), then it changed
for the one described by (29)–(34). There appeared a magnetic field (turn-
ing the magnetic needle, for example) and the electric field flattened in the
direction of motion (Fig. 8). No physicist would say that this is not a real
physical change in the electromagnetic field of the charge, only because
we can express the new electromagnetic field in terms of the variables
of another reference frame K in which it has the same form as the old
electromagnetic field expressed in the original variables – even if this K
happens to be the new co-moving frame of reference. Quite the contrary,
if the field remained unchanged it would have a different form in K ;
namely, the one obtained from (23)–(28) by a Lorentz transformation.
23. Thus, relativistic deformations are real physical deformations also in
special relativity theory. One has to emphasize this fact because it is an
important part of the physical content of relativity theory. It must be
clear, however, that this conclusion is independent of our main concern.
What is important is the following: The Lorentz theory and special rel-
and length,
ativity have identical assertions about length duration
and

duration, and shrinking,
shrinking etc. Consequently, whether or not
these facts provide enough reason to say that relativistic deformations are
real physical changes, the conclusion is common to both theories.
On the null result of the Michelson–Morley experiment

24. Consider the following passage from Einstein:
A ray of light requires a perfectly definite time T to pass from one mirror to
the other and back again, if the whole system be at rest with respect to the
aether. It is found by calculation, however, that a slightly different time T 1
is required for this process, if the body, together with the mirrors, be moving
relatively to the aether. And yet another point: it is shown by calculation that
for a given velocity v with reference to the aether, this time T 1 is different
when the body is moving perpendicularly to the planes of the mirrors from that
resulting when the motion is parallel to these planes. Although the estimated
difference between these two times is exceedingly small, Michelson and Morley
performed an experiment involving interference in which this difference should
have been clearly detectable. But the experiment gave a negative result – a fact
very perplexing to physicists. (Einstein 1920, p. 49)
The “calculation” that Einstein refers to is based on the Galilean “kine-

matics”, that is, on the invariance of “time” and “simultaneity”, on the
invariance of “distance”, on the classical addition rule of “velocities”, etc.
That is to say, “distance”, “time”, and “velocity” in the above passage
time,
mean the classical distance, and velocity
defined in (D3), (D4),
(D7), and (D8). The negative result was “very perplexing to physicists”
because their expectations were based on the traditional concepts of space
and they could not imagine other that if the speed
and time, of a light
of the same light
signal is c relative to one inertial frame then the speed
signal cannot be the same c relative to another reference frame.
25. On the other hand, Einstein continues this passage in the following
way:
Lorentz and FitzGerald rescued the theory from this difficulty by assuming
that the motion of the body relative to the aether produces a contraction of
the body in the direction of motion, the amount of contraction being just suf-
ficient to compensate for the difference in time mentioned above. Comparison
with the discussion in Section 11 shows that also from the standpoint of the
theory of relativity this solution of the difficulty was the right one. But on
the basis of the theory of relativity the method of interpretation is incompa-
rably more satisfactory. According to this theory there is no such thing as a
“specially favoured” (unique) co-ordinate system to occasion the introduction
of the aether-idea, and hence there can be no aether-drift, nor any experiment
with which to demonstrate it. Here the contraction of moving bodies follows
from the two fundamental principles of the theory, without the introduction of
⎫
[deformations] ⎪
⎪
⎪
⎪
& ⎪
⎡ ⎤ ⎪
⎪
⎪
⎡ ⎤

Galilean kinematics ⎪
⎪ the result of the
⎬ ⎢ Michelson-Morley ⎥
⎢ ⎥
⎢ for x̂, t̂ (the speed ⎥ ⇒ ⎢
⎣ experiment must ⎦
⎥
⎢
⎢ of light is NOT ⎥ ⎪
⎥ ⎪
⎪
⎣ ⎦ ⎪
⎪ be the null effect
the same in all ⎪
⎪
⎪
⎪
inertial frame) ⎪
⎭
# $% &

% &# ⎫
⎡ ⎤$ ⎪
⎪

Lorentz kinematics ⎪
⎪ ⎡ ⎤
⎢ ⎥ ⎪⎪
⎢ ⎥ ⎪⎪ the result of the
⎢ for x̃, t̃ (the speed ⎥ ⎬ ⎢ Michelson-Morley ⎥
⎣ of light IS the same ⎦ ⇒ ⎢
⎣ experiment must
⎥
⎦
⎪
⎪
in all inertial frame) ⎪
⎪
⎪
⎪ be the null effect
& ⎪
⎪
⎭
[deformations]
Schema 1: The null result of the Michelson–Morley experiment simulta-
for x̂ and
neously confirms both, the classical rules of Galilean kinematics

t̂, and the violation of these rules (Lorentzian kinematics)
for the space
tags x
and time ,
t.
particular hypotheses; and as the prime factor involved in this contraction we

find, not the motion in itself, to which we cannot attach any meaning, but the
motion with respect to the body of reference chosen in the particular case in
point. Thus for a co-ordinate system moving with the earth the mirror system
of Michelson and Morley is not shortened, but it is shortened for a co-ordinate
system which is at rest relatively to the sun. (Einstein 1920, p. 49)
What “rescued” means here is that Lorentz and FitzGerald proved, within
the framework of the classical theory and Galilean kinematics,
space-time
that if the assumed deformations of moving bodies exist then the expected
result of the Michelson–Morley experiment is the null effect. But we have
already clarified, what Einstein also confirms in the above quoted passage,
that these deformations also derive from the two basic postulates of special
relativity.
Putting all these facts together (see Schema 1), we must say that
the null result of the Michelson–Morley experiment simultaneously con-

firms both, the classical rules of Galilean kinematics for x̂ and t̂, and the

Lorentzian kinematics for the
space and time tags x,
t. It confirms the

classical addition rule of velocities, on the one hand, and, on the other
of light is the same in all frames of
hand, it also confirms that velocity
reference.
This actually holds for all other experimental confirmations of special
relativity. That is why the only difference Einstein can mention in the
quoted passage is that special relativity does not refers to the aether.
(As a historical fact, this difference is true. Although, as we will see in
Points 30–34, the concept of aether can be entirely removed from the
recent logical reconstruction of the Lorentz theory.)
The conventionalist approach

26. According to the conventionalist thesis, the Lorentz theory and Ein-
stein’s special relativity are two alternative scientific theories which are
equivalent on empirical level (see Friedman 1983, p. 293; Einstein 1983,
p. 35). Due to this empirical underdeterminacy, the choice between
these alternative theories is based on external aspects. (Cf. Zahar 1973;
Grünbaum 1974; Friedman 1983; Brush 1999; Janssen 2002.) Follow-
ing Poincaré’s similar argument about the relationship between geome-
try, physics, and the empirical facts, the conventionalist thesis asserts the
following relationship between the Lorentz theory and special relativity:
⎡ ⎤
⎡ ⎤ physical
classical + ,
⎢ content of ⎥ empirical
⎣ space-time ⎦ + ⎢ ⎥ =
⎣ Lorentz ⎦ facts
E3 × E1
theory
⎡ ⎤ ⎡ ⎤
relativistic special + ,
⎣ space-time ⎦ + ⎣ relativistic ⎦ empirical
=
facts
M4 physics
Continuing the symbolic notations we used in Points 1–2, denote Z those

objective features of physical reality that are described by the alternative
physical theories P1 and P2 in question. With these notations, the logical
schema of the conventionalist thesis can be described in the following way:
We cannot distinguish by means of the available experiments whether

G1 (M ) & P1 (Z) is true about the objective features of physical reality
M ∪ Z, or G2 (M ) & P2 (Z) is true about the same objective features
M ∪ Z. Schematically,
+ ,
empirical
[G1 (M )] + [P1 (Z)] =
facts
+ ,
empirical
[G2 (M )] + [P2 (Z)] =
facts
27. However, it is clear from the previous sections that the terms “space”
and “time” have different
meanings in the two theories. The Lorentz

theory claims G1 M̂ about M̂ and relativity theory claims G2 M
about some other features of reality M . Of course, this terminological

confusion also appears in the physical assertions. Let us symbolize with
of a rod, etc.,
Ẑ the objective features of physical reality, such as the length

described by physical theory P1 . And let Z denote some (partly) different
of a rod, etc. Now, as
features of reality described by P2 , such as the length
we have seen, both theories actually claim that G1 M̂ & G2 M . It is
also clear that, for example, within the Lorentz theory, we can legitimately
of a rod. For the Lorentz theory has complete description
query the length
of the behavior of a moving rigid rod, as well as the behavior of a moving
clock and measuring-rod. Therefore, it is no problem to predict, in the
Lorentz theory, the result of a measurement of the “length” of the rod, if
the measurement is performed with the co-moving measuring equipments,
according to empirical definition (D6). This prediction will be exactly
the same as the prediction of special relativity. And vice versa, special
of the rod as the
relativity would have the same prediction for the length
prediction of the Lorentz theory. That is to say, the physical contents of
the Lorentz theory and special relativity also are identical: both claim

. So we have the following:
that P1 Ẑ & P2 Z
+ ,
+ P1 Ẑ & P2 Z
empirical
G1 M̂ & G2 M =
facts
+ ,
+ P1 Ẑ & P2 Z
empirical
G1 M̂ & G2 M =
facts
In other words, there are no different theories; consequently there is no

choice, based neither on internal nor on external aspects.
Methodological remarks
28. It is to be noted that my argument is based on the following very
weak operationalist/verificationist premise: physical terms, assigned to
measurable physical quantities, have different meanings if they have dif-
ferent empirical definitions. This premise is one of the fundamental
pre-assumptions of Einstein’s 1905 paper and is widely accepted among
physicists. Without clear empirical definition of the measurable physical
quantities a physical theory cannot be empirically confirmable or discon-
firmable. In itself, this premise is not yet equivalent to operationalism or
verificationism. It does not generally imply that a statement is necessarily
meaningless if it is neither analytic nor empirically verifiable. However,
when the physicist assigns time and space tags to an event, relative to
a reference frame, (s)he is already after all kinds of metaphysical consid-
erations about “What is space and what is time?” and means definite
physical quantities with already settled empirical meanings.
29. In saying that the meanings of the words “space” and “time” are
different in relativity theory and in classical physics, it is necessary to
be careful of a possible misunderstanding. I am talking about something
entirely different from the incommensurability thesis of the relativist phi-
losophy of science (Kuhn 1970, Chapter X; Feyerabend 1970). How is it
that relativity makes any assertion about classical and
space and time,
vice versa, how can the Lorentz theory make assertions about quantities
which are not even defined in the theory? As we have seen, each of the
two theories is a sufficiently complete account of physical reality to make
predictions about those features of reality that correspond – according to
the empirical definitions – to the variables used by the other theory, and
it is no problem to compare these predictions. For example, within the
Lorentz theory, we can legitimately query the reading of a clock slowly
transported in K from one place to another. That exactly is what we
calculated in Point 14. Similarly, in special relativity theory, we can le-
gitimately apply formulas (1)–(2) to the tags of an event.
space and time
This is a fair calculation, in spite of the fact that the result thus obtained
is not explicitly mentioned and named in the theory. This is what we
actually did. And the conclusion was that not only are the two theories
commensurable, but they provide completely identical accounts of the
same physical reality; they are identical theories.
Privileged reference frame

30. Due to the popular/textbook literature on relativity theory, there
is a widespread aversion to a privileged reference frame. However, like
it or not, there is a privileged reference frame in both special relativity
and classical physics. It is the frame of reference in which the etalons
are at rest. This privileged reference frame, however, has nothing to do
with the concepts of “absolute rest” or the aether; it is not privileged by
nature, but it is privileged by the trivial semantical convention providing
meanings for the terms “distance” and “time”, by the fact that of all
possible measuring-rod-like and clock-like objects floating in the universe,
we have chosen the ones floating together with the International Bureau
of Weights and Measures in Paris. In Bridgman’s words:
It cannot be too strongly emphasized that there is no getting away from pre-
ferred operations and unique standpoint in physics; the unique physical opera-
tions in terms of which interval has its meaning afford one example, and there
are many others also. (Bridgman 1936, p. 83)
31. Many believe that one can avoid a reference to the etalons sitting

in a privileged reference frame by defining, for example, the unit of time

for an arbitrary (moving) frame of reference K through a cesium clock,
or the like, co-moving with K . In this way, it is claimed, one needs not
to refer to a standard clock accelerated from the reference frame of the
etalons into reference frame K .
In this view, however, there appears a methodological nonsense. For if

then we
this operation is regarded as a convenient way of measuring time,

still have time in the theory, together with the privileged reference frame
of the etalons. If, however, this operation is regarded as the empirical
definition of a physical quantity, then it must be clear that this quantity

but a new physical quantity, say time.
In order to establish
is not time

tags belonging to different reference frames,
any relationship between time
it is a must to use an “etalon cesium clock” as well as to refer to its
behavior when accelerated from one inertial frame into the other, or, in
some other way, to describe the other clocks’ behaviors in term of the
physical quantity defined with the etalon.
The aether
32. Many of those, like Einstein himself (see Point 25), who admit the
“empirical equivalence” of the Lorentz theory and special relativity argue
that the latter is “incomparably more satisfactory” (Einstein) because it
has no reference to the aether. As it is obvious from the previous sections,
we did not make any reference to the aether in the reconstruction of the
Lorentz theory. It is however a historic fact that, for example, Lorentz did.
In this section, I want to show that Lorentz’s aether hypothesis is logically
independent from the actual physical content of the Lorentz theory. In
other words, the concept of aether is merely a verbal ornament in Lorentz’s
theory, which can be interesting for the historians, but negligible from the
point of view of the recent logical reconstructions. (Actually the same
holds for the “denial of aether” by Einstein’s special relativity.)
Consider, for example, Lorentz’s aether-theoretic formulation of the
relativity principle – to touch on a sore point.
33. Let us introduce the following notation:
A (K , K ) := The laws of physics in K are such that the laws of physics

empirically ascertained by an observer in K , describing
the behavior of physical objects co-moving with K , ex-

pressed in variables x K
1 ,x K
2 ,x K
3 ,t
K , have the same
forms as the similar empirically ascertained laws of physics
in in K , describing the similar physical objects co-moving

with K , expressed in variables x K
1 ,x K
2 ,x K K
3 , t , if the
observer in K performs the same measurement opera-

tions as the observer in K with the same measuring equip-
ments transferred from K to K , ignoring the fact that the
equipments undergo deformations during the transporta-
tion.
Taking this statement, the usual Einsteinian formulation of the relativity

principle is the following:

Einstein’s Relativity Principle = (∀K ) (∀K ) A (K , K )
Many believe that this version of relativity principle is essentially differ-

ent from the similar principle of Lorentz, since Lorentz’s principle makes
explicit reference to the motion relative to the aether. Using the above
introduced notations, it says the following:

Lorentz’s Principle = (∀K ) A (aether, K )
It must be clearly seen, however, that Lorentz’s principle and Einstein’s
relativity principle are logically equivalent to each other. On the one hand,
it is trivially true that

Einstein’s Relativity Principle = (∀K ) (∀K ) A (K , K )
⇒ (∀K ) A (aether, K )

= Lorentz’s Principle
On the other hand, it follows from the meaning of A (K , K ) that
(∃K ) (∀K ) A (K , K ) ⇒ (∀K ) (∀K ) A (K , K )
The reason is that the laws of physics in K completely determine the

results of the measurements performed by a moving – relative to K –
observer on moving physical objects with moving measuring equipments.
Consequently,

Lorentz’s Principle = (∀K ) A (aether, K )
⇒ (∃K ) (∀K ) A (K , K )
⇒ (∀K ) (∀K ) A (K , K )

= Einstein’s Relativity Principle
Thus, it is Lorentz’s principle itself – which refers to the aether –

that renders any claim about the aether a logically separated hypothesis
outside of the scope of the factual content of both the Lorentz theory and
special relativity. It is Lorentz’s principle itself – again, which refers to
the aether – that implies that the role of the aether could be played by
anything else; the aether does not constitute a privileged reference frame.
As the Lorentz theory and special relativity unanimously claim, physical
systems undergo deformations when they are transferred from one inertial
frame K to another frame K . One could say, these deformations are
caused by the transmission of the system from K to K . You could say
they are caused by the “wind of aether”. By the same token you could
say, however, that they are caused by “the wind of anything”, since if
the physical system is transferred from K to K then its state of motion
changes relative to an arbitrary third frame of reference.
34. On the other hand, it must be mentioned that special relativity does
not exclude the existence of the aether. (Not to mention that already
in 1920 Einstein himself argues for the existence of some kind of aether.
See Reignier 2000.) Neither does the Michelson–Morley experiment. If
special relativity/Lorentz theory is true then there must be no indication
of the motion of the interferometer relative to the aether. Consequently,
the fact that we do not observe indication of this motion is not a challenge
for the aether theorist. Thus, the hypothesis about the existence of aether
is logically independent of both the Lorentz theory and special relativity.
Heuristic and explanatory values

35. The Lorentz theory and special relativity, as completely identical the-
ories, offer the same symmetry principles and heuristic power. As we

have seen, both theories claim that quantities x K ,
tK in an arbitrary

K and the similar quantities xK , tK in another arbitrary K are re-
lated through a suitable Lorentz transformation. This fact in conjunction
with the relativity principle (within the scope of validity of the principle)
implies7 that laws of physics are to be described by Lorentz covariant

equations, if they are expressed in terms of variables x and t, that is,
in terms of the results of measurements obtainable by means of the cor-
responding co-moving equipments – which are distorted relative to the
etalons. There is no difference between the two theories that this s pace-
symmetry provides a valuable heuristic aid in the search for new laws
time
of nature.
36. It is sometimes claimed that the main difference between the Lorentz
theory and Einstein’s special relativity is that the Lorentz theory is con-
structive, in the sense that it tries to explain the relativistic effects from
the laws of the detailed underlying physical processes, while special rela-
tivity deduces the same result from some basic principles (cf. Bell 1992,
p. 34; Brown and Pooley 2001). As we have seen in Point 17, the ba-
sic principles of the two theories are logically equivalent; both reduce to
(RP); the two theories are identical; the statements of “both theories”
can be derived from (RP). So, if the fact that the statements of special
relativity can be derived from (RP) provides enough reason to say that
special relativity is a principle theory, then the same fact provides enough
reason to say the same thing about the Lorentz theory. And vice versa, if
the statements of the Lorentz theory – all derivable from (RP) – provide
enough reason to say that it is a constructive theory, then the same fact
provides enough reason to say the same thing about special relativity.
Though, it is a historic fact that Lorentz, FitzGerald, and Larmor,
in contrast to Einstein, made an attempt to understand how these laws
actually come about from molecular forces. These are perfectly legitimate
additional questions.8
37. With these comments I have completed the argumentation for my

basic claim that special relativity and the Lorentz theory are completely
identical. Again, the historical questions are not important from the point
of view of our analysis. What is important is the logical possibility of a
7 In fact, the relativity principle does not necessarily imply the covariance of
the physical equations, except under some specific conditions (Gömöri and
Szabó 2009).
8 Moreover, a careful consideration of these details reveals that the principle
of relativity cannot be regarded as a universal principle: it does not neces-
sarily hold for the whole range of validity of the Lorentz covariant laws of
relativistic physics (Szabó 2004; Gömöri and Szabó 2009).
Lorentz-type theory: the classical Galileo-invariant spatio-temporal con-

ceptions + deformations of moving objects, governed by the relativity
principle. And, what we proved is that such a theory is completely iden-
tical to special relativity in both senses, as theories about space and time
and as theories about the behavior of moving physical objects. They are
not only “empirically equivalent”, as often claimed, but they are identical
in all sense; they are identical physical theories.
Consequently, in comparison with the classical Galileo-invariant
conceptions, special relativity theory tells us nothing new about the
spatiotemporal features of the physical world. As we have seen, the
longstanding belief that it does is the result of a simple but subversive
terminological confusion.
Acknowledgments
The research was partly supported by the OTKA Foundation, No. K
68043. I am grateful to the Netherlands Institute for Advanced Study
(NIAS) for providing me with the opportunity, as a Fellow-in-Residence,
to work on this project.
References
J. S. Bell, Speakable and unspeakable in quantum mechanics. Cambridge:
Cambridge University Press 1987.
J. S. Bell, “George Francis FitzGerald”, in: Physics World, 5, 1992, pp.
31–35.
P. Bridgman, The Logic of Modern Physics. New York: MacMillan 1927.
H. R. Brown and O. Pooley, “The origin of space-time metric: Bell’s
‘Lorentzian pedagogy’ and its significance in general relativity”, in:
C. Calleander and N. Huggett (eds.), Physics meets philosophy at the
Planck scale. Contemporary theories in quantum gravity. Cambridge:
Cambridge University Press 2001.
H. R. Brown, “The origins of length contraction: I. The FitzGerald-
Lorentz deformation”, http://philsci-archive.pitt.edu/archive/
00000218 (2001)
H. R. Brown, “Michelson, FitzGerald and Lorentz: the origins of
relativity revisited”, http://philsci-archive.pitt.edu/archive/00000987
(2003)
S. G. Brush, “Why was Relativity Accepted?”, in: Physics in Perspective,

1, 1999, pp. 184-214.
A. Einstein, “Zur Elektrodynamik bewegter Körper”, in: Annalen der
Physik, 17, 1905, p. 891.
A. Einstein, Relativity: The Special and General Theory. New York: H.
Holt and Company 1920.
A. Einstein, Sidelights on relativity. New York: Dover 1983.
P. K. Feyerabend, “Consolation for the Specialist”, in: I. Lakatos and A.
Musgrave (eds.), Criticism and the Growth of Knowledge. Cambridge:
Cambridge University Press 1970, pp. 197-230.
M. Friedman, Foundations of Space-Time Theories – Relativistic Physics
and Philosophy of Science. Princeton: Princeton University Press
1983.
M. Gömöri and L. E. Szabó, “Is the relativity principle consistent with
electrodynamics? Towards a logico-empiricist reconstruction of a
physical theory”, arXiv:0912.4388 (2009)
A. Grünbaum, Philosophical Problems of Space and Time. Dordrecht: D.
Reidel 1974.
L. Jánossy, Theory of relativity based on physical reality. Budapest:
Akadémiai Kiadó 1971.
M. Janssen, “Reconsidering a Scientific Revolution: The Case of Einstein
versus Lorentz”, in: Physics in Perspective, 4, 2002, pp. 421-46.
T. S. Kuhn, The Structure of Scientific Revolution. Chicago: University
of Chicago Press 1970.
H. A. Lorentz, “Electromagnetic phenomena in a system moving with any
velocity less than that of light”, in: Proc. R. Acad. Amsterdam, 6,
1904, p. 809.
H. A. Lorentz, The theory of electrons and its applications to the phe-
nomena of light and radiant heat; a course of lectures delivered in
Columbia University, New York, in March and April 1906. New York:
Columbia University Press 1909/1916.
J. Reignier, “The birth of special relativity. ‘One more essay on the
subject’”, arXiv:physics/0008229 (2000).
L. E. Szabó, “On the meaning of Lorentz covariance”, in: Foundations of
Physics Letters, 17, 2004, p. 479.
L. E. Szabó, “Empirical Foundation of Space and Time”, in: M. Suárez et

al. (eds.), EPSA Philosophical Issues in the Sciences. Launch of the
European Philosophy of Science Association. Berlin: Springer 2009.
E. Zahar, “Why did Einstein’s Programme Supersede Lorentz’s?”, in:
British Journal for the Philosophy of Science, 24, 1973, pp. 95-123,
223-262.
András Máté
Die Rezeption der neuen Logik in Ungarn1
Ich möchte hier keine vollständige Rezeptionsgeschichte geben, sondern

die Frage nur im Zusammenhang mit dem Einfluss des logischen Empi-
rismus untersuchen. Das heißt, dass meine Untersuchung vor allem auf
die ungarische Geschichte der symbolischen Logik als philosophische Lo-
gik, als Grundlage einer allgemeinen Methodologie gerichtet ist. Es folgt
aber nicht, dass die symbolische Logik als Gegenstand von besonders er-
folgreichen mathematischen Forschungen in Ungarn außer Betracht lassen
werden kann. Im Gegenteil: Die Ansichten der Mathematiker und der aus
der Mathematik kommenden Philosophen sind für uns besonders inter-
essant, können sogar als Ausgangspunkt dienen; es können aber eben nur
die philosophisch relevanten Gedanken und Beziehungen abhandelt wer-
den.
Die symbolische Logik wurde relativ spät zu einem Forschungsbereich
in der ungarischen Mathematik. Der erste bedeutende ungarische Mathe-
matiker auf dem Gebiet der Grundlagenforschung war Julius König, der
in den ersten Jahren des 20. Jahrhunderts wichtige Ergebnisse und ein
noch interessanteres falsches Ergebnis in der Mengenlehre publiziert hat.2
Er hat in seinen letzten Lebensjahren einen philosophischen Artikel über
Logik geschrieben und noch ein posthumes Buch über die Grundlagen der
Mathematik hinterlassen,3 seine Tätigkeit fand aber keine Nachfolger auf
1 Dieser Artikel wurde durch die Stiftung OTKA, Projekt No. 68043
gefördert.
2 Gyula Kõnig (in seinen deutschen Publikationen: Julius König, 1849–1913),
Professor der Technischen Universität in Budapest, weltweit anerkannte
Funktionentheoretiker. Er hat im Jahre 1905 eine Widerlegung von Cantors
Kontinuum-Hypothese veröffentlicht, sein Beweis war aber fehlerhaft. Siehe
darüber die MacTutor History of Mathematics, http://www-groups.dcs.st-
and.ac.uk/∼history/Biographies/Konig Julius.html.
3 Die Formalisierung des Wahrheitsbegriffs in der synthetischen Lo-
”
gik“(ung.), A Magyar Filozófiai Társaság Közleményei, 34 (1910),
S. 98-106; Neue Grundlagen der Logik, Arithmetik und Mengenlehre.
230 András Máté
dem Gebiet der Logik. John von Neumann, der in der zwanziger Jahren
grundlegende Arbeiten zur Mengenlehre publiziert hatte, schaltete sich in
die Forschung über die Grundlagen der Mathematik im Hilbert-Kreis in
Göttingen ein. László Kalmár hörte auch zum ersten Mal von Neumann
über die moderne Logik und wurde während seines Studienaufenthaltes in
Göttingen im Jahre 1929 zum Experten und Forscher auf diesem Gebiet.4
Er und seine lebenslange Mitarbeiterin Rózsa Péter5 haben die mathema-
tische Logik als Disziplin in die ungarische Mathematik eingeführt und
schon in den dreißiger Jahren internationales Ansehen erreicht.
Kalmár und Péter waren philosophisch sehr interessierte Mathemati-
ker; bei Kalmár kann man sogar von einer eigenen Philosophie der Ma-
thematik sprechen.6 Es wird sich lohnen, einige seiner Ansichten hier an-
Leipzig: Veit und Co. 1914. Auf diese heute schon ganz vergesse-
nen Schriften hat mich die Arbeit von Lilla Ignácz ( Die Rezep-
”
tion der symbolischen Logik in Ungarn“(ung.), http://www.geocities.
com/elianaszabo/AszimblogMon.htm# ftnref32) aufmerksam gemacht.
4 László Kalmár (1905–1976), Professor an der Universität von Szeged. Ne-
ben der mathematischen Logik war er auch in der Computerwissenschaft
tätig und hat die Universitätslehre in Informatik in Ungarn begründet; er
war auch ein besonders aktiver Befürworter der Modernisierung des Ma-
thematikunterrichts in Grund- und Mittelschulen. Wichtigste Ergebnisse in
Logik: On the Reduction of the Decision Problem I.“Journal of Symbolic
”
Logic 4 (1939). (Fortsetzungen mit János Surányi, dieselbe Journal of Sym-
bolic Logic 1947, 1950); Der Kalmársche Widerspruchsfreiheitbeweis“[der
”
Arithmetik], publiziert von Paul Bernays in: David Hilbert – Paul Bernays,
Grundlagen der Mathematik, Bd. II, 2. Aufl., Berlin: Springer 1970, S. 513-
536. Diesen Beweis hat Kalmár selber nie publiziert. Bernays hörte darüber
am von Imre Lakatos organisierten Londoner Symposium über die Philo-
sophie der Mathematik in 1965, (siehe unten, Angaben der Proceedings in
Fn. 8.) und Kalmár hat sein Manuskript von 1938 nach langer Überredung
seitens Bernays und Lakatos Bernays zur Verfügung gestellt. Quellen: das
Manuskript und die Korrespondenz zwischen Kalmár, Lakatos und Bernays
im Kalmár-Archiv der Universität von Szeged und im Lakatos-Archiv der
London School of Economics.
5 Rózsa Péter (1905–1977), Professorin der Pädagogischen Hochschule, später
der Eötvös Universität in Budapest, board member des Journal of Symbolic
Logic von seiner Gründung (1936) an. Péters Hauptwerk: Rekursive Funk-
tionen, Budapest: Akadémiai, 1951. In den sechziger und siebziger Jahren
nahm sie führend an der Reform des Mathematikunterrichts teil.
6 Wichtigste philosophische Schriften: Die Entwicklung der mathematischen
”
Exaktheit von der Anschauung bis zur axiomatischen Methode“(ung., Vor-
Die Rezeption der neuen Logik in Ungarn 231
zuführen, um die Parallelen und Unterschiede zur Logik- und Mathema-

tikauffassung des Wiener Kreises bewerten zu können.
Kalmár ist vor allem Fallibilist in der Mathematik; seiner Meinung
nach ist es ein verbreitetes aber falsches Vorurteil, dass die Theoreme
der Mathematik absolut sichere Wahrheiten sind. Die Exaktheit und Zu-
verlässigkeit der mathematischen Beweise ist zwar grundsätzlich wichtig,
ist aber keineswegs zeitlos und absolut. Die Methoden und Normen haben
sich in der Geschichte mehrmals geändert und werden sich auch weiter-
hin entwickeln. Die Mathematiker sind ebensowenig unfehlbar wie andere
Wissenschaftler; mathematische Beweise werden oft – teilweise oder sogar
gänzlich – widerlegt. Im Hintergrund der Arbeit der falliblen Mathema-
tiker steht keine unfehlbare ideale Mathematik. Im Zusammenhang mit
dieser Auffassung steht Kalmárs Ansicht über die Bedeutung der For-
malisierung in der Mathematik. Die Mathematik, oder gar eine mathe-
matische Theorie ist nicht identisch mit ihrer Formalisierung innerhalb
einer logischen Sprache. Die formal-axiomatische Methode ist nicht ein-
mal eine neue Stufe in der Entwicklung der mathematischen Exaktheit,
weil sie tatsächlich für sich betreiben wäre bloßes Spiel und keine Mathe-
”
matik“.7 Dieser Standpunkt ist grundverschieden von dem der Klassiker
des Logizismus, Frege und Russell. Frege in Grundgesetze der Arithme-
tik oder Russell und Whitehead in den Principia Mathematica haben
eben versucht, die Grundlagen der Mathematik mit einer rein formali-
sierten Sprache darzustellen, um die vollkommene Sicherheit, also Infal-
libilität zu erreichen. Hinsichtlich der Bedeutung der Formalisierung der
mathematischen Theorien vertritt Kalmár eigentlich die Ansichten der
Hilbert-Schule, und diese Ansichten stehen im Gegensatz zu jeder Form
von Logizismus, einschließlich des Logizismus des Wiener Kreises. Die
Formalisierung ist nach der Auffassung der Hilbertianer Formalisten“
”
ein Mittel der Metamathematik, die Theorien selbst zum Gegenstand der
mathematischen Untersuchung zu machen. Die Metamathematik ist nach
Kalmár ebenso inhaltlich wie jede andere mathematische Theorie, wenn
auch sie ganz einfach formalisiert werden könnte. Die Metamathematik
bringt jedoch eine neue Stufe der Exaktheit mit sich, es ist aber nicht
lesung, 1941 – englische Übersetzung in diesem Band unter dem Titel The
”
Development of Mathematical Rigor from Intuition to Axiomatic Method“.)
und die unten (Fn. 8., 9.) zitierten Konferenzvorlesungen.
7 The Development of Mathematical Rigor . . .“, S. ???.
”
die Formalisierung, sondern das Modellieren der mathematischen Theori-

en innerhalb einfacher, mehr zuverlässiger Theorien, insbesondere in der
Theorie der natürlichen Zahlen.
Der andere charakteristische Zug von Kalmárs Ansichten ist sein ma-
thematischer Empirismus, was auf dem Kongress über die Philosophie
der Mathematik in London 1965 beträchtliches Aufsehen erregte.8 Er
steht in einem natürlichen Zusammenhang mit seinem Fallibilismus, ist
aber nicht dasselbe. Die mathematischen Begriffe und einige mathemati-
sche Aussagen sollen nach Kalmár empirische Grundlagen haben. So sind
zum Beispiel die Church-These oder die Konsistenz der Zermelo-Fraenkel
Mengenlehre Behauptungen, auf denen unsere mathematische Tätigkeit
allgemein gründet, sie sind aber deduktiv nicht beweisbar. Wir nehmen
diese an, weil 1) unsere Erfahrungen sie nicht widerlegt haben, 2) wir auch
theoretische Argumente für sie haben und 3) wir sie benötigen, denn sie
sind sogar unentbehrlich für die moderne Mathematik. Das Eigentümliche
an Kalmárs Auffassung ist nicht, dass diese Behauptungen deduktiv nicht
beweisbar sind; das ist wohlbekannt. Das Neue und Seltsame ist, dass nach
ihm diese Aussagen zur Mathematik gehören – und nicht zur Interpreta-
tion der Mathematik oder zu mathematisch irrelevanten philosophischen
Behauptungen bezüglich der Mathematik. Es folgt aber aus Kalmárs be-
tont inhaltlicher, antiformaler Auffassung der Mathematik. Man kann das
Theorem von Church beweisen, ohne sich auf seine These zu stützen. Es
sagt dann aber nicht die Unentscheidbarkeit der Logik erster Stufe aus,
sondern nur soviel, dass es keine rekursive Funktion gibt, welche jeder
gültigen Formel der Logik erster Stufe 1, jedes ungültige 0 zuordnet. Es
ist wahr, aber uninteressant. Das Theorem wird inhaltsleer, wenn wir
die These außer Acht lassen wollen. Ebensogut kann man sagen, dass
der Mengentheoretiker in seinen Beweisen die Konsistenz des Zermelo-
Fraenkel Systems nie benutzt. Wenn aber die Konsistenz fällt, verlieren
seine Theoreme ihren Sinn im strengsten Sinne: sie werden nichtssagend.
Die These, dass die Mathematik nicht rein formal, nicht bloß eine
strukturierte Menge von Formeln ist, bedeutet in Kalmárs Auffassung
nicht nur, dass die Mathematik nicht auf eine oder mehrere formale Theo-
rien reduziert werden kann (was schon aus Gödels Theoremen folgt), son-
8 S. die Diskussion seines Vortrags Foundations of Mathematics – whither

”
now?“, in: Imre Lakatos (ed.), Problems in the Philosophy of Mathematics.
Amsterdam: North-Holland 1967, S. 186-194., Diskussion: ebd., S. 195-207.
dern auch, dass die Behauptungen, die den Sinn, den Inhalt der Theo-
reme sichern, ebensogut zur Mathematik gehören. Seine Argumentation
bezüglich der Church-These9 gibt ein Beispiel davon, wie solche Behaup-
tungen durch mathematische Mittel untersucht – wenn auch nicht ent-
schieden – werden können. Er will die These nicht widerlegen; er sagt
aber, dass Churchs These, ebenso wie andere empirische Verallgemeine-
rungen in der Wissenschaft, ständig überprüft und die Pros und Kontras
erwogen werden sollen.
Kalmárs Philosophie der Mathematik zieht radikale aber wohlbegrün-
dete Folgerungen aus Gödels Theoremen, insbesondere aus der Unbeweis-
barkeit der Konsistenz. Den Traum von den endgültigen und absolut siche-
ren Grundlagen der Mathematik muss man aufgeben. Die mathematische
Grundlagenforschung kann aber zu neuen, mehr und mehr sicheren und
zuverlässigen Grundlagen führen. In diesem Sinne ist Gentzens Beweis der
Konsistenz der Arithmetik ein Schritt vorwärts, wenn er auch kein streng
finitistischer Beweis ist. Kalmár wurde in der Diskussion seines Londoner
Vortrags gefragt, was sich in der Grundlagenforschung ändern sollte, wenn
man seinen mathematischen Empirismus akzeptiert. Seine Antwort war,
dass wir dasselbe tun werden wie bis jetzt, aber mit besserem Gewissen.10
Wir werden nie in der Lage sein, uns gänzlich von deduktiv unbewiesenen,
nur durch Erfahrung und mittelbare Argumente unterstüzten Prämissen
befreien zu können; es ist aber kein Fehler, sondern gehört zur wahren
Natur der Mathematik.
Rózsa Péters philosophische Gedanken kann man am besten aus ih-
rem populärwissenschaftlichen Buch Das Spiel mit dem Unendlichen 11
kennenlernen. Sie bespricht die mathematische Logik Hilbert’scher Auf-
fassung und konzentriert sich auf die Beweistheorie, auf die Selbstkritik
”
der reinen Vernunft“, wie sie sie apostrophiert. Sie nimmt auch Stellung
in der Diskussion, ob nach Gödels zweitem Unvollständigkeitstheorem das
Hilbert-Programm aufgegeben werden muss. Ihr Standpunkt ist dem von
Neumanns entgegengesetzt und verwandt mit Kalmárs Meinung: Die ur-
9 An Argument against the Plausibility of Church’s Thesis“, in: Arendt Hey-

”
ting (ed.), Constructivity in Mathematics, Amsterdam: North-Holland, S.
72-80.
10 Foundations of Mathematics – whither now?“(Diskussion), S. 203.
”
11 Erste Ausgabe, ungarisch: Budapest: Dante 1945, erste deutsche Ausgabe:
Teubner, Leipzig–Berlin 1956. Übersetzt bis heute in 13 Sprachen.
sprünglichen Zielsetzungen des Programms sind zwar übertrieben, die Ar-

beit kann und muss aber fortgesetzt werden. In diesem Sinne gibt sie eine
besonders klare Darstellung von Gentzens Beweis der Widerspruchslosig-
keit der Arithmetik – gestützt auf Kalmárs Analyse und Umformulierung
des Beweises.
In der Philosophie der Zwischenkriegperiode in Ungarn gilt der im
Jahre 1933 verstorbene Ákos von Pauler als großer Logiker. Er hat, in An-
lehnung an Husserl und die von Husserl angeregte Bolzano-Renaissance,
eine Theorie der Gegenstände der Logik ausgearbeitet, wonach die logi-
schen Gegenstände objektiv-ideale, von Sprache und Denken vollkommen
unabhängige zeitlose Entitäten sein sollen – er vertrat also eine starke
Form des logischen Platonismus.12 Man kann zwar nicht sagen, dass Pau-
ler vom Ausbau der mathematischen Logik gar nichts zur Kenntnis ge-
nommen hat, hat sie aber nie ausführlicher behandelt und sie offensichtlich
nicht als eine grundsätzlich wichtige Entwicklung in der Logik angesehen.
In den dreißiger Jahren erschienen aber in der führenden philosophischen
Zeitschrift Athenaeum und auch in Buchform die Schriften einiger Ver-
fasser, die Schüler und Anhänger von Pauler waren, aber zugleich be-
trächtliche Kentnisse über die moderne Logik hatten und eine gewisse,
obwohl nicht unbegrenzte Sympathie mit der Logikauffassung und Wis-
senschaftsphilosophie des Wiener Kreises zeigten: Béla Bencsik,13 Ferenc
Lehner14 und Frigyes Pozsonyi.15 Pozsonyis diesbezügliche Tätigkeit ist
die umfangreichste, und dem Niveau nach gehört ihm auch der Vorrang.
Ich werde mich also auf seine Schriften konzentrieren.
Pozsonyi rezensiert zwischen 1936 und 1947 in der führenden philo-
sophischen Zeitschrift der Zwischenkriegsperiode Athenaeum 7 Bücher,
die die symbolische Logik und/oder die Philosophie des Wiener Kreises
12 Logik. Versuch einer Theorie der Wahrheit, ung. Budapest:Eggenberger

1925, deutsch Berlin–Leipzig:de Gruyter 1929.
13 Béla Bencsik (1906–1943), Buch zum Thema: Paradoxien in der Logik und
Mathematik (ung.), Budapest: Sárkány Nyomda 1932.
14 Ferenc Lehner (1911–1988, nach 1945: Lénárd), Buch über die Logik:
Hauptrichtungen in der heutigen Logik (ung.), Budapest 1936. Nach 1945
als Professor der pädagogischen Psychologie tätig. Scharfer Gegner der
von Péter und ihren Schüler Tamás Varga geleiteten und von Kalmár un-
terstützten Reform des Mathematikunterrichts.
15 Keine dokumentierte Personalangaben; er hatte ganz gewiß keine akademi-
sche Position. Nach mündliche Informationen arbeitete er als Patentanwalt.
behandeln. Unter den rezensierten Verfassern finden wir Carnap, Tarski

und Hilbert; in der letzten Rezension der letzten Nummer von Athenaeum
bespricht er Victor Krafts Mathematik, Logik und Erfahrung. Er berichtet
auch den Lesern von Athenaeum von Churchs A Bibliography of symbolic
logic und von Kalmárs und Péters Übersichten ihrer eigenen Forschungs-
gebiete in einer ungarischen mathematischen Zeitschrift. In der Rezension
der englischen Ausgabe von Carnaps Logische Syntax der Sprache bewer-
tet er nach ausführlicher und klarer Darlegung der Grundgedanken das
Buch als eines der wichtigsten Ereignisse der Literatur der modernen
”
Logik“. Bezüglich des letzten Teils des Buches, welcher philosophische
Fragen der logischen Syntax erörtert, ist aber Pozsonyis Meinung nicht so
günstig. Als Begründung behauptet er, dass die grundlegenden Untersu-
chungen des Buches
eigentlich unabhängig von der nicht nur antimetaphysischen, sondern sogar an-
tiphilosophischen Philosophie des wiener (sic!) Kreises [sind]. Der Leser, der
das vollkommene Unverständnis der Geisteswissenschaften und der geisteswis-
senschaftlichen Philosophie mit den Mitgliedern des wiener Kreises nicht teilen
will, kann diese Untersuchungen als die Methodologie der Mathematik und der
Naturwissenschaften ansehen. Carnaps Wissenschaftslogik“, mit der er die tra-
”
ditionelle Philosophie supplieren will, ist in der Tat nicht mehr als dies. Alles,
was er sagt, hat einen Sinn nur auf dem Gebiet der Mathematik und der Natur-
wissenchaften, in der Philosophie auf dem Gebiet der Erkenntnistheorie und der
naturwissenschaftlich begründeten Naturphilosophie. Sein Werk hat aber dort
auch einen unbestreitbar großen Wert.16
Ich habe aus dieser Rezension so viel zitiert, weil dieses doppelte
Verhältnis zu den Ideen Carnaps und des Wiener Kreises nicht nur für
Pozsonyi und die anderen erwähnten Schüler von Pauler charakteristisch
ist. Ganz verschiedene Bewertungen des Wiener Kreises und der modernen
Logik haben auch ähnliche Züge. Die Logik und die formale Methodolo-
gie der Naturwissenschaften sind akzeptiert und sogar hoch bewertet, das
Programm des Wiener Kreises, die traditionelle Philosophie durch logi-
sche Analyse abzulösen, wird aber scharf abgelehnt. Am 12. Dezember
1939 hielt die Ungarische Philosophische Gesellschaft eine Diskussionssit-
16 Frigyes Pozsonyi: Rudolf Carnap: The logical syntax of language“

”
(Rezension, ung.), Athenaeum 24 (1938), S. 115-117.
zung über symbolische Logik;17 das einführende Referat hielt Pozsonyi.

Diskussionsredner waren Bencsik, Lehner, der Geometer Béla Kerékjártó
und der mir ganz unbekannte Tivadar Szabó, der die Logistik“ als die
”
umfassendste logische Theorie klassifiziert, welche (im Gegensatz zum Re-
ferat Pozsonyis) gar nicht in ein Teilgebiet der Logik einzuordnen ist
und in jeder Hinsicht einen Fortschritt im Verhältnis zur klassischen“
”
Logik bedeuten soll. Pozsonyi betont in seinem Referat als Hauptvorteil
der symbolischen Logik gegenüber der klassischen Syllogistik ihre An-
wendbarkeit, vor allem als Methodologie der deduktiven Wissenschaften.
Er behauptet, dass die Logistik nominalistisch ist, indem sie die Aus-
drucksformen und nicht die ausgedrückten Inhalte untersucht; diese Theo-
rie ist konsistent mit einer idealistischen Interpretation der Inhalte, setzt
sie aber nicht voraus. Er sieht eine Verwandtschaft zwischen Husserls Pro-
gramm und der Mathematikauffasung und den Ideen der Logistiker. Zur
Möglichkeit von mehreren Logiken behauptet er, dass es unmöglich sei,
verschiedene Logiken aufzustellen, in denen die logischen Konstanten die-
selben Bedeutungen haben, weil die Gesetze der Logik inhaltlich wahre
Sätze sind, die sich auf die Bedeutungen der Konstanten gründen. (Dies
ist ja eben Carnaps Auffassung.) Nach den Ergebnissen von Gödel und
Tarski ist es aber unmöglich, alle logischen Sätze in einem einzigen wi-
derspruchsfreien System zusammenzufassen, sondern wir benötigen eine
unendliche Hierarchie von aufeinander gebauten formalisierten Sprachen
und axiomatisierten logischen Systemen.
Béla Bencsik versucht in seinem Beitrag, dieses Problem zu beant-
worten, indem er vorschlägt, den Begriff der Struktur anzuwenden. Er be-
nutzt diesen Begriff ganz im modernen Sinne und hofft, dass die Logik als
Strukturwissenschaft eine absolute, von Unvollständigkeitserscheinungen
unabhängige Theorie werden kann. Lehner meint, dass bei Carnap die Lo-
gistik ihre Kompetenz überschreitet; besonders scharf kritisiert er die Idee
der Einheit der Wissenschaft wegen ihres Reduktionismus und die Auf-
fassung der Logistiker bezüglich Wissenschaft und Pseudowissenschaft.
Im Schlußwort betont Béla von Brandenstein, der Vorsitzende der Gesell-
schaft, den mathematischen Charakter der Logistik und die sich daraus
17 Frigyes Pozsonyi: Symbolische Logik. Diskussionssitzung der Ungarischen

”
Philosophischen Gesellschaft am 12. Dezember 1939. Mit Beiträge von Béla
Kerékjártó, Béla Bencsik, Ferenc Lehner, Tivadar Szabó und Béla von Bran-
denstein“, Athenaeum 26 (1940), S. 57-71.
ergebenden Schwierigkeiten für die meisten Philosophen, der Diskussion

auch nur zu folgen. Er erklärt Carnaps Versuch, die symbolische Logik
zu einer universalen Methode auszuweiten, als von vornherein nutzlos:
”
Wir können nicht die ganze Wirklichkeit mit allen ihren Facetten auf die
Weise und nach dem Muster des mathematischen Denkens behandeln.“
Pozsonyi gibt in seinem Buch Aufgabe und Sachgebiet der Logik, mit
einem historischen Überblick 18 eine kurze, aber sehr klare und korrekte
populäre Darstellung der modernen Logik, in die er die damals neuesten
Ergebnisse – wie die Sätze von Gödel und Tarski – einbezieht.
Bezüglich der philosophischen Beziehungen und der Anwendungen der
modernen Logik beruft er sich vor allem auf Carnap. Im Buch nehmen
aber die Erörterungen bezüglich des metaphysischen Status der logischen
Gegenstände viel mehr Platz ein, und in diesen Kapiteln kann der Le-
ser auch die Sympathie des Verfassers für einen phänomenologisierenden
Platonismus spüren. Ein anderes wiederkehrendes Interesse Pozsonyis gilt
der Frage der Möglichkeit von mehreren Logiken. Im Jahre 1943 publiziert
er im Athenaeum einen Aufsatz mit dem Titel Relativismus und nicht-
”
aristotelische Systeme in der modernen Logik“.19 Als Ausgangspunkt wi-
derlegt er die Ablehnung des logischen Relativismus mit der Berufung
auf die eine und einzige objektive Logik. Sein Gegenargument ist, dass
wir nicht a priori sicher sind, die Gesetze dieser objektiven Logik mit
unserem Verstand auffassen und beim Aufbau unserer wissenschaftlichen
Theorien anwenden zu können. Er behauptet, dass wir
. . . bei der Untersuchung der Frage von der Untersuchung der Logik der wissen-
schaftlichen Theorien ausgehen sollen. Die moderne Logik konnte eben deshalb
so große Ergebnisse innerhalb einiger Jahrzehnten erreichen, weil sie mit den sich
ausschließlich auf der prinzipiellen Ebene bewegenden philosophischen Analy-
sen der klassischen Logik . . . brach und für sich Mittel schuf, durch welche sie
konkrete, für die theoretischen Wissenschaften wichtige Probleme zu lösen fähig
wurde.
Der einzige zulässige Relativismus der Logik ist jener, welcher uns von den
Paradoxien der Mengenlehre und den Sätzen von Gödel und Tarski auf-
gezwungen wird, nämlich dass wir unsere Logik nur in einer unendlichen
Reihe von Sprachen und Theorien und nicht innerhalb einen einzigen Sys-
18 Ung., Budapest: Ausgabe des Verfassers, 1942.

19 Athenaeum 29 (1943), S. 15-33.
tems darlegen können. Er lehnt aber den Standpunkt von Antal Schütz20
ab, nach welcher die paradoxen Aussagen nicht von der objektiven Logik
handeln und daher etwas Objektives nicht zu fassen vermögen. Pozsonyi
bringt erneut seine frühere Argumentation vor; den Mathematiker beru-
higt diese Ansicht gar nicht, denn er hat keine Kriterien, um zu entschei-
den, welche Definitionen etwas Objektives fassen und welche nicht. Als
Konklusion behandelt er bejahend das Toleranzprinzip von Carnap.
Die akademische Antrittsrede von Antal Schütz trägt den Titel Lo-
”
giken und Logik“.21 Schütz argumentiert hier für den Monologismus“,
”
nämlich für die Ansicht, dass es nur eine einzige wahre Logik gibt, die
klassische, aristotelische, und widerlegt die polilogistischen“Ansichten.
”
Er bringt viele und meistens korrekte mathematische Beispiele zur Un-
terstützung seiner Meinung, beruft sich oft auf Hilbert, zitiert Russell,
L
ukasiewicz, Reichenbach, Schlick, den Letzteren sogar eindeutig zustim-
mend. Die ganze Schrift aber, die Autoren, welche bezüglich der Grund-
frage am meisten und umfangreichsten zitiert werden, die Auffassung des
Verfassers von einer logischen Theorie, all dies erweckt den Eindruck, dass
es keine symbolische Logik gebe. Schütz ist gar nicht uninformiert, be-
trachtet aber alle Entwicklungen seit Frege als unwesentlich für sein The-
ma. Logik ist für ihn eben nur die Fülle von sich ausschließlich auf der
”
prinzipiellen Ebene bewegenden philosophischen Analysen“, wie Pozsonyi
es apostrophiert hat.
Im Jahre 1938 publizierte der junge Kulturphilosoph der geistesge-
schichtlichen Richtung László Mátrai ein kleines populärwissenschaftliches
Buch mit dem Titel Modernes Denken“.22 Er gibt einen Überblick der
”
modernen Theorien vom Menschen, Geschichte, Kultur, physischen und
biologischen Welt. Er verteidigt die Idee des Fortschritts und kritisiert die
irrationalistischen, kulturpessimistischen Denker; er gibt auch eine korrek-
te Darstellung einiger Grundgedanken der Relativitätstheorie. Im Kapitel
über die moderne Logik“spricht er von der gefährlichen Entwicklung,
”
dass die Kluft zwischen Geistes- und Naturwissenschaften immer weiter
20 Antal Schütz (1880–1953), Professor der Dogmatik an der Theologischen

Fakultät der Péter Pázmány Universität Budapest, wohl der bedeutendste
Philosoph des ungarischen Katholizismus jener Zeit. Über seine logische
Ansichten siehe unten.
21 Ung., Budapest: Magyar Tudományos Akadémia 1941.
22 Ung., Budapest: Magyar Szemle Társaság 1938.
wird und behauptet, dass das logische Denken der gemeinsamen Grund
sein sollte, auf dem die beiden sich wieder treffen könnten. Unter mo-
derner Logik ist aber bei Mátrai die Logik von Pauler und Husserl zu
verstehen, die Logistik ist nach ihm nur eine Abzweigung. In der logi-
schen Formelsprache von Russell sieht er nur eine Erneuerung der Utopie
der Universalsprache. Über Carnap schreibt er:
anfänglich hat er sich damit begnügt, logistischen Untersuchungen dort zu fol-

gen, wo es in der Tat nötig ist (d.h. im Rahmen der Axiome . . . der Mathematik
und Geometrie) . . . die Kurzsichtigkeit und der Selbstwiderspruch, mit welchen
die Anhänger Carnaps die Einheit der Wissenschaften dadurch verwirklichen
wollen, dass sie ihre eigene Wissenschaft jedem Anderen beibringen, streifen
schon an das Komische.23
Ein Kommentar erübrigt sich; man muss aber bemerken, dass die An-
sichten des größten Anhängers der modernen Logik, Pozsonyi, und des
schärfsten Gegners, Mátrai, wichtige gemeinsame Züge haben. Erstens
lehnen beide die Zuständigkeit der symbolischen Logik auf dem Gebiet der
Human- und Sozialwissenschaften ab. Zweitens will Pozsonyi ebensowenig
wie Mátrai die Logik, wie sie im 20. Jh. besteht, also die moderne Logik,
mit der symbolischen Logik und ihrer Metatheorie identifizieren. Drittens
erkennt Mátrai die Leistungen der symbolischen Logik oder Logistik auf
dem Gebiet der Mathematik ebenso an wie Pozsonyi. Die Unterschiede
zwischen den beiden Autoren liegen vor allem in der Beurteilung der Be-
deutung dieser Leistungen für die Philosophie, und dabei zieht Pozsonyi
die Grenzen der relevanten Anwendung der symbolischen Logik freilich
wesentlich weiter als Mátrai.
Der Standpunkt der marxistisch-leninistischen dialektischen Logiker
ist eigentlich nicht so weit entfernt von dem des Geisteshistorikers Mátrai.
Der Initiator der dialektischen Logik in Ungarn war Béla Fogarasi24 – ein
Mitglied des Galileo-Kreises und im Jahr 1919 kompetenter Übersetzer
von Russells Problems of Philosophy. Von seiner sowjetischen Emigrati-
on kehrte er aber als dogmatischer Marxist-Leninist zurück und wurde
dann zum Machthaber des philosophischen Lebens. Er kritisiert in sei-
nem Logik-Lehrbuch Carnaps logische Semantik, weil sie . . . das Denken
”
23 Op. cit., S. 69.

24 Béla Fogarasi (1891–1958), in den fünfziger Jahren Professor der Lenin-
Institute in Budapest
mit der Sprache identifiziert, die Logik mit der Grammatik bzw. mit ei-
nem Zweig der Grammatik, der Syntax.“25 Über die symbolische Logik
im Allgemeinen sagt er:
. . . die Unzulänglichkeiten der Sprache kann man nicht auf eine grobe mecha-
”
nistische Weise dadurch eliminieren, dass wir die allgemeine Sprache durch eine
Symbolik substituieren. . . . Die symbolische Logik ist, indem sie nicht auf das
mathematische Denken bezogen ist, wo sie ihre Berechtigung hat, nichts ande-
res als eine Abkürzung der Ausdrücke der allgemeinen Sprache . . . Es ist ein
schwerer und gefährlicher Irrtum zu glauben, dass die symbolische Logik die
Sprache substituieren kann.“26
Man kann in diesen Worten die üblichen Argumente der Gegner der philo-
sophischen Anwendung der symbolischen Logik wiedererkennen. Der Un-
terschied zur Situation in den dreißiger Jahren ist, dass diese Ansichten
in den fünfziger Jahren als durch die Staatsmacht vorgeschriebene Mei-
nung galten. In den sechziger Jahren ändert sich aber die Lage allmählich.
Zwar setzt Fogarasis wesentlich weniger gebildeter Schüler László Erdei
den Kampf gegen den Einbruch der mathematischen Logik in die Philoso-
phie fort und hat auch mächtige Befürworter – unter anderen den schon
marxistisch-leninistischen Mátrai –, sein Standpunkt wird aber nicht mehr
die verordnetee, in der Öffentlichkeit allein mögliche Ansicht. Es kommt
zu zwei Entwicklungen im philosophischen Leben, die sich bis heute ent-
scheidend ausgewirkt haben.
Erstens gewinnt im akademischen Philosophieunterricht neben dem
offiziellen Sowjetmarxismus der Neomarxismus des Kreises von Georg
Lukács gewissen Raum. Das dauert zwar nicht lange; 1973 werden die
politisch unbotmäßigen Lukács-Schüler aus allen akademischen Positio-
nen entlassen und die meisten zur Emigration gezwungen. Diese kurze
Blütezeit hatte aber weitgehende Konsequenzen für die ungarische Phi-
losophie. Einerseits war es nicht einmal nach der Entlassung und Berufs-
verbot vieler Mitglieder des Lukács-Kreises möglich, die monolithische
Einheit des philosophischen Lebens und des Philosophieunterrichts wie-
der herzustellen, andererseits brachten die Lukács-Schüler ein neues Ni-
veau der Kenntnis und des Verständnisses des zeitgenössischen philoso-
phischen Denkens. Von den Mitgliedern des Lukács-Kreises ist für uns die
25 Logik (ung.), Budapest: Akadémiai 1953, S. 80.

26 Ebd., S. 85. Heraushebungen im Original.
Lehrtätigkeit von György Márkus27 von besonderer Bedeutung. Márkus

war nicht Schüler von Lukács im wortwörtlichen Sinn, hatte aber sich
nach seiner Rückkehr aus der Sowjetunion dessen Kreis angeschlossen.
Er hatte in Moskau Philosophie und bei dem späteren Dissidenten Sino-
wjew auch Logik sowie moderne Wissenschaftsphilosophie studiert. Wie
er später sagte, waren dort diese Zweige der Philosophie die brauchbars-
ten und interessantesten. Es ist nicht Thema dieses Artikels, die Tätigkeit
von Márkus und seiner Schüler zu untersuchen; ich möchte aber mindes-
tens zwei Mitglieder dieses Kreises nennen, deren Tätigkeit eine unmit-
telbare Beziehung zur Philosophie des Wiener Kreises und auch einen
gewissen Berührungspunkt mit der modernen Logik hatte: Márta Fehér
und Ferenc Altrichter.28 Sie veröffentlichten den dicken Band Die Philo-
sophie des Wiener Kreises,29 mit Übersetzungen von Schriften von Car-
nap, Schlick, Reichenbach und Neurath (Altrichter versah ihn mit einem
umfangreichen Vorwort), der vielleicht bis heute die wichtigste ungarische
Sekundärliteratur zum Wiener Kreis darstellt. In seinen anderen Schriften
hat Altrichter die Methode der logischen Rekonstruktion in der Geschichte
der Philosophie erfolgreich angewendet: Seine Analyse von Argumentatio-
nen, manchmal mit den Mitteln der modernen Logik, war ein Vorbild für
mehrere Philosophen der nächsten Generation, die aber auch einen ande-
ren Meister hatten, dessen Tätigkeit ausführlicher behandelt werden soll:
Imre Ruzsa. Sein Auftreten in der Philosophie ist also die zweite wichtige
Entwicklung der sechziger Jähre.
Imre Ruzsa30 absolvierte seine mathematischen Studien im Jahre 1956
als Schüler von Rózsa Péter und wurde ihr Assistent im Jahre 1960. Er
war also Mitarbeiter eines mathematischen Lehrstuhls, begann aber seine
Publikationstätigkeit mit Themen der philosophischen Logik – deontische
Logik und andere modale Systeme. Seine Hauptidee in der Logik war
27 György/George Márkus (1934–), heute Professor emeritus der Universi-

ty of Sidney. Wichtigste Schrift zum Thema: Richtungen in der heutigen
bürgerlichen Philosophie (mit Zádor Tordai, ung.), 2., verb. Aufl., Buda-
pest: Gondolat 1972.
28 Márta Fehér (1943–), Professor der Technischen Universität Budapest; Fe-
renc Altrichter (1943–), Professor der University of North Carolina.
29 Budapest: Gondolat 1972.
30 Imre Ruzsa (1921–2008), Professor, Gründer des heutigen Lehrstuhls für
Logik der Philosophischen Institut der Eötvös Universität Budapest.
die Weiterentwicklung von Arthur N. Priors Konzeption der Wahrheits-

wertlücken, nach welcher wir neben wahren und falschen auch wahrheits-
wertlose Aussagen in der Logik erlauben sollten. Ruzsa arbeitete während
seiner ganzen wissenschaftlichen Laufbahn an der Verallgemeinerung die-
ser Idee auf quantifizierte und typentheoretische modale und intensiona-
le Systeme.31 Anfang der sechziger Jahre übernahm er den Mathema-
tikunterricht für Philosophie-Studenten. 1965 wurde er Mitarbeiter des
Philosophischen Instituts der Akademie und begann für die ungarische
philosophische Öffentlichkeit zu publizieren. Seine Schriften über die Phi-
losophie der Mathematik32 haben einen beträchtlichen Einfluss auf die
jüngere Generation ausgeübt, obwohl Ruzsa sich selbst für keinen Spe-
zialisten auf diesem Gebiet hielt; er wollte nur eine vermittelnde Rolle
übernehmen. Demzufolge finden wir in diesen Büchern eher Darlegungen
von philosophischen Meinungen als Stellungnahmen, abgesehen von der
damals verbindlichen materialistischen Kritik der zeitgenössischen idea-
”
listischen“ Tendenzen (was auf das unvermeidliche Minimum reduziert
ist). Was man dennoch von der persönlichen Meinung des Verfassers er-
kennen kann, ist eine gewisse Sympathie mit der Hilbertschen Auffassung,
welche er von dem Formalismus im engeren Sinne sorgfältig unterscheidet.
Im Jahre 1971 wurde an der Eötvös Universität Budapest ein Lehr-
stuhl für Logik gegründet, unter der Leitung des oben erwähnten dialekti-
schen Logikers László Erdei, und unter dem Prinzip dass die wahre Logik
der marxistisch-leninistischen Philosophie die dialektische Logik sei, die
formale Logik aber nur eine untergeordnete Propädeutik dazu – wonach
auch der Logikunterricht organisiert wurde. Ruzsa wurde eingeladen, als
diesem Lehrstuhl zugeordneter Dozent formale Logik zu unterrichten, und
er nahm die Einladung an. Ruzsa ertrug mit schweigender Ironie Erdeis
oft betonte Ansichten, wonach die mathematische Logik, diese neoposi-
tivistische Erfindung, als formale Logik von Haus aus primitiver als die
dialektische sei. Seine Lehrtätigkeit bedeutete eine Wende von nicht zu
31 Monographien: Modal Logic with Descriptions, Amsterdam: Martinus Nij-

hoff 1982; Klassische, modale und intensionale Logik (ung.), Budapest:
Akadémiai Kiadó 1984; Logische Syntax und Semantik (ung.), Budapest:
Akadémiai Kiadó I. Bd. 1988, II. Bd. 1989).
32 Zu einigen philosophischen Problemen der Mathematik (ung.), Budapest:
Tankönyvkiadó, 1969; Die Begriffswelt der Mathematik (ung.), Budapest:
Gondolat 1973. Deutsch: Berlin: Volk und Wissen 1976).
überschätzender Bedeutung für den Logikunterricht an der Fakultät der

Humanwissenschaften. Neben dem Unterricht für Philosophie-Studenten
hielt Ruzsa Vorlesungen und Seminare in Logik und Mathematik für Stu-
denten der allgemeinen Linguistik. Diese Tätigkeit fiel glücklich zusammen
mit seinem neuen Interesse an der logischen Analyse natürlicher Sprachen.
Der neueste Fortschritt auf diesem Gebiet war in den siebziger Jahren die
auf die Kripke-Semantik gegründete intensionale Logik von Richard Mon-
tague. Ruzsa hat eine neue Fassung der intensionalen Logik ausgearbeitet,
in welcher nicht jeder (extensionale) Ausdruck eine Extension hat (d.h.
die Idee der Wahrheitswertlücken wurde verallgemeinert). Ruzsas Kompe-
tenz in den neuesten Methoden der formalen Semantik und seine eigenen
Ergebnisse sicherten ihm Ansehen und Popularität unter der neueren Ge-
neration der Linguisten; seine ersten persönlichen Schüler waren eigentlich
nicht Philosophen, sondern Linguisten. In der zweiten Hälfte der siebziger
Jahre bildete sich um Ruzsa ein Kreis von jungen Philosophen, Linguis-
ten und mathematischen Logikern. 1982 wurde es möglich, die Zwangsehe
mit der dialektischen Logik aufzulösen und eine neue Einheit unter den
Namen Fachgruppe (später Lehrstuhl) für symbolische Logik und Wis-
”
senschaftsmethodologie“ zu gründen. Diese Fachgruppe gab von 1984 an
ein Jahrbuch mit dem Titel Tertium non datur heraus, in welchem die
logischen und auf die moderne Logik gestützten philosophischen und lin-
guistischen Schriften der Mitglieder und zahlreichen Sympathisanten des
Ruzsa-Kreises erscheinen konnten.
Zum Zentenarium von Freges Begriffsschrift veröffentlichte Ruzsa mit
seinen Mitarbeitern einen Sammelband von Freges Schriften33 und meh-
rere Artikel über Frege und die moderne Logik. Diese Betonung der Rolle
Freges war gegen die oft vorgebrachte Anklage, nach welcher die moder-
ne Logik in der Philosophie ein Danaergeschenk des Neopositivismus sei,
gerichtet, sie zeigt aber auch den Standpunkt von Ruzsa in der Philo-
sophie der Logik. Er teilte zwar nicht den radikalen Platonismus Freges,
doch immerhin sympathisierte er viel mehr mit Frege als mit Carnap.
Er hatte sich auf eine realistische Auffassung der Logik festgelegt, nach
der die Grundsätze und Theoreme der Logik inhaltliche Wahrheiten und
keineswegs bloße Konventionen sind. Er betonte die Priorität der (nicht
33 G. Frege: Logika, szemantika, matematika (redigiert und kommentiert von

Imre Ruzsa, übersetzt von András Máté), Budapest: Gondolat, 1980.
formalen) semantischen Überlegungen bei der Begründung eines logischen

Systems.
In seinem am wenigsten formalen und am meisten philosophischen
Buch Klassische, modale und intensionale Logik 34 gehört Carnap – neben
Frege, Kripke und Quine – zu den am häufigsten zitierten Autoren. Ruzsa
erörtert ausführlich die L-Semantik von Carnaps Introduction to Seman-
tics. Er analysiert und würdigt die Ergebnisse von Carnap in der Klärung
des Verhältnisses zwischen Quantifikationstheorie und Modallogik, stellt
jedoch auch die Grenzen von Carnaps Theorie dar und führt diese teilwei-
se auf philosophisch problematische Ansichten zurück. Er bezeichnet Car-
naps L-Semantik als Quasi-Semantik, weil sie keine Denotations-Semantik
(mengentheoretische Semantik) ist, sondern sich die semantische Defini-
tion statt auf die Denotationen der Ausdrücke auf Mengen von anderen
Ausdrücken der Sprache bezieht. Er behauptet, dass die modalen Para-
doxien der Identität in Carnaps Theorie gar nicht gelöst, sondern durch
die Begrenzung der erlaubten (syntaktisch wohlgeformten) Deskriptionen
ausgeschaltet sind.35 Der Charakter dieser kritischen Bemerkungen ist je-
denfalls grundverschieden von dem der meisten früheren Verfasser. Zwar
sind Ruzsas Bemerkungen teilweise durch seine verschiedenen philosophi-
schen Einstellungen motiviert, bewegen sich aber gar nicht ausschließlich
”
auf der prinzipiellen Ebene“, sondern berühren vor allem den logischen
Inhalt von Carnaps Lehre. Die Kritik beabsichtigt vor allem, die durch
Carnap (und andere Verfasser) ungelösten Probleme aufzuzeigen, welche
Ruzsas eigene Theorie lösen will. Sie gehört also zur üblichen philosophi-
schen Diskussion.
Durch Ruzsas Tätigkeit ist also die moderne Logik normales Fachge-
biet der ungarischen Philosophie geworden. Sie wird vor allem auf dem
von ihm gegründeten Lehrstuhl – welcher nach der stillen Aufhebung der
dialektischen Logik“ den adjektivfreien Namen Lehrstuhl für Logik“
” ”
übernahm –, aber auch in manchen anderen philosophischen Institutio-
nen überwiegend von Ruzsas Schülern kultiviert. Was bezüglich des Ein-
flusses des Wiener Kreises heute von besonderer Bedeutung ist, ist die
wissenschaftsmethodologische Anwendung der Logik. Mathematiker und
Physiker, die auch einigermaßen am Ruzsa-Kreis anknüpfen, haben in den
34 Siehe Fn. 31.

35 Klassische, modale und intensionale Logik, S. 208-213.
letzten Jahren in der logischen Formalisierung und Untersuchung physika-

lischer Theorien wertvolle Ergebnisse erreicht. Darüber werden aber einige
Artikel dieses Bandes mehr sagen.
Zusammenfassend muss man sagen, dass die Rezeption der moder-
nen Logik in der ungarischen Philosophie kaum (genauer gesagt: nur bei
Frigyes Pozsonyi) in positivem Sinn mit der Rezeption der Ideen des Wie-
ner Kreises verknüpft war. Vielmehr war die Philosophie des Wiener Krei-
ses oft ein Grund (manchmal vielleicht nur ein Scheingrund) für die Geg-
ner, die neue Logik abzulehnen. Die aus der Mathematik gekommenen
moderne Logiker waren Anhänger anderer Philosophien der Mathema-
tik und Logik; Imre Ruzsa konnte aber trotz aller Meinungsunterschiede
Carnaps Gedanken ihrer Wichtigkeit gemäß verhandeln und anerkennen.
Diese Einstellung ist auch maßgebend für die heutigen Logiker: Die Ideen
der Denker des Wiener Kreises bleiben wichtige Elemente der philosophi-
schen Tradition, auf welche man besonders in der Philosophie der Logik
und in der Wissenschaftsmethodologie Bezug nehmen soll.
H. Andréka, J. X. Madarász, I. Németi, P. Németi and G. Székely
Vienna Circle and Logical Analysis

of Relativity Theory
Introduction
In this paper we will present some of our school’s results in the area of
building up relativity theory (RT) as a hierarchy of theories in the sense
of logic. We use plain first-order logic (FOL) as in the foundation of
mathematics (FOM) and we build on experience gained in FOM. The
main aims of our school are the following: We want to base the theory on
simple, unambiguous axioms with clear meanings. It should be absolutely
understandable for any reader what the axioms say and the reader can
decide about each axiom whether he likes it. The theory should be built
up from these axioms in a straightforward, logical manner. We want to
provide an analysis of the logical structure of the theory. We investigate
which axioms are needed for which predictions of RT. We want to make
RT more transparent logically, easier to understand, easier to change,
modular, and easier to teach. We want to obtain a deeper understanding
of RT.
Our work can be considered as a case study showing that the Vienna
Circle’s (VC) approach to doing science is workable and fruitful when
using the insights and tools of mathematical logic acquired since its for-
mative years at the very time of the VC activity. We think that logical
positivism was based on the insight and anticipation of what mathemati-
cal logic is capable of when elaborated to some depth. Logical positivism,
in great part represented by VC, influenced and took part in the birth of
modern mathematical logic. The members of VC were brave forerunners
and pioneers.
Let’s see what was available before or during the period the VC was
active, and what was not available for the members of VC but available for
us now. The VC activities in the strict sense fall into the period of 1922–
1936. The beginning of intensive development of FOL coincides with this
248 H. Andréka et al.
period. The elements of first-order language FOL (propositional logic,

quantifiers) were available based on the works of Boole, Peirce, Schröder,
Frege, Russell (roughly 1860–1910). However, the completeness and in-
completeness theorems, compactness theorem, semantics for FOL, model
theory, proof theory and definability theory were not there before VC.
Many of these became available as works of people influenced by or work-
ing in VC. Here in brief chronological order some of the turning-points
in this development: 1929 Gödel’s completeness theorem, 1930 Tarski’s
decision method for the elementary theory of reals, 1931 Gödel’s incom-
pleteness theorem, 1933 Tarski mathematical definition of truth, 1936
Tarski concept of semantical consequence relation, definition of model
theoretic semantics. Proof theory developed only later than around 1940,
and model theory developed only later than around 1950. Beth’s defin-
ability theorem was anticipated by Hans Reichenbach 1924 (motivated
by RT), Tarski 1936, but became available only in 1952. The authors of
the standard Model Theory book, Chang and Keisler, were both students
of Tarski. Two chapters of model theory, the theory of semantics and
the theory of definability are of great importance for our work. So what
was/is FOL used for? Some dates marking the emergence of paradoxes
in mathematics necessitating an axiomatic approach, experiments with
formal languages to describe, found and unify mathematics and then all
science are as follows: 1897 Burali-Forti Paradox, 1900 Hilbert’s Program,
1901 Russell’s Paradox, 1908 Zermelo’s Axiom system for Set Theory,
1910 Russell-Whitehead Principia Mathematica, 1918 Löwenhein-Skolem
“Paradox,” 1922 Fraenkel’s addition to Zermelo-Fraenkel Set theory, 1924
Tarski-Banach “Paradox,” 1926 Tarski’s axiomatization of Euclidean ge-
ometry in FOL, 1930 the addition of the Axiom of Regularity to Set
Theory, beginning with 1935 the Bourbaki group’s work formalizing and
uniting mathematics, 1937 Tarski role of logic in scientific studies. The
success story of FOM. Gödel-Bernays and von Neumann’s Set Theory. We
will see in this paper many of the above results used in our work. Gödel,
Tarski, Reichenbach, Hilbert, Russell, and Einstein were all connected to
VC in some way or other.
Why relativity theory? Logical positivism is a philosophy which holds
that the only authentic knowledge is that based on observation, experi-
ence, experiment through formal logic. Einstein’s relativity theory trans-
forms space and time from being a priori Euclidean and absolute things to
Vienna Circle and Logical Analysis of Relativity Theory 249
something which emerges from experience and experiments, thus claiming

the subjects of space and time from the realm of metaphysics for science.
Einstein’s thought-experiments served to bring logic into the picture. For-
malizing relativity theory in logical language was a primary interest in
logical positivism, see Reichenbach’s works. Relativity theory also led to
Modern High-Precision Cosmology as a branch of hard-core physics (and
not part of metaphysics). Since space-time is the arena in which most
processes studied by modern science unfold, a logic based foundation for
RT (the theory of space-time) might be a natural starting point for a
foundation and unification of all science (a VC goal (8), (29)).
Our group investigates a hierarchy of relativity theories, weaker and
stronger theories. We not only propose axiom systems and prove their
completeness with respect to their intended models, we also derive RT’s
main predictions, asking ourselves which axioms play the key role in their
derivations. We make “reverse relativity” in analogy with “reverse math-
ematics,” and we analyze the theories in many ways. In this paper we
present three of our main axiom systems (i.e., theories), just stating some
of their most important properties. These three theories have the same
language. We begin by introducing this language.
The Common Language of the Theories Presented here

We will use FOL. FOL can be viewed as a fragment of natural language
with unambiguous syntax and semantics. One of the acknowledged bene-
fits of using FOL is that it helps eliminate tacit assumptions, one of VC’s
maxim. The most important decision in writing up an axiom system in
FOL is to choose the vocabulary, or primitive symbols of our language,
i.e., what objects and what relations between them will belong to the
language we will use.
We want to talk about space and time as relativity theory conceives
them. We will talk about space-time as experienced through motion. We
represent motion as changing spatial location in time. We will call the
entities that do the motion “test-particles.” Sometimes, to use a shorter
word, we will call them “bodies” but in reality they can be anything that
moves, e.g., they can be coordinate systems or electromagnetic waves, or
light signals or centers of mass.1 To talk about spatial locations and time
we will use quantities arranged in a (space-time) coordinate system, and
we will have a basic relation, the so-called world-view relation, which tells
us which test-particles are present in which locations at which instants.
We will think of the quantities as the real numbers (i.e., the number-
line), so we will use a “less than” relation and two operations, addition
and multiplication, on them. In this paper to axiomatize special relativity
theory, we will use two more primitive notions, namely that of “inertial
test-particles” and “light-signals” which we will simply call photons.2
To illustrate what we said so far, let us consider the following two-
sorted first-order language:
{ B , IB, Ph, Q , +, ·, <, W },
where B (test-particles or bodies) and Q (quantities) are the two sorts,

IB (inertial bodies) and Ph (light signals or photons) are unary relation
symbols of sort B , · and + are binary function symbols and < is a binary
relation symbol of sort Q , and W (the worldview relation) is a 6-ary
relation symbol of sort BBQQQQ . B and Q can be thought of as the
physical and as the mathematical universes.
Atomic formulas IB(c) and Ph(p) are translated as “c is an inertial
body,” and “p is a photon,” respectively. We use the world-view rela-
tion W to speak about coordinatization by translating W(o, b, x, y, z, t) as
“observer o coordinatizes body b at space-time location x, y, z, t,” (i.e.,
at space location x, y, z and at instant t). We sometimes use the more
intuitive expressions “sees” or “observes” for coordinatation. We will use
the letters, and their variants, o, b, p, m, k for variables of sort B , and the
1 Regarding extended bodies: We concentrate on test-particles and see test-

particles as spatially point-like, i.e., of size zero. As far as we are aware of it,
this idealization is harmless from the point of view of the goals of relativity
theory. If we want to treat an extended body in our theory (as we do in
the theory AccRel of accelerated observers), we represent it as a “cloud” of
test-particles. This is consistent with the spirit of standard physical world
view of regarding extended bodies as clouds of elementary particles.
2 To talk about light-signals is not necessary for building up SR. One simple
way of avoiding them is defining light-signal as anything that moves with
“speed of light.” There are deeper ways of avoiding the use of light-signals
in building up relativity theory, see, e.g., (5, sec.5).
letters x, y, z, t and their variants for variables of sort Q . For easier read-
ability, we will use x̄, ȳ for sequences of four variables x1 , x2 , x3 , x4 and
y1 , y2 , y3 , y4 .
We have not introduced the concept of observers as a basic one because
it can be defined as follows: an observer is nothing else than a body who
“observes” (coordinatizes) some other bodies somewhere, this property
can be captured by the following first-order formula of our language:
def
Ob(o) ⇐⇒ ∃bx̄ W(o, b, x̄);
and inertial observers can be defined as inertial bodies which are ob-
servers, formally:
def
IOb(o) ⇐⇒ IB(o) ∧ Ob(o).
To abbreviate formulas of FOL we often omit parentheses according
to the following convention. Quantifiers bind as long as they can, and
∧ binds stronger than →. For example, we write ∀x ϕ ∧ ψ → ∃y δ ∧ η

instead of ∀x (ϕ ∧ ψ) → ∃y(δ ∧ η) .
Axioms of Special Relativity

Having specified the language, let us turn to the axioms of our first theory.
This will be an axiom system for Special Relativity theory (SR).
AxField: The quantity part Q ; +, ·, < is an ordered field.
For the FOL definition of a linearly ordered field see, e.g., (9, p.41); this
is a formulation of some of the most basic properties of addition and
multiplication of real numbers. One of these properties is that there is a
unique neutral element for addition (∃z∀x z + x = x), we call this element
z zero and we denote it with 0.
The next axiom simply states that each inertial observer assumes that
it rests at the origin of the space part of its coordinate system. It also
can be thought of as expressing that we identify a coordinate system (or
reference frame) with a test-particle “sitting” at the origin.
AxSelf: Any inertial observer coordinatizes (observes) itself as “living on

the time-axis,” i.e., it coordinatizes itself at a coordinate point if
and only if the space component of this point is the origin:

∀oxyzt IOb(o) → W(o, o, x, y, z, t) ↔ x = y = z = 0 .
Our next axiom is on the constancy of the speed of light. For convenience,
we choose 1 for this speed. This choice physically means using units of
distance compatible with units of time, such as light-year, light-second,
etc.
AxPh: The speed of light signals is 1 and it is possible to “send out” a

photon in any direction, according to any inertial observer:

∀ox̄x̄ IOb(o) → ∃p(Ph(p) ∧ W(o, p, x̄) ∧ W(o, p, x̄ ))

↔ (x1 − x1 )2 + (x2 − x2 )2 + (x3 − x3 )2 = (x4 − x4 )2 .
This is the most important axiom of SR, it is its “physical” axiom. Ax-
iom AxPh is very well confirmed by experiments, such as the Michaelson–
Morley experiment and its variants. The next axiom establishes connec-
tions between different coordinate systems. It expresses the idea that all
observers “observe” the same outside reality.
AxEv: All inertial observers coordinatize the same “meetings of bodies:”
∀oo x̄ IOb(o) ∧ IOb(o ) → ∃x̄ ∀b W(o, b, x̄) ↔ W(o , b, x̄ ).
We call “meetings of bodies” events. By our next axiom, we assume

that inertial observers use the same units of measurement. This is only a
“simplifying” axiom.
AxSymd: Inertial observers agree as to the spatial distance between events

if these events are simultaneous for both of them, formally:
∀oo x̄x̄ ȳ ȳ IOb(o) ∧ IOb(o ) ∧ x4 = y4 ∧ x4 = y4 ∧

∀b (W(o, b, x̄) ↔ W(o , b, x̄ )) ∧ ∀b (W(o, b, ȳ) ↔ W(o , b, ȳ ))
→ (x1 −y1 )2 +(x2 −y2 )2 +(x3 −y3 )2 = (x1 −x2 )2 +(y1 −y2 )2 +(z1 −z2 )2 .
Let us now introduce our axiom system of SR as the set of the axioms
above:
SpecRel = { AxField, AxSelf, AxPh, AxEv, AxSymd } .
The reader is invited to check that all the axioms of SpecRel are sim-
ple, comprehensible and observationally oriented. In setting up an axiom
system, we want the axioms be streamlined, economical, transparent and
few in number. On the other hand, we want to have all the surprising,
shocking, paradoxical predictions of RT as theorems (and not as axioms).
We want the price-value ratio to be good, where the axioms are on the
“cost”-side, and the theorems are on the “gain”-side.
Let us see what theorems we can prove from SpecRel. We will see that
we can prove everything from our five axioms that “usual” SR can, but
let us proceed more slowly. In the axioms we did not require explicitly,
but it can be proved from SpecRel with the rigorous methods of FOL
that inertial observers see each other move on a straight line, uniformly
(covering the same amount of distance in the same amount of time). For a
“fancy theorem” from “plain axioms,” let us prove from SpecRel that “no
inertial observer can move faster than light.” Below, denotes derivability
in one of FOL’s standard proof systems.
Theorem 12. (NoFTL) In an inertial observer m’s world view, any in-
ertial observer k moves slower than any light-signal p, i.e., if both k and
p move from spatial locations x1 , x2 , x3 to y1 , y2 , y3 , then for the ob-
server k this trip took more time than for the photon p. Formally:
SpecRel ∀mkpx̄ȳt IOb(m) ∧ IOb(k) ∧ Ph(p)∧

W(m, k, x̄) ∧ W(m, p, x̄) ∧ W(m, k, ȳ) ∧ W(m, p, y1 , y2 , y3 , t) → y4 > t.
For proof see, e.g., (22, Thm.3.2.13). What the average layperson usu-
ally knows about the predictions of relativity is that “moving clocks slow
down, moving spaceships shrink, and moving clocks get out of synchro-
nism, i.e., the clock in the nose of a fast moving spaceship is late (shows
less time) when compared with the clock in the rear.” See Figure 9. Let’s
call these three predictions the “paradigmatic effects” of SR. Now, SpecRel
implies all the paradigmatic effects quantitatively, too.3 From this it fol-
lows that the so-called worldview transformations are Poincaré-functions,
thus everything follows from our SpecRel what follows from “usual” special
relativity theory.
Different observers may observe different spatial distance between the
same two events. This is so in Newtonian Kinematics (NK), too. (For
example, if I ate a sandwich and later drank a coffee on a train, these
two events were at the same place according to me, but according to a
coordinate system attached to Earth I ate the sandwich in Budapest and
drank the coffee in Vienna.) However, in NK the time-difference between
two events is the same for all observers, it is “absolute”. According to the
paradigmatic effects, in RT even the time-difference between two events
depends on the state of motion of the observer! (The observer moving
relative to m will observe less time passed between e and e because his
clock “slowed down”.) In this respect, space and time in RT are “more
alike” than in NK. Our next theorem states that a certain combination of
spatial distance and time-difference is “absolute” in RT, too. The proof
of this theorem can be found in, e.g., (4, p.650). Let us define
μ(x̄, ȳ) := (x1 − y1 )2 + (x2 − y2 )2 + (x3 − y3 )2 − (x4 − y4 )2 .
Thus μ(x̄, ȳ) is squared spatial distance minus squared time-difference

between events e and e if these events took place at x̄ and ȳ, respectively.
This quantity is called (squared) relativistic distance between events e
and e (or squared Minkowski-distance between space-time locations x̄
and ȳ). According to the next theorem, relativistic distance is “absolute”
in RT. In RT relativistic distance plays the same role as absolute time
in NK. Minkowski geometry is based on relativistic distance (in place of
Euclidean distance).
Theorem 13. SpecRel ∀oo x̄x̄ ȳ ȳ IOb(o) ∧ IOb(o ) ∧

∀b (W(o, b, x̄) ↔ W(o , b, x̄ )) ∧ ∀b (W(o, b, ȳ) ↔ W(o , b, ȳ ))
→ μ(x̄, ȳ) = μ(x̄ , ȳ ).
3 They follow from our next theorem. However, in our works we usually
prefer proving the paradigmatic effects one-by-one, directly from the ax-
ioms of SpecRel because this illuminates or illustrates how we perform our
conceptual analysis. These proofs can be found, e.g., in (4, sec.2.4).
1 km = 1 light-second
my spaceship
is 1km long (in this picture)
√ √
0 −v 1 − v2 −v + 1 − v2
√
it’s only 1 − v 2
km long
0
m
1
11111111111111111111111111111111111111
00000000000000000000000000000000000000
00000000000000000000000000000000000000
11111111111111111111111111111111111111
now (m) 1 second later (m)
√
Figure 9: According to m, the length of the spaceship is 1 − v 2 km, it is
1 km wide and tall, and the clocks in the nose show v less time than those
in the rear, where v is the relative velocity of m and k. According to k,
the length of the ship is 1 km, it is 1 km wide and tall, and the clocks in
the nose and the ones in the rear all show the same time.
According to the theorem above, relativistic distance between events is

an “absolute” property. Clearly, any property defined from it is absolute,
too. By the use of modern rigorous logic, it can be stated and proved
that the properties definable from relativistic distance are the only abso-
lute properties; and moreover all of SpecRel can be re-constructed from
Minkowski geometry (i.e., from the “pseudo-metric” μ). How can one
formulate such a statement rigorously in formal logic?
Definability theory is one of the most beautiful parts of modern logic,
see (9), (21), (25), (30). It is about investigating connections between
theories formulated in FOL with completely different vocabularies (such
as, e.g., Finite Set Theory based on the relation and Arithmetic based on
+ and ∗). What would happen if we did not consider, e.g., the quantities
and operations on them as primitives of the language? What happens if
we are curious about where these primitive notions come from, if we want
to give them “operational” meanings? What happens if we choose the so-
called causality relation as the only primitive symbol of our language (as,
e.g., in Robb (32), Mundy (27))? Can we compare then these theories, can
we say that one of these is stronger or weaker than the other, or that two
such theories express the same amount of “knowledge” about the world?
Definability theory is strongly related to relativity theory and to pos-
itivist ideas. In fact, its existence was initiated by Hans Reichenbach in
1924 (31). Reichenbach in his works emphasized the need for definability
theory and made the first steps in creating it. It was Alfred Tarski who
later founded this branch of mathematical logic. Since then it developed
to a well-used and powerful theory, in largely due to the works of Michael
Makkai.
Very briefly, the reason why definability theory of logic is needed in rel-
ativity theory, as explained by Reichenbach, is as follows. When one sets
up a physical theory T h, one wants to use only so-called observational4
concepts, such as, e.g., “meeting of two particles”. While investigating
the (observationally based) theory T h (such as our SpecRel), one defines
new, so-called “theoretical” concepts, such as, e.g., “relativistic distance”
μ. Some defined concepts then prove to be so useful that one builds a new
theory T h based on the most useful theoretical concepts, and investigates
4 This observable/theoretical hierarchy is not perfectly well defined and is

known to be problematic, but as Friedman (13) puts it, it is still better
than nothing.
this new theory T h on its own merits. The new theory T h usually is
simple, streamlined, elegant - built in such a way that we satisfy our aes-
thetic desires. This is the case with Minkowski geometry. The original
theory T h contains its own interpretation, because we tried to use ob-
servational concepts. The physical interpretation of the new streamlined
theory T h is its connection with T h. The strongest relationship between
two theories in different first-order languages is called definitional equiv-
alence. When two theories are definitionally equivalent, in the rigorous
sense of definability theory of FOL, the observational oriented theory T h
can be recaptured completely from the theoretical-oriented streamlined
theory T h (and vice versa).
As examples, we can take our SpecRel for T h, and we can take
Minkowski Geometry for T h . Goldblatt (16, App. A) gave a com-
plete FOL axiom system MG for Minkowski geometry. His theory is
based on and is nicely analogous to Tarski’s FOL axiomatization for Eu-
clidean geometry (see, e.g., (37)). The definitional equivalence of our
present “observationally-oriented” theory SpecRel and the FOL theory of
Minkowski geometry MG given by Goldblatt is proved by Madarász (22,
Chap. 6.2).5 It is noteworthy to mention that in this application, relativ-
ity theory contributed to definability theory once again: for the precise
formulation of the equivalence of the two theories we had to elaborate a
methodology for how to define new “entities” (such as “events”) in ad-
dition to the old methods which are about how to define new relations
on already existing entities (such as “observer”). This definitional equiv-
alence of the two theories can also be expressed by saying that SpecRel is
complete with respect to the Minkowskian model of SR generalized over
ordered fields. Hence everything which can be formulated in our language
and true in these Minkowskian models can be proved from our axiom sys-
tem SpecRel. This is a kind of completeness theorem for the streamlined
theory SpecRel with respect to Minkowskian Geometry as the intended
model for SR.
Less tight relationships than definitional equivalence between theories
are also very useful, these kinds of relationships are called interpretability
and duality connections. For an illustration, let us turn to the question
5 For proving this equivalence, one has to add extensionality axioms for ob-
servers and light-signals to SpecRel, and one has to enrich MG with a “meter-
rod”.
of where quantities and coordinate systems come from. The axiom sys-
tem AxSR of James Ax (1) for SR is based on a first-order language that
contains only two unary relation symbols P,S for “particles” and “signals”
(corresponding to our “bodies” and “photons”), and two binary relation
symbols T,R for “transmitting a signal” and “receiving a signal”. One
can give an interpretation of our FOL theory SpecRel in Ax’s FOL theory
AxSR (see (3, proof-outline of Thm.2.1)). This amounts to defining the
primitive relations of the language of our SpecRel in terms of the primi-
tives of Ax’s AxSR, and then proving from AxSR the translated axioms of
SpecRel as theorems. This is an interpretation in the sense of definability
theory. Now, this interpretation also can be thought of as giving a kind of
operational “definition” for how to set up “operationally” the coordinate
systems appearing in SpecRel as primitives. The question of how to give
algorithms for setting up coordinate systems in this context is treated in
more detail and depth in Szabó (33).
Theories form a rich structure when we investigate their interconnec-
tions. Gödel’s incompleteness theorem pointed already in the direction of
investigating hierarchies of theories rather than single theories. (There is
no “strongest” theory for the interesting subjects, there are only stronger
and stronger theories.) Answering “why-questions”, “reverse mathemat-
ics”, modularizing our knowledge all point to the study of weaker and
weaker theories, and also to studying the interpretations between theories
(see (35)). Algebraic logic, developed by Tarski and his followers, is a
branch of definability theory which establishes a duality between hierar-
chies of theories and between classes of algebras (cf., e.g., (18, Chap. 4.3),
(19), (25)). In modern approaches to logic, theories are considered as dy-
namic objects as opposed to the more traditional “eternally frozen” idea
of theories. For approaches to the dynamic trend in mathematical logic
cf., e.g., van Benthem (38), Gabbay (15), and (26). This new “plurality of
theories” or “hierarchy of small theories (as opposed to a single monolithic
one)” approach can help realize the central or essential VC-aims without
the old stumble blocks of the original VC attempts. This is a wisdom
gained from FOM; see (12).
In the rest of the paper we briefly indicate how to arrive at a transpar-
ent FOL axiomatization of general relativity from our SpecRel. By this
we realize Einstein’s original program formally and literally.
First Step Toward GR: Extending the Theory to

Accelerated Observers
As a first step toward General Relativity theory (GR), we are going to
extend our SpecRel theory with accelerated observers. By accelerated
observer we mean any not necessarily inertial observer. Let us first note
that none of the axioms of SpecRel speaks about noninertial observers.
Since in the language we have already introduced the concept of ar-
bitrary observer, the only thing we have to do is to assume some axioms
about them. Our key axiom to assume about arbitrary observers is the
following:
AxCmv: At each moment of its life, any observer coordinatizes (“sees”)

the nearby world for a short while in the same way as some inertial
observer does.
For precise formulation of this axiom in the spirit we formulated the ax-
ioms of SpecRel see (4), (23), (35). Let AccRel− be the axiom system
consisting of AxCmv and all the axioms of SpecRel.
Let us see how strong our theory AccRel− is. To test its strength
we are going to investigate whether the Twin Paradox (TwP) and the
gravitational time dilation (“gravity causing slow time”) are provable from
it.
According to TwP, if a twin makes a journey into space (accelerates),
he will return to find that he has aged less than his twin brother who
stayed at home (did not accelerate). However surprising TwP is, it is not
a contradiction. It is only a fact showing that the concept of time is not
as simple as it seems at first.
A more optimistic consequence of Twp is the following. Suppose you
would like to visit a distant galaxy 200 light years away. You are told it
is impossible because even light travels there for 200 years. But you do
not despair, you accelerate your spaceship nearly to the speed of light.
Then you travel there in 1 year subjective time. When you arrive back,
you aged only 2 years. So you are happy, but of course you cannot tell
the story to your brother, who stayed on Earth. Alas, you can tell it to
your grand-. . . -grand-children only.
In the FOL language introduced in this paper we can formulate TwP,
see (23), (35). Let us denote the formulated version of TwP as TwP.
AccRel− is not yet strong enough to imply TwP. One would think that
this is so because we did not state enough properties of the real numbers
for speaking about curved lines. However, even assuming Th(R), i.e., all
the FOL formulas valid in the real numbers, together with AccRel− is not
sufficient to prove TwP, see (23), (35):
Theorem 14. AccRel− ∪ Th(R) TwP.
We note that the above theorem is a theorem stating that one cannot
prove TwP from AccRel− ∪ Th(R), it is not only the case that we are not
“clever enough” to find a proof but there is none. Its proof goes via using
the completeness theorem of FOL, namely we find a model in which all
the formulas in AccRel− ∪ Th(R) are true, but in which TwP is not true.
This theorem states that even assuming every first-order formula
which is true in R is not enough for our purposes. At first sight this
result suggests that our programme of FOL axiomatization of GR breaks
down at the level of TwP. It would be depressing if we were not able to
keep our axiomatization within FOL, because there are weighty method-
ological reasons for staying within it; see, e.g., (5, Appendix), (35, sec.
11). However, we are saved: in our language there is a FOL axiom scheme
(nice set of axioms) called IND which is sufficient for our purposes. Axiom
scheme IND expresses that every non-empty and bounded subset of the
quantities which is parametrically definable in our language has a least
upper bound (i.e., supremum). IND is a first-order logic approximation of
the second-order logic continuity axiom of the real numbers, and it belongs
to the methodology developed in FOM and in reverse mathematics that
AxField strengthened with IND are strong enough for a FOL treatment of
areas involving the real numbers.
Together with this scheme AccRel− implies TwP, i.e., the following
theorem can be proved, see (23), (35):
Theorem 15. AccRel− ∪ IND TwP.
How can a FOL axiom scheme be stronger than all the FOL formulas
valid in R? The answer is that IND is formulated in a richer language than
that of the reals, hence it can state more than the whole FOL theory of R.
If we assume IND only for formulas in the language of ordered fields, we
get an axiom schema equivalent to Th(R), see (35).6 Let us now introduce
our axiom system for accelerated observers as:
AccRel = SpecRel ∪ {AxCmv} ∪ IND .
Let us continue with the gravitational time dilation. By Einstein’s

equivalence principle, we can also formulate the statement “gravity causes
slow time” (usually called “gravitational time dilation” GTD) in our lan-
guage. Moreover, the formulated version of this statement is provable
from the theory AccRel, see (24), (35). The AccRel formulation of GTD
basically says that in any accelerated spaceship the clocks in the rear run
slower than those in the nose. (The effect is increasing with increasing
acceleration. Moreover, it approaches infinity as acceleration does.) So
we are able to derive nontrivial predictions about gravity before we have
introduced any axiom system of GR.
The theory AccRel is halfway between SR and GR. Einstein used a
non-formalized version of AccRel as a heuristic in introducing GR, e.g.,
when he made predictions about the influence of gravitation on the prop-
agation of light (10), (11, §§18-22).
Second Step: “Emancipating” Noninertial Observers

We are going to modify the axioms of SpecRel and AxCmv one by one and
get an axiomatic theory of general relativity. The modification consists
of “eliminating the privileged class of inertial reference frames,” which
was a central idea of Einstein’s, see (13), (11, §§18-22). We replace each
axiom of SpecRel by a new one which does not speak about inertiality but
otherwise the content of which tries to approximate that of the old one.
All the new axioms will be motivated by our theory AccRel. Roughly,
each axiom of AccRel will be replaced by a “generalized” version which
does not mention inertiality and which is still in the spirit of AccRel.
The generalized version of AxSelf is the following:
6 Actually, the restriction of IND to fields Q, +, ∗, < coincides with Tarski’s
FOL version of Hilbert’s continuity axiom for geometry, cf. (16, p.71, axiom
B5).
AxSelf − : An observer coordinatizes itself on a subset of the time axis:
∀oxzyt W(o, o, x, y, z, t) → x = y = z = 0.
The modified version AxEv− of AxEv contains the following two state-
ments: (1) any observer coordinatizes the events in which it was observed
by some other observer, and (2) if observer o coordinatizes an event which
is coordinatized by observer o , then o also coordinatizes the events which
are near this event according to o . This can be summarized as follows:
AxEv− : Any observer coordinatizes the events in which it was observed;
and the domains of worldview transformations are open.
The modified versions of AxPh and AxSymd are achieved by localizing
and generalizing them, i.e., we get the modified versions by restating
these axioms only in infinitesimally small neighborhoods, but for every
observer. The idea that “GR is locally SR” also goes back to Einstein.
Our symmetry axiom AxSymd has many equivalent versions, see (5, sec.s
2.8, 3.9, 4.2). We can localize any of these versions and use it in a FOL
axiom system for GR. For aesthetic reasons here we localize AxSymt, the
version stating that inertial observers see each others’ clocks behave the
same way. So AxPh− and AxSymt− are the formalized versions of the
following statements:
AxPh− : The instantaneous velocity of photons is 1 in the moment when
they “meet” the observer who coordinatizes them, and any observer
can send out photons in any direction with this instantaneous ve-
locity.
AxSymt− : Meeting observers see each other’s clocks behaving the same
way, at the event of meeting.
For formulation of these axioms and the corresponding concepts in our

first-order language, see (35).
Now all the four axioms of SpecRel are modified according to the
above requirements. Strictly following these guidelines, AxCmv− would
state that the worldview transformation between observers are differen-
tiable in their meeting-point. To avoid baroque wording, we state simply
differentiability of the worldview transformations. A natural generaliza-
tion is n-times differentiability (which is natural to consider in view of our
wanting to speak about location, speed and acceleration). Each axiom of

this series of potential axioms can be formulated in the language above
by the techniques used in (4), (23), (35).
AxDiffn : The worldview transformations are n-times differentiable func-

tions.
Let us introduce the following simple axiom systems for general relativity:

GenReln := AxField, AxSelf − , AxPh− , AxEv− , AxSymt− , AxDiffn ∪ IND .
The following theorem illustrates that our axiom system GenReln captures
the n-times differentiable standard models of usual GR well.
Lorentzian manifolds are the intended models of GR, much the same
way as Minkowski geometry was the intended model of SR. Roughly, a
Lorentzian manifold is a geometry which at every of its points locally
looks like the Minkowski geometry, cf., e.g., (39, p.23).
Theorem 16. GenReln is complete with respect to the n-times differen-

tiable Lorentzian manifolds over real-closed fields.
There are many interesting GR space-times, black holes, worm-holes,

time-warps, etc. The physical relevance of these so called exotic space-
times increases with time. For instance, there is a rapidly growing number
of experimental evidence for huge slowly rotating black holes, which are
the simplest examples of time-warps. According to Theorem 16 even the
most exotic model of GR is also a model of our GenRel theory. Hence,
within GenRel we can investigate the properties of these exotic models.
To ensure that we can do indeed physics in the framework of GenReln
(n ≥ 3) we defined in (4), (35) the notion of time-like geodesics in terms
of GenRel. These serve as world-lines of inertial bodies. So, though we
abandoned inertial observers as primitives, inertial motion becomes acces-
sible/definable as a derived notion (in terms of the primitives of GenRel).
Space-time curvature is defined from geodesics the usual way. So, in
particular, the outcomes of experiments involving inertial motion can be
predicted (e.g., computing the trajectories of bullets or photon geodesics)
on the basics of the new, streamlined theory GenRel in a purely logical
way.
Concluding Remarks
As it was the case with SpecRel, cf., (2)-(6), having obtained the stream-
lined axiomatization GenRel and its completeness for “usual” GR is only a
first step towards a logic based conceptual analysis of GR, its predictions,
alternatives or variants, answering the why-questions in a spirit which is
a natural continuation of the VC programme.
Acknowledgements: We thank the organizers and the participants of

the workshop “Vienna Circle and Hungary” held in Vienna in March
2008, for their hospitality, questions, and remarks. We also thank Rainer
Tiemeyer for inspiring letters about formalizing scientific theories. Re-
search supported by the National Foundation for Scientific Research grant
No T73601 as well as by a Bolyai Grant for Judit X. Madarász.
Bibliography
[1] Ax, J., The elementary foundations of spacetime, Foundations of Physics

8, 7-8 (1978), pp.507-546.
[2] Andai, A., Andréka, H., Madarász, J. X. and Németi, I., Visualizing some
ideas about Gödel-type rotating universes, in: M. Scherfner and M. Plaue,
eds., Gödel-type spacetimes: history and new developments, Springer, to
appear. arXiv:0811.2910v1.
[3] Andréka, H., Madarász, J. X. and Németi, I., Logical axiomatizations of

space-time. Samples from the literature, in: A. Prékopa and E. Molnár,
eds., Non-Euclidean geometries, volume 581 of Mathematics and Its Ap-
plications. Springer, 2006, pp.155-185.
[4] Andréka, H., Madarász, J. X. and Németi, I., Logic of space-time and
relativity theory, in: M. Aiello, I. Pratt-Hartmann and J. van Benthem,
eds., Handbook of Spatial Logics. Springer, 2007, pp.607-711.
[5] Andréka, H., Madarász, J. X. and Németi, I., On the logical structure of rel-
ativity theories. E-book, Alfréd Rényi Institute of Mathematics, Budapest,
2002. With contributions from Andai, A., Sági, G., Sain, I. and Tőke, Cs.
http://www.math-inst.hu/pub/algebraic-logic/Contents.html. 1312 pp.
[6] Andréka, H., Madarász, J. X., Németi, I. and Székely, G., Axiomatizing
relativistic dynamics without conservation postulates, Studia Logica 89, 2
(2008), pp.163-186.
[7] Andréka, H., Németi, I. and Wüthrich, C., A twist in the geometry of
rotating black holes: seeking the cause of acausality, General Relativity
and Gravitation 40, 9 (2008), pp.1809-1823.
[8] Carnap, R., Frank, P., Hahn, H., Neurath, O., Joergensen, J. and Morris,
C. (eds)., Unified Science. These works are translated in Unified Science:
The Vienna Circle Monograph Series Originally Edited by Otto Neurath,
Kluwer, 1987.
[9] Chang, C. C. and H. Keisler, J., Model theory. North-Holland 1973, 1977,
1990.
[10] Einstein, A., Über den Einfluss der Schwercraft auf die Ausbreitung des
Lichtes, Annalen der Physik 35, 10 (1911), pp.898-908.
[11] Einstein, A., Über die Spezielle und die Allgemeine Relativitätstheorie,
Verlag von F. Vieweg and Son, Braunschweig, 1921.
[12] Friedman, H., On foundational thinking 1, Posting in FOM (Foundations

of Mathematics), Archives www.cs.nyu.edu, January 20 2004.
[13] Friedman, M., Foundations of Space-Time Theories. Relativistic Physics

and Philosophy of Science, Princeton University Press 1983.
[14] Friedman, M., Reconsidering logical positivism, Cambridge University

Press 1999.
[15] Gabbay, D. M., Labelled Deductive Systems, Clarendon Press, 1996.
[16] Goldblatt, R., Orthogonality and spacetime geometry, Springer-Verlag,

1987.
[17] Gödel, K., An Example of a New Type of Cosmological Solutions of Ein-

stein’s Field Equations of Gravitation, Reviews of Modern Physics 21, 3
(1949), pp.447-450.
[18] Henkin, L., Monk, J. D., and Tarski, A., Cylindric Algebras Part II, North-
Holland, 1985.
[19] Henkin, L., Monk, J. D., Tarski, A., Andréka, H., and Németi, I., Cylindric
Set Algebras, Lecture Notes in Mathematics Vol 883, Springer Verlag, 1981.
[20] Hirsch, R., Einstein: Logic, Philosophy, Politics. Abstract for talk in “Logic
in Hungary, 2005”, Budapest, Hungary August 5-10, 2005. http://atlas-
conferences.com/cgi-bin/abstract/caqb-48
[21] Hodges, W., Model Theory, Cambridge University Press, 1993.
[22] Madarász, J. X., Logic and Relativity (in the light of definability theory).
PhD thesis, Eötvös Loránd Univ., Budapest, 2002. http://www.math-
inst.hu/pub/algebraic-logic/diszi0226.pdf.gz
[23] Madarász, J. X., Németi, I. and Székely, G., Twin paradox and the log-
ical foundation of relativity theory, Foundations of Physics 36, 5 (2006),
pp.681-714.
[24] Madarász, J. X., Németi, I. and Székely, G., First-order logic foundation of
relativity theories, in: D. M. Gabbay, S. Goncharov and M. Zakharyaschev,
eds., Mathematical problems from applied logic II. Springer, 2007, pp.217-
252.
[25] Makkai, M., Duality and definability in first order logic, Memoirs of the
American Mathematical Society No 503, Providence, Rhode Island, 1993.
[26] Mikulás, Sz., Németi, I. and Sain, I., Decidable Logics of the Dynamic
Trend, and Relativized Relation Algebras. In: L. Csirmaz, D. M. Gabbay
and M. de Rijke, eds., Logic Colloquium’92, Studies in Logic, Language
and Computation, CSLI Publications, 1995. pp.165-175.
[27] Mundy, B., Optical axiomatization of Minkowski space-time geometry,

Philosophy of Science 53, 1 (1986), pp.1-30.
[28] Németi, I. and Dávid, Gy., “Relativistic computers and the Turing bar-
rier”, in: Applied Mathematics and Computation 178, 1 (2006), pp.118-142.
[29] Neurath, O., Carnap, R. and Morris, C. (eds), Foundations of the Unity of
Science. The university of Chicago Press, Series: Foundations of the Unity
of Science: Toward an International Encyclopedia of Unified Science. Vol.s
1 and 2. 1971.
[30] Pambuccian, V., Alexandrov-Zeeman type theorems expressed in terms of

definability, Aequationes Mathematicae 74, 3 (2007), pp.249-261.
[31] Reichenbach, M., Axiomatization of the theory of relativity, Univeversity of

California Press, Berkeley, 1969. Translated by M. Reichenbach. Original
German edition published in 1924.
[32] Robb, A. A., A Theory of Time and Space. Cambridge University Press,
1914.
[33] Szabó, L. E., Empirical foundation of space and time. In: M. Suárez, M.
Dorato and M. Rédei, eds., EPSA; Epistemology and Methodology: Launch
of the European Philosophy of Science Association Proceedings of the First
Conference of the European Philosophy of Science Association, Madrid,
Spain, November 14-17, 2007, Springer, 2009.
[34] Székely, G., On why-questions in physics, in this volume.
[35] Székely, G., First-Order Logic Investigation of Relativity Theory with an

Emphasis on Accelerated Observers, PhD thesis, Eötvös Loránd Univ., Bu-
dapest, 2009.
[36] Tarski, A., What is elementary geometry?, In: L. Henkin, P. Suppes and
A. Tarski, eds., The axiomatic method with special reference to geometry
and physics, North-Holland, 1959, pp.16-29.
[37] Tarski, A., and Givant, S. R., Tarski’s system of geometry, Bulletin of
Symbolic Logic, 5, 2 (1999), pp.175-214.
[38] van Benthem, J. F. A. K., Exploring logical dynamics, Sudies in Logic,

Language and Information, CSLI Publications, 1996.
[39] Wald, R. M., General Relativity, The University of Chicago Press, 1984.
László Kalmár1
The Development of Mathematical Rigor from

Intuition to Axiomatic Method
It is proverbial to place one’s trust in the infallibility of mathematics: “as

sure as two plus two makes four”, one says about something when there is
not a shadow of doubt about it. We, mathematicians delight in basking in
the luster of trust and confidence, even though we know that most often,
that light shining on us is due not so much to convictions of certainty, but
instead to the mighty influence of the proverb. In addition, we might even
have created a reputation of formidable impenetrability for our scholarly
field – so much so that non-professionals would rather believe everything
just to keep us from explaining it all. And we show aristocratic disdain
for the “everyman”, because his mode of expression lacks precision, and
his views are not as enduring as ours: often we even scorn philosophers
and theoretical physicists, even though they are striving to implement
the methods they had learned from us, in order to make their own field
of inquiry more rigorous. Yet we balk at mentioning that mathematical
rigor is also the product of development. Whence the reluctance? Are we
perhaps concerned about losing outsiders’ trust? But what is honest in
this trust can only grow stronger once we make its grounds explicit. As for
trust unaccompanied by conviction, let it crumble so its place can be taken
over by honest trust. Instead of sheer (often unstated) criticism leveled at
the scholars striving for rigor, we would be of better service to them, too,
1 Kalmár’s lecture was first published in the 1942 yearbook of the Exodus
Working Group (a group of young Calvinist intellectuals influenced by the
ideas of Sándor Karácsony a professor of pedagogy). It was reprinted in a
volume of Kalmár’s popular and philosophical writings entitled Integrállevél
[Letter on Integral] (ed. Antal Varga, Budapest: Gondolat, 1986). The
present translation is based on the 1986 edition and is published here with
the kind permission of Éva and Zoltán Kalmár. One of Kalmár’s remarks,
concerning Hungarian mathematical terminology, has been omitted and
marked with . . . Note by the editor.
270 László Kalmár
if we explained how we, mathematicians have attained the level of rigor

that they also have come to admire and regard as a paradigm. Lastly,
it could well be of use to us, mathematicians as well, if we confronted
the fact that there is such a process of development; and if, in addition,
we draw the didactic consequences of this fact, we can even advance the
cause of teaching mathematics.
These considerations have led me to address this topic in a lecture
I gave on November 6, 1941 at the Loránd Eötvös Society for Mathe-
matics and Physics. And now, accepting the invitation of the Exodus
Working Group, I shall attempt to write up my presentation so as to take
into account what I have learned from comments that colleagues have
contributed since; I aim to explain things so that the major points (at
least those) can be readily understood by a broader audience: everyone
interested in this issue, not just mathematicians. Whether I succeed in
this endeavor is up to the reader to decide; if I have failed, then I shall
take home the lesson: before I write for others about pedagogical issues,
I myself have to become a better teacher.
Let me mention another point at the outset: I will not discuss the
issues from an historical perspective, I leave this task to someone with
a thorough knowledge of the history of mathematics. Instead, I will de-
scribe the road that individual mathematicians travel while constructing
for themselves a rigorous system of mathematical concepts and theorems.
And I will describe this path as I see it in hindsight. I realize that often I
am no longer seeing the road that I actually traveled, but see instead the
shortest path that I could have taken to get to where I am standing now.
I also realize that other mathematicians may view their own paths differ-
ently – it is to this fact that I will attribute objections from colleagues,
rather than blaming the phenomenon that we tend to balk at admitting
that each and every one of us must travel a distance before reaching our
present vantage point (although I do admit that I, for one, would not find
it an easy task to concede this).
1.
The point of departure for our journey is the intuition. Everyone accepts
that our geometrical concepts – like point, line, surface, direction, angle,
length, area, volume, etc. – derive from the contents of intuition. If we
The Development of Mathematical Rigor 271
consider things closely, we realize that the same holds for the concepts
of arithmetic, too: five chalks, half an apple – these denote clear con-
tents of intuition. But there is general agreement among experts that
certain rather abstract concepts of mathematics have nothing whatsoever
to do with intuition. Set theory is perhaps the most abstract branch of
mathematics; it deals with collections, sets of arbitrary objects, elements;
nonetheless, at the most rudimentary level of concept formation, we imag-
ine sets intuitively, as though they were like sacks into which someone has
put their elements. Those who still remain unconvinced that every math-
ematical concept arises out of intuition – who else could these doubters
be but mathematicians dealing with a very abstract subject matter, like
mine? – should recall when they understood a concept that had been con-
strued abstractly, or when they created a new concept: they must have
caught themselves – at least temporarily – considering the concepts intu-
itively, “heuristically”, only later dressing them up again in their abstract
form.
The intuitive level consists in concepts’ being associated with a vivid,
transparent picture; various properties of the concepts can be read off
this picture. The “laypeople” content themselves with recognizing only
the properties that are directly evident; mathematicians are different –
they want to explore even those properties that have greater complexity,
that are not directly transparent. To this end, they employ their logic:
via logical steps, they strive to trace back the more complex aspects to
directly transparent facts. At this stage, mathematical proof consists
precisely in this process of tracing back the concepts.
Through employing logic in this way, mathematicians come to realize
that the concepts that provide an exact match for the intuitive picture
lack sufficient logical manageability; through abstraction, they therefore
construct new, logically simpler concepts. Laypeople also accept that
the concept of a point without dimensions is far more manageable than
the ever-different chalk mark that the point had originally meant at the
intuitive level They likewise opt for the number five itself (that is, the
picture of five arbitrary objects, objects whose nature is of no interest
to us) over five apples or my five fingers. For the concept we arrived
at through abstraction, we still retain an associated picture, even if it is
less colorful than the original, many properties of which we have since
disregarded. But at this stage, we are still careful not to break away
from our picture, we can continue to read off it the characteristics of the

concept. As soon as we recognize, via logical steps, a property we could
not read off the picture originally, we return to the picture, coloring it with
the newly unveiled property. Thus the picture becomes more colorful and
vivid, so we can read off it the new, hitherto hidden properties as well.
For mathematicians, this development of intuition amply makes up for
the fading effect of the abstraction process; they are even emboldened
enough to carry out another round of abstraction on the newly re-colored
concepts gotten through abstraction. This is how mathematicians get
from the concept of 1, 2, 3, 4 and 5 to the concept of an arbitrary integer,
for example. Again, they search for the hidden properties of the new
concepts, using these to color the concepts again, and so on.
We see therefore that in and of itself, development in mathematics
does not force us to break away from the intuitive level. Why do we
abandon it then, when this is the level at which mathematics is at its
clearest and most beautiful?
2.
The major incentive that prompts us to break away from the intuition
is, I think, the fact that humans, including mathematicians, are social
creatures. They like to communicate to others what strikes them as in-
teresting and notable. This is when they are in for the first round of
disappointments. It turns out that what is obvious to me based on my
intuition might inspire a puzzled look from others. This is perhaps due to
the fact that their intuition is not as developed, or is more developed in a
different direction than mine; perhaps it is because they are more acutely
critical with respect to logical reasoning. This is how I come to realize
that my intuition is filled with subjective elements.
The easiest way to handle this is by listing, before presenting a cer-
tain idea, the concepts and the properties of those concepts to which I will
refer as evidently given by my intuition. Those to whom I am presenting
my proof can examine these one by one, check them against their own
intuition, to see whether they likewise find clear what these “basic con-
cepts” mean, and whether they likewise find these “basic truths” evident.
Only after they have accepted all this will I consider the more complicated
concepts on which my idea relies, tracing them back via logical steps to
the basic concepts, thus defining the complex concepts; only then will I
consider the more complicated truths that are my target, tracing them
back to the basic truths (also called axioms), thus proving the complex
truths. This process is the simplest version of the axiomatic method.
If some basic concept of mine is not immediately clear to my “part-
ners”, I try to trace it back, via logical steps, to even simpler concepts
that they do find clear; if they do not accept an axiom of mine as evident,
I try to trace it back, via logical steps, to truths that they do find evident.
If I do not succeed in this, I have no choice but to give up on getting my
idea across. (Perhaps it was such experiences that led Euclid, the Greek
mathematician to axiomatize geometry for the first time, to distinguish
between axioms and postulates. The former – which Euclid called koinai
ennoiai, that is, shared contents of intuition – are the basic truths that are
evident to everyone; the latter – aitemata, that is, requirements – impose,
in a way, a measure on those who strive to understand Euclid’s proofs;
the proofs are not intended for those who fail to meet the requirements,
that is to say, do not accept the postulates.)
It is well to observe that even in this most rudimentary form, the
axiomatic method requires that I silence the call of my intuition in two
ways. First: if my partners do not accept one or another of my axioms (or
basic concepts), I provide a proof (or definition) for my audience’s sake,
even though my intuition suggests that no proof (definition) is needed.
Second: after we have enumerated the basic concepts and axioms I give
further definitions and proofs in a purely logical fashion, still without
reference to the intuition. For example, if I have agreed with my partners
that we find it mutually clear what we mean by the area of a polygon, and
that we also find it mutually evident that (i) the area is always a positive
number; (ii) the area of the part is smaller than the area of the whole;
(iii) if I cut a polygon into two parts with a straight line, then the area of
the whole equals the sum of the areas of the two parts; (iv) two congruent
figures have the same area; (v) the area of the unit square is one (because
we have selected this as our area measurement unit), then I deduce from
these axioms the equations for the area of the rectangle, the triangle, and
the trapezoid without making reference to further intuition-based truths,
however forthcoming these might seem.
Nonetheless, at this stage, I still remain in close contact with the in-
tuition. I am being led by my own intuitive pictures (and my partners
are led by their own) during the process of selecting the axioms and car-
rying out the proofs; basically, we are still thinking intuitively, but with a
common starting point: with the system of axioms and the shared path –
with logic, that is – we ensure that we are proceeding in parallel, arriving
at the same place. We can therefore call this developmental stage the
intuitive axiomatic approach.
At this stage, the process of abstraction continues, fostered by the
axiomatic method, because through the axioms, we can provide conve-
nient and clear characterizations of the concepts developed through the
abstraction process.
System construction is also fostered by the axiomatic method. At the
intuitive level, we have already compiled theorems and problems based
on the outward similarity of their subject matter; this is how we have
talked about theorems of geometry, questions of arithmetic. In connection
with the axiomatic method, we realize that while proving theorems about
similar subject matters, by and large, the same axioms are needed, and
from then on, the theorems are compiled into individual systems (for
example, elementary geometry, arithmetic, calculus), each of which we
prove with a common set of axioms.
3.
There is yet another reason – besides the urge to communicate our ideas –
keeping us from remaining at the intuitive level, from regarding our intu-
itive concepts as final. In order to set aside the urge to communicate, let
us imagine a mathematician living a Robinson Crusoe-like existence on a
deserted island. His intuitive concepts have developed, he seeks out their
hidden aspects, he abstracts away. He thus reaches general concepts and
about these, general theorems. Another characteristic of mathematicians
is that they like applying their general theorems to special cases. While
they are thinking of the general concept, what they observe is the clear
“core” of the concept. For example, as a general triangle, they think of
various triangles with sides not too great or too small; as a continuous
curve, they think of a nice and smooth line. Sometimes they have to apply
the theorems to an unusual case, for example, to an extremely elongated
triangle, or to a spiral line that rotates around a point infinitely many
times. In the former case, it is beyond doubt that the triangle with the
very acute angle is still a triangle; in the latter case, the mathematicians
might be puzzled whether they really have at hand a continuous line.
They thus realize that specific intuitive concepts possess varying degrees
of distinctness. People experience the arithmetic concepts as the most
clearly distinct ones; these are what provide paradigms for distinguishing
our other concepts.
With the axiomatization itself, we have not yet obtained distinct con-
cepts; but the axiomatization suggests a path towards a solution – and
now recall yet again the “social” mathematicians, who are also striving
for neatly distinguished concepts. If we want to apply an axiomatically
proven property of a general concept to a special case (perhaps we want
to introduce the square root based on the theorem that an increasing
continuous function can always be inverted), we do not necessarily have
to prove that the special case belongs under the general concept, which
is perhaps not yet distinguished clearly (for example, whether x2 is a
continuous function in the intuitive sense); instead, we will only need to
see if the special case has the properties of the general concept that are
expressed in the axioms used within the proof. In the case of numerous
applications, it is worth giving a name to the concept characterized by
these axioms; ordinarily, the name that sticks is the one that was origi-
nally used for the intuitive concept. (An example of this is the concept
of a continuous function. Originally, this concept designated an intuitive
concept for us: a function whose curve can be drawn up without a break.
This intuitive concept has several other properties beyond those that we
require today of the “rigorous” concept of a continuous function, that is,
those properties that we need in constructing a theory of continuous func-
tions. Today, mathematicians call those functions continuous that have
these properties.)
With this, we have already covered the next segment of the road
ahead; we are at the stage of the abstract axiomatic approach. Our basic
concepts are no longer based on intuition, but they are simply undefined
concepts about which the only knowledge we have is that they satisfy the
axioms; the axioms do not serve to express evident truths, instead, they
specify conditions that characterize the basic concepts. This way, the
basic concepts are, to a certain extent, indeterminate; they could mean
anything that satisfies the axioms. For example, area could mean any-
thing that satisfies the list of five properties. Or consider the concept of
direction; through this, we can easily trace the road covered thus far. At
the intuitive level, we could, without giving it a second thought, use the
concept of direction, because everyone knows already what the direction
of a straight line is supposed to mean. . . . At the stage of the intuitive ax-
iomatic approach, we still understand direction as being this concept that
is clear to everyone, but we also establish that it is evident to everyone
that (i) parallel straight lines have the same direction, and that (ii) inter-
secting straight lines have distinct directions; in our thoughts concerning
lines, we employ these two properties of direction only, and they prove to
be sufficient. At the stage of the abstract axiomatic approach, direction
no longer carries for us its original intuitive content, but anything that
is the same in the case parallel straight lines, and different in the case of
intersecting straight lines; for everything that we have proven based on
these two properties will be valid for anything of the sort. We already
see from these overly simple examples that the indeterminacy of the basic
concepts is partly intentional: the more concepts satisfy the axioms, the
wider the scope of application of the theorems we have proven based on
those concepts. On the other hand, modern research on axiomatic method
has brought the result that this indeterminacy is far more extensive than
it might appear at first glance: the axiomatic systems are satisfied by sys-
tems of concepts (models) that have diverse contents, or that have diverse
structures; by themselves, the axioms are never sufficient to help us re-
solve each and every problem associated with the concepts characterized
by those axioms. But a detailed account of this line of research would
take us too far afield.
At the stage of the abstract axiomatic approach, new abstraction pos-
sibilities emerge. Mathematicians often experience the surprising phe-
nomenon when concepts deriving from quite diverse contents of intuition
turn out to share properties. (Consider, for example, the analogy between
arithmetic equality and geometric congruence.) The axiomatic method
often reveals the reason behind this phenomenon: the shared properties
could be due to the fact that there are shared axioms characterizing the
concepts in question, and to prove the shared properties, these shared
axioms are already sufficient. Because we no longer insist on deriving our
concepts from intuitive pictures, we can construct an abstract concept
that is exactly what the shared axioms characterize. This is how we get
the concept of an “equivalence relation”: through abstraction from the
concepts of equality, congruence, similarity, parallelism. The concepts of

modern algebra (group, ring, field etc.), and the concepts of the theory of
abstract space were created in a similar way.
At this level, one of the characteristic activities of mathematicians:
generalization also kicks in. Using the axiomatic systems that charac-
terize the intuition-based concepts, we construct new axiomatic systems
for which there is no longer an intuitive picture. As an example, I will
mention the geometry of bodies with more than three dimensions, which
was created as an analog of the geometry of straight lines, planes, and of
space.
The previously mentioned process of gradually coloring, at the in-
tuitive level, the gray picture associated with abstract concepts, is also
present in the minds of mathematicians with respect to the abstract con-
cepts constructed at the axiomatic level. There is somewhat of a picture,
albeit a vague one, associated with these concepts as well; after all, the
specific concepts whose shared properties we started out from provide ex-
amples of the concepts characterized by the shared axioms; and to the
concepts created through generalization, we associate, to some extent,
a picture based on the analogy. As we discover hidden properties and
aspects of these concepts, we subsequently begin to color the picture in
such a way that eventually, it will approximate the vividness of the initial
picture of the concept derived from intuition. This is how it is possible
for those working on multi-dimensional geometry to slowly begin to “see”
in multi-dimensional space as well, that is to say, they can readily survey
geometric aspects of that realm, too.
In principle, it is possible for us to select our axioms arbitrarily, inde-
pendently of any intuitive picture, and construct a theory of the concepts
characterized by those axioms. The “everyman” usually imagines mathe-
maticians’ work in this way: they sit at their desk and “draw inferences”.
I have never met a mathematician who does this. Setting out to solve
problems without any intuitive picture would be like feeling one’s way in
the dark: we would have no clue as to which direction to search for the
solution. In addition, we would have no standard for assessing whether
it is a worthwhile project to examine the properties of the concept thus
characterized, whether it presents an interesting problem or not. For this
reason, this sort of axiomatic approach that is completely divorced from
intuition is carried out at best as a game, a curiosity, but no-one considers

it serious mathematics.
4.
In principle, however, we could go even further in breaking away from
the intuition. So far, even when we characterized our concepts in terms
of axioms rather than contents of intuitions, we still continued to work
with contentful concepts; our theorems expressed properties of these con-
cepts. For this reason, the sort of axiomatic approach characterized up to
this point can be called content-based axiomatic approach. By contrast,
the standpoint of the formal axiomatic approach is that our concepts are
without content, names for them (or symbols – for often, mathematicians
introduce symbols rather than words for their concepts) are toys, just like
chess pieces; the axioms are not intended for characterizing the meanings
of concepts, but for specifying which are the theorems of our system. For
example, according to the content-based (abstract) axiomatic approach,
the Peano axioms of arithmetic serve to characterize the basic concepts
of 1, of (natural) number, and of succession; for these mean things for
which it holds that (i) 1 is a number; (ii) the successor of any number
is itself a number; (iii) distinct numbers have distinct successors; (iv) 1
is not a successor of any (natural) number;2 (5) any property of 1 that
is “inherited” from a number to its successor is a property of all natural
numbers. According to the formal axiomatic approach, 1, number, and
succession are symbols without meaning, the five axioms listed are combi-
nations of these that likewise lack meaning; theorems are any meaningless
combination of symbols gotten from the axioms via certain procedures
(that is, through logical inference). The axiomatic system thus plays a
role similar to the initial position in a chess game, while the logical rules
are similar to the rules of chess. It is easier to formulate them precisely if
we use symbols instead of words for the concepts. As a result, the axioms
and the theorems become formulae; for this, we just need to add a cou-
ple of symbols, especially for logical relations, in addition to the symbols
generally used in mathematics.
It holds even more so for the formal axiomatic approach that it is
given in principle only; in reality, pursuing it for its own sake would be
2 Kalmár does not regard 0 a natural number. Note by the editor.
a game only, not mathematics. Its significance resides in its utility as a

working principle, when it comes to examining various questions within
Hilbert-style proof theory: whether arithmetic is free of contradiction, or
whether all problems of arithmetic (or some other system) can be solved.
For we cannot even begin to answer such questions until we have formu-
lated precisely what it means for a theorem to be provable; and from the
perspective of the proof theory, it is an extremely significant fact that this
can be formulated, regardless of what the axioms might mean. Hilbert
himself, who was the first to formulate the formal axiomatic approach,
did not regard it as anything other than a working hypothesis; on several
occasions, he stressed that he did not view mathematics as a logical game
with formulae void of meaning.
5.
For this reason, we can exclude the formal axiomatic approach from the
developmental path; one segment of our journey is still ahead of us, how-
ever. We can reach it by following the principle of the economy of thought.
This principle is already prevalent at the intuitive axiomatic level: via
logic, I aim to trace back some of my axioms to others so that my “part-
ner” has fewer axioms to check. This means that I have to prove some
things that are evident to me as well as my partner. For example, in the
case of area, via a logical proof based on several axioms, I can prove that
the area of the part is smaller than the area of the whole; this way, I can
spare one axiom. It also follows from the just mentioned principle that if
possible, I define a basic concept with the help of the others, even if it is
clear to both me and my partner what the concept means.
The abstract axiomatic approach affords new opportunities for econo-
mizing on our axioms. If we manage to use the concepts of an already set
up system to construct new concepts that could replace the basic concepts
of a second system, satisfying the axioms of that system, then we have
economized on the basic concepts and the axioms of this second system:
instead of its basic concepts, we use the concepts we have constructed with
the help of the concepts of the first system, so the axioms of the second
system become theorems provable in the first system. For we have al-
ready accepted the view that the basic concepts can mean (independently
of their original intuitive content) anything that satisfies the axioms; they
can therefore mean the concepts in question, which we have constructed

with the concepts of the first system, for those constructed concepts do
satisfy the axioms. This procedure is what we call the construction of a
model within the first system for the basic concepts of the second system.
For example, if, in accordance with the abstract axiomatic approach,
we are willing to mean anything by the direction of a straight line that is
the same in the case of parallel lines and different in the case of intersecting
lines, then we can define the direction of the line in such a way that it
means the set consisting of the line and lines parallel to it, even thought
a completely different intuitive content is associated with this set than
with direction; but our approach tells us that the only thing that matters
is that if we start with two parallel lines, we end up with the same set
(the set of lines parallel to the first line is the same as the set of lines
parallel to the second line, always including the line in question), whereas
if we start with intersecting lines, we get distinct sets. In this case, based
on the concept of a set, we have constructed a model for the concept of
direction.
Generally, we would select arithmetic as the basis for model construc-
tion, hence the label ‘arithmetization’. It is via such arithmetization that
we define the points in a plane as pairs of numbers, the points in space
as triplets of numbers, and the irrational numbers as partitions (so called
Dedekind cuts) of the set of rational numbers such that the numbers in one
partition are all smaller than the numbers in the other partition. Laypeo-
ple and especially philosophers tend to object to such model-construction-
based definitions; they reproach us that the definiens (for example, the
pair of numbers) is associated with an altogether different intuitive content
than the definiendum is (for example, the point). This is quite natural,
after all, only those can understand and accept such definitions who have
previously accepted the abstract axiomatic approach and along with that,
the idea that the only thing that matters is that the definiens satisfy the
axioms with which we characterized the definiendum.
Besides economizing on basic concepts and axioms, model construc-
tion also mitigates the indeterminacy of the concepts characterized by
the axioms, in so far as the arithmetized concepts are determinate to the
extent that the concepts of arithmetic are. (The extent of this depends
again on how much we have axiomatized arithmetic; I do not want to
dwell on this question).
Generalization based on analogy is even more easily done at the level

of model construction than at the abstract axiomatic level; it is not the
axiomatic system, but the model itself that we generalize. For example, we
can introduce four-dimensional space simply as the totality of quadruples
of numbers.
6.
This developmental path is covered roughly the same way by every math-
ematician, some of them do so more consciously than others. The trigger-
ing causes for rising from one level to the next can vary by individual, of
course. If I have reached a certain level with respect to one system, and
there I see the advantages of the axiomatic approach, I can then axioma-
tize another system based on that example, even if I do not have a specific
triggering cause for it; this is how mathematicians arrived at the thought
of axiomatizing arithmetic based on the example of geometry. And we
were driven to axiomatize set theory because the intuitive, so-called naı̈ve
set theory turned out to be contradictory.
We see that development leads to increasingly more rigorous systems
of concepts. At the intuitive level, I was still working with subjective
concepts. At the intuitive axiomatic level, I have already specified, in a
way acceptable to others, which of these concepts we are allowed to use,
but the subjective component inherent in the intuitive concepts remained.
At the abstract axiomatic level, I characterized concepts objectively, by
means of axioms, but their indeterminacy still permitted that within the
bounds of the axioms, our understanding of the concepts vary from one
person to the next. At the level of model construction, the indeterminacy
is already disappearing.
For the sake of rigor, however, we gradually sacrifice intuitiveness. At
the intuitive axiomatic level, once I have specified the axioms, I stifle the
call of intuition as I prove things that are evident based on my intuition
(either because they are not evident to my partner, or in order to mini-
mize the number of axioms needed). At the abstract axiomatic level, at
the outset, I already reject the guidance of intuition. And at the level of
model construction, I start out with definitions that go directly against my
intuition. Yet throughout, intuition has remained an indispensable tool in
research; even for our most abstract concepts, we subsequently developed
intuitive pictures, so we could navigate among those concepts. There are

no mathematicians, however abstract their subject matter might be, who
do not initially think intuitively, “heuristically” during their research;
subsequently, they would cast their results in axiomatic form, thereby
camouflaging how they arrived at them. And even at the level of the ex-
treme formal axiomatic approach, when we regard theorems as stripped
completely of their meanings, as mere formulae, we still conceive of these
formulae and the combinatorial operation over them in an intuitive fash-
ion. The proof theory relies precisely on this intuitiveness of the formulae.
Based on all this – after realizing that when it comes to rigor, by sac-
rificing the intuition, or more exactly, failing to acknowledge its role, we
have not gotten the anticipated results – I think it is likely that further
development will afford a greater role to intuition yet again. This is indi-
cated, among other things, by the intuitive reasoning in proof theory, and
by the so-called intuitionistic mathematics, which is based on arithmetic
intuition.
7.
Now, however far we have gotten in this developmental process, and what-
ever our opinion might be about the steps ahead, we must realize that if
we want to introduce others to mathematics, we must help them so they,
too, can follow along this path, for it is only through these levels that
they can reach our position. For we are the same way in that when we are
presented with something new in axiomatic style, we understand it only
after we have found an intuitive meaning to attach to the subject, and
how we get from that intuitive meaning to the axiomatic form presented.
In secondary school, we mostly stay at the intuitive level; a summary,
a revision, or preparation for the final examinations might provide op-
portunities for rising to the intuitive axiomatic level. In higher levels of
mathematics education, we do nonetheless have to walk the entire path
from intuition to the abstract axiomatic approach, and even to model
construction.
The easiest way to convey how I think about this is by describing a
method that is diametrically opposite to mine. In fact, there are two such
methods. One of them – remaining at the intuitive level – is generally
considered obsolete today; for this way, we would have to forgo conveying
to our students the intellectual achievements of Cauchy, Weierstrass and

Dedekind. But all the greater is the temptation to turn to the opposite
extreme: start with axiomatic theory straight away. But beware: we could
be axiomatizing even when we do not specify basic concepts and axioms.
Axiomatization with tacit axioms is even more difficult to understand
than explicitly declared axiomatization. It is precisely at the highest level
of axiomatic method, during model construction, that axioms disappear
and we can deceive ourselves, saying: we are not even axiomatizing, just
giving definitions, and we are surely within our rights to do that. But
it does count as axiomatization when we define the irrational numbers
with Dedekind cuts, or when we define function continuity with ε and δ
(or with xn and yn ). (I apologize for bringing in examples from teaching
calculus. This is where I am most at home, which is to say, it is here that
I have committed the most “crime” in practice sessions on differential
and integral calculus, especially when I would teach some of the lectures
myself, substituting for my professor while he was away. Perhaps readers
whose field is far removed from mathematics will also forgive me for using
examples from areas unfamiliar to them in order to illustrate my points.
It compounds the problem if at the first class, I am already starting with
the Dedekind cut: I thereby risk the first impression of my students. And
all teachers are aware of how profoundly first impressions affect the image
my student formulates about me; they must also be aware that I am not
the one educating the student, but instead it is this image, which is often
caricature-like that is doing the teaching; and I can influence my student
only insofar as my behavior and presentation can influence this image.
What impact will this extreme axiomatic method have on my stu-
dents? This depends on what their interests are. Some of them are unin-
terested in mathematics itself, they need its applications only; I will call
these students “chemists”, because in Szeged, where I teach, this group
mostly comprises doctoral candidates in chemistry. For others, math-
ematics will be their profession, but they have no special ambitions; I
call these students “teacher candidates”. The third, smallest group com-
prises those for whom mathematics is their passion; we expect the next
generation of mathematicians to come from this group; I will call them
“mathematicians”.
After a couple of classes, the “chemists” resign themselves to not
understanding what I am talking about. They quit listening, and come to
class only to fulfill the attendance requirement, and even then, they are
sitting in the last rows and are diligently studying (I am guided by good
faith here!) – some other subject. Once they have to apply mathematics,
it will turn out that they attended my lecture in vain.
The “teacher candidates” know all too well that mathematics lec-
tures are important to them, they are quite interested, too; I cannot hold
it against them that they are trying to understand my lecture. And this
consists in their trying to extract the intuitive contents from the axiomatic
concepts. What else would they do: we do the same when we are try-
ing to understand something. And once they succeed, they are in for
a disappointment: using a sizable axiomatic apparatus, I prove for them
something that would be evident based on the intuitive content extracted.
They think: maybe continuous function means something other than what
they had thought after all, for then no proof would be needed to show
that a continuous function that is positive at some point and negative at
another must in the interim have the value zero, for that much should be
evident! I end up in the same place if I adhere to the fashionable slogan of
“intuitive teaching”, divide up the blackboard, and on one side, I make the
theorem intuitive through a diagram (for example, Rolle’s theorem), and
on the other side of the blackboard, I axiomatize as though the diagram
was not there at all (telling the students that now I am proving the same
thing “rigorously”). Upon seeing the diagram, the students’ faces light
up, they understand what I am talking about; but upon hearing about
the axiomatization, they lose heart, thinking: apparently they were mis-
understanding something after all, for else why would all this be needed?!
And by the time they hear that there is a continuous curve that does not
have a tangent anywhere, they give up the battle, and give up on being
able to understand these concepts, and reconcile themselves to the idea
that abstract things have to be accepted in an abstract way, definitions
must be memorized, for those who do not do so will fail the exam. No
wonder then that they cannot subsequently apply the theorems they had
memorized so precisely; their major bogeyman is the specialized teaching
examination.3 And when they become teachers, they relearn from a good
3 Those who wanted to receive a teaching certificate besides their university

degree had to enroll in the so-called Teacher Training Institute as well,
where they had to take a specialized exam in front of a Teacher Supervising
Committee. Note by the editor.
or a bad textbook what they have to teach; or worse, they might pull out
their university notes and based on those, would try to teach secondary
school students about the Dedekind cut, in order to “teach rigor” to the
students.
The “mathematicians” enjoy my lectures, delighting in the splendid
logical construction. Of course they are, for mathematics has been a
passion of theirs in secondary school already, like soccer, stamp collection,
or detective novels were for others. In their spare time, and even in
place of studying other subjects, they would be absorbed in books about
mathematics, solving problems. They have gotten past the intuitive level
on their own. During their university career, we would present to them the
finished results of mathematics in an elegant form. They are extremely
fascinated by this and can hardly wait to create something similar for
themselves. And when they set out to do so, they are in for a huge
disappointment. However much they might be interested in a problem,
a neat, smooth solution just does not want to fall out of the sky. They
become disheartened, “it looks like I am not talented enough”, they think.
Out of exasperation, they take off and pursue a doctorate in philosophy.
Or, if they are so interested in mathematics that nowhere else do they feel
at home, they stick with the problem, but retain the bitter feeling that
they have been cheated by me, for I have not shown them how others
came to discover any of those things, how it is that one can make such
discoveries. At long last, after much difficulty, they reach an independent
result; they describe it as they see it, for I have not taught them how
you put into a rigorous format what is heuristically formulated. And
then of course I criticize them, saying their writing is not accurate, and
am surprised that their piece is not fit for print even after the third and
fourth draft; and I do not acknowledge that I am the one who caused all
this.
For the sake of clarity, I exaggerated, drew a caricature. I realize
that hardly anyone in Europe would give an extreme axiomatic lecture;
reality is a mixture of the axiomatic approach and the method I described
above. I also realize that even with the worst method, there are still good
chemists and good teachers being trained, and the born mathematicians
still become scholars. But what is the point in mixing bad into the good,
when we could present the good in its pure form? Why should our students
learn despite our method when we could educate them correctly instead?
Yet again, I will use elements from calculus to illustrate how I conceive
of the correct method in practice. Dividing the material into chapters
(limits, continuity, differential calculus, integral calculus) is quite useful
for educational purposes; we can keep it roughly the same. But we must
begin each chapter with a plethora of examples, and from these, we can
create the intuitive concepts: the real number is the measure of distance;
a continuous function can be drawn as a curve without a break (intersect-
ing each ordinate line at one point only); the value of the differential is
the slope of the tangent, etc. The theory of limits we can go over while
at the intuitive level still, because the axiomatization and arithmetiza-
tion of the concept of a real number is a difficult task for beginners; at
the early stages of the lecture, we would find it hard to explain what is
their purpose. We can return to this at the end of the lecture, once we
have seen the utility of axiomatization on other examples; then we can
pose the question: couldn’t we axiomatize the concept of a real number
in a similar fashion. But it is also quite alright if this is left to a sepa-
rate lecture in which we give an axiomatic construction for the concept of
number starting with the integers. In the theory of continuous functions,
axiomatization readily suggests itself; the intuition about continuity in-
cludes boundedness and there being values between any two values of the
function, as well as the idea that a sufficiently small change δ in an inde-
pendent variable can yield an arbitrarily small change ε in the value of the
function; but if we want to prove a specific function to be continuous, it
would be inconvenient to prove every one of these properties (and it is un-
certain whether these would then provide an exhaustive characterization
of the intuitive concept of continuity); it is therefore important to know
that from the last property, all others follow logically. Because for our
purposes, from the perspective of future applications, the only thing that
matters is whether the above mentioned properties (and a few others that
follow from them) obtain or not, we agree that we will henceforth call those
functions continuous that bear this property in terms of ε and δ. At this
point, it is worth making students realize that we have here replaced the
original intuitive concept with one that is intentionally broader but is also
more manageable logically and arithmetically. This way, at the very least,
our students will not be surprised upon hearing about functions that are
everywhere continuous and nowhere differentiable, but they will instead
consider this phenomenon for what it really is: namely, that continuous
functions are no longer just those curves that can actually be drawn. In
subsequent chapters, we remain in close contact with the intuition; for
example, for Rolle’s theorem, we extract a rigorous proof (which I would
prefer to call ‘axiomatic proof’) out of the intuitive proof. The concept of
a definite integral provides an excellent occasion for arithmetization: after
we have given an intuitive proof to the effect that the differential of the
area underneath the curve is the ordinate and we establish which intuitive
characteristics of the area have been used in this proof, we construct an
arithmetical expression that bears these characteristics (but to this end,
we still draw our ideas from the intuition).
In this manner, we grant intuition the place it deserves, without com-
promising rigor. We get even the “chemists” to understand what we are
talking about; if we schedule the available time cleverly, we can build up
weekly lectures so that part of each class is spent with intuitive discussion,
while the other part, with making concepts more rigorous, or, when this is
not possible, we provide interesting, but purely theoretical applications;
we can allow the “chemists” to skip these latter classes. The “teacher
candidates” will understand how the rigorous discussion grew out of the
intuitive content: they will recognize, from a higher perspective, the in-
tuitive facts that can be utilized in secondary schools as well. And the
“mathematicians” will learn how they can see concepts intuitively dur-
ing the creative process, and how they can present in a rigorous manner
what they have discovered intuitively. Our students will develop a greater
appreciation of the achievements of Cauchy, Weierstrass and Dedekind if
they see how these mathematicians arrived at rigorous and handy defi-
nitions of the basic concepts of calculus, if they recognize the amount of
intellectual energy that went into each definition, rather than beginning
the discussion with the finished definition.
In my examples, I have been one-sided, confining myself to elements of
calculus; mathematicians can individually think over how these principles
can be implemented in other introductory lectures. Of course, in higher
courses, one does not always have to or is able to go back to the intuition;
once our students have gotten to know the axiomatic method, they can
understand more easily in other cases, too, what it is about. However,
with a small amount of extra effort, we can always present things in such
a way that we honestly reveal how we came to realize those things, or how
we could have come to realize them, and we could wait until later to cast
the theory in its final form. It is not at all a problem – in fact, it is a good
thing – if our students eventually come away with the impression: this
was no big deal, I could have arrived at it myself. It also fits better with
the scientific perspective if we present the process of development rather
than the axiomatic theory in its finished form; for it is not the latter that
expresses the present state of science, but the fact that this is where the
developmental path has led us.
Translated by Zsófia Zvolenszky

Namenregister / Index of Names1
Ady, E., 55 Bharati, A., 37
Alexits, Gy. v., 18 Birkhoff, G., 19
Altrichter, F., 16, 241 Böhm, K., 49, 59
Ambrus, G., 19 Bohr, N., 90
Ambrus, Z., 55 Bolzano, B., 234
Andréka, H., 16, 187 Bonjour, L., 118, 120, 128
Anscombe, E., 19 Boole, G., 248
Archimedes, 136, 139 Boring, E. G., 13
Aristotle, 14 Boyle, R., 184
Armstrong, D., 115 Brandenstein, B. v., 236
Austeda, F., 76 Brentano, F., 9, 20
Avenarius, R., 103 Bridgman, P. W., 13, 193, 220
Ax, J., 258 Brown, H. R., 224
Brunswik, E., 18, 19
Baeyer, A. v., 57 Brush, S. G., 217
Baier, K., 115–117 Bühler, K., 18
Balázs, B., 12 Bunsen, R. W., 57
Baumgartner, W., 9 Butterfield, J., 17
Bavinck, B., 70
Beer, G., 12 Carnap, R., 12, 13, 17, 18, 28,
Bell, J. S., 224 30, 31, 33, 42, 67, 82, 87,
Beltrami, E., 148, 150 89, 100, 106, 107, 109–
Bencsik, B., 234, 236 111, 115, 125, 193, 235–
Benetka, G., 83 239, 241, 243–245
Benthem, J. F. van, 258 Cauchy, A.-L., 283, 287
Bergmann, G., 12 Cayley, A., 130
Bernoulli, J., 138–141, 146 Chalmers, D., 119, 120
Berthelot, M., 57 Chang, C. C., 248
Beth, E. V., 248 Church, A., 232, 233, 235
Comte, A., 49, 59
1 Nicht erfasst wurden Anmerkun- Curie, M., 54
gen, Literaturverzeichnisse, Tabel-
len und Graphiken. Notes, refer-
ences, tables and figures are not d’Alembert, J., 130, 138, 140–143
included. Davidson, D., 118, 120, 128
290 Namenregister / Index of Names
Dedekind, R., 283, 287 Gabbay, D. M., 258

Dempf, A., 69, 70 Gábor, D., 13
Descartes, R., 102 Gabriel, L., 69, 77, 79, 81, 83
Dierker, E., 18 Galilei, G., 138, 139
Du Bois-Reymond, E., 46 Gárdonyi, G., 55
Duczynska, I., 11 Gauss, C. F., 147, 148
Duhem, P., 52 Gentzen, G., 233, 234
Duncker, K., 107 Gödel, K., 12–14, 232, 233, 236,
237, 248, 258
Einstein, A., 57, 187, 191, 193–196,
Goldblatt, R., 257
199, 200, 206, 207, 209,
Goldstein, S., 19
215–217, 219, 221, 223,
Gombocz, Z., 55
224, 248, 249, 258, 261,
Goodman, N., 12
262
Grassmann, H., 130
Eötvös, L., 53, 54, 59, 200
Grünbaum, A., 193, 202, 217
Erdei, L., 240, 242
Gundel, A., 55
Erismann, T., 70
Gyenis, B., 13, 17
Euclid, 134, 273
Gyenis, Z., 17
Euler, L., 130, 134–137
Faragó, L., 38 Hahn, H., 12, 28

Fechner, G., 42 Hahn-Neurath, O., 12
Fehér, M., 241 Haller, R., 9, 69
Feigl, H., 12, 42, 43, 87, 113–115, Hamilton, W. R., 153
128 Harsany (Harsányi), J., 17, 20
Ferenczy, B., 12 Hartmann, N., 41
Feyerabend, P., 19, 20, 76, 219 Hauser, A., 12
Fischer, A., 20, 25–43 Heidelberger, M., 41, 42
Fischer, K. R., 18, 19 Hein, C., 37
FitzGerald, G. F., 191, 215, 216, Heintel, E., 69, 77–79, 81, 84
224 Heiss, G., 12
Flexner, A., 15 Helmholtz, H., 153
Fogarasi, B., 239, 240 Hempel, C. G., 108, 115, 181, 184
Fraenkel, A., 248 Hevesy, G., 58
Frank, P., 12, 19, 32, 47, 58 Heyting, A., 13
Frege, G., 82, 113, 114, 231, 243, Hilbert, D., 12–14, 155, 231, 233,
244, 248 235, 238, 248
Friedman, M., 184, 217 Hlawka, E., 76, 81
Namenregister / Index of Names 291
Hofer-Szabó, G., 17 König, J., 229

Hollitscher, W., 76 Korach, M., 11
Holton, G., 12 Korányi, F., 59
Horthy, M., 20 Korda, A., 12
Husserl, E., 234, 236, 239 Körner, T., 11, 80, 83
Kraft, V., 12, 19, 31, 34, 35, 65, 68–
Ignotus, 55 71, 73, 74, 76, 77, 81, 84,
Ilosvay, L., 57 85, 87, 88, 93, 235
Kripke, S., 243, 244
Jakab, A., 10
Kronfellner, T., 85
Jánossy, L., 203
Kuhn, T., 19, 219
Janssen, M., 217
Kvasz, L., 129
Juhos, B., 19, 20, 31, 32, 34, 65–
69, 71–76, 79–85, 87–91, Laczkó, G., 60
93, 94, 99–105, 108–113, Lagrange, J., 130, 141–145, 147,
115, 116, 118–120, 122– 148, 153
125, 127, 128 Lakatos, I., 19, 20, 87
Juhos, P., 84 Larmor, J., 224
Lehner, F., 38, 234, 236
Kac, M., 19 Leibniz, G. W., 103
Kaffka, M., 55 Leinfellner, W., 18, 85, 86, 94
Kainz, F., 69, 81 Lendvai, P., 21
Kalmár, L., 20, 230–235 Lenin, V. I., 10, 11
Kant, I., 28, 29, 42, 48, 49, 66 Limbeck-Lilienau, C., 94
Kardos, L., 18 Lorentz, H. A., 191, 192, 194, 196,
Karlik, B., 70, 71 200, 202, 204–207, 209,
Kármán, T., 19, 58 215–218, 221–224
Kasper, M., 12 Lukács, Gy. (G.), 10, 12, 55, 240,
Katzenberger, W., 84, 85, 94 241
Kaufmann, F., 12 L
ukasiewicz, J., 238
Keisler, H. J., 248
Kelsen, H., 10 Mach, E., 9–11, 20, 28, 29, 42, 46–
Kende, Zs., 11 48, 52, 57, 58, 103, 131–
Kepler, J., 136, 181, 184–186 134, 153
Kerékjártó, B., 236 Maclaurin, C., 136, 137
Köhler, E., 18, 94 Madarász, J. X., 16, 257, 264
Kókai, K., 12, 20, 21, 94 Majoros, I., 21
Kolb, E., 79 Makkai, M., 256
292 Namenregister / Index of Names
Mannheim, K., 10 Pauer, I., 38, 49

Manninen, J., 19 Pauler, Á., 49, 50, 234, 235, 239
Márkus, G., 241 Peano, G., 278
Maruzsa, Z., 21 Pecko, G., 85
Marx, K., 11 Peirce, C. S., 248
Mátrai, L., 238–240 Perecz, L., 48
McDowell, J., 118, 120, 128 Péter, R., 230, 233, 235, 241
Meister, R., 70, 81 Péter, Z., 12, 21
Menger, K., 12, 18 Pikler, Gy. (J.), 10
Mersenne, M., 138 Pittioni, V., 85, 94
Michelson, A. A., 215, 216 Planck, M., 75
Mikola, S., 51, 59 Poincaré, H., 47, 48, 52, 56, 57, 61,
Mises, R. v., 12, 19 153, 191, 206, 217
Montague, R., 243 Polányi, K., 9–11
Morgenstern, O., 13, 17 Polányi, M., 9–11
Morley, E., 215, 216 Pólya, Gy., 11, 13
Morris, C., 18, 32 Pooley, O., 224
Mundy, B., 256 Popper, K., 11, 19, 20, 31, 87, 100
Posch, J., 49
Naess, A., 18 Pozsonyi, F., 38, 234–239, 245
Nash, J., 17 Prager, W., 19
Neider, H., 12 Prandtl, L., 58
Németi, I., 16, 187 Prior, A. N., 242
Neumann, J. v., 12–15, 17, 20, 51,
230, 233, 248 Quine, W. V. O., 12, 87, 244
Neurath, O., 10, 12, 18, 46, 67,
Radakovic, T., 12, 70
85, 106–108, 111, 115, 118,
Radon, J., 81
125, 241
Rand, R., 12
Newton, I., 130–135, 147, 149, 153,
Rathkolb, O., 12, 21
185, 186
Rédei, M., 13, 14, 17
Novotny, H., 76
Reichenbach, H., 9, 16, 19, 193, 238,
Nyı́ri, J. K., 55
241, 248, 249, 256
Oppacher, F., 85, 86, 94 Reidemeister, K., 12
Reininger, R., 31, 66, 67
Palágyi, M., 49 Reiter, W., 19
Palló, G., 21 Révész, G., 9
Pap, A., 20 Riehl, A., 41–43, 103
Namenregister / Index of Names 293
Riemann, B., 130 Stifel, M., 137

Robb, A. A., 256 Stumpf, C., 9, 28, 29
Rohracher, H., 70, 81, 83 Szabó, T., 236
Rolle, M., 284, 287 Szabó, L. E., 16, 17, 196, 213, 258
Rorty, R., 122 Szegő, G., 19
Rosenblüth, A., 12 Szegedi, P., 10
Russell, B., 14, 82, 87, 114, 231, Székely, G., 16
238, 239, 248 Szilárd, B., 57
Ruzsa, I., 241–245
Tarski, A., 12, 235–237, 248, 256–
Schenk, E., 77 258
Schleichert, H., 85, 86, 93, 94 Than, K., 59
Schlick, M., 9, 12, 28–31, 33, 35, 38, Thirring, H., 70, 71, 81
41–43, 66, 67, 70, 73, 74, Tiemeyer, R., 264
76, 77, 85, 87, 99, 100, 102, Topitsch, E., 35, 69
108–114, 120, 122, 124,
Urban, P., 36
125, 127, 128, 238, 241
Schöndorfer, U., 69 Waismann, F., 12, 76–79
Schopenhauer, A., 66, 67 Weibel, P., 21
Schorner, M., 94 Weierstrass, K., 283, 287
Schröder, E., 248 Weizsäcker, K. F. v., 77–79
Schrödinger, E., 70, 71 Wessely, A., 21
Schumpeter, J., 13 Whitehead, A. N., 82, 87, 231, 248
Schurz, G., 93 Wigner, E., 17, 51
Schütz, A., 238 Wilde, J., 12
Sebestyén, Gy., 21 Wittgenstein, L., 19, 20, 28, 119,
Seeber, U., 12 121, 122, 125, 131, 143
Sellars, W., 118, 119, 128 Wright, G. H. v., 77
Selten, R., 17
Sexl, R., 36 Zahar, E., 217
Sigmund, K., 18 Zalai, B., 60
Smart, J., 115–118, 128 Zeman, V., 20
Somogyi, J., 38 Zemplén, Gy., 45, 48, 52–61
Spann, O., 11 Zermelo, E., 248
Stadler, F., 9, 11, 12, 18, 19, 65, 94 Zilsel, E., 107
Stegmüller, W., 83, 86 Zipernovszky, K., 59
Stevens, S. S., 13
Die Autorinnen und Autoren / The Authors
Gergely Ambrus
Associate professor of philosophy at the University of Miskolc, Hungary.
He studied physics, aesthetics and philosophy at the Eötvös Loránd Uni-
versity in Budapest. Besides Miskolc University, he also taught at the
Eötvös Loránd University and at the University of Liverpool. His re-
search centers on the philosophy of mind and the history of analytic phi-
losophy, with an emphasis on the Vienna Circle, logical empiricism, and
the origins of analytic philosophy; he is also interested in metaphysics and
the philosophy of science. He published a book on the current debates
on consciousness (The Metaphysics of Consciousness, Budapest, 2007) a
book-length essay on the history of analytic philosophy, as well as papers
on personal identity, the self, perception, the mind-body problem, the
metaphysics of modalities, and topics from Russell, Carnap, Feigl, Bran-
dom. Currently he is working on a larger project about the philosophy of
psychology of logical empiricism.
Hajnal Andréka
PhD with Faculty of Mathematics, Eötvös Loránd University Budapest
1975, Doctor of Science with the Academy 1992. Has been working in
Rényi Institute of Mathematics of the Hungarian Academy of Sciences
since 1977, presently as head of department. Member of the Executive
Board of European Foundation for Logic, Language and Information 1991-
1997, member at large of the Council of the Association of Symbolic Logic
1999-2001. Close working connection with Tarski’s research school in
algebraic logic and logic in general, joint books: [Henkin, L., Monk, J. D.,
Tarski, A., Andréka, H., Németi, I.: Cylindric Set Algebras, Lecture Notes
in Mathematics 1981], [Andréka, H., Givant, S. R., Németi, I.: Decision
problems for equational theories of relation algebras. Memoirs of Amer.
Math. Soc. 1997]. Over 150 research papers in leading scientific journals.
Research interests: mathematical logic, relativity theory, foundations of
mathematics.
Károly Kókai
Geboren 1959 in Budapest, lebt seit 1981 in Österreich, Studium der
Philosophie und Kunstgeschichte an der Universität Wien, Lehraufträge
296 Die Autorinnen und Autoren / The Authors
in Wien und Klausenburg, Mitarbeiter des Instituts Wiener Kreis. Pub-

likationen u.a. Művek, Kiállı́tások, Írások, Budapest, 1999; Im Nebel.
Der junge Georg Lukács und Wien, Wien 2002; Visual Culture (Heraus-
geber), Budapest 2005; Habitus, Identität und die exilierten Dispositionen
(Mitherausgeber), Budapest 2008.
Ladislav Kvasz
Graduated in 1986 in mathematics at the Comenius University in Brati-
slava. After 1989 he began graduate studies in philosophy at the Comenius
University in Bratislava. In May 1995 he defended his thesis “Classifica-
tion of Scientific Revolutions”. Since 1986 he has been employed at the
Faculty of Mathematics and Physics of Comenius University, Bratislava
as a lecturer. In 1993 he won the Herder Scholarship and spent the aca-
demic year 1993/94 at the University of Vienna studying philosophy of the
Vienna Circle and of Wittgenstein. In 1995 he won the Masaryk Schol-
arship of the University of London and spent the year academic 1995/96
at King’s College London working with professor Donald Gillies. In 1997
he won the Fulbright Scholarship and spent the summer term of the aca-
demic year 1998/99 at the University of California at Berkeley, working
with Professor Paolo Mancosu. In 2000 he won the Humboldt Scholarship
and spent the years 2001 and 2002 at the Technical University in Berlin
working with Professor Eberhard Knobloch. His research has been in the
history and philosophy of mathematics, with particular interest in its cul-
tural background in arts, literature, and theology. He was the co-editor
of Appraising Lakatos (Kluwer 2002) and author of Patterns of Change
(Birkhauser 2008).
Judit Madarász
Research fellow at Alfréd Rényi Institute of Mathematics (Budapest) since
1998. Main fields of interests: foundations of relativity theories, logic,
algebraic logic. University: Eötvös Loránd University Budapest, 1990–
1995. PhD Dissertation: Logic and Relativity in the light of definability
theory (written in 2002, degree awarded in 2003).
András Máté
Studied mathematics and philosophy at the Eötvös University Budapest
(Hungary). He began his research in logic and its history as an assis-
tant of Imre Ruzsa. He is currently associated professor of logic at the
Philosophical Institute of the Eötvös University. He made his PhD (CSc)
at the Hungarian Academy of Sciences about Plato and Frege. His re-
search interests include history of logic and semantics (semantical ideas
Die Autorinnen und Autoren / The Authors 297
in Plato’s dialogues, Stoic logic, medieval semantics, Leibniz, Bolzano,

Frege) and philosophy of mathematics (second-order logic as a framework,
philosophical ideas of 20th century Hungarian mathematicians). He wrote
four textbooks of logic and its history and several papers about different
topics including even aesthetics of music in Hungarian, 14 papers in Ger-
man and English mainly about the history of logic. He translated works
by Plato, Frege, Tarski, Kneale & Kneale.
István Németi
PhD with Faculty of Mathematics, Eötvös Loránd University Budapest
1978, Doctor of Science with the Academy 1987. Has been working in
Rényi Institute of Mathematics of the Hungarian Academy of Sciences
since 1977, presently as scientific advisor. Szécheny Professorship 2000–
2003. Longer scientific visits: University of Waterloo (Canada) 1983, Iowa
State University (USA), University of California at Berkeley (USA), Uni-
versity of Boulder (USA) 1988, Banach Mathematical Center in Warsaw,
Poland 1991, University of Amsterdam (The Netherlands) 1998. Close
working connection with Tarski’s research school, joint book: Henkin, L.
Monk, J. D., Tarski, A., Andréka, H., Németi, I.: Cylindric Set Algebras,
Lecture Notes in Mathematics, 1981. Over 150 research papers in leading
scientific journals. Research interests: logic, mathematical logic, relativity
theory, foundations of science.
Péter Németi
Active member of the Algebraic Logic seminar of the Rényi Institute of
Mathematics, Budapest. Background: information technology, computer
science. Recent research interests: relativistic hypercomputing, relativity
theory, cosmology, logic.
Some recent work:
General relativistic hypercomputing and foundation of mathematics. Nat-
ural Computing 8,3 (2009), 499-516. With Andréka, H. and Németi, I.
A logic road from special relativity to general relativity. Synthese, invited
paper. With Andréka, H., Madarász, J. X., Németi, I. and Székely, G.
Relativistic hypercomputing. Physical Realisticity. Presentation at the
conference Unconventional Computation 09, Azores, September 2009.
With Andréka, H. and Németi, I. 63 dias.
Breaking the Turing barrier via GR. Relativistic hyper computing. Invited
talk at Physics and Computation, Workshop in Unconventional Compu-
tation 08 (UC08), Vienna, August 25-28, 2008. With Andréka, H. and
Németi, I.
General relativity, black hole physics – a logical analysis. Invited talk at

Conference on Logic and the Foundations of Physics, Brussels, Belgium,
December 11-12, 2008. With Németi, I.
Miklós Rédei
Szechenyi Professor of Philosophy (ELTE Budapest) 1999–2002. Cur-
rently Lecturer in Philosophy at the Department of Philosophy, Logic
and Scientific Method, London School of Economics and Political Science.
Research interests: Foundational and philosophical problems of modern
physics, quantum logic, general issues in philosophy of science. Publica-
tions include: (Editor) John von Neumann: Selected Letters, History of
Mathematics, vol. 27 (American Mathematical Society and London Math-
ematical Society, 2005); (Editor with M. Stöltzner) John von Neumann
and the Foundations of quantum Physics, Kluwer Academic Publishers,
Dordrecht–Boston–London, 2001; Quantum Logic in Algebraic Approach
(Fundamental Theories of Physics, vol. 91), Kluwer Academic Publishers,
Dordrecht–Boston–London, 1998; Introduction to Quantum Logic, Eötvös
University Press, Budapest, 1995. Papers (selection): “Quantum Proba-
bility theory” (with S. J. Summers), Studies in the History and Philosophy
of Modern Physics (forthcoming); “The birth of quantum logic”, History
and Philosophy of logic 28 (2007), 107-122; “Reichenbachian Common
Cause Systems of arbitrary finite size exist” (with G. Hofer-Szabo), Foun-
dations of Physics Letters 35 (2006), 745-746.
Wolfgang L. Reiter
Born 1946 in Bad Ischl, Austria, Honorary Professor at the Faculty of
Historical and Cultural Studies of the University of Vienna, co-founder
and vice president of the Erwin Schrödinger International Institute for
Mathematical Physics, studied physics, mathematics and philosophy and
received his Ph.D. degree in nuclear physics from the Institut für Radi-
umforschung und Kernphysik at the University of Vienna in 1974. Until
recently he was director of the natural sciences unit at the Austrian Fed-
eral Ministry for Education, Science and Culture. His interests are in the
history of physics and the forced emigration of scientists from Austria.
Friedrich Stadler
Professor für History and Philosophy of Science an der Universität Wien.
Begründer und seitdem wissenschaftlicher Leiter des Instituts Wiener
Kreis. Gastprofessuren an der Humboldt-Universität Berlin und an der
University of Minnesota (Minneapolis), zuletzt 2006/07 Fellow am Helsinki
Die Autorinnen und Autoren / The Authors 299
Collegium for Advanced Studies der Universität Helsinki. Seit Oktober

2009 Präsident der European Philosophy of Science Association (EPSA).
Publikationen zur Wissenschaftsgeschichte und Wissenschaftstheorie so-
wie zur Intellectual History (Exil und Wissenschaftsemigration) und his-
torischen Wissenschaftsforschung. Publikationen: Vom Positivismus zur
wissenschaftlichen Weltauffassung (1982), Studien zum Wiener Kreis
(1997), engl.: The Vienna Circle (2001); Reihenherausgeber: ,,Veröffentli-
chungen des Instituts Wiener Kreis“ (HPT/Springer 1991ff.), ,,Vienna
Circle Institute Yearbook“ (Kluwer/Springer 1993ff.), ,,Moritz Schlick
Gesamtausgabe“ (Springer 2006ff.), ,,Emigration – Exil – Kontinuität“
(LIT Verlag 2004ff.), ,,Ernst Mach Studienausgabe“ (xenomoi 2008ff.).
Born 1954, Hungarian physicist and philosopher. He studied physics and
philosophy in Budapest; PhD (physics) in 1985. He received Habilitation
(philosohy) in 2004; DSc (philosophy) in 2005. For a long period he had a
position at the Department of Theoretical Physics at Eötvös University in
Budapest. Recently he is a professor of philosophy at the Department of
Logic in the Institute of Philosophy of Eötvös University. His main field
of interest is philosophy of science; in particular, the philosophy of space
and time, causality, the EPR–Bell problem and foundations of quantum
mechanics, determinism vs. indeterminism, interpretation of probability,
and a physicalist account of mathematics.
Gergely Székely
Master’s degree in Mathematics at Eötvös Loránd University, Budapest
in 2004. PhD in Mathematics at Eötvös Loránd University, Budapest
in 2010. Working in Rényi Institute of Mathematics of the Hungarian
Academy of Sciences since 2008, presently as an assistant research fellow.
Member of “Logic and Relativity theory” research school led by Haj-
nal Andréka and István Németi since 2003. Both his Master’s and PhD
dissertations are in this field with titles “A First-Order Logic Investiga-
tion of the Twin Paradox and Related Subjects” and “First-Order Logic
Investigation of Relativity Theory with an Emphasis on Accelerated Ob-
servers,” respectively. 8 publications, among others in Studia Logica and
Foundations of Physics, 11 talks in international conferences, 4 of which
were invited ones. Research interests: mathematical logic, foundations of
spacetime theories, relativity theory.
Gábor Á. Zemplén

PD, PhD. He holds a position at BME (Budapest University of Technol-
ogy and Economics), and held courses at ELTE (Budapest, HU), Univer-
sität Bern (CH), and MOME (Budapest). He spent two years in Munich
(Deutsches Museum) as a predoctoral fellow, a year and a half in Berlin
(Max Planck Institut für Wissenschaftsgeschichte) as a postdoctoral stu-
dent, and is currently junior fellow of the Collegium Budapest (Hungary).
His main research areas and specific research topics are history of science
(mostly 17–19th c. light-, colour-, and vision- theories), history of phi-
losophy of science (methodological development in the 17th c., history of
turn-of-the-century philosophy of science), the study of scientific contro-
versies (also as editorial board member of the journal Argumentation),
argumentation-theory (especially dialectical models),and science educa-
tion (focusing on teaching critical thinking, argumentation, and nature of
science in high school and university curricula).
SpringerPhilosophie
Veröffentlichungen des Instituts Wiener Kreis
K. R. Fischer, F. Stadler (Hrsg.) Friedrich Stadler (Hrsg.)

„Wahrnehmung Elemente moderner
und Gegenstandswelt“ Wissenschaftstheorie
Zum Lebenswerk von Egon Brunswik Zur Interaktion von Philosophie,
(1903–1955) Geschichte und Theorie
der Wissenschaften
1997. 187 Seiten. 15 Abb. 1 Frontispiz. Text: d/e
Brosch. € (D) 34,95 | € (A) 35,93 | *sFr 47,00 2000. XXVI, 220 Seiten. 16 Abb.
ISBN 978-3-211-82864-9. Bd. 4 Brosch. € (D) 44,95 | € (A) 54,44 | € (A) | *sFr 60,50
ISBN 978-3-211-83315-5. Bd. 8

Bausteine wissenschaftlicher C. Jabloner, F. Stadler (Hrsg.)
Weltauffassung Logischer Empirismus
Lecture Series/Vorträge des und Reine Rechtslehre
Instituts Wiener Kreis 1992–1995 Beziehungen zwischen dem
Wiener Kreis und der Hans Kelsen-Schule
1997. 231 Seiten. Text: d/e
Brosch. € (D) 33,00 | € (A) 33,00 | *sFr 44,50 2001. XXI, 339 Seiten.
ISBN 978-3-211-82865-6. Bd. 5 Brosch. € (D) 66,95 | € (A) 68,83 | *sFr 90,00
ISBN 978-3-211-83586-9. Bd. 10

Phänomenologie
und logischer Empirismus
Zentenarium Felix Kaufmann (1895–1949)
1997. 163 Seiten. 1 Frontispiz. Text: d/e

Brosch. € (D) 32,95 | € (A) 33,87 | *sFr 44,50
ISBN 978-3-211-82937-0. Bd. 7
W
Wien, Österreich, Fax +43.1.330 24 26, order@springer.at, springer.at
Heidelberg, Deutschland, Fax: +49.6221.345-4229, orders-HD-individuals@springer.com
Preisänderungen und Irrtümer vorbehalten. *Unverbindliche Preisempfehlung.
SpringerPhilosophie
Veröffentlichungen des Instituts Wiener Kreis
E. Timms, J. Hughes (Hrsg.) Anne Siegetsleitner (Hrsg.)

Intellectual Migration Logischer Empirismus, Werte
and Cultural Transformation und Moral
Refugees from National Socialism Eine Neubewertung
in the English-Speaking World
2010. 256 Seiten.
2003. VI, 267 Seiten. Text: englisch Brosch. € (D) 43,72 | € (A) 44,95 | *sFr 59,00
Brosch. *€ (D) 34,24 | € (A) 35,20 | sFr 46,00 ISBN 978-3-7091-0159-9. Bd. 15
ISBN 978-3-211-83750-4. Bd. 12
Friedrich Stadler
The Vienna Circle –
E. Nemeth, N. Roudet (Hrsg.)
Studies in the Origins,
Paris – Wien
Development, and Influence
Enzyklopädien im Vergleich
of Logical Empiricism
2005. 262 Seiten. Text: dt./fr./engl.
2001. XV, 984 Seiten.
Brosch. € (D) 39,00 | € (A) 39,00 | *sFr 52,50
Geb. * € (D) 96,25 | € (A) 98,95 | sFr 129,00
ISBN 978-3-211-21538-8. Bd. 13
ISBN 978-3-211-83243-1. Sbd.
F. Stadler, K. R. Fischer (Hrsg.)

A. Müller, K. H. Müller,
Paul Feyerabend
F. Stadler (Hrsg.)
Ein Philosoph aus Wien
Konstruktivismus und
2006. XXXIV, 160 Seiten. Text: dt/engl. Kognitionswissenschaft
Brosch. € (D) 36,95 | € (A) 37,98 | *sFr 49,50
ISBN 978-3-211-29759-9. Bd. 14 2., aktualisierte und erw. Aufl.
2001. 308 Seiten. 22 Abb.
Brosch. * € (D) 45,96 | € (A) 47,25 | sFr 62,00
ISBN 978-3-211-83585-2. Sbd.
W
Wien, Österreich, Fax +43.1.330 24 26, order@springer.at, springer.at
Heidelberg, Deutschland, Fax: +49.6221.345-4229, orders-HD-individuals@springer.com
Preisänderungen und Irrtümer vorbehalten. *Unverbindliche Preisempfehlung.

The Vienna Circle in Hungary/Der Wiener Kreis in Ungarn

Hochgeladen von

Dokumentinformationen

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

The Vienna Circle in Hungary/Der Wiener Kreis in Ungarn

Hochgeladen von

Copyright:

Verfügbare Formate

W

Der Wiener Kreis in Ungarn

Gedruckt mit Unterstützung der Aktion Österreich-Ungarn

Das Werk ist urheberrechtlich geschützt.

Produkthaftung: Sämtliche Angaben in diesem wissenschaftlichen Werk erfolgen trotz

SpringerWienNewYork ist ein Unternehmen von

Satz/Layout: Reproduktionsfertige Vorlage des Instituts Wiener Kreis (Robert Kaller)

Gedruckt auf säurefreiem, chlorfrei gebleichtem Papier

Bibliografische Information der Deutschen Nationalbibliothek

ISBN 978-3-7091-0176-6 SpringerWienNewYork

Miklós Rédei and Friedrich Stadler

Gábor Á. Zemplén

Judit X. Madarász and Gergely Székely

H. Andréka, J. X. Madarász, I. Németi, P. Németi, G. Székely

Index of Names / Namenregister . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289

Die Autorinnen und Autoren / The Authors . . . . . . . . . . . . . . . . . 295

This publication brings together the articles contributed to a bilateral

Vienna, Budapest and London February 2011

Austria-Hungary in Philosophy and Science:

Contrary to the well-known joke in football concerning teams on the match

shortcomings. In the centre stood the economist and physiologist Gyula

lat ), a Library of the Galilei Circle (a Galilei Kör könyvtára), as well as

In considering the Hungarian emigration to Austria after the First

Bridgman, E. G. Boring, S. S. Stevens, J. Schumpeter, and the Hungarian

completeness theorem. Von Neumann therefore decided not to publish

Thus I am today of the opinion that

Indeed Gödel was more careful:

I wish to note expressly that Theorem XI [the second incompleteness theorem]

Strongly formulated disagreements with Gödel can also be found in

in his letter to A. Flexner (September 27, 1939), in which von Neumann

While the von Neumann-Gödel debate is perhaps the most fascinating

claims a theoretical move that arguably might be regarded as the amal-

in this spirit on ‘A Preference-Based Theory of Well-Being and a Rule-

chology (1930), before he was given the task in 1931/32 of setting up an

• Before 1914 there already existed a concrete reception of philosophy

• A long neglected emigration after World War I from Hungary to the

• Between 1918 and 1933/34 we see mutual contacts and collabora-

• From 1933/34 both trends are already manifestations of ‘science in

• It is notable that after 1938 there are isolated instances of continu-

F. Altrichter, editor (1972), A Bécsi Kör ﬁlozóﬁája, Budapest, . Gondolat.

Wilhelm Baumgartner / Franz-Peter Burkhard / Franz Wiedmann / Chri-

G. Hofer-Szabó and M. Rédei. “Reichenbachian common cause systems

Miklós Rédei/Michael Stöltzner (2000) (Eds.), John von Neumann and

Anton Fischers philosophische Grundlagen der

Es gibt zwei Arten von wissenschaftlicher Erkenntnis. Zum einen ist es

wurde er an der Universitätsklinik in Budapest zum Dozenten, später

Fischers Zugang wirkt, angesichts der Tatsache, dass Physik in der

4 Fischer 1947, S. 151.

Da die Grundlagen als Einführung verfasst sind, systematisiert Fischer

Das Verhältnis zum Wiener Kreis

5 Fischer 1947, S. 56ﬀ.

An dieser Ansicht übt Fischer Kritik. Die Verkoppelung von Logistik

7 Fischer 1947, S. 58.

Insbesondere scheint ihm viel an der Positivismuskritik zu liegen. So

Zu einem System der Erkenntnis wurden die Relationen durch R. Carnap in

Nach M. Schlick (Allg. Erkenntnislehre 2. Auﬂ. 1925) müsste der Positivis-

10 Fischer 1947, S. 58.

Durchführung seines Programmes würde einen Verzicht auf Erkenntnis über-

Die vollkommene Erkenntnis des Seienden“ scheine in unerreichbarer Fer-

14 Fischer 1947, S. 165.

Fischer spricht hier das in einschlägigen philosophischen Diskussionen wie-

18 Fischer 1967, S. 133f.

So viel zu den Philosophischen Grundlagen, Auﬂagen 1 und 2. Es ist

22 Als Ausnahmen sind anzuführen: Friedrich Stadler, Paul Feyerabend. Ein