Diagonal Method and Dialectical Logic Bo

DIAGONAL METHOD
AND
DIALECTICAL LOGIC
Tools, Materials, and Groundworks

for a Logical Foundation of Dialectic
and Speculative Philosophy
BOOK TWO:
HISTORICAL-PHILOSOPHICAL
BACKGROUND MATERIALS
Uwe Petersen
Contents
BOOK TWO: HISTORICAL-PHILOSOPHICAL

BACKGROUND MATERIALS 705
Part C. DIALECTIC, IDEALISM, AND

THE TRANSCENDENTAL APPROACH 709
Chapter XIV. PREPARATIONS 712

§56. Methodological preliminaries 712
56a. General remarks 712
56b. Concerning sources 714
§57. Ancient roots of the notion of dialectic 722
57a. Parmenides on being and nothing 722
57b. Platon’s dialectic 726
57c. Aristoteles 729
§58. The British empiricists 740
58a. Locke 740
58b. Berkeley 744
58c. Hume 749
Chapter XV. KANT’S TRANSCENDENTAL PHILOSOPHY 760

§59. Introduction 760
59a. Kant’s starting point 761
59b. Kant’s assessment of logic and mathematics 764
59c. The master-plan 767
§60. Forging the tools 773
60a. A priori knowledge, transcendental philosophy, and metaphysics 773
60b. Transcendental idealism — empirical realism 777
60c. The analytic-synthetic distinction 781
§61. The conception of critique 786
61a. The general structure of knowledge and the division of logic 786
61b. The forms of intuition 790
61c. Understanding and the categories 794
61d. Reason and ideas 799
61e. The role of a transcendental deduction of the categories and the
transcendental unity of apperception 801
61f. Phenomena and noumena: Kant’s concept of the thing in itself 809
vii
viii CONTENTS
§62. The Antinomy of Pure Reason 817

62a. The system of possible antinomies 817
62b. The logical structure of the antinomies 819
62c. The critical decision regarding the conflict of reason with itself 820
62d. The different resolutions of the mathematical-transcendental ideas
and the dynamical ones 825
Chapter XVI. HEGEL’S SPECULATIVE PHILOSOPHY 830

§63. From Kant to Hegel 831
63a. Fichte’s reception of Kant’s project 831
63b. The assault on the concept of the thing-in-itself 833
63c. Some of Hegel’s roots in ancient Greek philosophy 836
63d. Hegel’s logical turn: logic as an organon 839
§64. Hegel’s speculative philosophy as an extension of Kant’s transcendental
philosophy 844
64a. Hegel on Kant and Fichte as regards the categories 844
64b. Hegel on Kant as regards the antinomies, their solution, and
dialectic 850
§65. Logic, truth, and the content of thought determinations 856
65a. The thing in itself as regards logic 856
65b. Hegel’s conception of logic and metaphysics 858
65c. Truth, logic, form, and content 861
65d. Content of thought determinations on their own 866
§66. Contradictions and their Aufhebung 871
66a. Hegel on contradictions 871
66b. Negative dialectic 873
66c. Aufhebung, positive dialectic, and the speculative step 878
66d. Dialectic, method, and encyclopedia 886
Chapter XVII. AFTER THE GREAT MASTERS 894

§67. General considerations 895
67a. Aspects of the historical reception of ancient dialectic 896
67b. The challenge of the ancient Greeks 898
67c. Aspects of metaphysics and epistemology 905
67d. The concept of dialectic 910
§68. Struggling with transcendental philosophy 918
68a. How many Kants? 918
68b. Transcendental idealism, the analytic-synthetic distinction, and
universally valid knowledge 922
68c. The transcendental unity of apperception and the system of
categories 925
68d. The phenomena-noumena distinction, the notion of the thing in
itself, and the antinomies 929
§69. Struggling with speculative philosophy 940
69a. Science of Logic, dialectic, and the idealist conception 941
69b. Logic and Hegel’s approach to contradictions and their Aufhebung 954
CONTENTS ix
69c. Forms of thought and the idealistic conception 962

69d. Thought determinations, truth, and logic 965
69e. The ‘non-metaphysical reading’ of Hegel 971
Part D. THE PHILOSOPHY OF MATHEMATICS AND LOGIC 979
Chapter XVIII. THE EMERGENCE OF ANTINOMIES IN THE

FOUNDATIONS OF MATHEMATICS 982
§70. Ways of axiomatics 982
70a. Eukleides’ elements, the idea of axiomatics, and the unique case of
mathematical knowledge 982
70b. Leibniz’ characteristica universalis 987
70c. The development of non-Euclidean geometries 995
70d. Formal axiomatics 998
§71. Logicism — the ultimate foundation 1003
71a. The rise of the reductionist conception of foundation 1003
71b. The idea of a logical foundation of arithmetic 1008
71c. The making of a new logic 1013
71d. The making of “Grundgesetz V” 1021
71e. On the philosophical significance of the logical turn in mathematics
and Frege’s logicism 1027
§72. The discovery of paradoxes 1037
72a. From transfinite cardinals to Russell’s antinomy 1038
72b. Paradoxes proliferating 1039
Chapter XIX. EARLY REACTIONS TO THE PARADOXES 1042

§73. First reactions 1042
73a. Frege’s response 1042
73b. Russell’s Principles 1047
73c. Cantor’s own view 1049
73d. Blaming the actual infinite 1052
§74. Actual infinity, immutable classifications, and non-predicative
definitions: Poincaré’s analysis of the paradoxes 1054
74a. Infinity, classifications, and logic 1054
74b. Immutable classifications, predicative definitions, and sets 1057
§75. The vicious circle principle: Russell’s analysis of the paradoxes 1060
75a. Generalized propositions 1060
75b. Russell’s conception of vicious circles 1061
75c. From totalities to functions 1062
75d. Theory of description and no-class theory 1063
75e. From the vicious circle principle to the hierarchy of propositions
and propositional functions 1065
§76. Axiomatic set theory, formalism, and proof theory 1067
76a. Zermelo’s axioms for set theory 1067
76b. Hilbert’s proof theory 1069
76c. Induction, recursion, and the problem of impredicativity 1077
x CONTENTS
76d. Brouwer’s neo-intuitionism 1079
Chapter XX. FORMALIZATION, METAMATHEMATICS, AND THE

ESTABLISHMENT OF THE FIXED POINT PROPERTY 1082
§77. Easing the grip of orders and types 1082
77a. Non vicious circles: Ramsey’s reconsideration of Russell 1082
77b. Quine’s New Foundations 1087
77c. Gödel on Russell’s theory of types 1088
§78. Hilbert’s province 1092
78a. Philosophical issues 1092
78b. Formalization 1093
78c. Normal forms, decidability, and completeness 1096
78d. Consolidation of Zermelo’s set theory 1097
78e. Type free logics 1099
§79. Undecidability, incompleteness, and the notion of computability 1101
79a. The rise of self-reference in formalized theories 1102
79b. Church, Rosser, Kleene 1106
79c. Semantics and the concept of truth in formalized theories 1108
§80. Proof theory, ordinal analysis, and truth definitions 1110
80a. From axiomatics to the calculus of sequents 1111
80b. Cut elimination and consistency proofs 1112
80c. ‘Finitary means’, Gödel’s second theorem and ordinal analysis 1114
Chapter XXI. PHILOSOPHY OF MATHEMATICS AND LOGIC TODAY 1117

§81. The general issue of foundation 1117
81a. Looking back 1118
81b. Mathematics, realism, and the external world 1124
81c. Logicism, intuitionism, and formalism 1130
§82. Self-reference, the vicious circle principle, and the theory of types 1134
82a. Preliminary remarks 1134
82b. Transfinite types 1135
82c. The vicious circle principle 1136
§83. Set theory 1143
83a. Aspects of the reception of the antinomies today 1143
83b. The iterative conception of set 1151
83c. Set theory, higher order logic, and the foundations of mathematics 1155
§84. Metamathematics 1160
84a. Quantification, theoretical constants, and some philosophy of logic 1160
84b. Model theory and completeness 1166
84c. Self-reference, the fixed point property, and incompleteness 1168
84d. Evaluating incompleteness and undecidability results 1172
§85. Proof theory 1176
85a. Looking back 1176
85b. Formalism and the conception of proof theory 1179
85c. Cut elimination, consistency, and the finitist standpoint 1184
85d. Cut elimination, completeness, and decidability 1191
CONTENTS xi
85e. Cut elimination and the ω-rule 1193

85f. Proof theory and meaning 1194
85g. Structural rules, computing, and the nature of logic 1195
Part E. OUTSIDE THE TRANSCENDENTAL TRADITION 1199
Chapter XXII. ANALYTIC EMPIRICISM 1202

§86. From logicism to logical empiricism 1202
86a. From logicism to logical atomism 1203
86b. The philosophical significance of logicism for empiricism 1211
86c. Antinomies, the doctrine of types, and logical spheres 1215
86d. Tarski’s concept of truth 1216
86e. Fear of metaphysics, inquisition of language, and worship of science 1217
§87. After euphoria 1223
87a. The analytic-synthetic distinction 1224
87b. The three dogmas of logical empiricism 1224
87c. Further down the track 1228
§88. The rise of common sense and ordinary language, and the decline of
logical empiricism 1230
88a. Psychology and the rule of objects 1230
88b. Common sense and intuition 1232
88c. Ordinary language and linguistic analysis 1232
§89. Truth and reference 1237
89a. The semantic notion of truth 1237
89b. Truth, correspondence, and the problem of reference 1242
§90. Logic, ontology, and flight from intension 1244
90a. Quantification, variables, and ontology 1244
90b. Possible worlds, referential transparency and extensional opacity 1248
90c. Description, identity, and substitutivity 1254
90d. Intentionality 1257
§91. The return of metaphysics 1258
91a. Notions of realism 1258
91b. Reference and realism 1263
91c. Epistemology 1268
Chapter XXIII. PARADIGMS OF METHOD AND CONFIRMATION 1270

§92. General considerations 1270
92a. Text, context, pretext 1271
92b. Common sense, ordinary language, and introspection 1273
92c. Authority, law, and custom 1277
92d. Performative inconsistency, psychiatry, and science fiction 1279
92e. Reductionism, scientific revolution, and academic reality 1282
§93. The paradigm of physics 1286
93a. Galileo Galilei 1287
93b. Absolute space and time: Newton’s mechanics 1289
93c. Relativity 1291
xii CONTENTS
93d. Uncertainty and complementarity 1294

§94. The foundations of mathematics as a philosophical challenge 1304
94a. The mathematical method: discovery vs. invention 1304
94b. The issue of foundation 1310
94c. Philosophical quibbles 1312
§95. The paradigm of logic 1315
95a. The resilience of the Aristotelian paradigm 1315
95b. The use of logical tools 1318
95c. Employing logical and mathematical results and/or concepts 1322
95d. Incompleteness, undecidability, and the interconnectedness of
everything — after dinner 1325
95e. Self-reference and diagonalization 1332
Chapter XXIV. TRUTH, PARADOXES, AND NON-CLASSICAL
LOGICS 1338
§96. General aspects 1338
96a. Bivalence and dissatisfaction with classical logic 1339
96b. Truth, paradoxes, and partial recursive functions 1341
96c. Many-valued logics 1345
96d. Questions of method 1347
§97. Paraconsistent logic and dialetheism 1352
97a. The ‘intuitive’ background 1352
97b. Claims 1355
§98. Quantum logic 1360
98a. Intuitive background 1360
98b. Aspects of the logic of quantum logic 1362
§99. Substructural logics 1363
99a. Intuitive background 1363
99b. Contraction free logics 1364
99c. Linear logic 1365
99d. Contraction-free logic and unrestricted abstraction 1368
BOOK TWO
HISTORICAL-PHILOSOPHICAL
BACKGROUND MATERIALS
INTRODUCTORY COMMENTS TO BOOK TWO
It is conceivable that a well-organized collection

of aptly selected quotations with appropriate comments
can be a reasonably effective way
to express one’s comprehensive view1
This second book of my study concerning Diagonal Method and Dialectical

Logic is dedicated to the historical background of ideas and questions which set
the frame for my attempt at bringing Cantor’s diagonal method and Hegel’s idea
of dialectic together.
The questions with which I am concerned in this study have deep and widely
spread roots in the various branches of philosophy. Out of these, I shall only pick
certain aspects; I am not aiming at completeness. I shall not try to be exegetical;
I am just trying to give an account of the ideas that have fed into my notion of
dialectic. Secondary literature is not used for the purpose of elaboration, but for
documentation, if only of a particularly silly position.
Unlike the tools of book one, the content of the present book is not provided
for employment in demonstration, like methods of proof in the metatheory of a logic,
but rather for illustration, or illumination. Its concern is original (or translated from
original) formulations of ideas, not a set of nicely applicable tools. It is a collection of
quotations as materials for the purpose of providing an historical-theoretical back-
ground of the different components: from empiricism to transcendental philosophy
and from logicism to proof theory. I shall come back to it in the groundworks
when I develop my reasons for focusing ultimately on as abstract and unfamiliar a
phenomenon as the fixed point property in higher order logic. What I present in the
groundworks is a new way of dealing with old problems. In order to understand
what I am doing, it is, at least, helpful to understand the many ways in which these
old problems have surfaced since the times of ancient Greek philosophy.
Philosophical questions seem to be more intimately linked to the course of
history than those of mathematics, if only because they are lacking the kind of the-
oretical development that is characteristic of exact disciplines like mathematics and
physics, for instance, and which removes a contemporary formulation of a law like
that of gravity from the way it was originally formulated. Philosophical questions
only seem to survive in the specific forms that authors have given to them. The
present book tries to preserve some of this proximity by presenting materials in a
roughly historical order. As a result, theoretical coherence in the presentation of
mathematical-logical ideas cannot always be maintained; e.g., Montague’s modal
antinomy, an antinomy based on the examination paradox, which was only given
a precise formulation in 1963, is presented elsewhere,2 although from a theoretical
1 Wang [1987], pp. 254 f. This was once my aim, but I have not achieved it to my full satis-
faction. See quotation 56.1 below for the continuation of this quote.
2 Cf. 48.33iv in the tools.
point of view it would fit well with the other so-called semantic paradoxes which,
however, were discussed 30–60 years earlier and which are here presented in §72.
The present materials consist of three parts, a division which represents — to
some extent, at least — the threefold philosophical background of my approach to
dialectical logic. Part C, the first part, is devoted to a presentation of the historical
development of the ideas which stand behind my approach to dialectical logic, i.e.,
in particular Kant’s transcendental idealism, the (attempted) explicit formulation
of antinomies in metaphysics and, in its wake, the rise of the modern notion of
dialectic in Hegel’s speculative philosophy.
In part D, the second part of these materials, I have collected materials from
the philosophy of mathematics which lead up to the emergence of antinomies within
an attempt to give a purely logical foundation of mathematics and follow their traces
to the fixed point property.
Part E, the third part, is less structured. One chapter takes up various topics
from analytic philosophy which arose out of its involvement with the foundations of
mathematics and its preference for the methods developed there. Although analytic
philosophy is in no way congenial to dialectical logic, it can’t help being haunted
by various annoying phenomena which I, predictably, take to be indications of sup-
pressed dialectics in the subject matter.
My decision to provide an historical-philosophical background in form of a col-
lection of quotations, rather than in the familiar philosophical style of an essay,
which often enough just means “to rewrite others’ work from one’s own perspec-
tive”,3 is to a good extent my reaction to the dominant style in philosophy: a mixture
of poor research, fiction, and facile writing.4 Differently put, these materials have
the form they do, because I have grown sick and tired of philosophical narratives.
Apart from that my reason is simply that the ideas presented in these materials
are not my own, so I am not trying to express them in my own words. What is mine,
is the choice and the arrangement of materials with regard to a theory of dialectic.
My aim was to compile — the silly and the sublime, a panopticon of philosophical
bric à brac that is related, in some way, at least, to my enterprise.
3 Cf. Wang in quotation 56.1 below. In German this is sometimes euphemistically called a
“kritische Auseinandersetzung”.
4 Predictably, the very champions of this style will find my way of presentation deficient
and want to see it replaced by a “berichtenden und argumentativen Darstellung”, as if the endless
stream of secondary literature in philosophy had not sufficiently established its impotence.
708
Part C
DIALECTIC, IDEALISM, AND

THE TRANSCENDENTAL
APPROACH
Was der Geist will, ist,
seinen eigenen Begriff zu erreichen,
aber er selbst verdeckt sich denselben1
1 Hegel [1837/40], p. 90 f.
PART C
DIALECTIC, IDEALISM, AND THE TRANSCENDENTAL APPROACH
Although the target of my study may look extremely abstract, it still makes its
appearance in every day life and even shows its presence in common sense. It does
not, however, readily reveal itself as an abstraction; and this is what common sense
is commonly taken in by. As Hegel, perhaps, would have had it, common sense it-
self is highly abstract; it is only unaware of this fact and considers itself concrete
instead, and “ist stolz und voll von Genuß, in dieser Entfremdung seiner selbst”.2
In the present part, I have collected material in an attempt at highlighting
certain aspects of the struggle of philosophical thought with its own power of ab-
straction; it focuses on tendencies of self-awareness of thought, rather than thought
that busies itself with the reality of the world and all that it itself is not. It does
not try to represent the views of various philosophers, but only highlight aspects
which stand in a relation to my project.
This philosophical background may hopefully be edifying and also be informa-
tive to someone who is interested in the genesis of ideas and accustomed to the
philosophical way of thinking. It must not, however, be mistaken for a relevant part
of the theory of dialectic to be developed in the groundworks.
2 Hegel [1837/40], p. 91. Fortsetzung und Schluß des im obigen Motto begonnenen Zitats. A
translation can be found in Sibree [1899], p. 55. My translation of the first part can be found in
confrontations 113.12 in the groundworks.
711
CHAPTER XIV
PREPARATIONS
The historical-philosophical background of my theory of dialectic does not come

as a simple coherent philosophical subject. This first chapter of my presentation of
materials somewhat artificially puts together methodological remarks, ancient Greek
roots of my notion of dialectic, and some aspects of British empiricist philosophy.
§ 56. Methodological preliminaries
A central issue in the present approach to dialectic and speculative philosophy is

that of representation. This issue becomes curiously doubled when dealing with
translations. Since most of the sources presented here were not originally written in
English, this deserves some attention.
I begin with some rather general remarks.
56a. General remarks. The word “dialectic” names one of the great ideas
of the ancient Greeks. As with so many others of their ideas, so with dialectic:
it developed through the ages and considerably changed its meaning. Take, as an
example, the atomistic conception of matter from Leukippos and Demokritos1 until
today. No doubt the connection is faint to today’s notion of the atom, we still use,
however, the ancient name. In that sense one may say that the notion of dialectic is
still — as that of the notion of the atom was for a long time — more a programme
than a fully developed concept.
When attempting to trace back the origins of our modern notion — respectively,
lack of notion — of dialectic, I shall not only look for what has been explicitly so
called. Words may change their meanings, while ideas prevail; they may often do so
under very different labels.
The making of the conception of dialectic that forms the background of the
present study began with Kant’s transcendental philosophy. The latter, in turn,
had strong commitments in the philosophy of the British empiricists. In order to
understand Kant’s enterprise and the role of dialectic therein, I shall devote some
space to the British empiricists as well as to some aspects of the notion of dialectic
in ancient Greek philosophy.
Thus, I acknowledge at least two historical roots of the modern conception of
dialectic: the ancient Greek notion of dialectic, which was taken up in the medieval
Scholastic, and the epistemological considerations of the British empiricists. This
leads a curious path from Parmenides via Berkeley to Kant and Hegel.
1 As regards the transcription of Greek names see my comment to quotation 56.8.
712
§ 56. METHODOLOGICAL PRELIMINARIES 713
I shall confine myself to dialectic in the tradition of transcendental philosophy,

i.e., a so-called idealistic conception of dialectic. I shall not consider Marxist or
materialist dialectic more than in passing. My reasons for this will hopefully become
clear in the course of chapter XXV in the groundworks.
Unlike mathematics, the disciplines of philosophy such as epistemology and
metaphysics in particular, are of poor theoretical standard. Their basic concepts
are agonizingly vague and hardly ever supported by anything more than crude
common sense ideas. According to different views regarding what is intuitively cor-
rect, there are endless and fruitless discussions filling books and journals. This may
cause considerable frustration when dealing with these questions. Yet, my enter-
prise, viz., bringing diagonal method and dialectical logic together, is inseparable
from epistemology and metaphysics; indeed, they make up its proper philosophical
background. So I have to enter into epistemological and metaphysical discussions,
in order to place my subject within its appropriate frame. I do this, however, with
an eye to a precise formulation. The reader is warned, therefore, that the consider-
ations in these materials are not constitutive of the main line of argumentation.
What I try to do here, is to provide a sort of intuitive background for later parts,
in particular, Part F in the groundworks.
In representing historical material selected from a period of more than two
thousand years, I find myself confronted with one particular problem, namely that
of showing the continuity of my chosen subject throughout the different times and
disciplines without this subject actually appearing under the same label. I am afraid
I have no solution to this problem at the present stage. In the third book I shall use
paraphrasing as one way of illustrating a continuity in the history of dialectic from
Parmenides to Hegel; variations will provide another one. These ways of dealing
with historical material, however, do not seem to me to belong into the present
compilation.
I close this section with a quotation from Hao Wang, a mathematical logician
who moved to philosophy.
Quotation 56.1. Wang [1987], pp. 254 f.2

It is conceivable that a well-organized collection of aptly selected quotations
with appropriate comments can be a reasonably effective way to express
one’s comprehensive view, but I am not aware of any serious attempts to do
so. What is more commonly done is to rewrite others’ work from one’s own
perspective; for example, W. Windelband’s A History of Philosophy may be
viewed as an account of his own philosophy. In contrast, Montaigne presents
informally a comprehensive view in his Essays by recording reflections on
his readings and personal experience.
Comment. This is very close to my attitude. But it seems to me that the Windel-
band style is still the dominant one, and — actually — the only acknowledged one,
at least in philosophy. The reason, it seems to me, is the endemic lack of original-
ity and skill in mainstream philosophy; there isn’t enough content to substantiate
claims without placing them into the context of someone else’s ideas, but at the
2 This is a continuation of the motto to these materials on p. 707.
714 XIV. PREPARATIONS
same time there is too much pride to be satisfied with just arranging quotations and
commenting on them. Contemporary German philosophy of Kantian and Hegelian
orientation is still squarely in this tradition.
56b. Concerning sources. Although I emphasized that my aim does not lie
in providing a faithful historical presentation of anything that may lay claim to
the label ‘dialectic’, some attention to its historical genesis and wanderings is still
worthwhile in the context of my enterprise. First of all, I start from the basis, without
making any further attempt at supporting it,3 that philosophy has not yet managed
to grasp the fundamental features of a theory of dialectic in a systematic way, i.e.,
independently of its particular historical forms, such as Platonian, Aristotelian,
scholastic, Kantian, Hegelian, or Marxist dialectic.4 Secondly, it is not just one,
more or less homogeneous, doctrine that comes under the label dialectic but a
cluster of sometimes quite different ideas, and the present approach to dialectical
logic is only concerned with one particular string within this cluster; and even this
I do not try to represent adequately. Thirdly, a short presentation of the historical
background of the concept of dialectic may help to supply the formal approach with
some intuitive content.
But things are worse than that; there is not even a set of ideas which could
rightfully be said to form the background of my notion of dialectic. To put it bluntly,
I have to foist my ideas onto the historical material. But then, this strikes me as
general custom anyway. I only want to make sure that I am not misinterpreted as
trying to interpret other authors.
Quotations 56.2. (1) R. Robinson [1941], p. 1.

There are at least five ways in which misinterpretation is very common,
and the first of them is (1) mosaic interpretation, or the habit of laying any
amount of weight on an isolated text or single sentence, without determining
whether it is a passing remark or a settled part of your author’s thinking,
whether it is made for a special purpose or is intended to be generally valid,
and so on.
Comment. What I want to ask here is why somebody uses a text. If it is for the
purpose of finding the correct interpretation, I am not interested; this is just not the
aim of my study. Personally, I prefer the challenge of Heidegger’s interpretation of
Parmenides to the lack of inspiration in Robinson’s account of Plato’s Earlier Di-
alectic. After all, the problem with Parmenides is not so much mosaic interpretation,
but mosaic records.
(2) Kaufmann [1972], p. 25 f.
Quilt quotations. This device, used by other writers [[than Popper]],
too, has not received the criticism it deserves. Sentences are picked from
various contexts, often even out of different books, enclosed by a single set
of quotation marks, and separated only by three dots, which are generally
3 Flach in quotation 69.7 (4) in these materials is sufficient, to my mind.
4 Cf. also McKeon in quotation 67.25 (1) in these materials.
taken to indicate no more than the omission of a few words. Plainly, this
device can be used to impute to an author views he never held.
Here, for example, is a quilt quotation about war and arson: “Do not
think that I have come to bring peace on earth; I have not come to bring
peace, but a sword. . . . I came to cast fire upon the earth. . . . Do you think
I have come to bring peace on earth? No I tell you. . . . Let him who has no
sword sell his mantle and buy one.” This is scarcely the best way to establish
Jesus’ views of war and arson. In the works of some philosophers, too—
notably, Nietzsche—only the context can show whether a word is meant
literally.
Comment. I take the point, in principle. But I am not sure if, regarding the par-
ticulars, there isn’t more truth to the above quilt quotation than to the official
PR-material by Jesus’ institutionalized heirs. After all, the way the message of
Jesus has been carried around the world is inseparable from (gun)fire and sword.
Whereas science seems to develop in the sense that new material is provided
with new methods, history seems to be bound to stagnate in the sense that the
discovery of new sources has become rare. The fragments of the pre-Socratics aren’t
terribly likely to have grown by tomorrow. Whatever sources there are, revealing a
bit of the origin of dialectic, have probably been discussed. It is not to be expected
that the view on the origin of dialectic will change thoroughly through the discovery
of new sources.
But there is still the translation and the interpretation. The sources may not
have grown by tomorrow, but interpretations will certainly have, and translations
seem to have the tendency to go out of date and be replaced by “fresh” ones; more-
over the meaning of words is subject to change, e.g., what is to count as dialectic.
In the conflict between translation and interpretation I tend to put more weight
on interpretation. The framework of thought of the Parmenides fragments will not
come back to life. But today we may see them in a new light.
Having said this, I hasten to declare that I do put considerable weight on the
translation too. In fact, some philosophers need to be reminded that Parmenides
never said nor wrote “For the same are thinking and being” nor did he say “For it
is the same thing that can be thought and that can be”. Nor did Kant ever write
“we intuit objects”, nor Hegel “What is rational is actual”. Neither Parmenides, nor
Kant, nor Hegel wrote in English. I suspect it is particularly necessary to remind
just those philosophers who find my remark to this effect little more than a cheap
joke. A few quotations may help to illustrate some of the difficulties.
Quotations 56.3. (1) Cooke [1938], p. vii.

With an eye to the English reader, who knows, perhaps, little of logic
and less in that case of Aristotle’s, I have tried in translating these texts
to bring out the philosopher’s meaning as clearly as was in my power.
How far I have succeeded in doing so, provided I interpret it rightly, the
reader alone can determine. I cannot, in consequence, pretend that I literally
translate the Greek, where it seemed that a literal translation would fail to
achieve this main purpose. Some scholars may possibly object that at times
I paraphrase Aristotle. I can in that case only plead that a more or less
intelligible paraphrase does convey something to the reader, unlike strict
adherence to the letter. Moreover, a literal translation might often repel
English readers and read like some alien jargon[[.]]
Comment. The emphasis, I suggest, lies on the little subordinate clause “provided
I interpret it rightly”. What this means in ‘plain English’, “provided I interpret it
rightly”, is: I have a certain view of what Aristoteles means and I shall spare readers
the trouble of making up their own minds by simply giving them my reading of
Aristoteles. I shall refer to this kind of position as the pampering-the-English-reader
attitude. I seriously doubt that it is suitable to bring out or further critical leanings.
(2) N. K. Smith [1929], p. vi.
Kant’s German, even when judged by German standards, makes dif-
ficult reading. The difficulties are not due merely to the abstruseness of
the doctrines which Kant is endeavouring to expound, or to his frequent
alteration between conflicting points of view. Many of the difficulties are
due simply to his manner of writing.
Comment. I don’t know where N. K. Smith gets his “German standards” from,
but he is certainly working hard at constructing his image as an heroic transla-
tor battling Kant’s obscurity.5 In the introduction to his translation of the Critique,
he claims that “even in regard to so important a philosophical term as Verhält-
niss, [[Kant]] alternates between the feminine and the neuter.” 6 I subsequently went
through the whole Critique, but I wasn’t able to find a single one, which may be due
to my own unconscious corrective reading. Or did I possibly fall prey to corrections
made by the editor of my Kant edition? Can anybody help me? I’d love to see one
so I could appreciate N. K. Smith’s achievements more fully. Compare also what the
editor of my German edition says in quotations 56.4 below.
5 I am not at all convinced of N. K. Smith’s competence as regards the understanding of the
German idiom. In the introduction to the Critique, A4, Kant says: “Durch einen solchen Beweis
von der Macht der Vernunft aufgemuntert, sieht der Trieb zur Erweiterung keine Grenzen.” In
the introduction to the second edition “aufgemuntert” is replaced by “eingenommen”. N. K. Smith
[1929], p. 47, translates “eingenommen” as “misled”. This could come from a literal translation from
German: “taken in by”. The proximity of the two languages can be very deceptive. (The standard
German joke is: I want to become a beefsteak.) In the case of “eingenommen” the corresponding
English phrase is “taken with”. I hope that even the non German speaking reader can see that it is
the preposition which makes all the difference in the world. If it is right what Rée says in quotation
56.6 (4) below that English philosophy is a creature of translations, then I think it is high time for
a decent English translation of Kant’s main work. (There is a translation by Meiklejohn from 1855
which seems to be available again. But I don’t think this is a great improvement on the situation.)
Addendum. Since this was written two new English translations of Kant’s first Critique have
appeared which I have included in the bibliography: Pluhar [1996] and Guyer and Wood [1998].
I have not yet found the time to look at them carefully, but at least some of N. K. Smith’s
more distorting mistranslations seem to have been avoided. There is a review by McLaughlin in
Erkenntnis 51 (1999), 357–363, which praises N. K. Smith’s work as “a model of philosophical
translation.”
6 N. K. Smith [1929], p. vii. There is a possibility that Smith actually meant the word “Er-
kenntniss”. Cf. the phrase “im menschlichen Erkenntnis” in Kant [1787], p. 40 (B 30).
(3) French [1967], p. 624.

The writings of Kant are often confusing and sometimes confused. He
employs a vocabulary that is not readily assimilated by the contemporary
ear.
Comment. Beauty, they say, is in the eye of the beholder. Have you ever heard of
the confusion that is in the ear of the reader? My concern regarding this quotation
is to issue a general warning of the complacency of “the contemporary ear”.
Quotations 56.4. (1) R. Schmidt [1926], editing Kant’s Kritik der reinen Vernunft,
p. VI.
Der Originaltext wurde auch in solchen Fällen beibehalten, wo er offen-
sichtlich fehlerhaft ist. Bei der Verschiedenheit und Unvereinbarkeit der
Versuche zur Textverbesserung durch mehrere Generationen von Kantin-
terpreten, war häufig keine Möglichkeit gegeben, sich rückhaltlos für die
eine oder für die andere Version einzusetzen, auch sollte dem Leser selbst
die Entscheidung über die notwendige Korrektur und die Art ihrer Aus-
führung überlassen bleiben, im Gegensatz zu allen bekannten kritischen
Ausgaben, die dem Leser ihre Lesart aufzwingen und Abweichungen davon
in einen schwer übersichtlichen Anhang verweisen.
(2) R. Schmidt [1926], editing Kant’s Kritik der reinen Vernunft, p. VII.
Der so gebotene Text unterscheidet sich also von dem kantischen nur in
der Anwendung einer moderneren Schreibweise. (Auch hier wurde vorsich-
tig alles geschont, was mit der kantischen Schreibweise den kantischen Sinn
und die kantische Wucht verlieren würde.) Die häufig als völlig unzulänglich
beklagte kantische Interpunktion wurde ebenfalls aus einem guten Grunde
beibehalten. — Wer die Langatmigkeit und Unübersichtlichkeit gewisser
kantischer Perioden beklagt, macht häufig die überraschende Entdeckung,
daß diese Perioden sich im Original gar nicht so schwierig und unübersicht-
lich ausnehmen. Der Grund ist in der für Kant überaus bezeichnenden und
im ganzen konsequent durchgeführten Interpunktion zu suchen, die seine
Sätze zwar nicht immer in unserem Sinne grammatisch richtig aber doch in
sinnvollem gedanklichen Rhythmus gliedert.
The next couple of quotations may serve as an example of what I have in mind
when speaking of the pampering-the-English-reader attitude.
Quotations 56.5. (1) Hegel [1830], p. 181.

§. 76.
In Beziehung auf den Ausgangspunkt, die oben sogenannte u n b e f a n -
g e n e Metaphysik, das Princip des unmittelbaren Wissens betrachtet, so
ergiebt sich aus der Vergleichung, daß dasselbe zu jenem Anfang, den diese
Metaphysik in der neuern Zeit als c a r t e s i s c h e Philosophie genommen
hat, z u r ü c k g e k e h r t ist.
(2) Wallace [1873], p. 109; ‘translating’ Hegel [1830], p. 181.

§76
If we view the maxims of immediate knowledge in connection with the
uncritical metaphysics of the past from which we started, we shall learn
from the comparison the reactionary nature of the school of Jacobi. His
doctrine is a return to the modern starting-point of this metaphysics in the
Cartesian philosophy.
Comment. Even non-German speaking readers will be able to confirm that there
is no explicit mention of Jacobi in the German original. In fact, the phrase “the
reactionary nature of the school of Jacobi” has no counter part at all in Hegel’s
text.
(3) U. Petersen [2002], p. 718; translating Hegel [1830], p. 181.
If the principle of immediate knowing is considered in relation to the start-
ing point, uninhibited metaphysics as it was called above, what results from
the comparison is that it has returned to that beginning which this meta-
physics has taken in the more recent times as Cartesian philosophy.
Comment. I was not able to preserve the structure of the German sentence.
Quotations 56.6. (1) Tredennick [1933], p. xxiv.

As regards the translation, my chief object has naturally been to make
Aristotle’s meaning as clear as possible without too great a sacrifice of
brevity or literalness; and in pursuing this object I have not scrupled to
vary the rendering of the same Greek words in different contexts, even
where it was not absolutely necessary to do so. Where the sense of the
Greek is really doubtful I have thought it best to be non-committal.
Comment. And, if I may add, there is, of course, no doubt that Tredennick knows
“Aristotle’s meaning” so as to be sure where to “vary the rendering of the same
Greek words in different contexts”.
(2) N. K. Smith [1929], p. vii.
Certain sentences, [[. . .]] occurring not infrequently, present the transla-
tor with another type of problem: how far he ought to sacrifice part of what
is said, or at least suggested, to gain smoothness in the translation. There
are sentences which, to judge by their irregular structure and by the char-
acter of their constituents, must have owed their origin to the combination
of passages independently written and later combined. In the “four to five
months” in which Kant prepared the Critique for publication, utilising, in
the final version, manuscripts written at various dates throughout the pe-
riod 1769–1780, he had, it would seem, in collating different statements of
the same argument, inserted clauses into sentences that were by no means
suited for their reception. In such cases I have not attempted to trans-
late the sentences just as they stand. Were the irregularities retained, they
would hinder, not aid, the reader in the understanding of Kant’s argument.
Comment. The German language has a word “Bevormundung” which suits perfectly
to describe my impression of what is going on here. Curiously enough, the English
language doesn’t seem to have a pithy word for it; only something like ‘making up
one’s mind for someone else’. Does it possibly tell of a particularly insidious form
of ‘Bevormundung’ not to have a word denoting it?7
(3) Long and White [1979], p. VI.
[[W]]e have parted company with all previous English translators of Frege
by rendering ‘bedeuten’ and ‘Bedeutung’ as ‘mean’ and ‘meaning’. We have
done this throughout, both before and after he formulated his celebrated
distinction between Sinn (sense) and Bedeutung (except of course where the
obvious translation of ‘Bedeutung’ is ‘importance’ or ‘significance’). And
cognate terms such as ‘bedeutungsvoll ’ and ‘gleichbedeutend ’ we have ac-
cordingly rendered by ‘meaningful’ and ‘having the same meaning’. ‘Mean-
ing’ is, after all, the natural English equivalent for ‘Bedeutung’, and render-
ings such as ‘reference’ and ‘denotation’ are strictly incorrect and have only
been adopted by other translators for exegetical reasons. We have thought
it better not to beg questions of exegesis by suggesting through translation
a certain view of what Frege meant in his later writings by ‘Bedeutung’,
leaving it rather to the reader to form his own judgement of the contrast
Frege intended by his Sinn–Bedeutung distinction. If his later use of ‘be-
deuten’ and ‘Bedeutung’ reads oddly in German, this oddness should be
reflected in translation and not ironed out by mistranslation.
Comment. The recipe: first you stipulate a natural equivalent, then you can call
everything else “begging the question” or “mistranslation”. What is lacking in this
consideration is an awareness of the possibility of an indeterminacy of translation.
The German word ‘bedeuten’ has as an integral part ‘deuten’, used, for instance, to
describe a bodily movement, typically with a finger but also a movement of the head,
towards an object or a direction; a procedure by means of which Augustine claims
with regard to the words his elders uttered, as he is translated in quotation 88.2 (3)
in these materials, “I gradually learnt to understand what object they signified”;
or what object they meant ? The “natural English equivalent for ‘Bedeutung” ’ does
not reflect this aspect of ‘deuten’ inherent in the German word. Frege’s use of the
word ‘Bedeutung’ looses its oddness, given a little sensitivity for language; but I
don’t see how any sensitivity for language will be able to find a root of ‘pointing’
or ‘indicating’ in ‘meaning’, like you can find ‘sign’ and ‘signify’ in ‘significance’,
for instance. What may read “oddly in German” for some people, to which I may
have become insensitive after engaging with philosophy for thirty years, may be
captured by rendering ‘Bedeutung’ as ‘significance’: dark clouds in the sky may
signify that it will rain soon, like “dunkle Wolken am Himmel bedeuten, daß es
bald regnen wird”. Anyway, to come to an end of this comment, I wish to mention
that one good translator has finally closed the circle by translating Russell in “On
7 At the time when I first arrived in Australia, a pharmacist would get a prescription ready
for you by hiding behind a screen and destroying everything that could give you a hint at the
medicine prescribed. All the information you would get was something like: three times a day, so
and so much. DAJS (Don’t Ask Just Swallow).
Denoting”: Frege “distinguishes . . . meaning and the denotation” as “unterscheidet

. . . Bedeutung und Gekennzeichnetes”.8
(4) Rée [1994], p. 42.
Philosophy in English (as in every other language, no doubt) is a creature
of translations. Its translators are its poets, the unacknowledged legisla-
tors of the English philosophical world. And so English philosophy remains
haunted by a repressed knowledge of its alien origins: after the death of the
author, perhaps the return of the translator?
Comment. What about the return of the multilingual philosopher ? Or is that too
radical for Radical Philosophy? Or has it become radically un-English since the
days of Bertrand Russell?9 Anyway, I find the bit in brackets almost insulting to
people who cultivate a multilingual tradition. At Oslo University, for instance, I have
witnessed a doctoral viva held in English (on polar bears, quite funny, actually),
and dissertations written in English or German seem to be quite common. Be this
as it may, I have decided to take Rée’s consideration seriously and quote (most)
translations under the name of their creator, and only then add the name of the
author of the original. In this way I hope to acknowledge the “legislators of the
English philosophical world”, in particular in those cases where the original cannot
be held responsible for the rubbish that a translation conveys.
(5) Avineri [1962], p. 125.
Hegel termed the last phase of historical development die germanische Welt
(the Germanic world) and not die deutsche Welt (the German world). It is
significant—and deplorable—that Sibree [[[1899]]] in his translation of the
Philosophy of History never distinguished between germanisch and deutsch
and translated both terms as “German.” The term “Germanic” is, in the
German usage, always used to connote a cultural sphere, and had no po-
litical implications, whereas the term “German” aroused in Anglo-Saxon
readers of the last half century every possible association with German
political domination.
Comment. So much for the poetry of the translator.
After these quotations, allow me to introduce my maxim of English common

sense, MOECS,10 which I regard as one of the basic principles of Anglocentrism,
although it might not have had as much attention as it deserves.
Maxim 56.7. Anything that can be said in any language can be said in such a way
that it translates without conceptual difficulties into plain English.
8 In: W. Stegmüller (ed.): Das Universalienproblem. Wissenschaftliche Buchgesellschaft,
Darmstadt 1978; p. 26.

9 It is my understanding that Bertrand Russell communicated in French with Poincaré and
in German with Frege.

10 Acronyms in these materials are only provided to make fun of readers who think they
are cool. I shall not use them.

Comment. Or, for those who value a familiar tone: what can be said at all can be
said in plain English, and what we cannot talk about in plain English we must pass
over in silence.
The issue of translation is different according to the language in question. Ger-

man is my mother tongue and words like “aufheben” of which the cognoscenti of
Hegel’s philosophy make such a lot of ‘Aufhebens’ (fuss) are familiar, everyday
words for me. I don’t feel secure enough to decide whether the English word ‘to lift’
with its different meanings (lift a ban, lift a book from the floor)11 is a good substi-
tute for the German word ‘aufheben’, but I certainly resent the halo surrounding it,
polished by pompous academics to spread some shine on themselves.12 I feel secure
enough, however, to decide against a ‘translator’s poetry’ such as N. K. Smith’s
rendering “Entscheidung” in Kant’s Critique as “solution”.13 The situation is com-
pletely different in the case of ancient Greek, for instance, when it comes to the
translation of the Parmenides fragments. But occasionally I still have confronted
my own translation with that of other scholars with whom I had better not try to
get into a competition as regards the mastering of ancient Greek. The point is that
I pursue a specific philosophical interest and there are many cases where a trans-
lator, perhaps unaware of even the possibility of the existence of such an interest,
seems to be blind towards the possibility of a different reading or just chooses to
neglect it; in other words, as regards translation, just as in the case of choice of
quotations, I make no claim as to being impartial, but freely and happily admit
that I am tendentious, in the sense that my aim is to provide a background for my
own philosophical ambitions, not to provide a set of standardized materials for, say,
an introductory course in philosophy.
Also: proficient speakers of a natural language have certain preferred ways. In
English it seems fashionable to use “it is natural” or “precisely”, for example. Some
translators seem to be so fond of these phrases that they use them in their English
translations of German phrases.14 For instance, “es liegt nahe” (literally: it lies close,
perhaps: it is a short step) in Frege [1892], p. 26, is translated in Geach and Black
[1952], p. 57,15 as “it is natural”. This may not look too bad, at first sight, because it
is a common phrase in English; but it has some underhanded coerciveness about it
which the German doesn’t have: not to think this is unnatural. Like it is unnatural to
be gay, or for a girl to play football, things which the petty minded can get so excited
about. There is something of a reinforced conformity; it is not just that someone
11 There are, of course, further meanings of ‘to lift’ like ‘face-lifting’ or ‘to give someone a
lift’ which, admittedly, do not coincide with the multiple use of ‘aufheben’ in German.
12 Cf. Pinkard [1994], p. 349, note 28, for a critical account of the term Aufhebung. Cf. also
my comment to quotation 66.14 (1).

13 Cf. remark 62.8 on p. 820 and more generally section 62c in the next chapter. Curiously
enough, it is translated as “decision” in N. K. Smith’s Commentary. I have come across other

occasions in which N. K. Smith replaces Kant’s text by suggestions of reformulations from com-
mentators. These occasions are marked but I still find this approach distorting.
14 To get a feeling for how this may sound, take the phrase “he is just being difficult” and
replace “just” by “precisely”.

15 Cf. quotation 71.28 (1) in these materials.
thinks or feels differently, it is unnatural, deviant, abnormal, actually insane.16 All

I want to get across is that a certain care has to be practiced when dealing with
language and translation, a care that I have often found missing. Since the present
study is explicitly concerned with language and translation, albeit formal languages
and translations (and so-called realizations and relativizations as well as coding and
decoding), issues of this kind will turn up again from time to time.
I close this preparatory paragraph with a quotation that has gained my full
sympathy, admittedly only after an initial phase of indifference.
Quotation 56.8. Burnet [1914], pp. v f.

I have not thought it well to present Greek names in a Latin dress. I see
no advantage, and many disadvantages, in writing Herakleitos as Heraclitus.
It often leads to his being called out of his name, as the Emperor Herakleios
usually is when disguised as Heraclius. On the other hand, the Latin titles
of Plato’s dialogues are English words. Theaitetos of Athens is best left
with the beautiful name chosen for him by his father Euphronios, but “the”
Theaetetus is as much English as Thessalonians. We shall never, it seems,
reach agreement on this matter; I only wish to explain my own practice.
Comment. It was only after I found the “k” in the title “the life of sokrates”
of chapter viii in a copy of Burnet’s book in the library of the University of Western
Australia crossed out and a “c” written above it that I decided to side with Burnet.17
§ 57. Ancient roots of the notion of dialectic
Dialectic, as I shall deal with it, has twofold roots in ancient Greek philosophy
and curiously enough, these two are to some extent quite separate: the word which
names a method (διαλέγεσθαι , διαλέγω : I select; somewhat like sorting out), and
a subject (being, εἶναι) which doesn’t seem to have been treated by means of this
method before Hegel. It is this word and the method which are most commonly
taken to represent dialectic.
57a. Parmenides on being and nothing. Prima facie, Parmenides was not
talking about dialectic, nor did he use the method of dialogues. Secunda facie, how-
ever, focusing on the opposition of being and nothing his poem gives the note for one
of the most eminent themes of modern dialectical thought and this is witnessed in
Hegel’s lectures on the history of philosophy.18 It is in this respect that Parmenides
takes his prominent place in the present materials.
What I find intriguing about the following quotations is not just the conceptual
struggle in the fragments that have come down to us but also how this struggle is
reflected in different translations.
16 For the German speaker it should be sufficient to recall the word “entartet”. US-Americans
might remember ‘Un-American activities’.

17 Accordingly I shall speak of the person Eukleides, and Euclidean geometry.
18 Hegel [1833a], p. 313; cf. quotation 63.9 in these materials.
§ 57. ANCIENT ROOTS OF THE NOTION OF DIALECTIC 723
Quotations 57.1. (1) Diels and Kranz [1952], p. 231; Parmenides fr.2 (4 in earlier
editions), v.7.
οὔτε γὰρ ἂν γνοίης τὸ γε μὴ έὸν (οὐ γὰρ ἀνυστόν)
οὔτε φρὰσαις.
(2) Hegel [1833], p. 310; translating Parmenides fr.2, v.7.
denn das Nichtseyn kannst Du nicht erkennen, noch erreichen, noch aus-
sprechen.
(3) Diels and Kranz [1952], p. 231; German translation of fr.2, v.7.
denn weder erkennen könntest du das Nichtseiende (das ist ja unausführbar)
noch aussprechen;
(4) Cornford [1939], p. 31; translating Parmenides fr.2, v.7.
for thou couldst not know that which is not—for that is impossible—nor
utter it.
(5) Haldane [1892], p. 252; translating Hegel’s translation of fr.2, v.7.
for thou canst neither know, or attain to, or express, non-being.
Comment. My personal reading: “For neither could you know that which is not
(since it is impracticable) nor put it in words.” The emphasis of my interpretation
lies on “φράζω ”; uttering may well give rise to something like a performative in-
consistency in the sense of Boyle [1972],19 but it is perfectly possible. For me, the
question is, what does it mean to say that it is impracticable. I take the problem ad-
dressed here to be related to that of Berkeley’s “contradiction to talk of conceiving
a thing which is unconceived”,20 and Frege’s “to wash the fur without wetting it”,
or “to judge without judging”.21 It is worthwhile having a look at quotations 63.10
for Hegel’s comments.
Quotations 57.2. (1) Diels and Kranz [1952], p. 231; Parmenides fr.3 (5 in earlier
editions).
τὸ γὰρ αὐτὸ νοεῖν ἐστίν τε καὶ εἶναι
(2) Diels and Kranz [1952], p. 231; German translation of fr.3.
denn dasselbe ist Denken und Sein.
Comment. “For the same are thinking and being.”
(3) Cornford [1939], p. 31; translating Parmenides fr.3.
‘For it is the same thing that can be thought and that can be.’
Comment. My way of making sense of this fragment: it is the same to think (a thing)
and to say that it is; differently put: there is no thinking without predicating.
19 Cf. quotations 92.22 in these materials.

21 Frege [1884], p. 36. Cf. quotation 71.27 (3) in these materials.
Quotations 57.3. (1) Diels and Kranz [1952], p. 232; Parmenides fr.6.
χρὴ τὸ λέγειν τε νοεῖν τ’ ἐὸν ἔμμεναι· ἔστι γὰρ εἶναι,
μηδὲν δ’ οὐκ ἐστι·
Comment. According to “Tutti i Verbi Greci”, ἔμμεναι is the present infinitive active
form of ἐιμί, marked as belonging to “poetic and dialectic forms”.
(2) Diels and Kranz [1952], p. 232; German translation of fr.6.
Nötig ist zu sagen und zu denken, daß nur das Seiende ist;
denn Sein ist, ein Nichts dagegen ist nicht;
Comment. This is, by and large, my preferred reading, “It is necessary to say and
think that which is as being. For being is, nothing, however, is not.”
(3) Hegel [1833], p. 310; translating Parmenides fr.6.
Es ist nothwendig, daß das Sagen und Denken das Seyende ist;
denn das Seyn ist, aber das Nichts ist gar nicht.
(4) Cornford [1939], p. 31; translating Parmenides fr.6.
‘What can be spoken of and thought must be; for it is possible for it to
be, but it is not possible for “nothing” to be.
Comment. As regards the “it is possible”-phrase in Cornford’s translation, cf. his
comment quoted in 67.7 (1). I still find it consistent with what I consider the spirit
of Parmenides, but more an interpretation than a translation.
(5) Haldane [1892], p. 252; translating Hegel’s translation of fr.6.
It is necessary that saying and thinking should be Being; for Being is, but
nothing is not at all.
(6) Heidegger [1953], p. 85; translating/interpreting fr.6.
Not tut das sammelnde Hinstellen sowohl als das Vernehmen: Seiend in
dessen Sein;
Das Seiend nämlich hat Sein; Nichtsein hat kein »ist«;
Comment. Observe the replacement of the auxiliary verb to be by another auxiliary
verb to have. Now being has turned into something that ‘that which is’ (Seiendes)
has.
(7) Heidegger [1953], p. 107; translating/interpreting the first part of fr.6.
Not ist das λέγειν sowohl als auch Vernehmung, n"amlich das Seiend in
dessen Sein.
Comment. Who would have expected a normal translation from Heidegger anyway?
Quotations 57.4. (1) Diels and Kranz [1952], p. 236; Parmenides fr.8, v.8.
οὐ γὰρ φατὸν οὐδὲ νοητόν ἔστιν ὅπως οὐκ ἔστι.
(2) Diels and Kranz [1952], p. 236; German translation of fr.8, v.8.
Denn unaussprechbar und undenkbar ist, daß NICHT IST ist.
(3) Cornford [1939], p. 36.
for it cannot be said or thought that ‘it is not’.
Quotations 57.5. (1) Diels and Kranz [1952], p. 238; Parmenides fr.8, vv.34 ff.
ταὐτὸν δ’ ἐστὶ νοεῖν τε καὶ οὕνεκεν ἔστι νόημα.
οὐ γὰρ ἄνευ τοῦ ἐόντος, ἐν ὧι πεφατισμένον ἐστίν,
εὑρήσεις τὸ νοεῖν· οὐδὲν γὰρ hἠi ἐστιν ἠ ἐσται
ἀλλο πὰρεξ τοῦ ἐόντος [[·]]
(2) Diels and Kranz [1952], p. 238; German translation of fr.8, vv.34 ff.
Dasselbe ist Denken und der Gedanke, daß IST ist ; (35) denn nicht
ohne das Seiende, in dem es als Ausgesprochenes ist, kannst du das Den-
ken antreffen. Es ist ja nichts und wird nichts anderes sein außerhalb des
Seienden[[.]]
(3) Hegel [1833], p. 312; German translation of fr.8, vv.34 ff.
„Das Denken und das um weswillen der Gedanke ist, ist dasselbe. Denn
nicht ohne das Seyende, in welchem es sich ausspricht“ (manifestirt [[. . .]]),
„wirst du das Denken finden; denn es ist nichts und wird nichts seyn, außer
dem Seyenden.“
Comment. A look at Hegel’s comment quoted in 63.10 (3) of these materials is
worth the trouble, I find.
(4) Cornford [1939], p. 43; translating Parmenides fr.8, vv.34 ff.
Thinking and the thought that it is are one and the same. For you will
not find thought apart from that which is, in respect of which thought is
uttered; for there is and shall be no other besides what is[[.]]
Comment. The dative case got lost. Cf. my comment to variations 108.8 in the
groundworks.
(5) Haldane [1892], p. 253; translating Hegel’s translation of fr.8, vv.34 ff.
Thought, and that on account of which thought is, are the same. For not
without that which is, in which it expresses itself [[. . .]], wilt thou find
Thought, seeing that it is nothing and will be nothing outside of that which
is.
Comments. (α) Here, Haldane translates “das Seyende” as “that which is”, whereas
in the foregoing fragment he translates it as “Being”. Readers are encouraged to try
this translation with “Being” in the place of “that which is”.
(β) There is nothing in Hegel’s German that sounds as archaic as Haldane’s “wilt
thou”; but then, what can you do when a language has abandoned a separate singular
form of address in favor of the plural form (“you”)?
(6) Heidegger [1953], p. 106; translating/interpreting fr.8, v.34.
Dasselbe ist Vernehmung und das, worumwillen Vernehmung geschieht.
Vernehmung geschieht umwillen des Seins.
Comment. More regarding Heidegger’s view on Parmenides in relation to German
Idealism can be found in quotations 67.8 (4); cf. also quotations 92.2 for Heidegger’s
effusions on “Sein”.
Quotations 57.6. (1) Diels and Kranz [1952], p. 238; Parmenides fr.8, vv.38–41.
τῶι πάντ’ ὄνομ(α) ἔσται, ὅσσα βροτοὶ κατέθεντο πεποιθότες εἶναι ἄληθῆ,
γίγνεσθαί τε καὶ ὄλλυσθαι, εἶναι τε καὶ οὐχὶ, καὶ τόπον ἀλλάσσειν διὰ τε
χρόα φανὸν ἀμέιβειν.
Comment. This sounds like nominalism to me.
(2) Diels and Kranz [1952], p. 238; German translation of fr.8, vv.38–41.
Darum wird alles bloßer Name sein, was die Sterblichen in ihrer Sprache
festgesetzt haben, überzeugt es sei wahr: (40) Werden sowohl als Vergehen,
Sein sowohl als Nichtsein, Verändern des Ortes und Wechseln der leuchten-
den Farbe.
(3) Cornford [1939], p. 43; translating Parmenides fr.8, vv.38–41.
Therefore all those (names) will be a mere word—all (the names) that
mortals have agreed upon, believing that they are true: becoming and per-
ishing, both being and not being, change of place, and interchange of bright
colour.
Remark 57.7. The reader might miss Zenon and his paradoxes here. My emphasis,
however, is on Parmenides’ notion of being, not on the argumentation against the
possibility of change such as Parmenides fr.8, vv.19–30 and Zenon’s fragments. The
reason for this has to be seen mainly in the context of variations 106.5 on p. 1459 in
the groundworks, featuring prominently Berkeley, Kant, Hegel, Frege, and Berry.
57b. Platon’s dialectic. Prima facie, Platon could be expected to be a prime

contributor to any theory of dialectic; he is not as regards my approach. Although
the word “dialectic” can be found in quotations 57.8 below, it is no longer the tone
of naive surprise that is so characteristic of the Parmenides fragments. Differently
put, I do not take Platon as facing the problem raised in Parmenides’ poem.
Quotations 57.8. Lee [1955], p. 342; translating Platon II.532A/B (Republic).

[[W]]hen one tries to get at what each thing is in itself by the exercise of
dialectic, relying on reason without any aid from the senses and refuses to
give up until one has grasped by pure thought what the good is in itself,
one is at the summit of the intellectual realm[[.]]
Comment. Note: “dialectic, relying on reason . . .”; this is a constitutive part of the
notion of dialectic that I want to track down in these materials.
(2) Lee [1955], p. 344; translating Platon II.533C (Republic).
‘Dialectic, in fact, is the only procedure which proceeds by the destruc-
tion of assumptions to the very first principle, so as to give itself a firm
base.
Comment. Notice the phrase “destruction of assumptions”.
Ideas (forms), universals, contradictions, and dialectic: This is the framework

in which Parmenides queries young Sokrates in Platon’s Parmenides:
Quotations 57.9. (1) Goold [1926/39], p. 210; editing Platon III.130B (Parmeni-
des).
‘‘ [[. . .]] αὐτὸς σὺ οὕτω διήι ρησαι ὡς λέγεις, χωρὶς μὲν εἴδη αυτὰ ἄττα, χωρὶς
δὲ τὰ τούτων αὖ μετέχοντα; καί τί σοι δοκεῖ εἶναι αὐτὴ ὁμοιότης χωρὶς ἧς
ἡμεις ὁμοιότητος ἔχομεν, καὶ ἕν δὴ καὶ πολλὰ καὶ πάντα ὅσα νῦν δὴ Ζήνωνος
ἤκουες;’’
‘‘ ῎Εμοιγε,’’ φάναι τὸν Σωκράτη.
(2) Fowler [1926/39], p. 211; translating Platon III.130B.
[[“D]]id you invent this distinction yourself, which separates abstract ideas
from the things which partake of them? And do you think there is such
a thing as abstract likeness apart from the likeness which we possess, and
abstract one and many, and the other abstractions of which you heard Zeno
speaking just now?”
“Yes, I do,” said Socrates.
Quotations 57.10. (1) Fowler [1926/39], pp. 217, 219 & 221; translating Platon
III. 132A–133A (Parmenides). Parmenides speaking first, addressing Sokrates.
“I fancy your reason for believing that each idea is one is something like
this; when there is a number of things which seem to you to be great you
may think, as you look at them all, that there is one and the same idea in
them, and hence you think the great is one.”
“That is true,” he said.
“But if with your mind’s eye you regard the absolute great and these
many great things in the same way, will not another great appear beyond,
by which all these must appear to be great?”
“So it seems.”
“That is, another idea of greatness will appear, in addition to absolute
greatness and the objects which partake of it; and another again in addition
to these, by reason of which they are all great; and each of your ideas will
no longer be one, but their number will be infinite.”
“But, Parmenides,” said Socrates, “each of these ideas may be only a
thought, which can exist only in our minds; then each might be one, without
being exposed to the consequences you have just mentioned.”
“But,” he said, “is each thought one, but a thought of nothing?”
“That is impossible,” he replied.
“Yes.”
“Of something that is, or that is not?”
“Of something that is.”
“A thought of some single element which that thought thinks of as
appertaining to all and as being one idea?”
“Yes.”
“Then will not this single element, which is thought of as one and as
always the same in all, be an idea?”
“That, again, seems inevitable.”
“Well then,” said Parmenides, “does not the necessity which compels
you to say that all other things partake in ideas, oblige you also to believe
either that everything is made of thoughts, and all things think, or that,
being thoughts, they are without thought?”
“That is quite unreasonable, too,” he said, “but Parmenides, I think the
most likely view is, that these ideas exist in nature as patterns, and the
other things resemble them and are imitations of them; their participation
in ideas is assimilation to them, that and nothing else.”
“Then if anything,” he said, “resembles the idea, can that idea avoid
being like the thing which resembles it, in so far as the thing has been
made to resemble it; or is there any possibility that the like be unlike its
like?”
“No, there is none.”
“And must not necessarily the like partake of the same idea as its like?”
“It must.”
“That by participation in which the like partake of the same idea as its
like?”
“Certainly.”
“Then it is impossible that anything be like the idea, or the idea like
anything; for if they are alike, some further idea, in addition to the first,
will always appear, and if that is like anything, still another, and a new
idea will always be arising, if the idea is like that which partakes of it.”
“Very true.”
“Then it is not by likeness that other things partake of ideas; we must
seek some other method of participation.”
“So it seems.”
“Do you see, then, Socrates, how great the difficulty is, if we maintain
that ideas are separate, independent entities?”
“Yes, certainly.”
Comment. The problem that is addressed here is that of abstraction: does abstrac-
tion create a new object, and if yes, what kind of object is that? Is it an object that
resembles those from which it is abstracted?
Quotations 57.11. (1) Goold [1926/39], p. 228; editing Platon III, 135A/B.
‘‘ Ταῦτα [[. . .]] καὶ ἔτι ἄλλα πρὸς τούτοις πάνυ πολλὰ ἀναγκαῖον ἔχειν τὰ εἴδη,
εἰ εἰσὶν αὗται αἱ ἰδέαι τῶν ὄντων καὶ ὁριεῖταί τις αὐτό τι ἕκαστον εἶδος· [[. . .]]’’
[[. . .]]
‘‘ ᾿Αλλὰ μέντοι,’’ εἶπεν ὀ Παρμενίδες, ‘‘ εἴ γέ τις δή, ὦ Σώκρατες, αὖ μὴ
ἐάσει εἴδη τῶν ὄντων εἶναι, εἰς πάντα τὰ νῦν δὴ καὶ ἄλλα τοιαῦτα ἀποβλέψας,
μηδέ τι ὁριεῖται εἶδος ἑνὸς ἑκάστου, οὐδὲ ὅπήι τρέψει τὴν διάνοιαν ἕξει, μὴ ἐῶν
ἰδέαν τῶν ὄντων ἕκάστου τὴν αὐτὴν ἀεὶ εἶναι, καὶ οὕτως τὴν τοῦ διαλέγεσθαι
δύναμιν παντάπασι διαφθερεῖ. [[. . .]]’’
(2) Fowler [1926/39], p. 229; translating Platon III.135A/B (Parmenides).

“[[T]]hese difficulties and many more besides are inseparable from the ideas,
if these ideas of things exist and we declare that each of them is an absolute
idea. [[. . .]]”
[[. . .]]
“But on the other hand,” said Parmenides, “if anyone, with his mind
fixed on all these objections and others like them, denies the existence of
ideas of things, and does not assume an idea under which each individual
thing is classed, he will be quite at a loss, since he denies that the idea of
each thing is always the same, and in this way he will utterly destroy the
power of carrying on discussion. [[. . .]]”
Comment. Note the translation of διαλέγεσθαι as “discussion”.
Quotations 57.12. (1) Goold [1926/39], pp. 10 & 12; editing Platon I, 385B (Kraty-
los).
ΣΩ. [[. . .]] καλεῖς τι ἀληθῆ λέγειν καὶ ψευδῆ;
ERM. ῎Εγωγε.
SW. Οὐκοῦν εἰη ἂν λόγος ἀληθής, ὁ δὲ ψευδής;
ERM. Πάνυ γε.
SW. ῏Αρ’ οὖν οὖτος ὃς ἂν τὰ ὄντα λέγήι ὡς ἔστιν, ἀληθής· ὃς δ’ ἂν ὡς
οὐκ ἔστιν, ψευδής;
ERM. Ναί.
SW. ῎Εστιν ἄρα τοῦτο, λόγωι λέγειν ὄντα τε καὶ μή;
ERM. Πάνυ γε.
(2) Fowler [1926/39], pp. 11 & 13; translating Platon I, 385B (Cratylus).
soc. [[. . .]] Is there anything which you call speaking the truth and
speaking falsehood?
herm. Yes.
soc. Then there would be true speech and false speech?
herm. Certainly.
soc. Then that speech which says things as they are is true, and that
which says them as they are not is false?
herm. Yes.
soc. It is possible, then, to say in speech that which is and that which
is not?
herm. Certainly.
Comment. Note the proximity of the theme, firstly to Parmenides, in particular as
quoted in 57.4, and secondly to Aristoteles as quoted in 57.24 (last sentence). The
dialogue goes on, of course, and it is worth reading on, at least for a while.
57c. Aristoteles.
Features 57.13. What interests me in the writings of Aristoteles are the following
topics:
(AR1) the categories
(AR2) metaphysics (first philosophy)

(AR3) logic
(AR4) epistemology
In view of what is to come it might be worthwhile pointing out here that it is

the close connection of logic, epistemology, metaphysics, and the deduction of the
categories in the philosophy from Kant to Hegel that accounts for my interest.22
Aristoteles seems even less concerned with the problem raised by Parmenides than
Platon. In quotation 57.24 (2) below he is adamant that “we even say that not-being
is not-being”.
I begin with the categories.
Quotations 57.14. (1) Goold [1938], p. 16; editing Aristoteles 1 b 9–15 (κατη-
γορίαι).
III. ῞Οταν ἕτερον καθ’ ἑτέρου κατηγορῆται ὡς καθ’ ὑποκειμένου, ὅσα
κατὰ τοῦ κατηγορουμένου λέγεται, πάντα καὶ κατὰ τοῦ ὑποκειμένου ῥηθήσε-
ται, οἷον ἄνθρωπος κατά τοῦ τινὸς ἀνθρώπου κατηγορεῖται, τὸ δὲ ζῶι ον κατὰ
τοῦ ἀνθρώπου· οὐκοῦν καὶ κατὰ τοῦ τινὸς ἀνθρώπου κατηγορηθήσεται τὸ
ζῶι ον ὁ γάρ τις ἄνθρωπος καὶ ἄνθρωπός ἐστι καὶ ζῶι ον.
(2) Cooke [1938], p. 17; translating Aristoteles 1 b 9–15 (categories).
III. [[A word upon predicates here.]] When you predicate this thing or
that of another thing as of a subject, the predicates then of the predicate
will also hold good of the subject. We predicate ‘man’ of a man; so of ‘man’
do we predicate ‘animal.’ Therefore, of this or that man we can predicate
‘animal’ too. For a man is both ‘animal’ and ‘man.’
Comment. The point worth noting is the translation of “κατηγορέω” as “to predi-
cate” (in view of the meaning of the word “categories”). The first sentence, which I
put in double square brackets, has no counterpart in the Greek original. I attribute
its existence to a ‘pamper-the-English-reader-attitude’.
(3) Rolfes [1925], p. 44; translating Aristoteles 1 b 9–15 (Kategorien).

Wenn etwas von Etwas als seinem Subjekt ausgesagt wird, so muß alles,
was von dem Ausgesagten gilt, auch von dem Subjekt gelten. So wird z. B.
Mensch von einem bestimmten Menschen und Sinnenwesen von Mensch
ausgesagt. Mithin muß auch von einem bestimmten Menschen Sinnenwesen
ausgesagt werden; denn der bestimmte Mensch ist ein Mensch und auch ein
Sinnenwesen.
22 Some readers will, no doubt, miss material concerning the syllogisms. In view of the modern
development of logic, however, this has escaped my interest. This does not mean that logical
questions like the validity of tertium non datur would not continue to be of relevance in the
context of my enterprise. It just means that I have never found anything helpful in syllogistic logic
for my approach to dialectic, not unlike the methods of logic cultivated in analytic philosophy.
Quotations 57.15. (1) Goold [1938], pp. 16 & 18; editing Aristoteles 1 b 25–27
(κατηγορίαι).
IV. Τῶν κατὰ μηδεμίαν συμπλοκήν λεγομένων ἕκαστον ἤτοι οὐσίαν ση-
μαίνει ἢ ποσὸν ἢ ποιὸν ἢ πρός τι ἢ ποῦ ἢ ποτὲ ἢ κεῖσθαι ἢ ἔχειν ἢ ποεῖν ἢ
πάσχειν.
Comment. Note that the second through to the sixth are (accusative cases of)
interrogative pronouns, roughly corresponding to the English ‘how much’, ‘what
sort’, ‘against what’, ‘where’, ‘when’, and the last four are the infinite forms of
verbs, roughly corresponding to the English ‘to lie’, ‘to have’, ‘to do’, ‘to suffer’
(more in the sense of ‘what happens to somebody’). If I understand correctly, the
first one is derived from a form of the present participle of ‘to be’, as the ‘ontos’
from ‘ontology’ is.
(2) Cooke [1938], pp. 17 & 18; translating Aristoteles 1 b 25–27 (categories).
IV. Each uncombined word or expression means one of the following
things:—what (or Substance), how large (that is Quantity), what sort of
thing (that is, Quality), related to what (or Relation), where (that is, Place),
when (or Time), in what attitude (Posture, Position), how circumstanced
(State or Condition), how active, what doing (or Action), how passive, what
suffering (Affection).
Comment. This translation reflects little of the grammatical aspect pointed out in
the preceding comment.
(3) Rolfes [1925], pp. 45 f; translating Aristoteles 1 b 25–27 (Kategorien).
Jedes ohne Verbindung gesprochene Wort bezeichnet entweder eine
Substanz oder eine Quantität oder eine Qualität oder eine Relation oder
ein Wo oder ein Wann oder eine Lage oder ein Haben, oder ein Wirken,
oder ein Leiden.
Comment. This German translation reflects even less of the grammatical structure
of the categories in the Greek original.
Quotation 57.16. Tredennick [1933], p. 35; translating Aristoteles 986 a 23–986 b

3 (metaphysics).
Others of this same school [[of Pythagoreans, as I understand it]] hold
that there are ten principles, which they enunciate in a series of correspond-
ing pairs (i.) Limit and the Unlimited; (ii.) Odd and Even; (iii.) Unity and
Plurality; (iv) Right and Left; (v.) Male and Female; (vi.) Rest and Motion;
(vii.) Straight and Crooked; (viii.) Light and Darkness; (ix.) Good and Evil;
(x.)Straight and Oblong. Apparently Alcmaeon of Croton speculated along
the same lines, and either he derived the theory from them or they from
him; for [Alcmaeon was contemporary with the old age of Pythagoras, and]
his doctrines were very similar to theirs. He says that the majority of things
in the world of men are in pairs; but the contraries which he mentions are
not, as in the case of the Pythagoreans, carefully defined, but are taken at
random, e.g., white and black, sweet and bitter, good and bad, great and
small. Thus Alcmaeon only threw out vague hints with regard to the other
instances of contrariety, but the Pythagoreans pronounced how many and

what the contraries are.
Comment. The part in single square brackets is “probably true, but a later addition”
according to footnote a in the Goold edition. The quotation was instigated by Hegel
[1833], p. 264.
Quotations 57.17. (1) Goold [1938], p. 120; editing Aristoteles 17 a 1–6 (περὶ
ἑρμηνείας).
῎Εστι δὲ λόγος ἅπας μὲν σημαντικός, [[. . .]] ἀποφαντικὸς δὲ οὐ πᾶς, ἀλλ’
ἐν ὧι τὸ ἀληθεύειν ἥ ψεύδεσθαι ὑπάρχει. οὐκ ἐν ἅπασι δε ὑπάρχει, οἷον ἡ εὐχή
λόγος μέν, ἀλλ’ οὐτε ἀληθὴς οὐτε ψευδής.
(2) Cooke [1938], p. 121; translating Aristoteles 17 a 1–6 (on interpretation).
But while every sentence has meaning, [[. . .]] not all can be called propo-
sitions. We call propositions those only that have truth or falsity in them.
A prayer is, for instance, a sentence but neither has truth nor has falsity.
(3) Rolfes [1925], p. 97 f; translating Aristoteles 17 a 1–6 (Lehre vom Satz ).
Es zeigt aber jede Rede etwas an [[. . .]]. Dagegen sagt nicht jede etwas
aus, sondern nur die, in der es Wahrheit oder Irrtum gibt. Das ist aber
nicht überall der Fall. So ist die Bitte zwar eine Rede, aber weder wahr
noch falsch.
Quotations 57.18. (1) Goold [1938], pp. 122 & 124; editing Aristoteles 17 a 25–36
(περὶ ἑρμηνείας).
VI. Κατάφασις δέ ἐστιν ἀπόφανσίς τινος κατά τινος. ἀπόφασις δέ ἐστιν
ἀπόφανσίς τινος ἀπό τινος.
᾿Επεὶ δὲ ἐστι καὶ τὸ ὑπάρχον ἀποφαίνεσθαι ὡς μὴ ὑπάρχον καὶ τὸ μὴ
ὑπάρχον ὡς ὑπάρχον καὶ τὸ ὑπάρχον ὡς ὑπάρχον καὶ τὸ μὴ ὑπάρχον ὡς μὴ
ὑπάρχον, καὶ περὶ τοὺς ἐκτὸς δὲ τοῦ νῦν χρόνους ὡσαύτως, ἅπαν ἄν ἐνδέχοιτο
καὶ ὅ κατέφησέ τις ἀποφῆσαι καὶ ὅ ἀπέφησέ τις καταφῆσαι. ὥστε δῆλον ὅτι
πάσηι καταφάσει ἐστὶν ἀπόφασις ἀντικειμένη καὶ πάσηι ἀποφάσει κατάφασις.
καὶ ἔστω ἀντίφασις τοῦτο, κατάφασις καὶ ἀπόφασις αἱ ἀντικείμεναι. λέγω δὲ
ἀντικεῖσθαι τὴν τοῦ αὐτοῦ κατὰ τοῦ αὐτοῦ, μὴ ὁμωνύμως δέ [[.]]
(2) Cooke [1938], pp. 123 & 125; translating Aristoteles 17 a 25–36 (on interpreta-
tion).
VI. We mean by affirmation a statement affirming one thing of another;
we mean by negation a statement denying one thing of another.
As men can affirm and deny both the presence of that which is present
and the presence of that which is absent and this they can do with a ref-
erence to times that lie outside the present, whatever a man may affirm, it
is possible as well to deny, and whatever a man may deny, it is possible as
well to affirm. Thus, it follows, each affirmative statement will have its own
opposite negative, just as each negative statement will have its affirmative
opposite. Every such pair of propositions we, therefore, shall call contra-
dictories, always assuming the predicates and subjects are really the same
and the terms used without ambiguity.
Comment. This may not sound too bad, but I suspect it marks the point of “origin”
of what fashionable people like to call the “binarism” in “Western thought”. I want
to steer clear from the superficiality of that fashion without, however, giving in to
Aristoteles’ doctrine.
Bejahung ist eine Aussage, die einem etwas zuspricht, Verneinung ist
eine Aussage, die einem etwas abspricht.
Da man aber vom Seienden aussagen kann, daß es nicht ist, und vom
Nichtseienden, daß es ist, und wiederum, vom Seienden, daß es ist, und
vom Nichtseienden, daß es nicht ist, und da das ebenso für die Zeiten au-
ßerhalb der Gegenwart Geltung hat, so läßt sich alles, was einer bejaht,
verneinen und alles, was einer verneint, bejahen, und demgemäß ist offen-
bar jeder Bejahung eine Verneinung und jeder Verneinung eine Bejahung
entgegengesetzt. Und dies, entgegengesetzte Bejahung und Verneinung, soll
Kontradiktion sein. Ich verstehe aber unter Gegensatz, daß dasselbe von
demselben bejaht und verneint wird, aber nicht homonymisch[[.]]
Comment. One curious point in this piece is the translation of τὸ ὑπάρχον. Cooke
speaks of “that which is present” and Rolfes speaks of the “Seienden” (being).
Τὸ μὲν οὖν εἶναι τὸ ὄν ὅταν ἦι , καὶ τὸ μὴ ὀν μὴ εἶναι ὅταν μὴ ἦι , ἀνάγκη·
οὐ μὴν οὐτε τὸ ὄν ἅπαν ἀνάγκη εἶναι οὔτε τὸ μὴ ὄν μὴ εἶναι. οὐ γὰρ ταὐτόν
ἐστι τό ὄν ἅπαν εἶναι ἐξ ἀνάγκης ὅτε ἔστι, καὶ τὸ ἀπλῶς εἶναι ἐξ ἀνάγκης.
ὁμοίως δὲ καὶ ἐπὶ τοῦ μὴ ὀντος. καὶ ἐπὶ τῆς ἀντιφάσεως ὁ αὐτὸς λόγος. εἶναι
μὲν ἠ μὴ εἶναι ἅπαν ἀνάγκη, καὶ ἔσεσθαί γε ἠ μὴ· οὐ μέντοι διελόντα γε εἰπεῖν
θάτερον ἀναγκαῖον.
(2) Cooke [1938], p. 139; translating Aristoteles 19 a 24–29 (on interpretation).
What is must needs be when it is; what is not cannot be when it is not.
However, not all that exists any more than all that which does not comes
about or exists by necessity. That what is must be when ‘it is’ does not
mean the same thing as to say that all things come about by necessity. And
so, too, with that which is not. And with two contradictory statements the
same thing is found to hold good. That is, all things must be or not be, or
must come or not come into being, at this or that time in the future. But
we cannot determinately say which alternative must come to pass.
Comment. What follows is the well known example of a sea-fight taking place to-
morrow.
Daß nun das Seiende ist, wann es ist, und das Nichtseiende nicht ist,
wann es nicht ist, ist notwendig. Gleichwohl ist nicht notwendig, weder
daß alles Seiende ist, noch daß alles Nichtseiende nicht ist, Denn es ist
nicht dasselbe, daß alles Seiende notwendig ist, wann es ist, und daß es
schlechthin notwendig ist, und gleiches gilt von dem Nichtseienden.
Und mit der Kontradiktion hat es dieselbe Bewandtnis. Es ist notwen-

dig daß alles entweder ist oder nicht ist und sein wird oder nicht sein wird.
Es ist aber nicht notwendig, daß man eins von beiden getrennt für sich
behauptet.
Quotations 57.20. (1) Goold [1933], p. 160; editing Aristoteles 1005 b 19–22 (τὰ
μετὰ τὰ φυσικά).
τὸ γὰρ αὐτὸ ἅμα ὑπὰρχειν τε καὶ μὴ ὑπὰρχειν ἀδὺνατον τῶι αὐτῶι και κατὰ
τὸ αὐτό (καὶ ὅσα ἄλλα προσδιορισαίμεθ’ ἄν, ἔστω προσδιορισμένα πρὸς τὰς
λογικὰς δυσχερείας)·
(2) Tredennick [1933], pp. 161; translating Aristoteles 1005 b 19–22 (metaphysics,
Γ, III).
“It is impossible for the same attribute at once to belong and not to belong
to the same thing and in the same relation”; and we must add any further
qualifications that may be necessary to meet logical objections.
Comment. This is it, the classic line: further qualifications must be made according
to needs. Why doesn’t he make them straight away? The task might have looked
too small for Aristoteles, it might look too small for analytic philosophers, and not
at all occur to philosophers in the Continental tradition, but it is what the present
study is all about: what makes ‘further qualifications’ necessary, and where do they
come from? It is helpful to realize that what is evoked here presupposes categories:
‘at once’ (ἅμα ) and ‘in the same relation’ (κατὰ τὸ αὐτό ), even though they don’t
show up explicitly.
(3) Schwarz [1970], p. 89; translating Aristoteles 1005 b 19–22 (Metaphysik, Γ, 3.
Die Axiome und der Satz vom Widerspruch).
[[E]]s ist nicht möglich, daß dasselbe demselben in derselben Beziehung zu-
gleich zukomme [20] und nicht zukomme (und fügen wir noch andere Be-
stimmungen dazu, so deshalb, um logische Einwände zurückzuweisen).
Quotations 57.21. (1) Goold [1933], pp. 198 & 200; editing Aristoteles 1011 b
23–30 (τὰ μετὰ τὰ φυσικά).
VII. ᾿Αλλὰ μὴν οὐδὲ μεταξὺ ἀντιφάσεως ἐνδέχεται εἶναι οὐθέν, ἀλλ’
ἀνάγκη ἢ φάναι ἢ ἀποφάναι ἓν καθ’ ἑνὸς ὁτιοῦν. δῆλον δὲ πρῶτον μὲν ὁρι-
σαμένοις τί τὸ ἀληθὲς καὶ ψεῦδος. τὸ μὲν γὰρ λὲγειν τὸ ὂν μὴ εἶναι ἢ τὸ μὴ
ὂν εἶναι ψεῦδος, τὸ δὲ τὸ ὂν εἶναι καὶ τὸ μὴ ὂν μὴ εἶναι ἀληθὲς, ὥστε καὶ ὁ
λέγων εἶναι ἢ μὴ ἀληθεύσει ἢ ψεύσεται. ἀλλ’ οὔτε τὸ ὂν λέγεται μὴ εἶναι ἢ
εἶναι οὔτε τὸ μὴ ὂν.
(2) Translation: ibid. pp. 199 & 201; translating Aristoteles 1011 b 23–30 (meta-
physics, Γ, VII).
VII. Nor indeed can there be any intermediate between contrary state-
ments, but of one thing we must either assert or deny one thing, whatever
it may be. This will be plain if we first define truth and falsehood. To say
what is is not, or that what is not is, is false; but to say that what is is,
and what is not is not, is true; and therefore also he who says that a thing
is or is not will say either what is true or what is false. But neither what is
nor what is not is said not to be or to be.
Comment. The ‘definition of truth and falsehood’ in this passage was to be em-
ployed by Tarski some two thousand and three hundred years later.23 Interestingly,
it is here employed to back up a claim of tertium non datur.
(3) Schwarz [1970], pp. 107 f; translating Aristoteles 1011 b 23–30 (Metaphysik, Γ ,
7. Der Satz vom ausgeschlossenen Dritten).
Und doch ist es nicht möglich, daß es ein Mittleres zwischen den beiden
Gliedern des Widerspruches gibt, sondern man muß eben eines von beiden
entweder bejahen oder verneinen. [25] Das aber wird klar, wenn man erst
einmal feststellt, was wahr ist und was falsch. Denn zu behaupten, das Sei-
ende sei nicht oder das Nichtseiende sei, ist falsch. Aber zu behaupten, daß
das Seiende sei und das Nichtseiende nicht sei, ist wahr. Es wird demnach
der, der behauptet, daß etwas sei oder nicht sei, die Wahrheit sagen oder die
Unwahrheit. Aber man sagt weder vom Seienden noch vom Nichtseienden,
es sei nicht oder es sei.
ομοίως οι λόγοι αληθεῖς ωσπερ τὰ πράγματα[[·]]
(2) Cooke [1938], pp. 139 & 141; translating Aristoteles 19 a 33–34 (on interpreta-
tion).
[[T]]he truth of propositions consists in corresponding with facts[[.]]
Comment. Nicely taken out of context; made a cornerstone of Tarski’s semantic
conception of truth.23
(3) Rolfes [1925], p. 105; translating Aristoteles 19 a 33–34 (Lehre vom Satz ).
Behauptungen [[sind]] in derselben Weise wahr [[. . .]] wie die Dinge[[.]]
Comment. The German translation doesn’t lend itself so neatly to Tarski’s purpose.
I now turn to metaphysics. Aristoteles himself never used the term ‘meta-
physics’, but spoke of ‘first philosophy’ (πρώτη φιλοσοφία ). According to Hancock
[1967], p. 289, the term “was apparently introduced by the editors (traditionally
by Andronicus of Rhodes in the fist century B.C.) who classified and catalogued”
Aristoteles’ works.
Quotations 57.23. (1) Goold [1933], pp. 146; editing Aristoteles 1003 a 19–33 (τὰ
I. ῎Εστιν ἐπιστήμη τις ἥ θεωρεῖ τὸ ὂν ιἧ ὂν καὶ τὰ τοὺτωι ὑπάρχοντα
καθ’ αὑτό. αὕτη δ’ ἐστὶν οὐδεμιᾶι τῶν ἐν μέρει λεγομένον ἡ αὐτή· οὐδεμία
γὰρ τῶν ἄλλων ἐπισκοπεῖ καθόλου περὶ τοῦ ὄντος ἧι ὂν, ἀλλὰ μέρος αὐτοῦ
τι ἀποτεμόμεναι περὶ τούτου θεωροῦσι τὸ συμβεβηκός, οἷον αἱ μαθηματικαὶ
τῶν ἐπιστημῶν. ἐπεὶ δὲ τὰς ἀρχὰς καὶ τὰς ἀκροτάτας αἰτίας ζητοῦμεν, δῆλον
ὡς φύσεώς τινος αὐτὰς ἀναγκαῖον εἶναι καθ’ αὑτήν. εἰ οὖν καὶ οἰ τὰ στοιχεῖα
23 Cf. quotation 89.1 (3) in these materials and footnote 2 in Tarski [1933], p. 155.
τῶν ὄντων ζητοῦντες ταύτας τὰς ἀρχὰς ἐζήτουν, ἀνάγκη καὶ τὰ στοιχεῖα τοῦ
ὄντος εἶναι μὴ κατὰ συμβεβηκός, ἀλλ’ ἧι ὂν· διὸ καὶ ἡμῖν τοῦ ὄντος ἧι ὂν τὰς
πρώτας αἰτίας ληπτέον.
(2) Tredennick [1933], pp. 147; translating Aristoteles 1003 a 19–33 (metaphysics
Γ).
I. There is a science which studies Being qua Being, and the properties
inherent in it in virtue of its own nature. This science is not the same as
any of the so-called particular sciences, for none of the others contemplates
Being generally qua Being; they divide off some portion of it and study the
attribute of this portion, as do for example the mathematical sciences. But
since it is for the first principles and the most ultimate causes that we are
searching, clearly they must belong to something in virtue of its own nature.
Hence if these principles were investigated by those also who investigated
the elements of existing things, the elements must be elements of Being not
incidentally, but qua Being. Therefore it is of Being qua Being that we too
must grasp the first causes.
Comment. Notice the rendering of “καθ’ αὑτό” as “in virtue of its own nature”. This
may seem plausible in view of the phrasing “φύσεώς τινος . . . εἶναι καθ’ αὑτήν” later
which is also translated as “in virtue of its own nature”, but it still strikes me as
too much talk about “nature”. The emphasis, as I understand it, lies on ‘in itself’,
or ‘an sich’, not on “nature”; its nature is exhausted, as it were, in its ‘an sich’.
(3) Schwarz [1970], p. 82; translating Aristoteles 1003 a 19–33 (Metaphysik Γ).
[21] Es gibt eine Wissenschaft, die das Seiende, insofern es seiend ist, be-
trachtet und das, was ihm an sich zukommt. Diese ist aber mit keiner der
sogenannten Einzelwissenschaften identisch; denn keine der anderen Wis-
senschaften betrachtet allgemein das Seiende, insofern es seiend ist, son-
dern indem sie sich einen Teil vom Seienden herausschneiden, betrachten
sie diesen hinsichtlich seines Akzidens, [25] wie das etwa die mathemati-
schen Wissenschaften tun. Da wir aber die Prinzipien und die höchsten
Ursachen suchen, so ist es klar, daß diese Ursachen einer an sich existieren-
den Natur sein müssen. Wenn nun diejenigen, die die Elemente der Dinge
suchten, eben diese Prinzipien suchten, so waren notwendigerweise auch
die Elemente nicht Elemente des Seienden in akzidentellem Sinne, [30] son-
dern des Seienden insofern es seiend ist. Also müssen auch wir die ersten
Ursachen des Seienden, insofern es seiend ist, erfassen.
Comment. Notice that “καθ’ αὑτό ” is translated as “an sich” und “φύσεώς τινος
. . . εἶναι καθ’ αὑτήν” als “einer an sich existierenden Natur”; this doesn’t satisfy me
either.
Quotations 57.24. (1) Goold [1933], pp. 146 & 148; editing Aristoteles 1003 b
5–11 (τὰ μετὰ τὰ φυσικά).
τὸ [[δὲ]] ὂν λέγεται πολλαχῶς μέν, ἀλλ’ ἄπαν πρὸς μίαν ἀρχήν· τὰ μὲν γὰρ
ὅτι οὐσίαι ὄντα λέγεται, τὰ δ’ ὅτι πάθη οὐσίας, τὰ δ’ ὅτι ὁδὸς εἱς οὐσίαν, ἢ
φθοραὶ ἢ στερήσεις ἢ ποιότητες ἢ ποιοτηκὰ ἢ γεννητικὰ οὐσίας, ἢ τῶν πρὸς
τὴν οὺσίαν λεγομένον, ἢ τούτων τινὸς ἀποφάσεις ἢ οὐσίας· διὸ καὶ τὸ μὴ ὂν

εἶναι μὴ ὂν φαμέν.
(2) Tredennick [1933], pp. 147 & 149; translating Aristoteles 1003 b 5–11 (meta-
physics IV, ii. 3).
“[[B]]eing” is used in various senses, but always with reference to one prin-
ciple. For some things are said to “be” because they are substances; oth-
ers because they are modifications of substance; others because they are
a process towards substance, or destructions or privations or qualities of
substance, or productive or generative of substance or of terms relating to
substance, or negations of certain of these terms or of substance. (Hence
we even say that not-being is not-being.)
Comment. The last sentence is interesting in view of Parmenides.
(3) Schwarz [1970], p. 83; translating Aristoteles 1003 b 5–11 (Metaphysik Γ, 5–11).
[[Man]] spricht [[. . .]] zwar in vielfacher Bedeutung vom Seienden, doch stets
im Hinblick auf ein Prinzip. Man spricht nämlich vom Seienden, weil es
entweder ein Wesen ist oder eine Affektion eines Wesens oder weil es ein
Weg zum Wesen oder weil es Vergehen, Privation, Qualität, Bewirkendes
oder Erzeugendes eines Wesens ist oder von etwas, das in Beziehung auf das
Wesen ausgesagt wird, oder gar, weil es Verneinung von etwas derartigem
oder eines Wesens ist. [10] Daher sagen wir ja auch, ein Nichtseiendes sei
nichtseiend.
Comment. Note the German translation of οὐσία: Wesen.
Quotations 57.25. (1) Goold [1933], pp. 296 & 298; editing Aristoteles 1026 a
29–1026 b 1 (τὰ μετὰ τὰ φυσικά).
εἰ μὲν οὖν μὴ ἔστι τις ἑτέρα οὐσία παρὰ τὰς φύσει συνεστηκυίας, ἡ φυσική ἄν
εἴη πρώτη ἐπιστήμη· ἐι δ’ ἐστι τις οὐσία ἀκίνητος, αὕτη προτέρα καὶ φιλοσοφία
πρώτη, καὶ καθόλου οὕτος ὅτι πρώτη· καὶ περὶ τοῦ ὄντος ἧι ὄν, ταύτης ἄν εἴη
θεωρῆσαι, καὶ τί ἐστι καὶ τὰ ὑπάρχοντα ιἧ ὄν.
II. ΄Αλλ’ ἐπεὶ τὸ ὂν τὸ ἁπλῶς λεγόμενον λέγεται πολλαχῶς, ὧν ἓν μὲν
ἦν τὸ κατὰ συμβεβηκός, ἕτερον δὲ τὸ ὡς ἀληθές, καὶ τὸ μὴ ὄν ἠς τὸ ψεῦδος,
παρὰ ταῦτα δ’ ἐστὶ τὰ σχήματα τῆς κατηγορίας, οἷον τὸ μέν τί, τὸ δὲ ποιόν,
τὸ δὲ ποσόν, τὸ δὲ πού, τὸ δὲ ποτέ, καὶ εἴ τι ἄλλο σημαίνει τὸν τρόπον τοῦτον·
ἔτι παρὰ ταῦτα πάντα τὸ δυνάμει καὶ ἐνεργείαι · — ἐπεὶ δὴ πολλαχῶς λέγεται
τὸ ὄν, πρῶτον περὶ τοῦ κατὰ συμβεβηκὸς λεκτέον, ὀτι οὐδεμία ἐστὶ περὶ αὐτὸ
θεωρία.
(2) Tredennick [1933], pp. 297 & 299; translating Aristoteles 1026 a 29–1026 b 1
(metaphysics, E, VI. I.12–II.1).
[[I]]f there is not some other substance besides those which are naturally
composed, physics will be the primary science; but if there is a substance
which is immutable, the science which studies this will be prior to physics,
and will be primary philosophy, and universal in this sense, that it is pri-
mary. And it will be the province of this science to study Being qua Being;
what it is, and what the attributes are which belong to it qua Being.
II. But since the simple term “being” is used in various senses, of which
we saw that one was accidental, and another true (not-being being used in
the sense of “false”); and since besides these there are the categories, e.g.,
the “what,” quality, quantity, place, time, and any other similar meanings;
and further besides all these the potential and actual : since the term “being”
has various senses, it must first be said of what “is” accidentally, that there
can be no speculation about it.
(3) Schwarz [1970], p. 157; translating Aristoteles 1026 a 29–1026 b 1 (Metaphysik,
E).
Wenn es nun neben den von Natur bestehenden Wesen nicht ein davon
verschiedenes Wesen gibt, so dürfte wohl die Naturwissenschaft die erste
Wissenschaft sein. Wenn es aber ein unbewegtes Wesen gibt, so ist wohl
dies das frühere und die Philosophie [30] die erste und allgemein, weil sie die
erste ist. Und es dürfte wohl ihre Aufgabe sein, das Seiende zu betrachten,
insofern es ein Seiendes ist, sowohl sein Was als auch das ihm Zukommende,
insofern es seiend ist.
2. [[. . .]] Da aber das Seiende, das man schlechthin das Seiende nennt,
in vielfachen Bedeutungen ausgesagt wird, von denen die eine das Akzi-
dentelle bezeichnete, die andere das Seiende im Sinne des Wahren [35] und
das Nichtseiende im Sinne des Falschen, außerdem noch die Figuren der
Aussageweise (wie etwa das Was, das Quale, das Quantum, das Wo, das
Wann und was sonst noch ähnliche Bedeutungen anzeigt) und weiter neben
alledem noch das dem Vermögen und der Verwirklichung nach Seiende —
da also das Seiende in vielfachen Bedeutungen ausgesagt wird, so muß man
zuerst von dem im akzidentellen Sinn Seienden sagen, daß es davon keine
wissenschaftliche Betrachtung gibt.
Comment. Note the translation of τὰ σχήματα τῆς κατηγορίας as “die Figuren der
Aussageweise”.
Quotations 57.26. (1) Goold [1933], pp. 312; editing Aristoteles 1028 b 2–8 (τὰ
καὶ δὴ καὶ τὸ πὰλαι τε καὶ νῦν καὶ ἀεὶ ζητούμενον καὶ ἀεὶ ἀπορούμενον, τί τὸ
ὄν, τοῦτό ἐστι, τίς ἡ οὐσία· τοῦτο γὰρ οἱ μὲν ἓν εἶναί φασιν, οἱ δὲ πλείω ἢ ἕν,
καὶ οἱ μὲν πεπερασμένα, οἱ δὲ ἄπειρα· διὸ καὶ ἡμῖν καὶ μάλιστα καὶ πρῶτον καὶ
μόνον ὡς εἰπεῖν περὶ τοῦ οὕτως ὄντος θεωρητέον τί ἐστιν.
(2) Tredennick [1933], pp. 313; translating Aristoteles 1028 b 2–8 (metaphysics, VII,
i, 7).
Indeed, the question which was raised long ago, is still and always will be,
and which always baffles us—“What is Being?”—is in other words “What
is substance?” Some say that it is one; others, more than one; some, finite;
others infinite. And so for us too our chief and primary and practically our
only concern is to investigate the nature of “being” in the sense of substance.
Comment. Note the translation of τί τὸ ὄν as “What is Being?”, and that of τίς ἡ
οὐσία as “What is substance?”.
(3) Schwarz [1970], p. 165; translating Aristoteles 1028 b 2–8.

Und die Frage, die bereits von alters her erhoben wurde, die auch heute
erhoben wird und immer erhoben werden und Gegenstand der Ratlosigkeit
sein wird, was nämlich das Seiende sei, bedeutet nichts weiter als, was das
Wesen sei (von dem nämlich sagen die einen, es sei Eines, [5] die anderen es
sei mehr als Eines, einige sagen, es sei begrenzt, andere, es sei unbegrenzt).
Deshalb müssen wir auch hauptsächlich und zuerst und sozusagen einzig
betrachten, was das in diesem Sinne Seiende ist.
I close this paragraph with a quotation regarding AR4 (epistemology). I do not

include the Greek original, but only add, in a few cases, the Greek word in double
square brackets.
Quotation 57.27. Cooke [1938], pp. 55, 57, and 59; translating Aristoteles 7 b 15
– 8 a 13 (categories).
Correlatives are commonly held to come into existence together, and
this for the most part is true, as, for instance, of double and half. [[. . .]] How-
ever, the view that correlatives come into being together does not appear
true at all times, for it seems that the object of knowledge is prior to, exists
before, knowledge. We gain knowledge, commonly speaking, of things that
already exist, for in very few cases or non can our knowledge have come
into being along with its own proper object.
Should the object of knowledge be removed, then the knowledge itself
will be cancelled. The converse of this is not true. If the object no longer
exists, there can no longer be any knowledge, there being now nothing to
know. If, however, of this or that object no knowledge has yet been acquired,
yet that object itself may exist. Take the squaring of the circle, for instance,
if that can be called such an object. Although it exists as an object, the
knowledge does not yet exist. If all animals ceased to exist, there would then
be no knowledge at all, though there might be in that case, notwithstanding,
be still many objects of knowledge.
The same may be said of perception [[αἰσθάσεως]] . The object [[αἰσθη-
τὸν]], I mean would appear to be prior to the act of perception. Suppose that
you cancel the perceptible [[αἰσθητὸν ]]; you cancel the perception as well.
Take away or remove the perception, the perceptible still may exist. For the
act of perception implies or involves, first, a body perceived, then a body in
which it takes place. Therefore, if you remove the perceptible, body itself is
removed, for the body itself is perceptible. And, body not being existence,
perception must cease to exist. Take away the perceptible, then, and you
take away also perception. But the taking away of perception does not take
such objects away. If the animal itself is destroyed, then perception is also
destroyed. But perceptibles yet will remain, such as body, heat, sweetness
and bitterness and everything else that is sensible.
Perception, further, comes into being along with the subject perceiv-
ing—that is, with the live thing itself. The perceptible, however, is prior
to the animal and to perception. For such things as water and fire, out of
which are composed living being, exist before any such beings and prior to
all acts of perception. The perceptible, so we conclude, would appear to be
prior to perception.
Comment. Contrast Berkeley, as for instance in quotation 58.11 (4) below.
§ 58. The British empiricists
The British empiricists were in no way explicitly concerned with dialectics; they
were concerned, however, with problems of the kind that have been addressed in
quotations 57.10 and 57.27 above. It is through their influence on Kant that the
British empiricists, in particular Locke and Hume, had a decisive impact on the
making of the modern notion of dialectic. The present paragraph attempts to illus-
trate this influence.
Locke did not show what Hegel called the hardness of the concept, or, as Rus-
sell put it, he was “sensible” enough not to draw the consequences from what he
proposed.24 But, it seems to me, he has provided the starting point for a curious
philosophical development which I find beautifully summed up in Kant’s formula-
tion that it is the “transcendental realist who afterwards plays the part of empirical
idealist.” 25 It strikes me as somewhat similar in character to Cantor’s diagonal
method. Berkeley’s conclusion can be found in quotation 58.14 below.
Because of my lack of familiarity with the philosophy of the British empiricists,
this paragraph can only give some hints.
58a. Locke. British empiricism is commonly said to have begun with Locke’s
Essay Concerning Human Understanding. In this essay, Locke puts forward an ex-
tremely simple, almost simplistic — and thus attractive, at least to some people —
view of how human understanding comes to achieve knowledge. To make my point
clear: if it weren’t for the difficulties this view encounters when taken to its logical
implications,26 there would be nothing interesting in it as regards my enterprise.
Features 58.1. The salient features in Locke’s empiricism which I shall concentrate
on are the following:
(LK1) sensation and reflection are the exclusive sources of knowledge;
(LK2) our knowledge is only conversant about the mind’s ideas;
(LK3) introspection as a method;
(LK4) distinction of primary and secondary properties.
Remark 58.2. The two Roman digits and one Arabic digit in brackets after page
numbers relating to Locke [1690] (An Essay Concerning Human Understanding)
indicate the book, chapter, and paragraph, respectively.
24 Cf. quotation 92.7 (2) on p. 92.7 of these materials.

25 Kant [1781] (A), p. 369; cf. quotation 60.5 (3) in these materials.
26 Cf. Russell’s comment in quotation 92.7 (2) in these materials.
§ 58. THE BRITISH EMPIRICISTS 741
Quotations 58.3. (1) Locke [1690], p. 43 (I.I.1).

The Understanding, like the Eye, whilst it makes us see and perceive all
other Things, takes no notice of itself; And it requires Art and Pains to set
it at a distance and make it its own Object.
Comment. This sounds to me like the dawning of an awareness of the relevance
of an investigation into what I want to call a medium of knowledge. While Locke
clearly sees “arts and pains” being required, he does not seem to expect fundamental
difficulties. It may be worthwhile at this early stage to indicate the direction this is
going to take: Kant’s critique, and Hilbert’s proof theory as two attempts to set a
tool — of some kind — “at a distance and make it its own Object.”
(2) Locke [1690], p. 43 (I.I.2).
§ 2. This, therefore, being my Purpose to enquire into the Original,
Certainty, and Extent of humane Knowledge[[.]]
Quotations 58.4. (1) Locke [1690], p. 47 (I.I.4).

I must [[. . .]] beg pardon of my Reader for the frequent use of the word Idea
[[. . .]]. It being that Term which, I think, serves best to stand for whatsoever
is the Object of the Understanding when a Man thinks, I have used it to
express whatever is meant by Phantasm, Notion, Species, or whatever it is
which the Mind can be employ’d about in thinking[[.]]
(2) Locke [1690], p. 105 (II.I.4).
External Material things, as the Objects of SENSATION, and the Oper-
ations of our own Minds within, as the Objects of REFLECTION, are to
me, the only Originals from whence all our Ideas take their beginnings.
(3) Locke [1690], p. 106 (II.I.5).
External Objects furnish the Mind with the Ideas of sensible qualities, which
are all those different perceptions they produce in us: And the Mind fur-
nishes the understanding with ideas of its own operations.
Remark 58.5. Diagram 103.5 in the groundworks aims at visualizing these re-
lations.
Quotations 58.6. (1) Locke [1690], p. 120 (II.II.2).

The Dominion of Man, in this little World of his own Understanding, being
muchwhat the same, as it is in the great World of visible things; wherein
his Power, however managed by Art and Skill, reaches no farther, than to
compound and divide the Materials, that are made to his Hand; but can do
nothing towards the making the least Particle of new Matter, or destroying
one Atome of what is already in Being. The same inability, will every one
find in himself, who shall go about to fashion in his Understanding any
simple Idea, not received in by his Senses, from external objects; or by
reflection from the operations of his own mind about them. I would have
any one try to fancy any taste, which had never affected his palate; or frame
the Idea of Scent he had never smelt: And when he can do this, I will also
conclude, that a blind Man hath Ideas of Colours, and a deaf Man true
distinct Notions of Sounds.
Comment. Note the emphasis on the non-constitutive role of understanding: “in-
ability”.
(2) Locke [1690], p. 120 (II.II.3).
I think, it is not possible, for any one to imagine any other Qualities in
Bodies, howsoever constituted, whereby they can be taken notice of, besides
Sounds, Tastes, Smells, visible and tangible Qualities. And had Mankind
been made but with four Senses, the Qualities then, which are the Object of
the Fifth Sense, had been as far from our Notice, Imagination, and Concep-
tion, as now any belonging to a Sixth, Seventh, or Eight Sense, can possibly
be[[.]]
Quotations 58.7. (1) Locke [1690], pp. 163 (II.XII.1).

[[A]]s the Mind is wholly Passive in the reception of all its simple Ideas, so
it exerts several acts of its own, whereby out of its simple Ideas, as the
Materials and Foundations of the rest, the others are framed. The Acts of
the Mind wherein it exerts its Power over its simple Ideas are chiefly these
three, 1. Combining several simple Ideas into one compound one, and thus
all Complex Ideas are made. 2. The 2d. is bringing two Ideas, whether
simple or complex, together; and setting them by one another, so as to take
a view of them at once, without uniting them into one; by which way it gets
all its Ideas of Relations. 3. The 3d. is separating them from all other Ideas
that accompany them in their real existence; this is called Abstraction: And
thus all its General Ideas are made.
Comment. Note: acts of mind as regards the manipulation of simple ideas, but
“wholly passive” in their reception.
(2) Locke [1690], p. 164 (II.XII.2).
§ 2. In this faculty of repeating and joining together its Ideas, the Mind
has great power in varying and multiplying the Objects of its Thoughts,
infinitely beyond what Sensation or Reflection furnished it with: But all
this still confined to those simple Ideas which it received from those two
Sources, and which are the ultimate Materials of all its Compositions. For
simple Ideas are all from things themselves, and of these the Mind can have
no more, nor other than what are suggested to it. It can have no other Ideas
of sensible Qualities, than what come from without by the Senses; nor any
Ideas of other kind of Operations of a thinking Substance, than what if
finds in itself[[.]]
(3) Locke [1690], p. 164 (II.XII.3).
§ 3. Complex Ideas, however compounded and decompounded, though
their number be infinite, and the variety endless, wherewith they fill, and
entertain the Thoughts of Men; yet, I think, they may be all reduced under
these three Heads.
1. Modes.
2. Substances.
3. Relations.
(4) Locke [1690], pp. 300 f (II.XXIII.9).
§ 9. The Ideas that make our complex ones of corporeal Substances, are
of these three sorts. First, The Ideas of the primary Qualities of things,
which are discovered by our Senses, and are in them even when we perceive
them not, such are the Bulk, Figure, Number, Situation, and Motion of the
parts of Bodies, which are really in them, whether we take notice of them or
not. Secondly, the sensible secondary Qualities, which depending on these,
are nothing but the Powers, those Substances have to produce several Ideas
in us by our Senses; which Ideas are not in the things themselves, otherwise
than as any thing is in its Cause. Thirdly, The aptness we Consider in any
Substance, to give or receive such alterations of primary Qualities, as that
the Substance so altered, should produce in us different Ideas from what
it did before, these are called active and passive Powers: All which Powers,
as far as we have any Notice or Notion of them, terminate only in sensible
simple Ideas. For whatever alteration a Load-stone has the power to make
in the minute Particles of Iron, we should have no Notion of any Power it
had at all to operate on Iron, did not its sensible Motion discover it; and I
doubt not, but there are a thousand Changes, that Bodies we daily handle,
have a Power to cause in one another, which we never suspect, because they
never appear in sensible effects.
Comment. The first part provides a good characterization of what Kant later called
transcendental realism: number (for instance) is in the things themselves.
Quotations 58.8. (1) Locke [1690], p. 525 (IV.I.1).

§ 1. Since the Mind, in all its Thoughts and Reasonings, hath no other
immediate Object but its own Ideas, which it alone does or can contemplate,
it is evident that our Knowledge is only conversant about them.
§ 2. Knowledge then deems to me to be nothing but the perception of
the connexion and agreement, or disagreement and repugnancy of any of
our Ideas. [[. . .]]
§ 3. But to understand a little more distinctly, wherein this agreement
or disagreement consists, I think we may reduce it all to these four sorts:
1. Identity, or Diversity.
2. Relation.
3. Co-existence, or necessary connexion.
4. Real Existence.
(2) Locke [1690], pp. 538 f (IV.III.1).
§ 1. Knowledge, as has been said, lying in the Perception of the Agree-
ment, or Disagreement, of any of our Ideas, it follows from hence, That,
First, We can have Knowledge no farther than we have Ideas.
§ 2. Secondly, That we can have no Knowledge farther, than we can have
Perception of that Agreement, or Disagreement: Which Perception being,
1. Either by Intuition, or the immediate comparing any two Ideas; or, 2. By

Reason, examining the Agreement, or Disagreement of two Ideas, by the
Intervention of some others: Or, 3. By Sensation, perceiving the Existence
of particular Things. Hence it also follows,
§ 3. Thirdly, That we cannot have an intuitive Knowledge, that shall
extend it self to all our Ideas, and all that we would know about them;
because we cannot examine and perceive all the Relations they have one to
another by juxta-position, or an immediate comparison one with another.
[[. . .]]
§ 4. Fourthly, It follows also, from what is above observed, that our
rational Knowledge, cannot reach to the whole extent of our Ideas. [[. . .]]
§ 5. Fifthly, Sensitive Knowledge reaching no farther than the Existence
of Things actually present to our Senses, is yet much narrower than either
of the former.
§ 6. From all which it is evident, that the extent of our Knowledge comes
not only short of the reality of Things, but even of the extent of our own
Ideas.
58b. Berkeley.27 What Berkeley targets is a peculiar inconsistency in con-

cepts such as substance, matter etc., i.e., essentially what Locke refuses to see.
What he seems to see is that on the one hand these concepts are trying to reach
at something beyond the realm of the conceptual, on the other hand, by their very
nature, they are vested in the conceptual. This proves Berkeley to be the most
dialectical thinker among the British empiricists.
Features 58.9. I am interested in the following features in Berkeley:

(BK1) esse is percipi; the existence of an idea consists in being perceived;28
(BK2) introspection as a method;29
(BK3) it is a contradiction to talk about conceiving what is unconceived.30
Quotations 58.10. (1) Berkeley [1710], p. 136 (Introduction, §1).

[[N]]o sooner do we depart from sense and instinct to follow the light of a su-
perior principle, to reason, meditate, and reflect on the nature of things, but
a thousand scruples spring up in our minds concerning those things which
before we seemed fully to comprehend. Prejudices and errors of sense do
from all parts discover themselves to our view, and, endeavoring to correct
these by reason, we are insensibly drawn into uncouth paradoxes, difficul-
ties, and inconsistencies, which multiply and grow upon us as we advance in
speculation, till at length having wandered through many intricate mazes,
we find ourselves just where we were, or, which is worse, sit down in a
forlorn scepticism.
27 The attention to Berkeley in these materials owes its existence to a joint undergraduate
course in metaphysics with Graham Priest at the University of Western Australia in 1988.
28 Cf. quotations 58.11, in particular 58.11 (4) below.
29 Cf. quotation 58.12 below.
30 Cf. quotation 58.13 (3) below.
Comment. I suggest that the mention of scepticism be seen in the light of Kant’s re-
mark that it is the “transcendental realist who afterwards plays the part of empirical
idealist”.31
(2) Berkeley [1710], p. 136 f (Introduction, §3).
It is a hard thing to suppose that right deductions from true principles
should ever end in consequences which cannot be maintained or made con-
sistent.
Comment. In this context, I suggest a comparison with Russell’s remark in quota-
tion 86.7 (1) in these materials.
(3) Berkeley [1710], p. 137 (Introduction, §3).
I am inclined to think that the far greater part, if not all, of those difficul-
ties which have hitherto amused philosophers, and blocked up the way to
knowledge, are entirely owing to ourselves—that we first raised a dust and
then complain we cannot see.
4. My purpose therefore is, to try if I can discover what those princi-
ples are which have introduced all that doubtfulness and uncertainty, those
absurdities and contradictions[[.]]
Comment. As regards the idea of us possibly fooling ourselves, cf. the motto to this
part taken from Hegel, and also Bohm in quotation 93.17 (2) in these materials;
also variations 100.4.
Quotations 58.11. (1) Berkeley [1710], pp. 151 f (§2).

[[B]]esides all that endless variety of ideas or objects of knowledge, there is
likewise something which knows or perceives them, and exercises divers op-
erations, as willing, imagining, remembering, about them. This perceiving,
active being is what I call mind, spirit, soul, or myself. By which words I
do not denote any one of my ideas, but a thing entirely distinct from them
wherein they exist, or, which is the same thing, whereby they are perceived;
for the existence of an idea consists in being perceived.
(2) Berkeley [1710], p. 164 (§35).
I do not argue against the existence of any one thing that we can apprehend,
either by sense or reflection. That the things I see with my eyes and touch
with my hands do exist, really exist, I make not the least question. The
only thing whose existence we deny is that which philosophers call Matter
or corporeal substance.
(3) Berkeley [1710], pp. 164 f (§37).
It will be urged that thus much at least is true, to wit, that we take away
all corporeal substances. To this my answer is, that if the word substance
be taken in the vulgar sense, for a combination of sensible qualities, such
as extension, solidity, weight, and the like—this we cannot be accused of
taking away: but if it be taken in a philosophic sense, for the support of
accidents or qualities without the mind—then indeed I acknowledge that
31 Cf. quotation 60.5 (3) in the next chapter.
we take it away, if one may be said to take away that which never had any
existence, not even in the imagination.
(4) Berkeley [1710], p. 152 (§3).
The table I write on I say exists—that is, I see and feel it; and if I were out
of my study I should say it existed—meaning thereby that if I was in my
study I might perceive it. There was an odor, that is, it was smelt; there
was a sound, that is it was heard; a colour or figure, and it was perceived
by sight or touch. This is all that I can understand by these and the like
expressions. For as to what is said of the absolute existence of unthinking
things without any relation to their being perceived, that seems perfectly
unintelligible. Their esse is percipi, nor is it possible they should have any
existence out of the minds or thinking things which perceive them.
Quotations 58.12. (1) Berkeley [1710], p. 159 (§22).

It is but looking into your own thoughts, and so trying whether you can
conceive it possible for a sound or figure, or motion, or colour to exist
without the mind or unperceived. This easy trial may perhaps make you
see that what you contend for is a downright contradiction. [[. . .]] [[I]]f you can
but conceive it possible for one extended movable substance, or in general
for any one idea, or anything like an idea, to exist otherwise than in a mind
perceiving it, I shall readily give up the cause[[.]]
(2) Berkeley [1710], p. 194.
108. Those men who frame general rules from the phenomena and after-
wards derive the phenomena from those rules, seem to consider signs rather
than causes. A man may well understand natural signs without knowing
their analogy, or being able to say by what rule a thing is so and so. And,
as it is very possible to write improperly, through too strict an observance
of general grammar rules; so, in arguing from general laws of nature, it is
not impossible we may extend the analogy too far, and by that means run
into mistakes.
109. As in reading other books a wise man will choose to fix his thoughts
on the sense and apply it to use, rather than lay them out in grammatical
remarks on the language; so, in perusing the volume of nature, it seems
beneath the dignity of the mind to affect an exactness in reducing each
particular phenomenon to general rules, or shewing how it follows from
them.
Comment. My point here is that what we run into is dialectic, which is by no means
mistakes.
(3) Berkeley [1710], p. 160 (§24).
It is very obvious, upon the least inquiry into our own thoughts, to know
whether it be possible for us to understand what is meant by the absolute
existence of sensible objects in themselves, or without the mind. To me it is
evident those words mark out either a direct contradiction, or else nothing
at all. And to convince others of this, I know no readier or fairer way than
to entreat they would calmly attend to their own thoughts; and by this
attention the emptiness or repugnancy of those expressions does appear,
surely nothing more is requisite for their conviction. It is on this therefore
that I insist, to wit, that the absolute existence of unthinking things are
words without a meaning, or which include a contradiction. This is what I
repeat and inculcate, and earnestly recommend to the attentive thoughts
of the reader.
Comment. I shall later replace a criterion such as “inquiry into our own thoughts”
by a criterion that doesn’t succumb to psychology and is closer to logical deduction,
such as consistency of a concept.
Quotations 58.13. (1) Berkeley [1713], p. 245.

Phil. How say you, Hylas, can you see a thing which is at the same time
unseen?
Hyl. No, that were a contradiction.
Phil. Is it not as great a contradiction to talk of conceiving a thing
which is unconceived?
Hyl. It is.
Comment. Given a certain dialetheist interpretation of the blackboard paradox,32 I

suppose, it would even be possible to “see a thing which is at the same time unseen”,
at least for a dialetheist.
(2) Berkeley [1713], p. 245. Philonous speaking.

[[W]]hat is conceived is surely in the mind.
(3) Berkeley [1713], pp. 245 f.
Hyl. [[. . .]] [[N]]ow I plainly see, that all I can do is to frame ideas in my
own mind. I may indeed conceive in my own thoughts the idea of a tree, or
a house, or a mountain, but that is all. And this is far from proving, that I
can conceive them existing out of the minds of all spirits.
Phil. You acknowledge then that you cannot possibly conceive how any
one corporeal sensible thing should exist otherwise than in a mind.
Hyl. I do.
Comment. This is a central point in my presentation of the British Empiricists

and in one or the other form it is the topic of my study. An attempt at a logical
reformulation of the gist of Berkeley’s reasoning can be found in §103.
(4) Berkeley [1713], p. 218.

[[T]]here is no such thing as what philosophers call material substance.
32 Cf. proposition 142.13 and remarks 142.14 (2) and 113.2 (3) in the groundworks.
Quotation 58.14. Berkeley [1710], p. 154 (§9).

9. Some there are who make a distinction betwixt primary and sec-
ondary qualities. By the former they mean extension, figure, motion, rest,
solidity or impenetrability, and number; by the latter they denote all other
sensible qualities, as colors, sounds, tastes, and so forth. The ideas we have
of these they acknowledge not to be the resemblances of anything existing
without the mind, or unperceived, but they will have our ideas of the pri-
mary qualities to be patterns or images of things which exist without the
mind, in an unthinking substance which they call matter. By matter, there-
fore, we are to understand an inert, senseless substance, in which extension,
figure, and motion do actually subsist. But it is evident from what we have
already shown, that extension, figure, and motion are only ideas existing
in the mind, and that an idea can be like nothing but another idea, and
consequently neither they nor their archetypes can exist in an unperceiving
substance. Hence, it is plain that the very notion of what is called matter,
or corporeal substance, involves a contradiction in it.

That the principles laid down by mathematicians are true, and their way
of deduction from those principles clear and incontestible, we do not deny;
but we hold there may be certain erroneous maxims of greater extent than
the object of mathematics, and for that reason not expressly mentioned,
though tacitly supposed throughout the whole progress of that science, and
that the ill effects of those secret unexamined errors are diffused through
all the branches thereof. To be plain, we suspect the mathematicians are
as well as other men concerned in the errors arising from the doctrine of
abstract general ideas, and the existence of objects without the mind.
Comment. The reader might guess that I shall later locate this “error” in some way
in the use of classical logic, more to the point, the use of contraction; as regards set
theory, this finds its expression, for instance, in the use of the axiom of choice.
(2) Berkeley [1710], p. 201 (§122).
In arithmetic [[. . .]] we regard not the things, but the signs, which neverthe-
less are not regarded for their own sake, but because they direct us how to
act with relation to things, and dispose rightly of them.
Comment. This strikes me as a curious merger of Hilbert (emphasis on signs: in
the beginning is the symbol, cf. quotation 76.9 (2) in these materials) and Frege
(emphasis of the aspect that symbols express thought, as documented in quotation
71.25 (1) of these materials).

[[S]]everal difficult and obscure questions, on which abundance of speculation
has been thrown away, are entirely banished from philosophy. “Whether
corporeal substance can think,” “whether matter be infinitely divisible,”
and “how it operates on spirit”—these and like inquiries have given infinite
amusement to philosophers in all ages; but, depending on the existence of

matter, they have no longer any place on our principles.
(2) Berkeley [1710], pp. 177 f (§65).
[[T]]he connection of ideas does not imply the relation of cause and effect,
but only of a mark or sign with the thing signified. The fire which I see is
not the cause of the pain I suffer upon my approaching it, but the mark
that forewarns me of it. In like manner the noise that I hear is not the
effect of this or that motion or collision of the ambient bodies, but the sign
thereof.
(3) Berkeley [1710], pp. 185 f (§87).
87. Color, figure, motion, extension, and the like, considered only as so
many sensations in the mind, are perfectly known, there being nothing in
them which is not perceived. But, if they are looked on as notes or images,
referred to things or archetypes existing without the mind, then are we
involved all in scepticism. We see only the appearances, and not the real
qualities of things. What may be the extension, figure, or motion of anything
really and absolutely, or in itself, it is impossible for us to know, but only
the proportion or relation they bear to our senses. Things remaining the
same, our ideas vary, and which of them, or even whether any of them at
all represent the true quality really existing in the thing, it is out of our
reach to determine. So that, for aught we know, all we see, hear, and feel
may be only phantom and vain chimera, and not at all agree with the real
things existing in our rerum natura. All this scepticism follows from our
supposing a difference between things and ideas, and that the former have
a subsistence without the mind or unperceived.
Comment. Note the anticipation of Kant: “We see only the appearances . . . ”. Cf.
also variations 106.8 in the groundworks.
(4) Berkeley [1710], p. 187 (§90).
90. Ideas imprinted on the senses are real things, or do really exist; this
we do not deny, but we deny they can subsist without the minds which
perceive them, or that they are resemblances of any archetypes existing
without the mind; since the very being of a sensation or idea consists in
being perceived, and an idea can be like nothing but an idea. Again, the
things perceived by sense may be termed external, with regard to their
origin: in that they are not generated from within by the mind itself, but
imprinted by a Spirit distinct from that which perceives them.
58c. Hume. Like Locke, Hume has little to offer with regard to my enterprise
besides providing a good example of what I consider an effort to find something in
the wrong place.33 But then there is Kant’s mention that it was the “remembering
David Hume” that awoke him from his “dogmatic slumber” 34 and it is sometimes
intriguing to see ideas in preparation that were to take a central position in Kant’s
33 Cf. paraphrase 109.6 on p. 1494 in the groundworks.
34 Cf. quotation 59.8 (1) in the next paragraph.
thinking. The inherent fallacy in Hume’s considerations may already become clear
from Kant’s criticism and further remark 118.12 in the groundworks, but I shall
dedicate a little more space to it in section 103b in the groundworks.
Quotations 58.17. (1) Hume [1739], p. xvii.

Principles taken upon trust, consequences lamely deduced from them, want
of coherence in the parts, and of evidence in the whole, these are every
where to be met with in the systems of the most eminent philosophers, and
seem to have drawn disgrace upon philosophy itself.
Comment. Nicely put, still valid, and perfectly self-applicable.
(2) Hume [1739], pp. xix f.
’Tis evident, that all the sciences have a relation, greater or less, to
human nature; [[. . .]] Even Mathematics, Natural Philosophy, and Natural
Religion, are in some measure dependent on the science of MAN; since they
lie under the cognizance of men, and are judged of by their powers and
faculties. ’Tis impossible to tell what changes and improvements we might
make in these sciences were we thoroughly acquainted with the extent and
force of human understanding, and cou’d explain the nature of the ideas
we employ, and of the operations we perform in our reasonings.[[. . .]]
[[. . .]]
Here then is the only expedient, from which we can hope for success in
our philosophical researches, to leave the lingring method, which we have
hitherto followed, and instead of taking now and then a castle or village of
the frontier, to march up directly to the capital or center of these sciences,
to human nature itself.; [[. . .]] There is no question of importance, whose
decision is not compriz’d in the science of man; and there is none, which
can be decided with any certainty, before we become acquainted with that
science. In pretending therefore to explain the principles of human nature,
we in effect propose a compleat system of the sciences. built on a foundation
almost entirely new, and the only one upon which they can stand with any
security.
And as the science of man is the only solid foundation for the other
sciences, so the only solid foundation we can give to this science itself must
be laid on experience and observation.
Comment. May I continue the last sentence: . . . and experience and observation are
never secure.
Quotations 58.18. (1) Hume [1739], p. 1

All the perceptions of the human mind resolve themselves into two
distinct kinds, which I shall call Impressions and Ideas. The difference
betwixt these consists in the degrees of force and liveliness with which they
strike upon the mind, and make their way into our thought or conscious-
ness. Those perceptions, which enter with most force and violence, we may
name impressions; and under this name I comprehend all our sensations,
passions and emotions, as they make their first appearance in the soul. By
ideas I mean the faint images of these in thinking and reasoning; such as,
for instance, are all the perceptions excited by the present discourse, ex-
cepting only, those which arise from the sight and touch, and excepting the
immediate pleasure or uneasiness it may occasion.
(2) Hume [1739], p. 2.
There is another division of our perceptions, which it will be convenient
to observe, and which extends itself both to our impressions and ideas. This
division is into Simple and Complex. Simple perceptions or impressions
and ideas are such as to admit of no distinction nor separation. The complex
are the contrary to these, and may be distinguished into parts. Tho’ a par-
ticular colour, taste, and smell are qualities all united together in this apple,
’tis easy to perceive they are not the same, but are at least distinguishable
from each other.
(3) Hume [1739], pp. 7 f.
Impressions may be divided into two kinds, those of Sensation and those
of Reflexion. The first kind arises in the soul originally, from unknown
causes. The second is derived in a great measure from our ideas, and that in
the following order. An impression first strikes upon the senses, and makes
us perceive heat or cold, thirst or hunger, pleasure or pain of some kind or
other. Of this impression there is a copy taken by the mind, which remains
after the impression ceases; and this we call an idea. This idea of pleasure or
pain, when it returns upon the soul, produces the new impressions of desire
and aversion, hope and fear, which may properly be called impressions of
reflexion, because derived from it. These again are copied by the memory
and imagination, and become ideas; which perhaps in their turn give rise to
other impressions and ideas. So that the impressions of reflexion are only
antecedent to their correspondent ideas; but posterior to those of sensation,
and deriv’d from them.
Comment. There seems to be a lot of copying to be done and the main question
for XeroxTM would probably be whether in colour, digital or what have you; for a
lawyer it would probably be the question of copyright ; for me the question will be
more along the line of asking whether all this copying is all that harmless. (I am
not suggesting that it is going to affect the content, but something that comes very
close to that.)
(4) Hume [1739], p. 84.
As to those impressions, which arise from the senses, their ultimate
cause is, in my opinion, perfectly inexplicable by human reason, and ’twill
always be impossible to decide with certainty, whether they arise immedi-
ately from the object, or are produc’d by the creative power of the mind,
or are deriv’d from the author of our being. Nor is such a question any
way material to our present purpose. We may draw inferences from the
coherence of our perceptions, whether they be true or false; whether they
represent nature justly, or be mere illusions of the senses.
(5) Hume [1739], p. 72.

[[A]]ll our ideas are copy’d from our impressions.
(6) Hume [1748], p. 350.
[[A]]ll our ideas are nothing but copies of our impressions, or, in other words,
that it is impossible for us to think of anything, which we have not an-
tecedently felt, either by our external or internal senses.
(7) Hume [1739], p. 29.
Wherever ideas are adequate representations of objects, the relations,
contradictions and agreements of the ideas are all applicable to the objects;
and this we may in general observe to be the foundation of all human
knowledge. But our ideas are adequate representations of the most minute
parts of extension; and thro’ whatever divisions and subdivisions we may
suppose these parts to be arriv’d at, they can never become inferior to
some ideas, which we form. The plain consequence is, that whatever appears
impossible and contradictory upon the comparison of these ideas, must be
really impossible and contradictory, without any farther excuse or evasion.
Quotations 58.19. (1) Hume [1739], p. 20.

When we have found a resemblance among several objects, that often occur
to us, we apply the same name to all of them, whatever differences we may
observe in the degrees of their quantity and quality, and whatever other
differences may appear among them. After we have acquired a custom of
this kind, the hearing of that name revives the idea of one of these objects,
and makes the imagination conceive it with all its particular circumstances
and proportions.
Comment. With a bit of fantasy, it seems to me, one can see a dawning of the idea
of set theoretical abstraction here.
(2) Hume [1739], p. 73.
All kinds of reasoning consist in nothing but a comparison, and a dis-
covery of those relations, either constant or inconstant, which two or more
objects bear to each other.
(3) Hume [1739], pp. 69 f.
There are seven different kinds of philosophical relation, viz. resem-
blance, identity, relations of time and place, proportion in quantity or num-
ber, degrees in any quality, contrariety, and causation. These relations may
be divided into two classes; into such as depend entirely on the ideas, which
we compare together, and such as may be chang’d without any change in
the ideas. ’Tis from the idea of a triangle, that we discover the relation of
equality, which its three angles bear to two right ones; and this relation
is invariable, as long as our idea remains the same. On the contrary, the
relations of contiguity and distance betwixt two objects may be chang’d
merely by an alteration of their place, without any change on the objects
themselves or on their ideas; [[. . .]] There is no single phænomenon, even the
most simple, which can be accounted for from the qualities of the objects,
as they appear to us; or which we cou’d foresee without the help of our
memory and experience.
It appears, therefore, that of these seven philosophical relations, there
remain only four, which depending solely upon ideas, can be the objects of
knowledge and certainty. These four are resemblance, contrariety, degrees
in quality, and proportions in quantity or number.

[[A]]s the power, by which one object produces another, is never discover-
able merely from their idea, ’tis evident cause and effect are relations, of
which we receive information from experience, and not from any abstract
reasoning or reflection.
(2) Hume [1748], p. 324.
[[C ]]auses and effects are discoverable, not by reason but by experience.
Comment. And what about the laws of causality?
(3) Hume [1748], p. 325.
[[T]]o convince us that all the laws of nature, and all the operations of bodies
without exception, are known only be experience, the following reflections
may, perhaps, suffice. Were any objects presented to us, and were we re-
quired to pronounce concerning the effect, which will result from it, with-
out consulting past observation; after what manner, I beseech you, must
the mind proceed in this operation? It must invent or imagine some event,
which it ascribes to the object as its effect; and it is plain that this invention
must be entirely arbitrary. The mind can never possibly find the effect in
the supposed cause, by the most accurate scrutiny and examination. For
the effect is totally different from the cause, and consequently can never be
discovered in it. Motion in the second billiard ball is a quite distinct event
from motion in the first; nor is there anything in the one to suggest the
smallest hint of the other. A stone or piece of metal raised into the air, and
left without any support, immediately falls: but to consider the matter a
priori, is there anything we discover in this situation which can beget the
idea of a downward rather than an upward, or any other motion, in the
stone or metal?
(4) Hume [1739], p. 74.
We readily suppose an object may continue individually the same, tho’
several times absent from and present to the senses; and ascribe to it an
identity, notwithstanding the interruption of the perception, whenever we
conclude, that if we had kept our eye or hand constantly upon it, it wou’d
have convey’d an invariable and uninterrupted perception. But this con-
clusion beyond the impressions of our senses can be founded only on the
connexion of cause and effect ; nor can we otherwise have any security, that
the object is not chang’d upon us, however much the new object may re-
semble that which was formerly present to the senses.
(5) Hume [1739], p. 78.

I shall proceed to examine [[. . .]]
First, For what reason we pronounce it necessary, that every thing
whose existence has a beginning, shou’d also have a cause?
Secondly, Why we conclude, that such particular causes must necessar-
ily have such particular effects; and what is the nature of that inference we
draw from the one to the other and of the belief we repose in it?
(6) Hume [1739], p. 82.
Since it is not from knowledge or any scientific reasoning, that we derive the
opinion of the necessity of a cause to every new production, that opinion
must necessarily arise from observation and experience. The next question,
then, shou’d naturally be, how experience gives rise to such a principle?
The next set of quotations is partly taken from the Enquiry and aims at illus-
trating the possible origin of Kant’s ideas although it is said that Kant has probably
not read the Enquiry.
Quotations 58.21. (1) Hume [1748], pp. 322 f (section iv).

All the objects of human reason or inquiry may naturally be divided
into two kinds, to wit, relations of ideas, and matters of fact. Of the first
kind are the sciences of geometry, algebra, and arithmetic; and in short, ev-
ery affirmation which is either intuitively or demonstratively certain. That
the square of the hypothenuse is equal to the squares of the two sides, is
a proposition which expresses a relation between these figures. That three
times five is equal to the half of thirty, expresses a relation between these
numbers. Propositions of this kind are discoverable by mere operation of
thought, without dependence on what is anywhere existent in the universe.
Though there never was a circle or triangle in nature, the truths demon-
strated by Euclid would for ever retain their certainty and evidence.
Matters of fact, which are the second objects of human reason, are
not ascertained in the same manner; nor is our evidence of their truth,
however great, of a like nature with the foregoing. The contrary of every
matter of fact is still possible; because it can never imply a contradiction,
and is conceived by the mind with the same facility and distinctness, as
if ever so conformable to reality. That the sun will not rise tomorrow is
no less intelligible a proposition, and implies no more contradiction than
the affirmation, that it will rise. We should in vain, therefore, attempt to
demonstrate its falsehood. Were it demonstratively false, it would imply a
contradiction, and could never be distinctly conceived by the mind.
Comment. Hume’s distinction of relations of ideas and matters of fact has been said
to “foreshadow” Kant’s distinction of synthetic and analytic statements by Quine
and is commonly parroted in analytic philosophy these days. The above quotation
is included here to be able to take issue with this view later. Note the criterion of
possibility: “it can never imply a contradiction”.
(2) Hume [1748], p. 323 (section iv).

All reasonings concerning matter of fact seem to be founded on the
relation of cause and effect. By means of that relation alone we can go
beyond the evidence of our memory and senses. [[. . .]] A man finding a
watch or any other machine in a desert island, would conclude that there
had once been men in that island. All our reasonings concerning fact are
of the same nature. And here it is constantly supposed that there is a
connection between the present fact and that which is inferred from it.
Were there nothing to bind them together, the inference would be entirely
precarious.
(3) Hume [1739], p. 170.
We may define a cause to be ‘An object precedent and contiguous to
another, and where all the objects resembling the former are plac’d in like
relations of precedency and contiguity to those objects, that resemble the
latter.’ If this definition be esteem’d defective, because drawn from objects
foreign to the cause, we may substitute this other definition in its place,
viz., ‘A cause is an object precedent and contiguous to another, and so
united with it, that the idea of the one determines the mind to form the
idea of the other, and the impression of the one to form a more lively idea
of the other.’
(4) Hume [1748], p. 362.
[[W]]e may define a cause to be an object, followed by another, and where
all the objects similar to the first are followed by objects similar to the
second. Or, in other words, where, if the first object had not been, the second
never had existed. [[. . .]] We say, for instance, that the vibration of this
string is the cause of this particular sound. But what do we mean by that
affirmation? We either mean that this vibration is followed by this sound,
and that all similar vibrations have been followed by similar sound : Or, that
this vibration is followed by this sound, and that upon the appearance of
one the mind anticipates the senses, and forms immediately an idea of the
other. We may consider the relation of cause and effect in either of these
two lights, but beyond these, we have no idea of it.
(5) Hume [1739], pp. 170 f.
When I examine with the utmost accuracy those objects, which are com-
monly denominated causes and effects, I find, in considering a single in-
stance that the one object is precedent and contiguous to the other; and in
inlarging my view to consider several instance, I find only that like objects
are constantly plac’d in like relations of succession and contiguity. Again,
when I consider the influence of this constant conjunction, I perceive, that
such a relation can never be an object of reason, and can never operate
upon the mind, but by means of custom, which determines the imagination
to make a transition from the idea of one object to that of its usual at-
tendant, and from the impression of one to a more lively idea of the other.
However extraordinary these sentiments may appear, I think it fruitless to
trouble myself with any farther enquiry or reasoning upon the subject, but
shall repose myself on them as on establish’d maxims.
’Twill only be proper, before we leave this subject, to draw some corol-
laries from it, by which we may remove sever prejudices and popular errors,
that have very much prevailed in philosophy. First, We may learn from the
foregoing doctrine, that all causes are of the same kind, and that in par-
ticular there is no foundation for that distinction which we sometime make
betwixt efficient causes, and causes sine qua non; or betwixt efficient causes,
and formal, and material, and exemplary, and final causes. For as our idea
of efficiency is deriv’d from the constant conjunction of two objects, wher-
ever this is observ’d, the cause is efficient; and where it is not, there can
never be a cause of any kind.
Comment. With hindsight, of course, the point is whether the idea of efficiency can
have no other origin than that of being derived “from the constant conjunction of
two objects”. The point is similar to that of the notion of natural numbers: we may
have learnt counting by using our fingers or pebble stones, but this doesn’t exhaust
the possibilities of a foundation of the notion of number. The way we acquire a
certain knowledge is not necessarily linked to the way this knowledge is justified.35
(6) Hume [1739], p. 139.
[[T ]]here is nothing in any object, consider’d in itself, which can afford us a
reason for drawing a conclusion beyond it; [[. . .]] even after the observation
of the frequent or constant conjunction of objects, we have no reason to
draw any inference concerning any object beyond those of which we have
had experience.
Comment. How do you consider an object “in itself”? Cf. Hegel in quotations 65.2
in these materials.
(7) Hume [1748], p. 351.
When we look about us towards external objects, and consider the operation
of causes, we are never able, in a single instance, to discover any power or
necessary connection; any quality, which binds the effect to the cause, and
renders the one an infallible consequence of the other. We only find that
the one does actually, in fact, follow the other. The impulse of one billiard
ball is attended with motion in the second. This is the whole that appears
to the outward senses. The mind feels no sentiment or inward impression
from this succession of objects: consequently there is not, in any single,
particular instance of cause and effect, anything which can suggest the idea
of power or necessary connection.
(8) Hume [1739], pp. 173 f.
1. The cause and effect must be contiguous in space and time.
2. The cause must be prior to the effect.
3. There must be a constant union betwixt the cause and effect. ’Tis
chiefly this quality, that constitutes the relation.
35 Cf. Frege in quotation 71.7 (2) on p. 1009 in these materials.
4. The same cause always produces the same effect, and the same ef-
fect never arises but from the same cause. This principle we derive from
experience, and is the cause of most of our philosophical reasonings. For
when by any clear experiment we have discover’d the causes of effects of
any phænomenon, we immediately extend our observation to every phæ-
nomenon of the same kind, without waiting for that constant repetition,
from which the first idea of this relation is derive’d.
5. [[W]]here several different objects produce the same effect, it must be
by means of some quality, which we discover to be common amongst them.
[[. . .]]
6. [[. . .]] The difference in the effects of two resembling objects must
proceed from that particular, in which they differ. [[. . .]]
Comment. There are two more points, which, however, I do not find relevant in the
context of my project.
The next set of quotations is concerned with Hume’s view on mathematics.

The great advantage of the mathematical sciences above the moral con-
sists in this, that the ideas of the former, being sensible, are always clear
and determinate, the smallest distinction between them is immediately per-
ceptible, and the same terms are still expressive of the same ideas, without
ambiguity or variation.
Comment. Note the location of mathematics in the realm of the ‘sensible’; an an-
ticipation of Kant’s view?
(2) Hume [1739], pp. 70 f.
[[G]]eometry, or the art, by which we fix the proportions of figures; tho’ it
much excels, both in universality and exactness, the loose judgments of the
senses and imagination; yet never attains a perfect precision and exactness.
Its first principles are still drawn from the general appearance of the objects;
and that appearance can never afford us any security[[.]] [[. . .]]
There remain, therefore, algebra and arithmetic as the only sciences, in
which we can carry on a chain or reasoning to any degree of intricacy, and
yet preserve a perfect exactness and certainty. We are possest of a precise
standard, by which we can judge of the equality and proportion of numbers;
and according as they correspond or not to that standard, we determine
their relations, without any possibility of error. When two numbers are so
combin’d, as that the one has always an unite answering to every unite of
the other, we pronounce them equal; and ’tis for want of such a standard of
equality in extension, that geometry can scarce be esteem’d a perfect and
infallible science.
[[. . .]] The reason why I impute any defect to geometry, is, because its
original and fundamental principles are deriv’d merely from appearances[[.]]
Comment. I can almost see Kant waiting in the background. Notice also the very
modern characterization of the equality of numbers in terms of the existence of a
one-one correspondence.
(3) Hume [1739], pp. 49 f.
’Tis true, mathematicians pretend they give an exact definition of a
right line, when they say, it is the shortest way betwixt two points. But in
the first place, I observe, that this is more properly the discovery of one
of the properties of a right line, than a just definition of it. For I ask any
one, if upon mention of a right line he thinks not immediately on such a
particular appearance, and if ’tis not by accident only that he considers
this property? A right line can be comprehended alone; but this definition
is unintelligible without a comparison with other lines, which we conceive
to be more extended. In common life ’tis establish’d as a maxim, that the
streightest way is always the shortest; which wou’d be as absurd as to say,
the shortest way is always the shortest, if our idea of a right line was not
different from that of the shortest betwixt two points.
Secondly, I repeat what I have already establish’d, that we have no
precise idea of equality and inequality, shorter and longer, more than of a
right line or a curve; and consequently that the one can never afford us a
perfect standard for the other. An exact idea can never be built on such as
are loose and undeterminate.
Comment. I include this a fine example of poor reasoning, characteristic not just
of Hume, but quite general of philosophy, and not just that of past times. It con-
sists mainly in not pursuing a concept, but preferring associations, no matter how
subjective. The problem, as I see it, is that philosophers are so full of themselves
that it doesn’t cross their minds that their ideas are just products of habitual as-
sociations grown in the swamp of subjective experience which do little to disclose
logical implications of concepts. This is particular important to emphasize in the
case of Hume since he is commonly taken to bring out the habitual character in
the notion of causality. But what I want to say is that Hume only discovers what
he has put into the notion of causality and that without realizing his authorship.
Whether the notion of causality is capable of a precise and coherent treatment can-
not be discovered by following Hume’s practice of introspection. The example of the
notion of the straight line in the quotation above may serve as a warning against
Hume’s treatment of causality, because geometry, as a conceptual science, not one of
what someone thinks “immediately on such an appearance”, has progressed in a way
unforeseen by philosophers, and unaffected by what they declared to be evident.
This gets me to the dogmatic side of Hume’s philosophy, a side which seems to
have survived well into the philosophy of his modern heirs.
Quotations 58.23. (1) Hume [1748], p. 428 (XII.iii).
It seems to me, that the only objects of the abstract science or of demon-
stration are quantity and number, and that all attempts to extend this more
perfect species of knowledge beyond these bounds are mere sophistry and
illusion.
Comment. Perhaps this is a good opportunity to emphasize that I propose to do

just that, i.e., extend the “objects of the abstract science” beyond that of quantity
and number. Even more, viz., that the notion of number itself is based on a notion
which Hume doesn’t include among the “objects of the abstract science”. Note also
that Hume does not include geometry.
(2) Hume [1748], p. 430 (XII.iii).
If we take in our hand any volume; of divinity or school metaphysics, for
instance; let us ask, Does it contain any abstract reasoning concerning quan-
tity or number ? No. Does it contain any experimental reasoning concerning
matter of fact and existence? No. Commit it then to the flames: for it can
contain nothing but sophistry and illusion.
Comment. A quotation much beloved by empiricists. Carnap [1935], pp. 218 f,36 is
worth a look in this context.
I’m afraid, this is where my quotations of David Hume come to an abrupt

end. I regret that yet another German has committed a book to the flames, sixty
years down the track.37 But I want to say in my own defence that I acted in a
mood of spontaneous approval of Hume’s proposition. But then, being sucked in by
demagoguery is part of the problem, isn’t it?

37 For all those who want to know it exactly: dies ater est 10 Maius 1993. What a coin-
cidence! Auf den Tag genau sechzig Jahre seit jener Nacht da in deutschen Universitätsstädten
Scheiterhaufen für Bücher errichtet wurden und einen milden Vorgeschmack dessen gaben wozu
sich deutscher Geist berufen fühlte. It might be worthwhile mentioning that some authors at least
were pretty miffed at their books not having been burned.
CHAPTER XV
KANT’S TRANSCENDENTAL PHILOSOPHY
With Kant’s Critique of Pure Reason, the concept of dialectic began its rise into a
central position in speculative thinking. At the same time it underwent a consider-
able change. The new role Kant assigned to dialectic was to become the starting-
point for Hegel’s speculative philosophy.
The present chapter is devoted to those basic ideas of Kant’s transcendental
philosophy which I regard as relevant for my enterprise of a logical foundation of
dialectic. By far the majority of quotations comes from the Critique of Pure Reason,
but I have also quoted from the Prolegomena, the Critique of Practical Reason, and
the Logic.
In referring to Kant’s Kritik der reinen Vernunft, I give the page numbers of the
first (A) or second (B) edition, not those of my actual copy listed in the bibliography.
§ 59. Introduction
My presentation of Kant’s transcendental philosophy comes as a sequel to my pre-

sentation of the philosophy of the ancient Greeks and that of the British empiricists,
more specifically, to certain questions already addressed in Parmenides and Berkeley,
for instance. In the center of my attention stands the conceptual struggle in dealing
with something on the level of conceptualization which by definition is beyond the
reach of, at least, perception, if not a certain conception.
In presenting Kant’s transcendental philosophy I shall try to keep apart three
aspects of his work which are to play different roles later, namely,
(i) the problem,
(ii) the plan, and
(iii) the realization.
The problem is rather traditional; the plan is fairly mind blowing; the realization
is poor, and it is certainly not for the standard of argumentation that I engage so
much more with Kant than with Hume; rather, it is the central position of the Antin-
omy of Pure Reason, differently put, the dialectic of reason and its epistemological
relevance, an idea for which there was no place in Hume’s set up.
The problem is located within the tradition of the question of how the mind
can know the world. The plan is to take the “endless controversies” 1 of metaphysics
towards a proof of a certain form of idealism. The realization centers around a
foundation of the faculties of the mind. The latter — which Fichte was quick to
1 N. K. Smith [1929], p. 8; see quotation 59.2 (1) below.
760
§ 59. INTRODUCTION 761
take up — may be seen as the most controversial one within the three aspects of
his work. Taking into account the poor methodological standards of his time, this
will hardly be surprising. It is the problem and the plan which shall provide the
link to the foundations of mathematics in chapter XXVII and thereby enable me to
formulate my notion of dialectic in chapter XXVIII.
The salient features in Kant’s transcendental philosophy on which I shall con-
centrate my attention are the following:
(KT1) the possibility of universally valid a priori knowledge;
(KT2) logic a canon, not an organon;
(KT3) the conception of Critique;
(KT4) the hierarchical structure of the faculties of the mind;
(KT5) phenomena and noumena;
(KT6) the transcendental deduction of the categories;
(KT7) the transcendental unity of apperception;
(KT8) the Antinomy of Pure Reason;
(KT9) the critical decision regarding the Antinomy of Pure Reason;
(KT10) the resolutions of the ideas of reason.
What I call Kant’s realization of his project is mostly covered by KT4–KT8.
Remark 59.1. Kant’s Critique of Pure Reason was not written in English, although
from the way it is sometimes cited one may get a different impression. I shall select
quotations from N. K. Smith’s translation of Kant’s Kritik der reinen Vernunft and
I shall identify them as such. N. K. Smith’s translation is not always true to the
text. Occasionally, I shall propose my own reading, but by and large I found it too
laborious to make out my own translation. There is one minor point, however, which
I would like to mention: as far as I could see N. K. Smith translates “Erkenntnis”
as “knowledge” throughout. There are many situations where I would prefer “cogni-
tion” as in Carus and Ellington [1977].2 To my mind “knowledge” corresponds more
to “Wissen”, which does not carry the active undertone of “Erkenntnis”. “Wissen”,
typically, is “abfragbar”;3 the schoolmaster’s paradigm.
59a. Kant’s starting point. The preface to the first edition of Kant’s Cri-
tique of Pure Reason begins by expounding the apparent deadlock in which meta-
physics finds itself, without initially mentioning metaphysics by name.
Quotations 59.2. (1) N. K. Smith [1929], pp. 7 f; translating Kant [1781] (A),
pp. vii ff.
Human reason has this peculiar fate that in one species of its knowledge
it is burdened by questions which [[. . .]] it is not able to ignore, but which
[[. . .]] it is also not able to answer.
2 Cf. Hartman and Schwarz [1974], p. ix, in their “note on the translation” of Kant’s Logic
regarding Abbott’s translation of Erkenntnis(se) as “knowledge” in an earlier translation of the

introduction to Kant’s Logic.
3 Cf. Hegel’s elaborations regarding “dogmatism as a way of thinking” in quotation 65.10 (3)
in these materials.
762 XV. KANT’S TRANSCENDENTAL PHILOSOPHY
[[. . .]] The battle-field of these endless controversies is called meta-

physics 4 .
[[. . .]] In more recent times, it has seemed as if an end might be put
to all these controversies and the claims of metaphysics receive final judg-
ment, through a certain physiology of the human understanding—that of
the celebrated Locke. But it has turned out otherwise.
Comment. This sets the framework, metaphysics and analysis (“physiology” 5) of
the human understanding, and it already indicates the direction Kant is going to
take, viz. to employ an analysis of the human understanding in order to get a grip
on metaphysics.
(2) Carus and Ellington [1977], p. 19; translating Kant [1783], p. 271.
[[I]] n all ages one metaphysics has contradicted another either in its asser-
tions or their proofs, and thus has itself destroyed its own claim to lasting
assent.
(3) N. K. Smith [1929], pp. 8 f; translating Kant [1781] (A), pp. x ff.
[[I]]t is idle to feign indifference to such enquiries, the object of which can
never be indifferent to our human nature. Indeed these pretended indif-
ferentists, however they may try to disguise themselves by substituting a
popular tone for the language of the Schools, inevitably fall back, in so far as
they think at all, into those very metaphysical assertions which they profess
so greatly to despise. None the less this indifference [[. . .]] is a phenomenon
that calls for attention and reflection. [[. . .]] It is a call to reason to under-
take anew the most difficult of all its tasks, namely that of self-knowledge,
and to institute a tribunal which will assure to reason its lawful claims, and
dismiss all groundless pretensions [[. . .]]. This tribunal is no other than the
critique of pure reason.
I do not mean by this a critique of books and systems, but of the faculty
of reason in general, in respect of all knowledge after which it may strive
independently of all experience. It will therefore decide as to the possibility
or impossibility of metaphysics in general, and determine its sources, its
extent, and its limits—all in accordance with principles.
Remarks 59.3. (1) This idea of Kant’s is prone to be interpreted in a fairly simple-
minded fashion as some sort of “investigation of the set of ideas which forms the
limiting framework of all our thought about the world and experience of the world”.6
Suffice it here to express a general warning. The project of an investigation into the
faculties of the mind within Kant’s project of the Critique is part of a broadly laid
out plan. It is within the framework of this plan that the two central features of
my study show up: the deduction of the categories and the antinomies. To give
an idea at this early stage: what I am aiming at is not an interpretation of the
limitations as limitations of senses or concepts, but as structural limitations of a
self-investigating reason, as a result of the phenomenon of self-reference, which are
4 Emphasized in my German edition, but not in N. K. Smith’s translation.
5 As regards the term “physiology”, cf. also the Prolegomena, p. 306
6 Strawson [1966], p. 15; quotation 68.12 in these materials.
surmountable in the sense that they give rise to new thought determinations. To
reduce Kant’s approach to a sort of stocktaking of available resources wouldn’t do
justice to his endeavour, particularly not to the issue of a transcendental deduction
of the categories. It would thus deprive me of one of the most important aspects
of speculative reasoning, and stand in the way of an appropriate understanding of
Fichte and Hegel.
(2) Kant seems to somewhat dissociate his approach from that of Locke’s by using
the term “physiology” though he seems to acknowledge the importance of the general
idea. Kant uses the term “physiology” again towards the very end of his Critique
in a chapter entitled The Architectonic of Pure Reason.7 I see Kant as standing at
the end of a development, viz., that from Locke to Hume, and consciously taking
advantage of surprises cropping up in what, at the beginning, may have looked like
a fairly unproblematic venture. In this sense Kant has already taken the first hurdle
and he is — unsuspectingly though — about to prepare the grounds for a new
surprise.
Quotations 59.4. (1) N. K. Smith [1929], p. 21; translating Kant [1787b] (B),
p. xiv f.
Metaphysics is a completely isolated speculative science of reason, which
soars far above the teachings of experience, and in which reason is indeed
meant to be its own pupil. Metaphysics rests on concepts alone—not, like
mathematics, on their application to intuition. But though it is older than
all other sciences, [[. . .]], it has not yet had the good fortune to enter upon
the secure path of science. [[. . .]]
What, then, is the reason why, in this field, the sure road to science has
not hitherto been found?
Comment. Note the phrase “reason is . . . meant to be its own pupil”.
[[T]]hough we cannot assume metaphysics to be an actual science, we can
say with confidence that certain pure a priori synthetic cognitions are ac-
tual and given, namely, pure mathematics and pure physics; for both con-
tain propositions which are everywhere recognized as apodeictically certain,
partly by mere reason, partly by universal agreement from experience, and
yet as independent of experience. We have therefore some, at least uncon-
tested, synthetic knowledge a priori, and need not ask whether it be possible
(for it is actual) but how it is possible, in order that we may deduce from
the principle which makes the given knowledge possible the possibility of
all the rest.
(3) N. K. Smith [1929], p. 56; translating Kant [1787b] (B), pp. 20 f.
How is pure mathematics possible?
How is pure science of nature possible?
Since these sciences actually exist, it is quite proper to ask how they
are possible; for that they must be possible is proved by the fact that they
7 Compare quotation 60.3 (5) below.
exist. But the poor progress which has hitherto been made in metaphysics,
and the fact that no system yet propounded can, in view of the essential
purpose of metaphysics, be said really to exist, leaves everyone sufficient
ground for doubting as to its possibility.
59b. Kant’s assessment of logic and mathematics. As the last quota-

tions already indicate, Kant places the desolate situation of metaphysics before the
background of the apparent success of logic, mathematics, and physics. I shall de-
vote some space to his views on these three subjects, beginning with logic. One
aspect is particularly important for me: Kant’s emphasis of logic as not being an
organon.
p. viii.
That logic has already, from the earliest times, proceeded upon this
sure path [[of a science]] is evidenced by the fact that since Aristotle it has
not required to retrace a single step[[.]]
(2) N. K. Smith [1929], pp. 18; translating Kant [1787b] (B), p. ix.
That logic should have been thus successful is an advantage which
it owes entirely to its limitations, whereby it is justified in abstracting—
indeed, it is under obligation to do so—from all objects of knowledge and
their differences, leaving the understanding nothing to deal with save itself
and its form. But for reason to enter on the sure path of science is, of course,
much more difficult, since it has to deal not with itself alone but also with
objects.
Comment. What seems to be assumed here is that if understanding has nothing to
deal with save itself, it has no serious problems. I regard this as a crucial fallacy
in Kant’s thought, one which I take Hegel to turn against and the paradoxes of
modern logic to support Hegel.
(3) Hartman and Schwarz [1974], pp. 14 f; translating Kant [1800], pp. 433 f (A 3 ff).
All rules according to which the understanding proceeds are either nec-
essary or contingent. The former are those without which no use of the
understanding would be possible at all; the latter are those without which
a certain use of the understanding would not take place. [[. . .]]
If, now, we set aside all cognition that we must borrow from objects and
reflect solely upon the use of the understanding in itself, we discover those of
its rules which are necessary throughout, in every respect and regardless of
any special objects, because without them we would not think at all. Insight
into these rules can therefore be gained a priori and independently of any
experience, because they contain, without discrimination between objects,
merely the conditions of the use of the understanding itself, be it pure or
empirical. And it also follows from this that the universal and necessary
rules of thought in general can concern solely its form, and not in any way
its matter. Accordingly, the science containing these universal and necessary
rules is a science of the mere form of one intellectual cognition or of thinking.
And we can therefore form for ourselves the idea of the possibility of such a
science, just as that of a general grammar which contains nothing beyond
the mere form of a language in general, without words, which belong to the
matter of the language.
Now this science of the necessary laws of the understanding and reason
in general, or—which is the same—of the mere form of thinking, we call
logic.
As a science concerning all thinking in general, regardless of objects as
the matter of thinking, logic is to be considered as
1) the basis of all other sciences and the propaedeutic of all use
of the understanding. For this very reason, however, because it abstracts
entirely from all objects, it can be
2) no organon of the sciences.
By organon namely we understand an instruction for bringing about a
certain cognition. This implies, however, that I already know the object of
the cognition that is to be produced according to certain rules. An organon
of the sciences is therefore not mere logic, because it presupposes the exact
knowledge of the sciences, of their objects and sources. Thus mathematics,
for example is an excellent organon as a science containing the ground of
the expansion of our cognition in regard to a certain use of reason. [[. . .]]
3) As a science of the necessary laws of thinking without which no
use of the understanding and of reason takes place at all[[.]]
Comment. What is important for me in this quotation is (α) Kant’s characterization
of logic as general grammar, (β) the relentless opposition of form and matter, and
(γ) the rejection of logic as an organon. (Kant’s specification of what logic is to be
considered as, contains two more points which, however, are of no particular interest
to me.)
(4) Hartman and Schwarz [1974], p. 22; translating Kant [1800], p. 441 (A 16).
[[Logic]] can not be a science of the speculative understanding. For as a
logic of speculative cognition or of the speculative use of reason it would
be an organon of other sciences and not merely a propaedeutic concerned
with all possible use of the understanding and of reason.
Comment. This is a decisive point as regards my approach to dialectic: I claim, of
course, that Kant went wrong here. The point, however, is to work out in which
way; this question will be taken up in the groundworks.
Logic [[. . .]] has not gained much in content since Aristotle’s times and
indeed it cannot, due to its nature.
[[. . .]] There are but few sciences that can come into a permanent state
beyond which they undergo no further change. To these belong logic, and
also metaphysics. Aristotle had omitted no moment of the understanding[[.]]
Comment. This must be one of the most favorite quotations for opening a book
on modern formal logic. It is just too tempting a piece of knowledge by hindsight.
What does Kant’s gross misjudgment of the state of logic as a complete science
indicate?8 It may, indeed, serve as a warning for our assessing an apparent stability
of a theory, but this is not what I quote it for. My interest lies in the view that
both, logic and metaphysics, are capable of reaching a “permanent state beyond
which they undergo no further change”.
In present times there has been no famous logician, and we do not
indeed need any new inventions for logic, because it contains merely the
form of thinking.
Comment. My emphasis lies on merely (bloß). There’ll be more ‘merelies’ making
their appearance in this book.
(7) N. K. Smith [1929], p. 111; translating Kant [1781] (A), p. 76.
General logic [[. . .]] abstracts from all content of knowledge, and looks
to some other source, whatever that may be, for the representations which
it is to transform into concepts by process of analysis.
Comment. What interests me in this quotation is the rigid separation of form and
content, enduring feature of a classical way of thinking: a dualism of some form or
another.
(7) N. K. Smith [1929], pp. 85 f; translating Kant [1781] (A), p. 47.
[[I]]t is evident that from mere concepts only analytic knowledge, is to be
attained.
Comment. This is the stubborn prejudice that has to be cracked, if there is to be
any hope for Hegel’s idea of dialectic.
Thus as regards Kant’s assessment of the situation of logic; then there is math-
ematics and physics.
Quotations 59.6. (1) N. K. Smith [1929], pp. 18 f; translating Kant [1787b] (B),
p. x.
Mathematics and physics, the two sciences in which reason yields theo-
retical knowledge, have to determine their objects a priori, the former doing
so quite purely, the latter having to reckon, at least partially, with sources
of knowledge other than reason.
Geometry is based upon the pure intuition of space. Arithmetic attains its
concepts of numbers by the successive addition of units in time, and pure
mechanics especially can attain its concepts of motion only by employing
the representation of time.
8 Examples of the incompleteness of Aristotelian logic can be found in Hilbert and Ackermann
[1928], pp. 65 f.
(3) N. K. Smith [1929], p. 588; translating Kant [1781] (A), p. 731 (B 759).
[[M]]athematical definitions can never be in error. For since the concept is
first given through the definition, it includes nothing except precisely what
the definition intends should be understood by it.
Comment. Cum grano salis this may be viewed as expressing a conceptualistic view
of mathematics. I shall engage with Kant’s view regarding mathematical definition
in the context of §112 in the groundworks. For those who want to know what
the German original for N. K. Smith’s “precisely” is: “gerade”, i.e., straight or even.
This will come up again when I discuss Frege’s use of “ungerade Bedeutung”. Suffice
it here to say that I take the rendering of “gerade” as “precisely“ to make as much
sense as rendering “ungerade Bedeutung” as “imprecise reference”.
(4) N. K. Smith [1929], p. 19, translating Kant [1787b] (B), p. xi.
In the earliest times to which the history of human reason extends,
mathematics, [[. . .]] had already entered upon the sure path of science. But
it must not be supposed that it was as easy for mathematics as it was for
logic—in which reason has to deal with itself alone[[.]]
[[M]]athematics has an advantage over philosophy, in that the cognitions of
the former are intuitive, those of the latter, on the contrary, only discursive.
The reason, however, why in mathematics we ponder more the quantities
lies in this, that quantities can be constructed a priori in intuition, whereas
qualities cannot be exhibited in intuition.
Comment. There seems to be some reminiscence of Hume [1748], p. 349, as quoted
in 58.22 (1) in these materials, in particular, in locating mathematics in a form of
sensibility, while the difference lies in the assessment of the nature of this sensibility
and its impact on mathematics.
Remark 59.7. Although Kant is quite explicit in his considering physics alongside
with mathematics and logic as regards the problematic of theoretical knowledge a
priori, I deliberately neglect this aspect, the reason being that it is only the founda-
tions of logic and mathematics that will play an eminent role in my investigations
later on. Kant’s considerations on physics would have to be added here, if the goal of
this treatise were to incorporate the problems of quantum mechanics and quantum
logic (which it was originally meant to do, but I gave up on this idea for purely
pragmatic reasons).
59c. The master-plan. Kant started from — what looked to him like — the
obvious existence, and security, of a priori sciences — such as mathematics and
pure physics — on the one hand, and, on the other hand, the constant failure to
establish a similarly respectable science of metaphysics.9 This is the situation he
set out to make sense of and towards which he designed his strategy. Here are its
salient features:
(i) a priori knowledge;
9 Cf. Kant [1783], §15.
(ii) the paradigm of Copernicus (Copernican revolution 10 );

(iii) the conception of transcendental philosophy.
In order to present them, I begin by taking up Kant’s assessment of Hume and carry
on from there.
Kant is quite clear that he regards his transcendental idealism and the distinc-
tion of analytic and synthetic judgments as a genuine achievement as against the
British empiricists.
Quotations 59.8. (1) Carus and Ellington [1977], pp. 5 f; translating Kant [1783],
pp. 260 f.
I openly confess that my remembering David Hume was the very thing
which many years ago first interrupted my dogmatic slumber and gave my
investigations in the field of speculative philosophy a quite new direction.
I was far from following him in the conclusions to which he arrived by
considering, not the whole of his problem, but a part, which by itself can
give us no information. If we start from a well-founded, but undeveloped,
thought which another has bequeathed to us, we may well hope by continued
reflection to advance further than the acute man to whom we owe the first
spark of light.
So I tried first whether Hume’s objection could not be put into a gen-
eral form, and soon found that the concept of the connection of cause and
effect was by no means the only concept by which the understanding thinks
the connection of things a priori, but rather that metaphysics consists al-
together of such concepts. I sought to ascertain their number; and when
I had satisfactorily succeeded in this by starting from a single principle,
I proceeded to the deduction of these concepts, which I was now certain
were not derived from experience, as Hume had tried, but sprang from
the pure understanding. This deduction (which seemed impossible to my
acute predecessor, which had never even occurred to any one else, though
no one had hesitated to use the concepts without investigating the basis
of their objective validity) was the most difficult task ever undertaken in
the service of metaphysics; and the worst was that metaphysics, such as it
then existed, could not assist me in the least because this deduction alone
can render metaphysics possible. But as soon as I had succeeded in solv-
ing Hume’s problem, not merely in a particular case, but with respect to
the whole faculty of pure reason, I could proceed safely, though slowly, to
determine the whole sphere of pure reason completely and from universal
principles, in its boundaries as well as in its contents. This was required for
metaphysics in order to construct its system according to a sure plan.
But I fear that the execution of Hume’s problem in its widest extent
(namely, my Critique of Pure Reason) will fare as the problem itself fared
when first proposed.
10 Kant never used this phase himself.

Comment. What Bertrand Russell thought of Kant’s relation to Hume can be found
in Russell [1946], p. 678; quotation 68.3 (1) in these materials. In this context, cf.
also quotation 68.3 (2).
(2) N. K. Smith [1929], p. 55; translating Kant [1787b] (B), p. 20.
[[Hume]] occupied himself exclusively with the synthetic proposition regard-
ing the connection of an effect with its cause (principium causalitatis),
and he believed himself to have shown that such an a priori proposition
is entirely impossible. If we accept his conclusions, then all that we call
metaphysics is a mere delusion whereby we fancy ourselves to have rational
insight into what, in actual fact, is borrowed solely from experience, and
under the influence of custom has taken the illusory semblance of necessity.
Hume was perhaps aware, although he never followed the matter out,
that in judgments of a certain kind we pass beyond our concept of the
object. I have entitled this kind of judgment synthetic.
(4) N. K. Smith [1929], p. 609; translating Kant [1781] (A), pp. 764 f (B 792 f).
Experience is itself a synthesis of perceptions, whereby the concept which
I have obtained by means of perception is increased through the addition
of other perceptions. But we suppose ourselves to be able to pass a priori
beyond the concept, and so to extend our knowledge. This we attempt to
do either through the pure understanding, in respect of that which is at
least capable of being an object of experience, or through pure reason, in
respect of such properties of things, or indeed even of the existence of such
things, as can never be met with in experience. [[Hume]] did not distinguish
these two kinds of judgments.
(5) Carus and Ellington [1977], pp. 17 f; translating Kant [1783], p. 270.
In Locke’s Essay, however, I find an indication of my division [[of judgments
into analytic and synthetic]]. For in the fourth book (chap. iii., §9, seq.), hav-
ing discussed the various connections of representations in judgments, and
their sources, one of which he makes “identity or contradiction” (analytic
judgments) and another the coexistence of representations in a subject (syn-
thetic judgments), he confesses (§10) that our (a priori) knowledge of the
latter is very narrow and almost nothing. But in his remarks on his species
of cognition, there is so little of what is definite and reduced to rules that
we cannot wonder if no one, not even Hume, was led to make investigations
concerning judgments of this kind.
The third one is well-documented by Kant himself in his Prolegomena.
pp. 257 f.
Hume started mainly from a single but important concept in meta-
physics, namely, that of the connection of cause and effect (including its
derivative concepts of force and action, etc.). He challenged reason, which
pretends to have given birth to this concept of herself, to answer him by
what right she thinks anything could be so constituted that if that thing
be posited, something else also must necessarily be posited; for this is the
meaning of the concept of cause. He demonstrated irrefutably that it was
entirely impossible for reason to think a priori and by means of concepts
such a combination as it involves necessity. We cannot at all see why, in
consequence of the existence of one thing, another must necessarily exist, or
how the concept of such a combination can arise a priori. Hence he inferred
that reason was altogether deluded with reference to this concept, which
she erroneously considered as one of her own children, whereas in reality
it was nothing but a bastard of imagination, impregnated by experience,
which subsumed certain representations under the law of association, and
mistook a subjective necessity (custom) for an objective necessity arising
from insight. Hence he inferred that reason had no power to think such
combinations, even in general, because her concepts would then be purely
fictitious and all her pretended a priori cognitions nothing but common
experiences marked with a false stamp. This is as much as to say that there
is not, and cannot be, any such thing as metaphysics at all.
(2) Abbott [1873], pp. 140 f; translating Kant [1788], pp. 60 f.
David Hume, of whom we may say that he commenced the assault

on the claims of pure reason, which made a thorough investigation of it
necessary, argued thus: the notion of cause is a notion that involves the
necessity of the connexion of the existence of different things, and that, in
so far as they are different, so that, given A, I know that something quite
distinct therefrom, namely B, must necessarily also exist. Now necessity
can be attributed to a connexion only in so far as it is known a priori; for
experience would only enable us to know of such a connexion that it exists,
not that it necessarily exists. Now, it is impossible, says he, to know a priori
and as necessary the connexion between one thing and another (or between
one attribute and another quite distinct) when they have not been given
in experience. Therefore the notion of a cause is fictitious and delusive,
and, to speak in the mildest way, is an illusion, only excusable inasmuch
as the custom (a subjective necessity) of perceiving certain things, or their
attributes as often associated in existence along with or in succession to one
another, is insensibly taken for an objective necessity of supposing such a
connexion in the objects themselves, and thus the notion of a cause has
been acquired surreptitiously and not legitimately: nay, it can never be
so acquired or authenticated, since it demands a connexion in itself vain,
chimerical, and untenable in presence of reason, and to which no object can
ever correspond.
(3) Abbott [1873], p. 143; translating Kant [1788], pp. 62 f.
If Hume took the objects of experience for things in themselves (as is almost
always done), he is quite right in declaring the notion of cause to be a
deception and false illusion; [[. . .]]
It resulted, however, from my inquiries, that the objects with which we

have to do in experience are by no means things in themselves, but merely
phenomena[[.]]
(4) Carus and Ellington [1977], pp. 53 f; translating Kant [1783], pp. 310 f.
[[Hume]] justly maintains that we cannot comprehend by reason the possi-
bility of causality, that is, of the reference or the existence of one thing to
the existence of another which is necessitated by the former. [[. . .]] But I am
very far from holding these concepts to be derived merely from experience,
and the necessity represented in them to be fictitious and a mere illusion
produced in us by long habit. On the contrary, I have amply shown that
they and the principles derived from them are firmly established a priori
before all experience and have their undoubted objective rightness, though
only with regard to experience.
Comment. The last subordinate clause plays a key role in understanding Kant’s
doctrine.
This is an important point in my presentation of Kant: to give an impression

of why Kant thinks he has shown that the objects with which we have to do in
experience are not things as they are in themselves but phenomena.
In order to explain the existence of a priori sciences in general, Kant proposed
a basic change in our way of thinking, which he considered a kind of Copernican
revolution: instead of trying to interpret knowledge in terms of the subject of knowl-
edge orienting itself (or “conform” as “sich richten nach” is translated in N. K. Smith
[1929], p. 22) to the constitution of the object, we may try, for a change, to conceive
any knowledge of an object as conforming to the constitution of the subject. In
Kant’s own (translated) words:
pp. xvi f.
Hitherto it has been assumed that all our knowledge must conform to ob-
jects. But all attempts to extend our knowledge of objects by establishing
something in regard to them a priori, by means of concepts, have, on this
assumption, ended in failure. We must therefore make trial whether we may
not have more success in the tasks of metaphysics, if we suppose that ob-
jects must conform to our knowledge. This would agree better with what is
desired, namely, that it should be possible to have knowledge of objects a
priori, determining something in regard to them prior to their being given.
We should then be proceeding [[. . .]] on the lines of Copernicus’ primary
hypothesis. Failing of satisfactory progress in explaining the movement of
the heavenly bodies on the supposition that they all revolved round the
spectator, he tried whether he might not have better success if he made the
spectator to revolve and the stars to remain at rest. A similar experiment
can be tried in metaphysics, as regards the intuition of objects. If intui-
tion must conform to the constitution of the objects, I do not see how we
could know anything of the latter a priori; but if the object (as object of
senses) must conform to the constitution of our faculty of intuition, I have

no difficulty in conceiving such a possibility.
Comment. For those who want to know what I left out in the place indicated by
the three dots in double square brackets: “precisely”. This time it is a pure addition
by the translator/interpreter N. K. Smith.
(2) N. K. Smith [1929], p. 19; translating Kant [1787b] (B), pp. xi f.
A new light flashed upon the mind of the first man (be he Thales or some
other) who demonstrated the properties of the isosceles triangle. The true
method, so he found, was not to inspect what he discerned either in the
figure, or in the bare concept of it, and from this, as it were, to read off its
properties; but to bring out what was necessarily implied in the concepts
that he had himself formed a priori, and had put into the figure in the
construction by which he presented it to himself. If he is to know anything
with a priori certainty he must not ascribe to the figure anything save what
necessarily follows from what he has himself set into it in accordance with
his concept.
(3) N. K. Smith [1929], p. 20; translating Kant [1787b] (B), p. xiii.
When Galileo caused balls, the weights of which he had himself [[chosen]],
to roll down an inclined plane; when Torricelli made the air carry a weight
which he had [[conceived of]] beforehand to be equal to that of a [[known
water column]]; [[. . .]] a light broke upon all students of nature. They learned
that reason has insight only into that which it produces after a plan of its
own, and that it must not allow itself to be kept, as it were, in nature’s
leading-strings, but must itself show the way with principles of judgment
based upon fixed laws, constraining nature to give answer to questions of
reason’s own determining. Accidental observations, made in obedience to
no previously thought-out plan, can never be made to yield a necessary law,
which alone reason is concerned to discover. Reason, holding in one hand
its principles, according to which alone concordant appearances can be ad-
mitted as equivalent to laws, and in the other hand the experiment which
it has devised in conformity with these principles, must approach nature in
order to be taught by it. It must not, however, do so in the character of a
pupil who listens to everything that the teacher chooses to say, but of an
appointed judge who compels the witnesses to answer questions which he
has himself formulated.
Comment. Compare quotations 93.1 for a more modern (historical) account of Ga-
lilei’s approach.
(4) U. Petersen [2002], p. 772; translating Kant [1787b] (B), p. 1.
That all our cognition begins with experience, this is not at all in
doubt[[.]] [[. . .]]
Although, however, all our cognition begins with experience, it does
not yet all spring, therefore, from experience. For it could well be that
even our cognition through experience is a composition of what we receive
through impressions, and that which our own capacity of cognition (merely
§ 60. FORGING THE TOOLS 773
occasioned by sensible impressions) produces out of itself, an addition which

we do not distinguish from that primary material, until long practice has
made us aware of it and skilled in separating it.
§ 60. Forging the tools
Kant’s approach to the problem of a priori knowledge is challenging and even though
there is quite a number of inconsistencies in his use of concepts, it is possible to
interpret him in such a way that everything has its carefully designed place within
the system; at least this is what I am trying to do when I am now turning to the
core of his transcendental philosophy.
60a. A priori knowledge, transcendental philosophy, and metaphys-

ics. According to Kant, a priori knowledge reveals itself because of its necessity
and universal validity.
pp. 3 f.
What we here require is a criterion by which to distinguish with cer-
tainty between pure and empirical knowledge. Experience teaches us that
a thing is so and so, but not that it cannot be otherwise. First, then, if we
have a proposition which in being thought is thought as necessary, it is an a
priori judgment; [[. . .]]. Secondly, experience never confers on its judgments
true or strict, but only assumed and comparative universality, through in-
duction. [[. . .]] Necessity and strict universality are thus sure criteria of a
priori knowledge, and are inseparable from one another.
Now it is easy to show that there actually are in human knowledge judg-
ments which are necessary and in the strictest sense universal, and which
are therefore pure a priori judgments. If an example from the sciences be
desired, we have only to look to any of the propositions of mathematics; if
we seek an example from the understanding in its quite ordinary employ-
ment, the proposition, ‘every alteration must have a cause’, will serve our
purpose. In the latter case, indeed, the very concept of a cause so manifestly
contains the concept of a necessity of connection with an effect and of the
strict universality of the rule, that the concept would be altogether lost if
we attempted to derive it, as Hume has done, from a repeated association
of that which happens with that which precedes, and from a custom of con-
necting representations, a custom originating in this repeated association,
and constituting therefore a merely subjective necessity.
This consideration then gives rise to Kant’s notion of metaphysics and specu-
lation. At the same time, it evokes great hopes for metaphysics under this modified
notion.
Quotations 60.2. (1) N. K. Smith [1929], p. 14; translating Kant [1781] (A), p. xx.
What reason produces entirely out of itself cannot be concealed, but is
brought to light by reason itself immediately the common principle has
been discovered. The complete unity of this kind of knowledge, and the fact
that it is derived solely from pure concepts, entirely uninfluenced by any
experience or by special intuition, such as might lead to any determinate
experience that would enlarge and increase it, make this unconditioned
completeness not only practicable but also necessary.
Comment. This is a crucial point. The two important features in Kant’s view are:
firstly, reason has immediate insight into its own creations, and, secondly, a certain
immutability of this realm. In view of what is to come, I just want to mention here
that these are what I consider the basic misjudgments in Kant’s approach.
(2) N. K. Smith [1929], p. 13; translating Kant [1781] (A), p. xx.

Metaphysics, on the view which we are adopting, is the only of all the
sciences which dare promise that through a small but concentrated effort it
will attain, and this in a short time, such completion as will leave no tasks
to our successors leave that of adapting it in a didactic according to their
own preferences[[.]]
[[H]]ere is an advantage upon which, of all possible sciences, metaphysics
alone can with certainty reckon: that it can be brought to such completion
and fixity as to require no further change or be capable of any augmentation
by new discoveries, because here reason has the sources of its knowledge in
itself, not in objects and their observation, by which its stock of knowledge
could be further increased. When, therefore, it has exhibited the funda-
mental laws of its faculty completely and so determinately as to avoid all
misunderstanding, there remains nothing more for pure reason to cognize
a priori; nay there is even no ground to raise further questions.
(4) N. K. Smith [1929], p. 10; translating Kant [1781], p. xiv.
I have to deal with nothing save reason itself and its pure thinking; and to
obtain complete knowledge of these, there is no need to go far afield, since
I come upon them in my own self. Common logic itself supplies an exam-
ple, how all the simple acts of reason can be enumerated completely and
systematically. The subject of the present enquiry is the [kindred] question,
how much we can hope to achieve by reason, when all the material and
assistance of experience are taken away.
In view of what is to come it is important to emphasize that Kant himself

did not look upon his Critique of Pure Reason as the final word as the following
quotation indicates.
pp. 13 f (B 27 ff).
Transcendental philosophy is only the idea of a science, for which the
critique of pure reason has to lay down the complete architectonic plan.
That is to say, it has to guarantee, as following from principles, the com-
pleteness and certainty of the structure in all its parts. It is the system
of all principles of pure reason. And if this critique is not itself to be en-
titled a transcendental philosophy, it is solely because, to be a complete
system, it would also have to contain an exhaustive analysis of the whole of
a priori human knowledge. Our critique must, indeed, supply a complete
enumeration of all the fundamental concepts that go to constitute such pure
knowledge. [[. . .]]
The critique of pure reason therefore will contain all that is essential in
transcendental philosophy. While it is the complete idea of transcendental
philosophy, it is not equivalent to that latter science; for it carries the
analysis only so far as is requisite for the complete examination of knowledge
which is a priori and synthetic.
What has to be kept in view in the division of such a science, is that
no concepts be allowed to enter which contain in themselves anything em-
pirical, or, in other words, that it consists in knowledge wholly a priori.
I entitle transcendental all knowledge which is occupied not so much with
objects as with the mode of our knowledge of objects in so far as this mode
of knowledge is to be possible a priori. A system of such concepts might be
entitled transcendental philosophy. But that is still, at this stage, too large
an undertaking.
Comment. The phrasing in the first edition was slightly different. Cf. also quota-
tion 60.5 (8) below.
Transcendental philosophy is [[. . .]] a philosophy of pure and merely specu-
lative reason.
The philosophy of pure reason is either a propaedeutic (preparation), which
investigates the faculty of reason in respect of all its pure a priori knowl-
edge, and is entitled criticism or secondly, it is the system of pure reason,
that is, the science which exhibits in systematic connection the whole body
(true as well as illusory) of philosophical knowledge arising out of pure
reason, and which is entitled metaphysics.
(5) N. K. Smith [1929], p. 661; translating Kant [1781] (A), pp. 845 f. (B 873)
All pure a priori knowledge, owing to the special faculty of knowledge in
which alone it can originate, has in itself a peculiar unity; and metaphysics is
the philosophy which has as its task the statement of that knowledge in this
systematic unity. Its speculative part, which has especially appropriated this
name, namely, what we entitle metaphysics of nature, and which considers
everything in so far as it is (not that which ought to be) by means of a

priori concepts, is divided in the following manner.
Metaphysics, in the narrower meaning of the term, consists of transcen-
dental philosophy and physiology of pure reason. The former treats only of
the understanding and of reason, in a system of concepts and principles
which relate to objects in general but take no account of objects that may
be given (Ontologia); the latter treats of nature, that is, of the sum of given
objects (whether given to the senses, or, if we will, to some other kind of
intuition) and is therefore physiology—although only rationalis.
(6) N. K. Smith [1929], p. 14; translating Kant [1781] (A), p. xxi.
[[A]] system of pure (speculative) reason I hope myself to produce under the
title Metaphysics of Nature. It will be not half as large, yet incomparably
richer in content than this present Critique, which has as its first task to
discover the sources and conditions of the possibility of such criticism[[.]]
[[. . .]] [[H]]owever completely all the principles of the system are presented
in this Critique, the completeness of the system itself likewise requires that
none of the derivative concepts be lacking. These cannot be enumerated
by any a priori computation, but must be discovered gradually. Whereas,
therefore, in this Critique the entire synthesis of the concepts has been
exhausted, there will still remain the further work of making their analysis
similarly complete, a task which is rather an amusement than a labour.
Comment. As regards Kant’s notion of metaphysics here, cf. quotations 60.2 (1)
and (2) above.
pp. xxxvii f.
[[S]]peculative reason has a structure wherein everything is an organ, the
whole being for the sake of every part, and every part for the sake of all
the others, so that even the smallest imperfection, be it a fault (error)
or a deficiency, must inevitably betray itself in use. This system will, as
I hope, maintain, throughout the future, this unchangeableness. It is not
self-conceit which justifies me in this confidence, but the evidence exper-
imentally obtained through the parity of the result, whether we proceed
from the smallest elements to the whole of pure reason or reverse-wise form
the whole (for this also is presented to reason through its final end in the
sphere of the practical) to each part.
[[P]]ure reason is a sphere so separate and self-contained that we cannot
touch a part without affecting all the rest. We can therefore do nothing
without first determining the position of each part and its relation to the
rest. For inasmuch as our judgment within this sphere cannot be corrected
by anything outside of pure reason, so the validity and use of every part
depends upon the relation in which it stands to all the rest within the
domain of reason, just as in the structure of an organized body the end of
each member can only be deduced from the full conception of the whole.
Although Kant never presented such a system of pure (speculative) reason, his
point seems clear enough. What is particularly interesting here, is the view which
shows up in quotations 60.2 that such a system could be presented in a more or
less final shape, i.e., this system would be static, not subject to scientific change or
any further extension as regards its content. As a result, the system of categories is
viewed as being complete. This is an important point, since Hegel claims that the
system of categories is dynamic in a certain sense, which is also the point of the
present study.
This is what I regard as the legacy of Kant’s philosophy.
60b. Transcendental idealism — empirical realism. Note that Kant’s

suggestion to follow the example of Copernicus gives rise to a transcendental ide-
alism, not a dogmatic one, i.e., Kant does not claim that there is no reality in
itself, but only that our knowledge is confined to appearances. Hence it is, in fact,
somewhat misleading to say that the objects have to conform to the subject. This,
however, is a frequently used phrase, even by Kant himself though he sometimes
is careful enough to make clear, within parentheses, that it is not the object itself,
but, e.g., the object as object of the senses.
In view of the many possible misunderstandings of the nature of this approach,
it seems worthwhile to devote some space to Kant’s idealism.
I begin with a classification which Kant added in the second edition.
p. 274.
Idealism—meaning thereby material idealism—is the theory which de-
clares the existence of objects in space outside us either to be merely doubt-
ful and indemonstrable or to be false and impossible. The former is the
problematic idealism of Descartes, which holds that there is only one em-
pirical assertion that is indubitably certain, namely, that ‘I am’. The latter
is the dogmatic idealism of Berkeley. He maintains that space, with all the
things of which it is the inseparable condition, is something which is in
itself impossible; and he therefore regards the things in space as merely
imaginary entities.
Comment. Is ‘I am’ an empirical assertion? See also Boos [1983] for a discussion.
The dictum of all genuine idealists, from the Eleatic school to Bishop
Berkeley, is contained in this formula: “All cognition through the senses and
experience is nothing but sheer illusion, and only in the ideas of the pure
understanding and reason is there truth.”
The principle that throughout dominates and determines my idealism
is, on the contrary: “cognition of things merely from pure understanding or
pure reason is nothing but sheer illusion, and only in experience is there
truth.”
But this is directly contrary to idealism proper. How came I then to
use this expression for quite an opposite purpose[[. . . .]]
[[. . .]] Space and time, together with all that they contain, are not things
in themselves or their qualities but belong merely to the appearances of
the things in themselves; up to this point I am one in confession with
the above idealists. But these, and among them more particularly Berkeley,
regarded space as a mere empirical representation that, like the appearances
it contains, is, together with its determinations, known to us only by means
of experience and perception. I, on the contrary, prove in the first place
that space (and also time, which Berkeley did not consider) and all its
determinations can be cognized a priori by us, because, no less than time, it
inheres in us a pure form of our sensibility before all perception of experience
and makes possible all intuition of sensibility, and therefore all appearances.
It follows from this that, as truth rests on universal and necessary laws as
its criteria, experience, according to Berkeley, can have no criteria of truth
because its appearances (according to him) have nothing a priori at their
foundation, whence it follows that experience is nothing but sheer illusion;
whereas with us, space and time (in conjunction with the pure concepts of
the understanding) prescribe their law a priori to all possible experience
and, at the same time, afford the certain criterion for distinguishing truth
from illusion therein.
(3) N. K. Smith [1929], pp. 345 f; translating Kant [1781] (A), p. 369.
By transcendental idealism I mean the doctrine that appearances are to be
regarded as being, one and all, representations only, not things in them-
selves, and that time and space are therefore only sensible forms of our in-
tuition, not determinations given as existing by themselves, nor conditions
of objects viewed as things in themselves. To this idealism there is opposed
a transcendental realism which regards time and space as something given
in themselves, independently of our sensibility. The transcendental realist
thus interprets outer appearances (their reality being taken as granted) as
things-in-themselves, which exist independently of us and of our sensibility
[[. . .]]. It is, in fact, this transcendental realist who afterwards plays the part
of empirical idealist. After wrongly supposing that objects of the senses, if
they are to be external, must have an existence by themselves, and inde-
pendently of the senses, he finds that, judged from this point of view, all
our sensuous representations are inadequate to establish their reality.
The transcendental idealist, on the other hand, may be an empirical
realist[[.]]
Comment. This passage can only be found in the first edition and I know of no
corresponding part in the second edition. This may indicate that Kant was no
longer happy with this formulation.
The transcendental idealist is [[. . .]] an empirical realist, and allows to
matter, as appearance, a reality which does not permit of being inferred,
but is immediately perceived. Transcendental realism, on the other hand, in-
evitably falls into difficulties, and finds itself obliged to give way to empirical
idealism, in that it regards the objects of outer sense as something distinct

from the senses themselves, treating mere appearances as self-subsistent
beings, existing outside us. On such a view as this, however clearly we may
be conscious of our representation of these things, it is still far from certain
that, if the representation exists, there exists also the object correspond-
ing to it. In our system, on the other hand, these external things, namely
matter, are in all their configurations and alterations nothing but mere
appearances, that is, representations in us, of the reality of which we are
immediately conscious.
Comment. As for the preceding quotation.
(5) N. K. Smith [1929], p. 220; translating Kant [1781] (A), p. 190; (B 235).
Our transcendental idealism [[. . .]] admits the reality of the objects of
outer intuition, as intuited in space, and of all changes in time, as repre-
sented by inner sense. For since space is a form of that intuition which we
entitle outer, and since without objects in space there would be no empiri-
cal representation whatsoever, we can and must regard the extended beings
in it as real; and the same is true of time. But this space and this time,
and with them all appearances, are not in themselves things!in themselves;
they are nothing but representations, and cannot exist outside our mind.
(6) N. K. Smith [1929], p. 220; translating Kant [1781] (A), p. 190; (B 235).
How things may be in themselves, apart from the representations through
which they affect us, is entirely outside our sphere of knowledge.
I find that [[. . .]] all properties which constitute the intuition of a body belong
merely to its appearance. The existence of the thing that appears is thereby
not destroyed, as in genuine idealism, but it is only shown that we cannot
possibly know it by the senses as it is in itself.
I have myself given this my theory the name of transcendental idealism,
but that cannot authorize anyone to confound it either with the empirical
idealism of Descartes [[. . .]], or with the mystical and visionary idealism
of Berkeley [[. . .]]. My idealism concerns not the existence of things (the
doubting of which, however, constitutes idealism in the ordinary sense),
since it never came into my head to doubt it, but it concerns the sensuous
representation of things to which space and time especially belong. [[. . .]]
[[T]]he word “transcendental,” which with me never means a reference of our
cognition to things, but only to the cognitive faculty, was meant to obviate
this misconception.
Comment. Bad luck; but what do you expect with all the idiots out there? Cf. also
quotation 61.16 (1) above.
(9) N. K. Smith [1929], p. 259; translating Kant [1783], p. 238 (B 298).

The transcendental employment of a concept in any principle is its appli-
cation to things in general and in themselves; the empirical employment
is its application merely to appearances; that is, to objects of a possible
experience.
Comment. There is a footnote in my German edition quoting Kant’s note from

his personal copy regarding the phrase “Gegenständ überhaupt und an sich selbst”
(“things in general and in themselves”) which runs in my translation: “objects which
are not given to us in any intuition, hence non-sensible objects”.
Remark 60.6. Kant’s distinction between transcendental idealism and empirical

realism is subtle and not always recognized.
p. 274.
Dogmatic idealism is unavoidable, if space be interpreted as a property that
must belong to things in themselves. For in that case space, and everything
to which it serves as condition, is a non-entity.
Comment. Recall that Kant labeled Berkeley’s idealism dogmatic.11 What Kant
says here is virtually that dogmatic idealism is a consequence of transcendental
realism.
Problematic idealism, which makes no such assertion [[as the dogmatic one]],
but merely pleads incapacity to prove, through immediate experience ex-
cept our own, is, in so far as it allows of no decisive judgment until sufficient
proof has been found, reasonable and in accordance with a thorough and
philosophical mode of thought. The required proof must, therefore, show
that we have experience, not merely imagination of outer things; and this,
it would seem, cannot be achieved save by proof that even our inner experi-
ence, which for Descartes is indubitable, is possible only on the assumption
of outer experience.
At this stage, Kant’s transcendental idealism is still a mere hypothesis. It has to

be confirmed by the further elaboration of his approach. This, indeed, is the subject
proper of the Transcendental Doctrine of Elements. This hypothetical character of
transcendental idealism and its confirmation by the Critique of Pure Reason as a
whole is emphasized by Kant in the introduction.
11 See quotation 60.5 (1). “Dogmatic idealism” does not occur in the index of N. K. Smith’s
translation.
Quotation 60.8. N. K. Smith [1929], p. 24; translating Kant [1787b] (B), pp. xix–
xxi.
[[W]]e are brought to the conclusion that we can never transcend the limits
of possible experience [[. . .]]. This situation yields, however, just the very
experiment by which, indirectly, we are enabled to prove the truth of this
first estimate of our a priori knowledge of reason, namely, that such knowl-
edge has to do only with appearances, and must leave the thing in itself
as indeed real per se, but as not known by us. For what necessarily forces
us to transcend the limits of experience and of all appearances is the un-
conditioned, which reason, by necessity and by right, demands in things in
themselves, as required to complete the series of conditions. If, then, on
the supposition that our empirical knowledge conforms to objects as things
in themselves, we find that the unconditioned cannot be thought without
contradiction, and that when, on the other hand, we suppose that our re-
presentation of things, as they are given to us, does not conform to these
things in themselves, but that these objects, as appearances, conform to
our mode of representation, the contradiction vanishes; and if, therefore,
we thus find that the unconditioned is not to be met with in things, so far
as we know them, that is, so far as they are given to us, but only so far as
we do not know them, that is, so far as they are things in themselves, we
are justified in concluding that what we at first assumed for the purposes
of experiment is now definitely confirmed.
Comment. In other words, Kant’s Critique of Pure Reason may be viewed as a
sort of reductio ad absurdum. As opposed to this quotation: in [1781] (A), pp. 490 f
(B 518 f) Kant claims that he has “sufficiently proved” the transcendental ideality
of space and time.
Remark 60.9. This layout of Kant’s Critique of Pure Reason as a reductio ad
absurdum does not always seem to be clear. It has been strongly emphasized in
Kowalewsky [1918] and more recently in Krausser [1988].
This then is what Kant sets out to do: to prove by reductio ad absurdum that
our knowledge can only be of things as they appear to us and not as they are in
themselves.
60c. The analytic-synthetic distinction. Kant’s turn to the cognitive fac-
ulties is accompanied by a distinction of two different kinds of judgment: so-called
analytic and synthetic judgments. In this section I shall try to give an impression
of how Kant himself dealt with this distinction.
Remarks 60.10. (1) A distinction of the sort analytic-synthetic which Kant laid at
the bottom of his transcendental idealism will be my concern in §115 from a purely
formal point of view. It is a key notion for any gnoseological theory in the tradition
of transcendental idealism.
(2) Kant’s analytic-synthetic distinction is commonly misinterpreted as something
of the kind logical-factual.12
12 Cf. quotations 87.4 (1) and (2) in these materials.
Quotation 60.11. (1) N. K. Smith [1929], p. 48; translating Kant [1781] (A), pp. 6 f
(B 10 f).
Either the predicate B belongs to the subject A, as something which is
(covertly) contained in this concept A; or B lies outside the concept A, al-
though is does indeed stand in connection with it. In the one case I entitle
the judgment analytic, in the other synthetic. Analytic judgments (affir-
mative) are therefore those in which the connection of the predicate with
the subject is thought through identity; those in which this connection
is thought without identity should be entitled synthetic. The former, as
adding nothing through the predicate to the concept of the subject, but
merely breaking it up into those constituent concepts that have all along
been thought in it, although confusedly, can also be entitled explicative.
The latter, on the other hand, add to the concept of the subject a predicate
which has not been in any wise thought in it, and which no analysis could
possibly extract from it; and they may therefore be entitled ampliative.
Comment. Notice the definition of “analytic judgement” in terms of being “thought”.
Cf. also the Prolegomena, § 2. Also: the “either–or”.
(2) N. K. Smith [1929], p. 49, translating Kant [1787b] (B), p. 12.
If I say, for instance, ‘All bodies are extended’, this is an analytic judgment.
For I do not require to go beyond the concept which I connect with ‘body’ in
order to find extension as bound up with it. To meet with this predicate, I
have merely to analyse the concept, that is, to become conscious to myself
of the manifold which I always think in that concept. The judgment is
therefore analytic. But when I say, ‘All bodies are heavy’, the predicate is
something quite different from anything that I think in the mere concept
of a body in general; and the addition of such a predicate therefore yields
a synthetic judgment.
Comment. Note the psychological criterion: “become conscious to myself”.
In this context it is worthwhile having a look at what Kant says about logical
predicates.
Quotation 60.12. N. K. Smith [1929], p. 504; translating Kant [1781] (A), p. 598
(B 626).
Anything we please can be made to serve as a logical predicate; the subject
can even be predicated of itself; for logic abstracts from all content. But a
determining predicate is a predicate which is added to the concept of the
subject and enlarges it. Consequently, it must not be already contained in
the concept.
‘Being’ is obviously not a real predicate; that is, it is not a concept of
something which could be added to the concept of a thing. It is merely the
positing of a thing, or of certain determinations, as existing in themselves.
Logically, it is merely the copula of a judgment.
Comments. (α) This is a crucial point in view of Hegel’s treatment of the concept
of Being. Apart from the well-known first chapter in the Science of Logic, see also
quotation 65.12 (2) in these materials.
(β) Allison [1983], p. 76 f, renders Kant as saying that ‘existence’ is not a real
predicate.13 I wonder if he is possibly confusing ‘existence’ and ‘being’.
(γ) Note the passage “the subject can even be predicated of itself”. Apparently,
Kant had no principal reservations against self-reference.
It may seem obvious that Kant’s formulations are pretty hopeless in the face of
modern logic, but this is no reason for me to dismiss the idea of such a distinction. It
needs a very careful exposition, and this will part of my endeavour later on. For the
moment I am only interested in presenting Kant’s original ideas, and the distinction
of synthetic and analytic judgments plays a central role in Kant’s enterprise.
Quotations 60.13. (1) Carus and Ellington [1977], p. 14; translating Kant [1783],
pp. 268 f.14
It must at first be thought that the proposition 7 + 5 = 12 is a mere
analytic judgment, following from the concept of the sum of seven and five,
according to the principle of contradiction. But on closer examination it
appears that the concept of the sum of 7 + 5 contains merely their union
in a single number, without its being at all thought what the particular
number is that unites them. The concept of twelve is by no means thought
by merely thinking of the combination of seven and five; and, analyze this
possible sum as we may, we shall not discover twelve in the concept. We must
go beyond these concepts, by calling to our aid some intuition corresponding
to one of them, i.e., either our five fingers or five points (as Segner has in
his Arithmetic); and we must add successively the units of the five given in
the intuition to the concept of seven. Hence our concept is really amplified
by the proposition 7 + 5 = 12, and we add to the first concept a second
one not thought in it. Arithmetical judgments are therefore synthetic, and
the more plainly according as we take large numbers; for in such cases it is
clear that however closely we analyze our concepts without calling intuition
to our aid, we can never find the sum by such mere analysis.
Comment. This is a famous passage as regards Kant’s view that mathematical
judgments are analytical judgments. Compared with Leibniz’ analysis of 2 + 2 =
4, Kant’s standard of argumentation is hopelessly poor. Cf. Frege in [1884], p. 7,
regarding Leibniz’ analysis.
The assertion that 7 + 5 is equal to 12 is not an analytic proposition. For
neither in the representation of 7, nor in that of 5, nor in the representation
of the combination of both, do I think the number 12. (That I must do so
in the addition of the two numbers is not to the point, since in the analytic
13 Cf. quotation 68.11 in these materials.
14 Cf. also Kant [1787b] (B), p. 15.
proposition the question is only whether I actually think the predicate in

the representation of the subject.)
Comment. This is, of course, a central point. Notice, first of all, the subjective na-
ture of the criterion of analytic proposition, the “I think”. Secondly, cf. Poincaré in
quotation 70.21 (3) (regarding induction), Frege in quotations 71.7, and my treat-
ment in §137 in the groundworks. At the present stage I only wish to mention that
under that treatment any specific numerical equation like, for instance, 7 + 5 = 12
may well be regarded as analytic, while a general equation involving variables, such
as x + y = y + x, for instance, requires a somewhat synthetic step, manifested in
the introduction of the schema of Z-inferences, somewhat comparable to that of the
axiom of infinity in the set theoretical approach from section 52c, which, in turn,
corresponds to the role of mathematical induction in the proof of proposition 44.33
on p. 539 in the tools.
That the straight line between two points is the shortest, is a synthetic
proposition. For my concept of straight contains nothing of quantity, but
only of quality. The concept of the shortest is wholly an addition, and
cannot be derived, through any process of analysis, from the concept of the
straight line. Intuition, therefore, must here be called in; only by its aid is
the synthesis possible.
(4) Abbott [1873], p. 142; translating Kant [1788], pp. 61 f.
Mathematics escaped well, so far, because Hume thought that its propo-
sitions were analytical; that is, proceeded from one property to another, by
virtue of identity, and consequently according to the principle of contradic-
tion. [[(]]This, however, is not the case, [[whilst]], on the contrary, they are
synthetical; and although geometry, for example, has not to do with the
existence of things, but only with their a priori properties in a possible
intuition, yet it proceeds just as in the case of the causal notion, from one
property (A) to another wholly distinct (B), as necessarily connected with
the former.[[)]] Nevertheless, mathematical science, so highly vaunted for
its apodictic certainty, must at last fall under this empiricism for the same
reason for which Hume put custom in the place of objective necessity in the
notion of cause, and, in spite of all its pride must consent to lower its bold
pretensions of claiming assent à priori, and depend for assent to the uni-
versality of its propositions on the kindness of observers, who, when called
as witnesses, would surely not hesitate to admit that what the geometer
propounds as a theorem they have always perceived to be the fact, and
consequently, although it be not necessarily true, yet they would permit us
to expect it to be true in the future. In this manner Hume’s empiricism
leads inevitably to scepticism, even with regard to mathematics[[.]]
Comment. I inserted round brackets in the above translation in those places where
they occur in the German original. I further replaced “since” by “whilst” to avoid
the impression that Kant is saying: it is not the case that mathematical truth is
analytic, since they are synthetic. The German original is “indem“.
The concept of magnitude in general can never be explained except by
saying that it is that determination of a thing whereby we are enabled to
think how many times a unit is posited in it. But this how-many-times is
based on a successive repetition, and therefore on time and the synthesis of
the homogeneous in time.
Comment. Kant touches here on a point which is quite crucial in the present study;
but his argumentation is appallingly poor.
The main application of the analytic-synthetic distinction is the formulation of
the proper problem of pure reason which is the key for Kant to tackle the problem
of why metaphysics failed so far. The next quotation is meant to illustrate this.
Quotation 60.14. N. K. Smith [1929], p. 55; translating Kant [1787b] (B), p. 19.
[[T]]he proper problem of pure reason is contained in the question: How are
a priori synthetic judgments possible?
That metaphysics has hitherto remained in so vacillating a state of un-
certainty and contradiction, is entirely due to the fact that this problem, and
perhaps even the distinction between analytic and synthetic judgments, has
never previously been considered. [[. . .]] Among philosophers, David Hume
came nearest to envisaging this problem, but still was very far from con-
ceiving it with sufficient definiteness and universality.
As regards the mentioning of Hume, it seems more appropriate to emphasize
the difference than the similarity, because it is this difference which Kant claims
enables him to develop his approach.
Now in which way is this question the key?
This question marks the point where Kant’s genuine contribution to philosophy,
the idea of a transcendental philosophy begins. This I leave to a new paragraph.
Before I close the present one, however, I quote a passage from Kant’s Prolegomena
which I find helpful for understanding in which way Kant thought he had overcome
the problem Hume was caught up with, as well as a quotation from the Critique
that is likely to shed some light on this question.
Quotations 60.15. (1) Carus and Ellington [1977], p. 48, footnote 15; translating
Kant [1783], p. 305.15
[[H]]ow does the proposition that judgments of experience contain necessity
in the synthesis of perceptions agree with my statement so often before
inculcated that experience, as cognition a posteriori, can afford contingent
judgments only? When I say that experience teaches me something, I mean
only the perception that lies in experience—for example, that heat always
follows the shining of the sun on a stone; consequently, the proposition of
experience is always so far contingent. That this heat necessarily follows
the shining of the sun is contained indeed in the judgment of experience
(by means of the concept of cause), yet is a fact not learned by experience;
15 Cf. also Kant [1787b] (B), p. 15.
for, conversely, experience is first of all generated by this addition of the

concept of the understanding (of cause) to perception.
(2) U. Petersen [2002], p. 786; translating Kant [1787b] (B), p. 125.
There are only two cases possible in which synthetic representation and
their objects (can) come together (zusammentreffen), (can) relate to one
another in a necessary way, and, as it were, can meet one another. Either if
it is the object alone which makes the representation possible, or if it is the
latter alone which makes the object possible. If it is the former, then this
relation is only empirical, and the representation is never possible a priori.
And this is the case with appearance, in view of that which is in them that
belongs to sensation. If it is the latter, however, then, because representation
in itself [[. . .]] does not produce its object as regards existence (dem Dasein
nach), the representation in view of the object is still determining a priori,
if through it alone it is possible to cognize something as an object.
§ 61. The conception of critique
I now come to the second feature in Kant’s set up, the conception of an investiga-
tion into the faculties (or properties) of the mind. This forms the core of Kant’s
transcendental philosophy.
61a. The general structure of knowledge and the division of logic.

In order to understand Kant’s argumentation, in particular why he considers his
transcendental idealism vindicated by the Antinomy of Pure Reason, I now shall
have a closer look at his view of how knowledge is constituted. This sophisticated
construction of mental faculties is designed to make sense, inter alia, of the four
cosmological antinomies. Later I shall argue — following Hegel — that Kant’s in-
sistence on the completeness of these four antinomies reveals that his approach —
though fascinating as it may be as a theoretical exercise — is inadequate.
Quotations 61.1. (1) N. K. Smith [1929], p. 92; translating Kant [1781] (A), p. 50
(B 74).
Our knowledge springs from two fundamental sources of the mind; the first
is the capacity of receiving representations (receptivity for impressions),
the second is the power of knowing an object through these representations
(spontaneity [in the production] of concepts). Through the first an object
is given to us, through the second the object is thought in relation to that
[given] representation (which is a mere determination of the mind).
(2) N. K. Smith [1929], p. 93; translating Kant [1781] (A) p. 51 (B 75).
If the receptivity of our mind, its power of receiving representations in
so far as it is in any wise affected, is to be entitled sensibility, then the
mind’s power of producing representations from itself, the spontaneity of
knowledge, should be called the understanding.
§ 61. THE CONCEPTION OF CRITIQUE 787
Comment. Apparently, Kant is quite traditional in his distinction of αἰσθητά and

νοητά .16 As regards his use of words: “sources of knowledge” are “capacities” of
“receptivity” or “spontaneity”, i.e., certain faculties, distinguished from the “elements
of knowledge”, “intuitions” and “concepts”, i.e., certain objects.
The situation is more complicated than that, however.
(B 377).
The genus is representation in general (representatio). Subordinate to it
stands representation with consciousness (perceptio). A perception which
relates solely to the subject as the modification of its state is sensation
(sensatio), an objective perception is knowledge (cognitio). This is either
intuition or concept (intuitus vel conceptus). The former relates immedi-
ately to the object and is single, the latter refers to it mediately by means
of a feature which several things may have in common. The concept is ei-
ther an empirical or a pure concept. The pure concept, in so far as it has
its origin in the understanding alone (not in the pure image of sensibility),
is called a notion. A concept formed from notions and transcending the
possibility of experience is an idea or concept of reason.
Comment. A “feature which several things have in common”. Compare with set.
Quotations 61.3. (1) N. K. Smith [1929], p. 92; translating Kant [1781] (A), pp. 50 f
(B 74 f).
Intuition and concepts constitute [[. . .]] the elements of all our knowledge,
so that neither concepts without an intuition in some way corresponding
to them, nor intuition without concepts, can yield knowledge. Both may be
either pure or empirical. When they contain sensation (which presupposes
the actual presence of the object), they are empirical. When there is no
mingling of sensation with the representation, they are pure. [[. . .]] Pure in-
tuition, therefore, contains only the form under which something is intuited;
the pure concept only the form of the thought of an object in general.
[[T]]he object cannot be given to a concept otherwise than in intuition;
for though a pure intuition can indeed precede the object a priori, even
this intuition can acquire its object, and therefore objective validity, only
through the empirical intuition of which it is the mere form.
Quotations 61.4. (1) N. K. Smith [1929], p. 300; translating Kant [1781] (A),
pp. 298 f (B 355).
All our knowledge starts with the senses, proceeds from thence to un-
derstanding, and ends with reason, beyond which there is no higher faculty
to be found in us for elaborating the matter of intuition and bringing it
under the highest unity of thought.
16 Cf. [1787b] (B), p. 35 f, footnote.
Comment. Reason is somewhat new. Cf. Hegel in quotation 64.5 (2) in these ma-
terials.
In whatever manner and by whatever means a mode of knowledge may
relate to objects, intuition is that through which it is in immediate relation
to them, and to which all thought as a means is directed. But intuition
takes place only in so far as the object is given to us. This again is only
possible, to man at least, in so far as the mind is affected in a certain way.
The capacity (receptivity) for receiving representations through the mode in
which we are affected by objects, is entitled sensibility. Objects are given to
us by means of sensibility, and it alone yields us intuitions; they are thought
through the understanding, and from the understanding arise concepts. But
all thought must, directly or indirectly, by way of certain characters, relate
ultimately to intuitions, and therefore, with us, to sensibility, because in no
other way can an object be given to us.
Comment. Kant is not at all consistent in his use of notions. Here he first speaks
of intuition as “stattfinden” (transl. “taking place”), which seems to suggest that
intuition is some sort of process, and later of sensibility as “liefern” intuition (transl.
“yield”), which seems to suggest that it is some sort of object. In view of other
passages I prefer the objectual interpretation of intuition.
[[T]]he categories do not afford us any knowledge of things; they do so only
through their possible applications to empirical intuition. In other words,
they serve only for the possibility of empirical knowledge; and such knowl-
edge is what we entitle experience. Our conclusion is therefore this: the
categories, as yielding knowledge of things, have no kind of application,
save only in regard to things which may be objects of possible experience.
Comment. This strikes me as a crucial point. I try to accommodate for it in dia-
gram 104.4 and again 104.17 in the groundworks. It marks a significant difference
to Hume.
Let me emphasize this claim: All thought must relate ultimately to intuition.
This is a fairly strong claim and Kant gives no support for it, apart from adding the
further claim that in no other way can an object be given to us; a claim which is as
unfounded as the first one. It indicates, however, how much Kant is still committed
to the dominating role of sense perception in the domain of knowledge.
There is intuition, understanding, and reason; all three of them have synthetic
a priori applications.
Quotation 61.5. N. K. Smith [1929], pp. 97; translating Kant [1781] (A), pp. 58 ff
(B 83 ff).
If truth consists in the agreement of knowledge with its object, that
object must thereby be distinguished from other objects; for knowledge
is false, if it does not agree with the object to which it is related, even
although it contains something which may be valid of other objects. Now
a general criterion of truth must be such as would be valid in each and

every instance of knowledge, however their objects may vary. It is obvious
however that such a criterion [being general] cannot take account of the
[varying] content of knowledge (relation to its [specific] object). But since
truth concerns just this very content, it is quite impossible, and indeed
absurd, to ask for a general test of the truth of such content. A sufficient
and at the same time general criterion of truth cannot possibly be given.
Since we have already entitled the content of knowledge its matter, we must
be prepared to recognise that of the truth of knowledge, so far as its matter
is concerned, no general criterion can be demanded. Such a criterion would
by its very nature be self-contradictory.
But on the other hand, as regards knowledge in respect of its mere form
(leaving aside all content), it is evident that logic, in so far as it expounds
the universal and necessary rules of the understanding, must in these rules
furnish criteria of truth. Whatever contradicts these rules is false. For the
understanding would thereby be made to contradict its own general rules
of thought, and so to contradict itself. These criteria, however, concern
only the form of truth, that is, of thought in general; and in so far they
are quite correct, but are not by themselves sufficient. For although our
knowledge may be in complete accordance with logical demands, that is,
may not contradict itself, it is still possible that it may be in contradiction
with its object. The purely logical criterion of truth, namely, the agreement
of knowledge with the general and formal laws of the understanding and
reason, is a conditio sine qua non, and is therefore the negative condition
of all truth. But further than this logic cannot go. It has no touchstone for
the discovery of such error as concerns not the form but the content.
General logic resolves the whole formal procedure of the understanding

and reason into its elements, and exhibits them as principles of all logi-
cal criticism of our knowledge. This part of logic, which may therefore be
entitled analytic, yields what is at least the negative touchstone of truth.
Its rules must be applied in the examination and appraising of the form of
all knowledge before we proceed to determine whether their content con-
tains positive truth in respect to their object. But since the mere form of
knowledge, however completely it may be in agreement with logical laws,
is far from being sufficient to determine the material (objective) truth of
knowledge, no one can venture with the help of logic alone to judge re-
garding objects, or to make any assertion. We must first, independently of
logic, obtain reliable information; only then are we in a position to enquire,
in accordance with logical laws, into the use of this information and its
connection in a coherent whole, or rather to test it by these laws. There
is, however, something so tempting in the possession of an art so specious,
through which we give to all our knowledge, however uninstructed we may
be in regard to its content, the form of understanding, that general logic,
which is merely a canon of judgment, has been employed as if it were an
organon for the actual production of at least the semblance of misapplied.

General logic [[now, as a putative]] organon, is called dialectic.
Comment. Isn’t this a challenging characterization of dialectic? One just has to find
the guts to take it up and challenge the “putative”. That’s what I take Hegel to have
tried, and it is certainly what I understand myself to be doing in Part G of this
study.
61b. The forms of intuition. Following Kant’s view of the structure of the
faculties of the mind, I begin with intuition; not, however, without a certain caveat
regarding the term “intuition” which is the topic of the next remark.
Remark 61.6. “Intuition” is being used as the English translation of the German
word “Anschauung” and “to intuit” as that of “anschauen“. It might be said that
this translation is symptomatic of the misunderstandings between philosophers of
the two language areas. The German word “anschauen” is basic; it is part of the
vocabulary of a four year old: a little boy having painted his face with his mother’s
lipstick would typically say to his mother: “schau mich an, Mami”. Now if you want
to translate that into English following the way in which Kant has been translated,
what comes out is: “intuit me, mummy”. Given some basic insensitivity towards
language, like a failure or an incapacity to distinguish between an original and a
translation, this may well lead to the idea that Kant had a strange jargon when
talking about the way we look at objects.17 I wish to point out, however, that I don’t
think there is much a translator can do: the English language just doesn’t lend itself
to expressing ideas grouping around the notion of Anschauung. This then, to get
back to the beginning of this remark, might be said to indicate that the reasons for
the misunderstandings between philosophers of the two language areas lies deeper
than that of an unfortunate translation.18

(B 37).
By means of outer sense, a property of our mind, we represent to our-
selves objects as outside us, and all without exception in space.

(B 37 f).
What, then, are space and time? Are they real existences? Are they only
determinations or relations of things, yet such as would belong to things
even if they were not intuited? Or are space and time such that they belong
only to the form of intuition, and therefore to the subjective constitution
of our mind, apart from which they could not be ascribed to anything
whatsoever?
17Cf. quotation 68.23 (2) in these materials.
18 Problems of this kind extend to the reception of Hilbert’s proof theory as well; Gödel’s
remark regarding Hilbert’s notion of Anschauung in quotation 94.12 in these materials is of
interest in this respect.
Comments. (α) I shall use this passage for paraphrasing and linking in Frege in
§107.
(β) The German original for “real existences” is “wirkliche Wesen”.
Kant provides his answer to this question a few pages later.
Quotations 61.9. (1) N. K. Smith [1929], p. 71; translating Kant [1781] (A), p. 26
(B 42).
Space does not represent any property of things in themselves, nor does
it represent them in their relation to one another. That is to say, space
does not represent any determination that attaches to the objects them-
selves, and which remains even when abstraction has been made of all the
subjective conditions of intuition.
(2) N. K. Smith [1929], pp. 71 f; translating Kant [1781] (A), pp. 26 f (B 42 f).
It is, therefore, solely from the human standpoint that we can speak
of space, of extended things, etc. If we depart from the subjective condi-
tion under which alone we can have outer intuition, namely, liability to be
affected by objects, the representation of space stands for nothing whatso-
ever. This predicate can be ascribed to things only insofar as they appear
to us, that is, only to objects of sensibility. The constant form of this recep-
tivity, which we term sensibility, is a necessary condition of all the relations
in which objects can be intuited as outside us; and if we abstract from
these objects, it is a pure intuition, and bears the name of space. Since we
cannot treat the special conditions of sensibility as conditions of the possi-
bility of things, but only of their appearances, we can indeed say that space
comprehends all things that appear to us as external, but not all things
in themselves, by whatever subject they are intuited, or whether they be
intuited or not. For we cannot judge in regard to the intuitions of other
thinking beings, whether they are bound by the same conditions as those
which limit our intuition and which for us are universally valid.
If we add to the concept of the subject of a judgment the limitation under
which the judgment is made, the judgment is then unconditionally valid.
The proposition, that all things are side by side in space, is valid under the
limitation that these things are viewed as objects of our sensible intuition.
If, now, I add the condition to the concept, and say that all things, as outer
appearances are side by side in space, the rule is valid universally and with-
out limitation. Our exposition therefore establishes the reality, that is, the
objective validity, of space in respect of whatever can be presented to us
outwardly as object, but also at the same time the ideality of space in re-
spect of things when they are considered in themselves through reason, that
is, without regard to the constitution of our sensibility. We assert, then, the
empirical reality of space, as regards all possible outer experience; and yet
at the same time we assert its transcendental ideality—in other words, that
it is nothing at all, immediately we withdraw the above condition, namely,
its limitation to possible experience, and so look upon it as something that

underlies things in themselves.
Comments. (α) This quotation is actually the continuation of the preceding one,
but the thought expressed in it has a new and different relevance in the context of
my enterprise and it will play an eminent role later independently of that of the
preceding quotation. The point that I want to draw the reader’s attention to is that
Kant’s suggestion only works, if “the limitation under which the judgment is made”
can be expressed in a statement; in logical terms: in a finitely axiomatizable theory.
This is not the case, for instance, in a first order formulation of arithmetic, i.e., what
cannot be expressed is a limitation of the kind “if this variable takes only natural
numbers as values“, because the notion of natural number cannot be expressed as a
first order property; in logical terms: first order arithmetic requires infinitely many
induction axioms. It is the case, however, in a second order formulation of arithmetic
and this has been used by Takeuti for the purpose of a relativization which enabled
him to reduce the consistency of second order theories of arithmetic and analysis to
that of cut elimination in second order logic with different stages of comprehension.
This is partly sketched in the tools in sections 40b and 47d. I shall employ this
kind of approach in the groundworks for the purpose of expressing limitations
such as “this variable takes only truth-definite propositions as values”.
(β) I suggest this should also be seen in the context of Montague’s antinomy19 and
the proof of results such as 48.28 in the tools.
(γ) Note the phrase “reality, that is the objective validity”.
Against this theory, which admits the empirical reality of time, but
denies its absolute and transcendental reality, I have heard men of intelli-
gence so unanimously voicing an objection, that I must suppose it to occur
spontaneously to every reader to whom this way of thinking is unfamiliar.
The objection is this. Alterations are real, this being proved by change of
our own representations—even if all outer appearances, together with their
alterations, be denied. Now alterations are possible only in time, and time
is therefore something real. There is no difficulty in meeting this objection.
I grant the whole argument. Certainly time is something real, namely the
real form of inner intuition. It has therefore subjective reality in respect of
inner experience; that is, I really have the representation of time and of my
determinations in it. Time is therefore to be regarded as real, not indeed
as object but as the mode of representation of myself as object. If without
this condition of sensibility I could intuit myself, or be intuited by another
being, the very same determinations which we now represent to ourselves
as alterations would yield knowledge into which the representation of time,
and therefore also of alteration, would in no way enter.
Comment. The last part is quoted in Gödel [1949], p. 557, footnote 3; see also quo-
tation 93.9 (1) in these materials.
19 See Montague [1963], p. 292, theorem 1; cf. lemma 48.32 and 48.33iv in the tools.
[[E]]verything intuited in space or time, and therefore all objects of any
experience possible to us, are nothing but appearances, that is, mere repre-
sentations, which, in the manner in which they are represented, as extended
beings, or as series of alterations, have no independent existence outside our
thoughts. This doctrine I entitle transcendental idealism. The realist in the
transcendental meaning of this term, treats these modifications of our sensi-
bility as self-subsistent things, that is, treats mere representations as things
in themselves.
It would be unjust to ascribe to us that long-decried empirical idealism,
which, while it admits the genuine reality of space, denies the existence of
the extended beings in it, or at least considers their existence doubtful[[.]]
pp. 38 f (B 55 f).
Time and space are [[. . .]] two sources of knowledge, from which bod-
ies of a priori synthetic knowledge can be derived. [[. . .]] Time and space,
taken together, are the pure forms of all sensible intuition, and so are what
make a priori synthetic propositions possible. But these a priori sources of
knowledge, being merely conditions of our sensibility, just by this very fact
determine their own limits, namely, that they apply to objects only in so
far as objects are viewed as appearances, and do not present things as they
are in themselves.
Space and time are indeed a priori representations, which dwell in us
as forms of our sensible intuition, before any real object, determining our
sense through sensation, has enabled us to represent the object under those
sensible relations.
I close this section with what I consider an intriguing point without really
knowing where to place it.
Quotation 61.11. (1) N. K. Smith [1929], p. 76; translating Kant [1787b] (B),
pp. 48 f.
[[T]]he concept of alteration, and with it the concept of motion, as alteration
of place, is possible only through and in the representation of time; and that
if this representation were not an a priori (inner) intuition, no concept, no
matter what it might be, could render comprehensible the possibility of
alteration, that is, of a combination of contradictorily opposed predicates
in one and the same object, for instance, the being and the not-being of
one and the same thing in one and the same place. Only in time can two
contradictorily opposed predicates meet in one and the same object, namely,
one after the other.
Comment. The contradiction of being at the same place and not being at the same
place is lifted (aufgehoben) in time.
(2) N. K. Smith [1929], pp. 77 f; translating Kant [1781] (A), pp. 34 f.

If we abstract from our mode of inwardly intuiting ourselves—the mode
of intuition in terms of which we likewise take up into our faculty of re-
presentation all outer intuitions—and so take objects as they may be in
themselves, then time is nothing. It has objective validity only in respect
of appearances, these being things which we take as objects of our senses.
It is no longer objective, if we abstract from the sensibility of our intui-
tion, that is, from that mode of representation which is peculiar to us, and
speak of things in general. Time is therefore a purely subjective condition
of our (human) intuition (which is always sensible, that is, so far as we are
affected by objects), and in itself, apart form the subject, is nothing. Never-
theless, in respect of all appearances, and therefore of all the things which
can enter into our experience, it is necessarily objective. We cannot say
that all things are in time, because in this concepts of things in general we
are abstracting from every mode of their intuition and therefore from that
condition under which alone objects can be represented as being in time. If,
however, the condition be added to the concept, and we say that all things
as appearances, that is, as objects of sensible intuition, are in time, then
the proposition has legitimate objective validity and universality a priori.
What we are maintaining is, therefore, the empirical reality of time,
that is, its objective validity in respect of all objects which allow of ever
being given to our senses. And since our intuition is always sensible, no
objects can ever be given to us in experience which does not conform to
the condition of time. On the other hand, we deny to time all claim to
absolute reality; that is to say, we deny that it belongs to things absolutely,
as their condition or property, independently of any reference to the form
of our sensible intuition; properties that belong to things in themselves can
never be given to us through the senses. This, then, is what constitutes the
transcendental ideality of time.
Comment. This is essentially an adaptation to the notion of time of what was said
in quotation 61.9 (3) above regarding space.
61c. Understanding and the categories. Whereas the first part of the
transcendental doctrine of elements, the transcendental aesthetic, dealt with the
principles of sensibility, the second part, the transcendental logic, deals with the
principles of conceptuality. It is, in turn, divided into two sections, the first of
which deals with the concepts of understanding.
p. 62 (B 87).
In a transcendental logic we isolate the understanding—as above, in the
Transcendental Aesthetic, the sensibility—separating out from our knowl-
edge that part of thought which has its origin solely in the understanding.
The employment of this pure knowledge depends upon the condition that
objects to which it can be applied be given to us in intuition. In the absence
of intuition all our knowledge is without objects, and therefore remains en-
tirely empty.
Comment. The main question for my enterprise is, of course, how can this “separat-
ing out” be done? It will probably not come as a surprise that I regard Kant’s own
attempt at this problem as insufficient; but this is not really the point here since I
am not looking for tools in Kant’s Critique of Pure Reason, but for a philosophical
background.

[[W]]e can reduce all acts of the understanding to judgments, and the under-
standing may therefore be represented as a faculty of judgment. [[. . .]] The
functions of the understanding can, therefore, be discovered if we can give
an exhaustive statement of the functions of unity in judgment.
(3) Hartman and Schwarz [1974], pp. 106 f; translating Kant [1800], pp. 532 f (A
156 f).
§ 17. Explanation of a Judgment as Such
A judgment is the presentation of the unity of the consciousness of sev-
eral presentations, or the presentation of their relation so far as they make
up one concept.
§ 18. Matter and Form of Judgments

To every judgment belong, as its essential components, matter and
form. The matter of judgment consists in given cognition that are joined in
judgment into unity of consciousness; in the determination of the manner
in which various presentations as such belong to one consciousness consists
the form of judgment.
§ 19. Object of Logical Reflection:

The mere Form of Judgments
Since logic abstracts from all real or objective differences of cognition,
it can deal with the matter of judgments as little as with the content of
concepts. It therefore has to ponder solely the difference of judgments in
respect of their mere form.
§ 20. Logical Forms of Judgments: Quantity,

Quality, Relation, and Modality
The differences of judgments in regard to their form may be reduced
to the four chief moments of quantity, quality, relation, and modality[[.]]
(4) N. K. Smith [1929], pp. 106 f; translating Kant [1781] (A), p. 70 (B 95).
If we abstract from all content of a judgment, and consider only the
mere form of understanding, we find that the function of thought in judg-
ment can be brought under four heads, each of which contains three mo-
ments. They may be conveniently represented in the following table:
I
Quantity of Judgments
Universal
Particular
II Singular III
Quality Relation
Affirmative Categorical
Negative Hypothetical
Infinite Disjunctive
IV
Modality
Problematic
Assertoric
Apodeictic
Comment. The problem here, at least for me, lies in the formulation “if we abstract
from all content of a judgment and consider only the mere form of understanding”;
how is one to do that? What Kant presents is devoid of all analysis and the criticism
it has attracted is part of an important tradition in the development of Hegel’s
speculative philosophy.20
From judgments Kant moves to categories.
Quotations 61.13. (1) N. K. Smith [1929], p. 112 f; translating Kant [1781], p. 79

(B 105).
The same function which gives unity to the various representations in
a judgment also gives unity to the mere synthesis of various representations
in an intuition; and this unity, in its most general expression, we entitle
the pure concept of the understanding. The same understanding, through
the same operations by which in concepts, by means of analytical unity, it
produced the logical form of a judgment, also introduces a transcendental
content into its representations, by means of the synthetic unity of the
manifold in intuition in general. On this account we are entitled to call
these representations pure concepts of the understanding, and to regard
them as applying a priori to objects—a conclusion which general logic is
not in a position to establish.
In this manner there are precisely the same number of pure concepts of
the understanding which apply a priori to objects of intuition in general,
as, in the preceding table [[preceding quotation]] there have been found to be
logical functions in all possible judgments. For these functions specify the
understanding completely, and yield an exhaustive inventory of its powers.
These concepts we shall, with Aristotle, call categories, for our primary
20 Cf.quotation 64.3 (3) in the next paragraph.
purpose is the same as his, although widely diverging from it in manner of

execution.
Table of Categories
I
Of Quantity
Unity
Plurality
Totality
II III
Of Quality Of Relation
Reality Of Inherence and Subsistence
Negation (substantia et accidens)
Limitation Of Causality and Dependence
(cause and effect)
Of Community (reciprocity
between agent and patient)
IV
Of Modality
Possibility—Impossibility
Existence—Non-existence
Necessity—Contingency
This then is the list of all original pure concepts of synthesis that the
understanding contains within itself a priori.
(2) N. K. Smith [1929], p. 128; translating Kant [1787], p. 129.
[[Categories]] are concepts of an object in general, by means of which the
intuition of an object is regarded as determined in respect of one of the log-
ical functions of judgment. Thus the function of the categorical judgment
is that of the relation of subject to predicate; for example, ‘All bodies are
divisible‘. But as regards the merely logical employment of the understand-
ing, it remains undetermined to which of the two concepts the junction of
the subject, and to which the function of predicate, is to be assigned. For we
can also say, ‘Something divisible is a body’. But when the concept of body
is brought under the category of substance, it is thereby determined that
its empirical intuition in experience must always be considered as subject
and never as mere predicate. Similarly with all the other categories.
Remarks 61.14. (1) The idea of a transcendental logic as distinct from formal
logic is peculiar to Kant’s Critique of Pure Reason. The label “logic” is important
here, since it is taken up by Hegel later, although Hegel explicitly remarked on the
“little influence” that the “complete transformation which philosophical thought in

Germany has undergone” has had “as yet on the structure of logic”.21
(2) The last quotation which concerns the schematism of the table of categories is
included because it contains a few clues for Kant’s approach to the antinomies, but
as for its content it is of little interest for my enterprise.
pp. 82 f (B 108).
In this treatise, I purposely omit the definitions of the categories, al-
though I may be in possession of them. I shall proceed to analyse these con-
cepts only so far as is necessary in connection with the doctrine of method
which I am propounding. In a system of pure reason, definitions of the cat-
egories would rightly be demanded, but in this treatise they would merely
divert attention from the main object of the enquiry, arousing doubts and
objections which, without detriment to what is essential to our purposes,
can very well be reserved for another occasion. Meanwhile, from the little
that I have said it will be obvious that a complete glossary, with all the
requisite explanations, is not only possible, but an easy task.
Comment. Cf. Fichte in quotation 63.2 (2) and Hegel in quotations 64.3 (6) and (7).
In general: it is the easy tasks which prove the greatest obstacles, except, of course,
for those who leave them to others.
The first of the considerations suggested by the table is that while it
contains four classes of the concepts of understanding, it may, in the first
instance be divided into two groups; those in the first group being concerned
with objects of intuition, pure as well as empirical, those in the second group
with the existences of these objects in their relation either to each other or
to the understanding.
The categories in the first group I would entitle the mathematical, those
in the second group the dynamical. The former have no correlates; these
are to be met with only in the second group. This distinction must have
some ground in the nature of the understanding.
Secondly, in view of the fact that all a priori division of concepts must
be by dichotomy, it is significant that in each class the number of the
categories is always the same, namely three. Further, it may be observed
that the third category in each class always arises from the combination of
the second category with the first.
Comment. The last sentence must sound curiously familiar to anyone acquainted
with common presentations of Hegel’s dialectic as a triad.
It was an enterprise worthy of an acute thinker like Aristotle to make
search for these fundamental concepts. But as he did so on no principle,
he merely picked them up as they came his way and at first procured ten
21 Hegel [1812], p. 13; see quotation 63.12 (1) in these materials.
of them, which he called categories (predicaments). Afterwards he believed

that he had discovered five others, which he added under the name of post-
predicaments. But his table still remained defective. Besides, there are to
be found in it some modes of pure sensibility (quando, ubi, situs, also prius,
simul ), and an empirical concept (motus), none of which have any place in a
table of the concepts that trace their origin to the understanding. Aristotle’s
list also enumerates among the original concepts some derivative concepts
(actio, passio); and of the original concepts some are entirely lacking.
61d. Reason and ideas. I now turn to the third faculty: Reason.
p. 302 (B 359).
Understanding may be regarded as a faculty which secures the unity of
appearances by means of rules, and reason as being the faculty which se-
cures the unity of the rules of understanding under principles. Accordingly,
reason never applies itself directly to experience or to any object, but to
understanding, in order to give to the manifold knowledge of the latter an
a priori unity by means of concepts, a unity which may be called the unity
of reason, and which is quite different in kind from any unity that can be
accomplished by the understanding.
Can we isolate reason, and is it, so regarded, an independent source
of concepts and judgments which springs from it alone, and by means of
which it relates to objects; [[. . .]] In a word, the question is, does reason in
itself, that is, does pure reason, contain a priori synthetic principles and
rules, and in what may these principles consist?
(3) N. K. Smith [1929], p. 306; translating Kant [1781] (A), pp. 307 f (B 364).
[[T]]he principle peculiar to reason in general, in its logical employment, is:—
to find for the conditioned knowledge obtained through the understanding
the unconditioned whereby its unity is brought to completion.
But this logical maxim can only become a principle of pure reason
through our assuming that if the conditioned is given, the whole series of
conditions, subordinated to one another—a series which is therefore itself
unconditioned—is likewise given, that is, is contained in the object and its
connection.
Such a principle of pure reason is obviously synthetic; the conditioned
is analytically related to some condition but not to the unconditioned.
(4) N. K. Smith [1929], pp. 308 f; translating Kant [1781] (A), p. 311 (B 367 f).
If the concepts of reason contain the unconditioned, they are concerned with
something to which all experience is subordinate, but which is never itself
an object of experience—something to which reason leads in its inferences
from experience, and in accordance with which it estimates and gauges the
degree of its empirical employment, but which is never itself a member of
the empirical synthesis.
The transcendental concept of reason is [[. . .]] none other than the concept
of the totality of the conditions for any given conditioned. Now since it is
the unconditioned alone which makes possible the totality of conditions,
and, conversely, the totality of conditions is always itself unconditioned, a
pure concept of reason can in general be explained by the concept of the
unconditioned, conceived as containing a ground of the synthesis of the
conditioned.
(6) N. K. Smith [1929], p. 318; translating Kant [1781] (A), pp. 326 f (B 383).
Reason concerns itself exclusively with absolute totality in the employment
of the concepts of the understanding, and endeavours to carry the synthetic
unity, which is thought in the category, up to the completely unconditioned.
We may call this unity of appearances the unity of reason, and that ex-
pressed by the category the unity of understanding. Reason accordingly
occupies itself solely with the employment of understanding, not indeed in
so far as the latter contains the ground of possible experience [[. . .]], but
solely in order to prescribe to the understanding its direction towards a
certain unity of which it has itself no concept, and in such manner as to
unite all the acts of the understanding in respect of every object, into an
absolute whole. The objective employment of the pure concepts of reason
is, therefore, always transcendent, while that of the pure concepts of un-
derstanding must, in accordance with their nature, and inasmuch as their
application is solely to possible experience, be always immanent.
Concepts are based on the spontaneity of thought, sensible intuitions on
the receptivity of impressions.
(B 435 f).
Reason does not really generate any concept. The most it can do is to free
a concept of understanding from the unavoidable limitations of possible
experience, and so to endeavour to extend it beyond the limits of the em-
pirical, though still, indeed, in terms of its relation to the empirical. This is
achieved in the following manner. For a given conditioned, reason demands
on the side of the conditions—to which as the conditions of synthetic unity
the understanding subjects all appearances—absolute totality, and in so
doing converts the category into a transcendental idea.
pp. 334 f (B 391 f).
All transcendental ideas can [[. . .]] be arranged in three classes, the first
containing the absolute (unconditioned) unity of the thinking subject, the
second the absolute unity of the series of conditions of appearance, the third
the absolute unity of the condition of all objects of thought in general.
The thinking subject is the object of psychology, the sumtotal of all

appearances (the world) is the object of cosmology, and the thing which
contains the highest condition of the possibility of all that can be thought
(the being of all beings) the object of theology. Pure reason thus furnishes
the idea for a transcendental , doctrine of the soul (psychologia rationalis),
for a transcendental science of the world (cosmologia rationalis), and, fi-
nally, for a transcendental knowledge of God (theologia transzendentalis).
There are, then, only three kinds of dialectical syllogisms—just so many
as there are ideas in which their conclusions result. In the first kind of
syllogism I conclude from the transcendental concept of the subject itself,
of which, however, even in so doing, I possess no concept whatsoever. This
dialectical inference I shall entitle the transcendental paralogism. The second
kind of pseudo-rational inference is directed to the transcendental concept
of the absolute totality of the series of conditions for any given appearance.
From the fact that my concept of the unconditioned synthetic unity of the
series, as thought in a certain way, is always self-contradictory, I conclude
that there is really a unity of the opposite kind, although of it also I have
no concept. The position of reason in these dialectical inferences I shall
entitle the antinomy of pure reason. Finally in the third kind of pseudo-
rational inference, from the totality of the conditions under which objects
in general, in so far as they can be given to me, have to be thought, I
conclude to the absolute synthetic unity of all conditions of the possibility
of things in general, i.e. from things which I do not know through the merely
transcendental concept of them I infer an ens entium, which I know even
less through any transcendental concept, and of the unconditioned necessity
of which I can form no concept whatsoever. This dialectical syllogism I shall
entitle the ideal of pure reason.
Of these three kinds of dialectical syllogisms I shall only concentrate on the

second one, the Antinomy of Pure Reason, the one whose concern is the sumtotal
of all appearances. Before I do this in the next paragraph, I focus on two aspects
of Kant’s approach, the notion and the role of a transcendental deduction and the
distinction of all “Gegenstände überhaupt” into phenomena and noumena.
61e. The role of a transcendental deduction of the categories and the

transcendental unity of apperception. So far I have presented Kant’s views
about the structure of the mind; what I have not yet presented are his ideas about
how he wants to determine the sources of knowledge. Recall quotation 59.2 (1) where
Kant labeled Locke’s enterprise a “certain physiology of the human understanding”
which he doesn’t seem to consider as quite sufficient. Why? By its very nature an
empirical method is inappropriate for a transcendental investigation. The present
section is devoted to a selection of what Kant says with regard to a deduction of
the categories.
p. 38.
By exposition (expositio) I mean the clear, though not necessarily exhaus-
tive, representation of that which belongs to a concept: the exposition is
metaphysical when it contains that which exhibits the concept as given a
priori.
Comment. The concept of exposition, German: “Erörterung”, and, in particular,
the distinction of metaphysical and transcendental exposition is new in the second
edition.
I understand by a transcendental exposition 22 the explanation of a con-
cept, as a principle from which the possibility of other a priori synthetic
knowledge can be understood.
Comment. This is new in the second edition. Compare it to Kant [1781] (A), p. 85,
re: transcendental deduction.23
(3) N. K. Smith [1929], p. 70; translating Kant [1781] (A), p. 24 (B 38 f).
Space is a necessary a priori representation, which underlies all outer intui-
tion. We can never represent to ourselves the absence of space, though we
can quite well think it as empty of objects. It must therefore be regarded as
the condition of the possibility of appearances, and not as a determination
dependent upon them.
Comment. As an aside: Kant seems to have changed his view radically on this issue
since his early work Gedanken von der wahren Schätzung der lebendigen Kräfte
(1747). I translate from §9 of that work which I found in Becker [1964], p. 176: “It
is easy to show that there would be no space and no extension, had the substances
not the power to act outside of themselves. For without this power there is no
connection, without that no order, and without that finally no space.”
First of all, it seems appropriate to point out some cause for notational confu-
sion. Kant distinguishes three forms of deductions ‘metaphysical’, ‘transcendental’,
and ‘empirical deductions’.
p. 84 (B 116).
Jurists, when speaking of rights and claims, distinguish in a legal action
the question of right (quid juris) from the question of fact (quid facti); and
they demand that both be proved. Proof of the former, which has to state
the right or the legal claim, they entitle the deduction.
22 These italics are mine; they take account of a spaced out print in my German edition.
23 See quotation 61.23 (2) below.

In the metaphysical deduction the a priori origin of the categories has
been proved through their complete agreement with the general logical
functions of thought; in the transcendental deduction we have shown their
possibility as a priori modes of knowledge of objects of an intuition in
general[[.]]
Remark 61.21. The index of N. K. Smith’s translation refers the reader to pp. 106–
119 regarding ‘metaphysical deduction’. I couldn’t find a single mention of ‘meta-
physical deduction’ there, not even of ‘deduction’. It is only on p. 120 of N. K.
Smith’s translation that the term ‘deduction’ is introduced. The only occurrence of
‘metaphysical deduction’ I know of is on p. 170.
p. 86 (B 118 f).
[[A]]n investigation of the first strivings of our faculty of knowledge, whereby
it advances from particular perceptions to universal concepts, is undoubt-
edly of great service. We are indebted to the celebrated Locke for opening
out this line of enquiry. But a deduction of the pure a priori concepts can
never be obtained in this manner[[.]]
(2) N. K. Smith [1929], pp. 127 f; translating Kant [1787b] (B), pp. 127 f.
The illustrious Locke failing to take account of these considerations,
and meeting with pure concepts of the understanding in the experience,
deduced them also from experience, and yet proceeded so inconsequently
that he attempted with their aid to obtain knowledge which far transcends
all limits of experience. David Hume recognised that, in order to be able to
do this, it was necessary that these concepts should have an a priori origin.
But since he could not explain how it can be possible that the understanding
must think concepts, which are not in themselves connected in the under-
standing, as being necessarily connected in the object, and since it never
occurred to him that the understanding might itself, perhaps through these
concepts, be the author of the experience, in which its objects are found, he
was constrained to derive them from experience, namely, from a subjective
necessity (that is, from custom), which arises from repeated association
in experience, and which comes mistakenly to be regarded as objective.
But from these premisses he argued quite consistently. It is impossible, he
declared, with these concepts and the principles to which they give rise,
to pass beyond the limits of experience. Now this empirical derivation, in
which both philosophers agree, cannot be reconciled with the scientific a
priori knowledge which we do actually possess, namely, pure mathematics
and general science of nature; and this fact therefore suffices to disprove
such derivation.
Being pure concepts the categories have to be deduced. This is a main feature
of the Critique of Pure Reason.
p. 67 (B 92).
Transcendental philosophy, in seeking for its concepts, has the advan-
tage and also the duty of proceeding according to a single principle. For
these concepts spring, pure and unmixed, out of understanding which is an
absolute unity, and must therefore be connected with each other accord-
ing to one concept or idea. Such a connection supplies us with a rule, by
which we are enabled to assign its proper place to each pure concept of
the understanding, and by which we can determine in an a priori manner
their systematic completeness. Otherwise we should be dependent in these
matters on our own discretionary judgment or merely on chance.
(2) N. K. Smith [1929], pp. 120 f; translating Kant [1781] (A), p. 85 (B 117).
[[A]]mong the manifold concepts which form the highly complicated web of
human knowledge, there are some which are marked out for pure a priori
employment, in complete independence of all experience; and their right
to be so employed always demands a deduction. For since empirical proofs
do not suffice this kind of employment, we are faced by the problem how
these concepts can relate to objects which they do not obtain from any
experience. The explanation of the manner in which concepts can thus relate
a priori to objects I entitle their transcendental deduction; and from it I
distinguish empirical deduction, which shows the manner in which a concept
is acquired through experience and through reflection upon experience, and
which therefore concerns, not its legitimacy, but only its de facto mode of
origination.
Quotation 61.24. N. K. Smith [1929], p. 121; translating Kant [1781] (A), pp. 85 f
(B 118).
We are already in possession of concepts which are of two quite different
kinds, and which yet agree in that they relate to objects in a completely a
priori manner, namely, the concepts of space and time as forms of sensibil-
ity, and the categories as concepts of understanding. To seek an empirical
deduction of either of these types of concept would be labour entirely lost.
For their distinguishing feature consists just in this, that they relate to
their objects without having borrowed from experience anything that can
serve in the representation of these objects. If, therefore, a deduction of
such concepts is indispensable, it must in any case be transcendental.
Remarks 61.25. (1) In [1781] (A), p. 87 (B 119), Kant claims that he has traced
back the “concepts” of space and time to their sources by means of a “transcendental
deduction”. This calls for two comments on my part: firstly, in the Transcendental
Aesthetic, space and time are not introduced as concepts but as forms of sensibil-
ity (although later Kant commonly refers to them as “concepts”, in particular in
the additions to the second edition), and it is clearly stated that “from the under-
standing arise concepts”; secondly, in the Transcendental Aesthetic Kant speaks of
a transcendental exposition, but not of a transcendental deduction.
(2) The transcendental deduction of the categories is probably the most difficult and
controversial part in Kant’s Critique of Pure Reason. It is remarkable, at least, that
the section which is concerned with it, viz., §§15–27 in the second edition, has been
completely rewritten for the second edition.24 Kant himself talks about the “obscu-
rity of the deduction of the concepts of understanding”.25 Quotation 61.27 (2) below
indicates that Kant thought of it highly. Nevertheless, the details of its argumenta-
tion are of no interest for my enterprise. The interested reader who wonders what
is going to take the place of the transcendental unity of apperception is referred to
chapter XXXIII in the groundworks, more specifically to §132, the introduction
of Z-inferences, as well as to chapter XXX which discusses the philosophical import
of the introduction of Z-inferences.
(3) Are these considerations by Kant sufficient to establish why a category such
as causality is necessarily applicable? Kant has indicated how to discover the cat-
egories, but to what extent does this establish a foundation in the sense of an
explanation why the categories apply necessarily?
According to Kant’s view, there is a basic difference between the level of in-
tuition and that of understanding as regards their objective validity. This view is
relevant for my enterprise, since I distinguish the ‘tautological’ character of purely
logical constants from that of those constants which are introduced by Z-inferences,
for instance.
p. 89 f (B 121 ff).
We have already been able with but little difficulty to explain how the
concepts of space and time, although a priori modes of knowledge, must
necessarily relate to objects, and how independently of all experience they
make possible a synthetic knowledge of objects. For since only by means
of such pure forms of sensibility can an object appear to us, and so be
an object of empirical intuition, space and time are pure intuitions which
contain a priori the condition of the possibility of objects as appearances,
and the synthesis which takes place in them has objective validity.
The categories of understanding, on the other hand, do not represent
the conditions under which objects are given in intuition. Objects may,
therefore, appear to us without their being under the necessity of being
related to the functions of understanding; and understanding need not,
therefore contain their a priori conditions. Thus a difficulty such as we did
not meet with in the field of sensibility is here presented, namely, how sub-
jective conditions of thought can have objective validity, that is, can furnish
conditions of the possibility of all knowledge of objects. For appearances
can certainly be given in intuition independently of functions of the under-
standing. Let us take, for instance, the concept of cause, which signifies a
24 Apart from these sections, major changes have only been made in the introduction and
the section on the paralogisms of pure reason.

25 See Kant [1787b] (B), p.xxxvii.
special kind of synthesis, whereby upon something, A, there is posited some-

thing quite different, B, according to a rule. It is not manifest a priori why
appearances should contain anything of this kind [[. . .]]; and it is therefore
a priori doubtful whether such a concept be not perhaps altogether empty,
and have no object anywhere among appearances. That objects of sensible
intuition must conform to the formal conditions of sensibility which lie a
priori in the mind is evident, because otherwise they would not be objects
for us. But that they must likewise conform to the conditions which the
understanding requires for the synthetic unity of thought, is a conclusion
the grounds of which are by no means so obvious. Appearances might very
well be so constituted that the understanding should not find them to be
in accordance with the conditions of its unity.
(2) N. K. Smith [1929], p. 126; translating Kant [1781] (A), p. 93 (B 125 f).
All appearances necessarily agree with this formal condition of sensibility,
since only through it can they appear, that is, be empirically intuited and
given. The question now arises whether a priori concepts do not also serve
as antecedent conditions under which alone anything can be, if not intuited,
yet thought as object in general. In that case all empirical knowledge of
objects would necessarily conform to such concepts, because only as thus
presupposing them is anything possible as object of experience.
The transcendental deduction of all a priori concepts has thus a principle
according to which the whole enquiry must be directed, namely, that they
must be recognised as a priori conditions of the possibility of experience,
whether of the intuition which is to be met with in it or of the thought.
p. 323.
[[I]]n order to discover such a principle I looked about for an act of the under-
standing which comprises all the rest and is differentiated only by various
modifications or moments, in bringing the manifold of representation under
the unity of thinking in general. I found this act of the understanding to
consist in judging.
(2) N. K. Smith [1929], pp. 11 f; translating Kant [1781] (A), p. xvi.
I know no enquiries which are more important for exploring the faculty
which we entitle understanding, and for determining the rules and limits of
its employment, than those which I have instituted in the second chapter of
the Transcendental Analytic under the title Deduction of the Pure Concepts
of Understanding. They are also those which have cost me the greatest
labour[[.]]
pp. 129 f.
The manifold of representations can be given in an intuition which is purely
sensible, that is, nothing but receptivity; and the form of this intuition
can lie a priori in our faculty of representation, without being anything

more than the mode in which the subject is affected. But the combina-
tion (conjunctio) of a manifold in general can never come to us through
the senses, and cannot therefore, be already contained in the pure form of
sensible intuition. For it is an act of spontaneity of the faculty of repre-
sentation; and since this faculty, to distinguish it from sensibility, must be
entitled understanding, all combination—be we conscious of it or not, be it
a combination of the manifold of intuition, empirical or non-empirical, or
of various concepts—is an act of the understanding. To this act the general
title ‘synthesis’ may be assigned, as indicating that we cannot represent to
ourselves anything as combined in the object which we have not ourselves
previously combined, and that of all representations combination is the only
one which cannot be given through objects. Being an act of the self-activity
of the subject, it cannot be executed save by the subject itself.
In the case of the transcendental deduction of the categories, the principle of

the deduction is the so-called “transcendental unity of apperception” which is just
a principle of understanding.
pp. 134 f.
Combination does not [[. . .]] lie in the objects, and cannot be borrowed
from them [[. . .]]. On the contrary, it is an affair of the understanding alone,
which is nothing but the faculty of combining a priori, and of bringing the
manifold of representations under the unity of apperception. The princi-
ple of apperception is the highest principle in the whole sphere of human
knowledge.
This principle of the necessary unity of apperception is itself, indeed,
an identical, and therefore analytic, proposition; nevertheless it reveals the
necessity of a synthesis of the manifold given in intuition, without which
the thoroughgoing identity of self-consciousness cannot be thought.
Comment. Readers may find it worthwhile comparing Schelling’s formulation in

quotation 63.7 (3), last paragraph.
(2) N. K. Smith [1929], pp. 155 f; translating Kant [1787b] (B), p. 136.
The supreme principle of the possibility of all intuition in its relation
to sensibility is, according to the Transcendental Aesthetic, that all the
manifold of intuition should be subject to the formal conditions of space
and time. The supreme principle of the same possibility, in its relation to
understanding, is that all the manifold of intuition should be subject to
conditions of the original synthetic unity of apperception. In so far as the
manifold representations of intuition are given to us, they are subject to
the former of these two principles; in so far as they must allow of being
combined in one consciousness, they are subject to the latter.
(3) N. K. Smith [1929], pp. 192 f; translating Kant [1781] (A), pp. 155 ff (B 194 ff).
If knowledge is to have objective reality, that is, to relate to an ob-
ject, and is to acquire meaning and significance in respect to it, the object
must be capable of being in some manner given. Otherwise the concepts are
empty; through them we have indeed thought, but in this thinking we have
really known nothing; we have merely played with representations. That an
object be given (if this expression be taken, not as referring to some merely
mediate process, but as signifying immediate presentation in intuition),
means simply that the representation through which the object is thought
relates to actual or possible experience. Even space and time, however free
their concepts are from everything empirical, and however certain it is that
they are represented in the mind completely a priori, would yet be without
objective validity, senseless and meaningless, if their necessary application
to the objects of experience were not established. Their representation is
a mere schema which always stands in relation to the reproductive imag-
ination that calls up and assembles the objects of experience. Apart from
these objects of experience, they would be devoid of meaning. And so it is
with concepts of every kind.
The possibility of experience is, then, what gives objective reality to all
our a priori modes of knowledge. Experience, however, rests on the syn-
thetic unity of appearances, that is, on a synthesis according to concepts of
an object of appearances in general. Apart from such synthesis it would not
be knowledge, but a rhapsody of perceptions that would not fit into any
context according to rules of a completely interconnected (possible) con-
sciousness, and so would not conform to the transcendental and necessary
unity of apperception. Experience depends, therefore, upon a priori prin-
ciples of its form, that is, upon universal rules of unity in the synthesis of
appearances. Their objective reality, as necessary conditions of experience,
and indeed of its very possibility, can always be shown in experience. Apart
from this relation synthetic a priori principles are completely impossible.
For they have then no third something, that is, no object, in which the
synthetic unity can exhibit the objective reality of its concepts.
The transcendental unity of apperception is that unity through which
all the manifold given in an intuition is united in a concept of the object. It
is therefore entitled objective, and must be distinguished from the subjective
unity of consciousness, which is a determination of inner sense—through
which the manifold of intuition for such [objective] combination is empiri-
cally given.
It seems worthwhile emphasizing that Kant thought of these laws as being

invariable, i.e., as not being submitted to anything like change, development, or
improvement.
I close this section with a short quotation regarding the situation of the third
and last kind of synthetic apriorical concepts.
(B 393).
No objective deduction, such as we have been able to give of the categories,
is, strictly speaking, possible in the case of these transcendental ideas.
61f. Phenomena and noumena: Kant’s concept of the thing in itself.

In the last section I concentrated on how Kant dealt with the problem of explaining
the nature of necessity in the categories. This problem did not arise in the realm
of intuition, because, according to Kant, it is through intuition only that objects
appear to us. The situation is different, however, with regard to the categories. This
does not just create a problem with regard to foundation but offers the possibility
of forming a certain hypothetical concept of what is beyond the senses.
Quotation 61.31. N. K. Smith [1929], p. 266; translating Kant [1787b] (B),

pp. 305 f.
The categories are not, as regards their origin, grounded in sensibility, like
the forms of intuition, space and time; and they seem, therefore, to allow
of an application extending beyond all objects of the senses. As a matter
of fact they are nothing but forms of thought, which contain the merely
logical faculty of uniting a priori in one consciousness the manifold of a
given intuition[[.]]
Comment. Let me emphasize this: forms of thought, which contain a merely logical
faculty. This is particularly interesting in view of Hegel.26
p. 306.
[[I]]f we entitle certain objects, as appearances, sensible entities (phenom-
ena), then since we thus distinguish the mode in which we intuit them from
the nature that belongs to them in themselves, it is implied in this distinc-
tion that we place the latter [[. . .]] in opposition to the former, and that in
so doing we entitle them intelligible entities (noumena). The question then
arises, whether our pure concepts of understanding have meaning in respect
of these latter, and so can be a way of knowing them.
(2) N. K. Smith [1929], p. 269; translating Kant [1781] (A), p. 252
[[T]]he word appearance must be recognised as already indicating a relation
to something[[.]]
Comment. This may sound suspect, but not when taken on the background of
conceptual activity, i.e., not as referring.
This question is my concern in the present section. It will be of considerable

importance in later parts of my study. Here, I shall only be concerned with a pre-
sentation of Kant’s ideas of such a hypothetical “thing in itself”.
26 See section 65d in the next chapter.

Quotation 61.33. N. K. Smith [1929], p. 267; translating Kant [1787b] (B),

pp. 306 f.
At the very outset, however, we come upon an ambiguity which may
occasion serious misapprehension. The understanding, when it entitles an
object in a [certain] relation mere phenomenon, at the same time forms,
apart from that relation, a representation of an object in itself, and so comes
to represent itself as also being able to form concepts of such objects. And
since the understanding yields no concepts additional to the categories,
it also supposes that the object in itself must at least be thought through
these pure concepts, and so is misled into treating the entirely indeterminate
concept of an intelligible entity, namely, of a something in general outside
our sensibility, as being a determinate concept of an entity that allows
of being known in a certain [purely intelligible] manner by means of the
understanding.
Comment. The dawning of the dialectic of form and content (use and mention).
More carefully than in Berkeley.
Kant is, in some parts, at least, very careful regarding the concept of something
beyond the faculties of the mind. It is often tempting to treat such a concept as
referring to something. Observe, however, that Kant is quite adamant that it is only
negatively defined. Being negatively defined, it is important to take into account in
regard to what it is negatively defined. The understanding may form the concept of
something beyond the senses.
p. 307.
[[B]]y ‘noumenon’ we mean a thing so far as it is not an object of our sensible
intuition, and so abstract from our mode of intuiting it[[.]]
That, therefore, which we entitle ‘noumenon’ must be understood as being
such only in a negative sense.
Since Kant’s distinction between phenomena and noumena is prone to misun-

derstandings, I add some further quotations.
p. 307.
[[T]]his is a noumenon in the negative sense of the term. But if we understand
by it an object of a non-sensible intuition, we thereby presuppose a special
mode of intuition, namely, the intellectual, which is not that which we
possess, and of which we cannot comprehend even the possibility. This
would be ‘noumenon’ in the positive sense of the term.
If the objective reality of a concept cannot be in any way known, while

yet the concept contains no contradiction and also at the same time is con-
nected with other modes of knowledge that involve given concepts which it
serves to limit, I entitle that concept problematic. The concept of a noume-
non—that is, of a thing which is not to be thought as object of the senses
but as a thing in itself, solely through a pure understanding—is not in any
way contradictory. For we cannot assert of sensibility that it is the sole pos-
sible kind of intuition. Further, the concept of a noumenon is necessary, to
prevent sensible intuition from being extended to things in themselves, and
thus to limit the objective validity of sensible knowledge. The remaining
things, to which it does not apply, are entitled noumena, in order to show
that this knowledge cannot extend its domain over everything which the
understanding thinks. But none the less we are unable to comprehend how
such noumena can be possible, and the domain that lies out beyond the
sphere of appearances is for us empty. That is to say, we have an under-
standing which problematically extends further, but we have no intuition,
indeed not even the concept of possible intuition, through which objects
outside the field of sensibility can be given, and through which the under-
standing can be employed assertorically beyond that field. The concept of
a noumenon is thus a merely limiting concept, the function of which is to
curb the pretensions of sensibility; and it is therefore only of negative em-
ployment. At the same time it is no arbitrary invention; it is bound up with
the limitation of sensibility, though it cannot affirm anything beyond the
field of sensibility.
The division of objects into phenomena and noumena, and the world
into a world of the senses and a world of the understanding, is therefore
quite inadmissible in the positive sense, although the distinction of concepts
as sensible and intellectual is certainly legitimate.
(3) N. K. Smith [1929], pp. 268 f; translating Kant [1787b] (B), pp. 307 f.
The doctrine of sensibility is likewise the doctrine of the noumenon in

the negative sense, that is, of things which the understanding must think
without this reference to our mode of intuition, therefore not merely as ap-
pearances but as things in themselves. At the same time the understanding
is well aware that in viewing things in this manner, as thus apart from our
mode of intuition, it cannot make any use of the categories. For the cate-
gories have meaning only in relation to the unity of intuition in space and
time[[.]] [[. . .]] In cases where this unity of time is not to be found and there-
fore in the case of the noumenon, all employment, and indeed the whole
meaning of the categories, entirely vanishes[[.]]
Comment. Note the specific claim that it is the categories which can’t be applied
to things regarded as being in themselves.
What our understanding acquires through this concept of a noumenon, is
a negative extension; that is to say, understanding is not limited through
sensibility by applying the term noumena to things in themselves (things
not regarded as appearances). But in doing so it at the same time sets limits
to itself, recognising that it cannot know these noumena through any of the
categories, and that it must therefore think them only under the title of an
unknown something.
(5) Carus and Ellington [1977], p. 82; translating Kant [1783], pp. 341 f.
When I speak of objects in time and in space, it is not of things in
themselves, of which I know nothing, but of things in appearance, i.e., of
experience, as a particular way of cognizing objects which is only afforded to
man. I must not say of that which I think in time or in space, that in itself,
and independent of these my thoughts, it exists in space and in time, for in
that case I should contradict myself; because space and time, together with
the appearances in them, are nothing existing in themselves and outside
of my representations, but are themselves only modes of representation,
and it is palpably contradictory to say that a mere mode of representation
exists outside our representation. Objects of the senses therefore exist only
in experience, whereas to give them a self-subsisting existence apart from
experience or prior to it is merely to represent to ourselves that experience
actually exists apart from experience or prior to it.
Comment. Note the similarity to Berkeley’s formulation in [1713], p. 245.27 The
whole thing is of decisive importance for the interpretation of the Antinomy of
Pure Reason.
The concept of the noumenon is [[. . .]] not the concept of an object, but is a
problem unavoidably bound up with the limitation of our sensibility—the
problem, namely, as to whether there may not be objects entirely disengaged
from any such kind of intuition. [[. . .]]
Understanding accordingly limits sensibility, but does not thereby ex-
tend its own sphere. In the process of warning the latter that it must not
presume to claim applicability to things-in-themselves but only to appear-
ances, it does indeed think for itself an object in itself, but only as tran-
scendental object, which is the cause of appearance and therefore not itself
appearance[[.]] [[. . .]] If we are pleased to name this object noumenon for the
reason that its representation is not sensible, we are free to do so. But since
we can apply to it none of the concepts of our understanding, the repre-
sentation remains for us empty, and is of no service except to mark the
limits of our sensible knowledge and to leave open a space which we can fill
neither through possible experience nor through pure understanding.
27 See quotation 58.13 (1) in these materials. Cf. also variations 106.5 in the groundworks.
Comment. The reader who puzzles over the beginning of this quotation, the “con-
cept of the noumenon is . . . not the concept of an object”, in relation to the phrasing
in quotation 2 above, the “distinction of objects into phenomena and noumena” may
find it helpful to know that the distinction that Kant has made in German between
“Objekt” in the first case and “Gegenstände” in the second is not preserved in the
English translation.
[[W]]e indeed, rightly considering objects of sense as mere appearances, con-
fess thereby that they are based upon a thing in itself, though we know
not this thing as it is in itself but only know its appearances, viz., the way
in which our senses are affected by this unknown something. The under-
standing therefore, by assuming appearances, grants also the existence of
things in themselves, and thus far we may say that the representation of
such things as are the basis of appearances, consequently of mere beings of
the understanding, is not only admissible but unavoidable.
Our critical deduction by no means excludes things of that sort (nou-
mena), but rather limits the principles of the Aesthetic in such a way that
they shall not extend to all things (as everything would then be turned into
mere appearance) but that they shall hold good only of objects of possible
experience. Hereby, then, beings of the understanding are admitted, but
with the inculcation of this rule which admits of no exception: that we nei-
ther know nor can know anything determinate whatever about these pure
beings of the understanding, because our pure concepts of the understand-
ing as well as our pure intuitions extend to nothing but objects of possible
experience, consequently to mere things of sense; and as soon as we leave
this sphere, these concepts retain no meaning whatever.
(8) Kant, Opus postumum, II, 26; quoted after Werkmeister [1981], p. 304.
[[T]]he distinction of the conceptions of a thing in itself and of one in ap-
pearance is not objective but subjective only.
(9) U. Petersen [2002], p. 813; translating Kant [1783], pp. 312 f.
To make an attempt at Hume’s problematic concept (this [[being]] his
crux metaphysicorum), namely the concept of cause, what is given to me a
priori is firstly by means of the logic the form of a conditional judgment
in general, namely to employ a given knowledge as basis and the other as
consequence. It is possible, however, that a rule of relation is met with in
perception, which says: that one particular phenomenon is followed con-
stantly by another one (although not vice versa); and this is a case where I
employ the hypothetical judgment and say, for instance: if a body is sunlit
for long enough, then it will get warm. Admittedly, here is not yet a neces-
sity of connection, hence [[not yet]] the concept of cause. However, I continue
and say: if the statement above, which is merely a subjective connection of
perceptions, is to be a statement of experience, then it must be looked upon
as necessary and universally valid. Such a statement, however, would be:
sun by virtue of its light is the cause of warmth. The above empirical rule is
now being looked upon as a law, and that not as holding merely of appear-
ances, but of them for the purpose of a possible experience which requires
general and thus necessarily valid rules. Thus I well see the concept of a
cause as a concept necessarily belonging only to the mere form of experience
and its possibility, as a synthetic bringing together of perceptions in one
consciousness in general; the possibility of a thing in general, however, as a
cause I do not see at all, and that because the concept of a cause does in no
way indicate a condition adhering to things, but only to experience, namely,
that these could only be an objectively valid knowledge of appearances and
their sequence in time, so far as the previous can be connected with the
subsequent according to the rule of hypothetic judgments.
§ 30
Therefore, the pure concepts of understanding have no meaning at all, if
they leave objects of experience and want to be related to things in them-
selves (noumena). They only serve, as it were, to spell appearances, so they
can be read as experiences; the principles which arise from their relation
to the world of senses, only serve our understanding for use in experience;
further to that they are arbitrary connections without objective reality, the
possibility of which one can neither know a priori nor confirm their rela-
tion to objects by any example, or only make it comprehensible, because all
examples are borrowed from some possible experience, hence the objects of
those concepts too can be met nowhere else but in a possible experience.
I close this chapter with a set of quotations regarding the notion of the tran-
scendental object, a notion which I feel bears some strong similarity to that of the
noumenon and which also provides a clue for understanding Kant’s critical decision
of the Antinomy of Pure Reason.
The first three of the following set of quotations regarding the transcendental
object are taken from the chapter on the “Paralogisms of Pure Reason” in the first
edition. The chapter has been largely rewritten in the second edition, apparently
without any mention of the transcendental object. These first three quotations are
included here to give an impression of what Kant excluded in the second edition.
Quotation 61.36. (1) N. K. Smith [1929], p. 339; translating Kant [1781] (A),
p. 358.
[[T]]he something which underlies the outer appearances and which so affects
our sense that it obtains the representations of space, matter, shape, etc.,
may yet, when viewed as noumenon (or better, as transcendental object),
be at the same time the subject of our thoughts.
[[I]]f we regard outer appearances as representations produced in us by their
objects, and if these objects be things existing in themselves outside us, it
is indeed impossible to see how we can come to know the existence of the
objects otherwise that by inference from the effect to the cause; and this
being so, it must always remain doubtful whether the cause in question be
in us or outside us. We can indeed admit that something, which may be (in
the transcendental sense) outside us, is the cause of our outer intuitions,
but this is not the object of which we are thinking in the representations
of matter and of corporeal things; for these are merely appearances, that
is, mere kinds of representation, which are never to be met with save in
us, and the reality of which depends on immediate consciousness, just as
does the consciousness of my own thoughts. The transcendental object is
equally unknown in respect to inner and to outer intuition. But is not of
this that we are here speaking, but of the empirical object, which is called
an external object if it is represented in space, and an inner object if it is
represented only in its time-relations. Neither space nor time, however, is
to be found save in us.
The expression ‘outside us’ is thus unavoidably ambiguous in meaning,
sometimes signifying what as thing in itself exists apart from us, and some-
times what belongs solely to outer appearance. In order, therefore, to make
this concept, in the latter sense—in the sense in which the psychological
question as to the reality of our outer intuition has to be understood—quite
unambiguous, we shall distinguish empirically external objects from those
which may be said to be external in the transcendental sense, by explicitly
entitling the former ‘things which are to be found in space’[[.]]
(3) N. K. Smith [1929], p. 352; translating Kant [1781] (A), pp. 379 f.
Neither the transcendental object which underlies outer appearances nor
that which underlies inner intuition, is in itself either matter or a thinking
being, but a ground (to us unknown) of the appearances which supply to
us the empirical concept of the former as well as of the latter mode of
existence.
Comment. Notice that this passage is only to be found in the first edition.
Whatever in an object of the senses is not itself appearance, I entitle
intelligible. If, therefore, that which in the sensible world must be regarded
as appearance has in itself a faculty which is not an object of sensible in-
tuition, but through which it can be the cause of appearances, the causality
of this being can be regarded from two points of view. Regarded as the
causality of a thing in itself, it is intelligible in its action; regarded as the
causality of an appearance in the world of sense, it is sensible in its effects.
We should therefore have to form both an empirical and an intellectual
concept of the causality of the faculty of such a subject, and to regard both
as referring to one and the same effect, This twofold manner of conceiving
the faculty possessed by an object of the senses does not contradict any of
the concepts which we have to form of appearances and of a possible expe-
rience. For since they are not things in themselves, they must rest upon a
transcendental object which determines them as mere representations; and
consequently there is nothing to prevent us from ascribing to this transcen-

dental object, besides the quality in terms of which it appears, a causality
which is not appearance, although its effect is to be met with in appearance.
The faculty of sensible intuition is strictly only a receptivity, a capacity
of being affected in a certain manner with representations, the relation of
which to one another is a pure intuition of space and of time (mere forms of
our sensibility), and which, in so far as they are connected in this manner
in space and time, and are determinable according to laws of the unity of
experience, are entitled objects. The non-sensible cause of these representa-
tions is completely unknown to us, and cannot therefore be intuited by us as
object. For such an object would have to represented as neither in space nor
in time (these being merely conditions of sensible representation), and apart
form such conditions we cannot think any intuition. We may, however, enti-
tle the purely intelligible cause of appearances in general the transcendental
object, but merely in order to have something corresponding to sensibility
viewed as a receptivity. To this transcendental object we can ascribe the
whole extent and connection of our possible perceptions, and can say that
it is given in itself prior to all experience. But the appearances, while con-
forming to it, are not given in themselves, but only in this experience, being
mere representations, which as perceptions can mark out a real object only
in so far as the perception connects with all others according to the rules of
the unity of experience. Thus we can say that the real things of past time
are given in the transcendental object of experience; but they are objects
for me and real in past time only in so far as I represent to myself (either
by the light of history or by the guiding clues of causes and effects) that a
regressive series of possible perceptions in accordance with empirical laws,
in a word, that the course of the world, conducts us to a past time-series as
condition of the present time—a series which, however, can be represented
as actual not in itself but only in the connection of a possible experience.
Many forces in nature, which manifest their existence through certain
effects, remain for us inscrutable; for we cannot track them sufficiently
far by observation. Also, the transcendental object lying at the basis of
appearances (and with it the reason why our sensibility is subject to certain
supreme conditions rather than to others) is and remains for us inscrutable.
The thing itself [[die Sache selbst]] is indeed given, but we can have no insight
into its nature. But it is quite otherwise with an ideal of pure reason; it
can never be said to be inscrutable. For since it is not required to give
any credentials of its reality save only the need on the part of reason to
complete all synthetic unity by means of it; and since, therefore, it is in
no wise given as thinkable object, it cannot be inscrutable in the manner
in which an object is. On the contrary it must, as a mere idea, find its
place and its solution in the nature of reason, and must therefore allow of
investigation. For it is of the very essence of reason that we should be able
§ 62. THE ANTINOMY OF PURE REASON 817
to give an account of all our concepts, opinions, and assertions, either upon
objective or, in the case of mere illusion upon subjective grounds.
Comment. The last sentence reasserts a kind of self-transparency already mentioned

in quotation 60.2 (1).
If it be the transcendental object of our idea that we have in view, it is
obvious that we cannot thus, in terms of the concepts of reality, substance,
causality, etc., presuppose its reality in itself, since these concepts have not
the least application to anything that is entirely distinct from the world of
sense.
§ 62. The Antinomy of Pure Reason
This paragraph is devoted to KT8, the Antinomy of Pure Reason, KT9, the critical
decision regarding the Antinomy of Pure Reason, and KT10, the resolution of the
ideas of reason.
The Antinomy of Pure Reason lies at the heart of Kant’s transcendental philos-
ophy. It is for this antinomy that transcendental idealism provides the key, as Kant
put it in the headline to section 6 of the chapter on the Antinomy of Pure Reason.
As regards KT8, I shall distinguish the following two features in Kant’s treat-
ment:
(i) the system of possible antinomies;
(ii) the logical structure of the antinomies.
I begin with the first one.
62a. The system of possible antinomies. In view of the table of the cate-
gories, there are, according to Kant, only four antinomies possible.
Quotation 62.1. N. K. Smith [1929], pp. 422 f; translating Kant [1781] (A), p. 463
(B 491).
Whether the world has a beginning [in time] and any limit to its extension
in space; whether there is anywhere, and perhaps in my thinking self, an
indivisible and indestructible unity, or nothing but what is divisible and
transitory; whether I am free in my actions or, like other beings, am led
by the hand of nature and of fate; whether finally there is a supreme cause
of the world, or whether the things of nature and their order must as the
ultimate object terminate thought—an object that even in our speculations
can never be transcended[[.]]
These particular questions have their origin in the table of categories:

(B 442).
When we thus select out those categories which necessarily lead to a
series in the synthesis of the manifold, we find that there are but four
cosmological ideas, corresponding to the four titles of the categories:
1. Absolute completeness
of the Composition
of the given whole of all appearances.
in the Division
of a given whole in the [field of] appearance.
in the Origination
of an appearance.
as regards Dependence of Existence
of the alterable in the [field of] appearance.
Comment. Note the mathematical/logical character coming into play: concept of

infinity, completeness of a series, absoluteness.
Quotation 62.3. N. K. Smith [1929], pp. 391 f; translating Kant [1781] (B),
pp. 417 f.
This unconditioned may be conceived in either of two ways. It may
be viewed as consisting of the entire series in which all the members with-
out exception are conditioned and only the totality of them is absolutely
unconditioned. This regress is to be entitled infinite. Or alternatively, the
absolutely unconditioned is only a part of the series—a part to which the
other members are subordinated, and which does not itself stand under any
other condition. On the first view, the series a parte priori is without lim-
its or beginning, i.e. is infinite, and the same time is given in its entirety.
But the regress in it is never completed, and can only be called potentially
infinite. On the second view, there is a first member of the series which
in respect of past time is entitled, the beginning of the world, in respect of
space, the limit of the world, in respect of the parts of a given limited whole,
the simple, in respect of causes, absolute self-activity (freedom), in respect
of the existence of alterable things, absolute natural necessity.
(B 490).
We have now completely before us the dialectic play of cosmological
ideas. The ideas are such that an object congruent with them can never
be given in any possible experience, and that even in thought reason is
unable to bring them into harmony with the universal laws of nature. Yet
they are not arbitrarily conceived. Reason, in the continuous advance of
empirical synthesis, is necessarily led upon them whenever it endeavours to
free from all conditions and apprehend in its unconditioned totality that
which according to the rules of experience can never be determined save
as conditioned. These pseudo-rational assertions are so many attempts to
solve four natural and unavoidable problems of reason. There are just so
many, neither more nor fewer, owing to the fact that there are just four
series of synthetic presuppositions which impose a priori limitations on the
empirical synthesis.
Comment. Important points: number of antinomies, being determined by the four
titles of the categories.28
62b. The logical structure of the antinomies. As regards the general

logical frame of the Antinomy of Pure Reason, it will be clear that Kant adheres
to classical logic and, therefore, cannot accept a provable contradiction of the form
A∧¬A. This, then, sets the first problem: how can, of two contradictory propositions,
both be false? Kant discusses this question at two occasions, in [1781] (A), pp. 504 f
(B 531 f) and in [1783], pp. 341 f, concentrating on different aspects. The strategy
put forward in the Critique of Pure Reason runs as follows:
Quotation 62.5. N. K. Smith [1929], p. 446; translating Kant [1781] (A), pp. 502 f
(B 532).
If two opposed judgments presuppose an inadmissible condition, then in
spite of their opposition, which does not amount to a contradiction strictly
so-called, both fall to the ground, inasmuch as the condition, under which
alone either of them can be maintained, itself falls.
Comment. I shall formulate in a formal fashion what I take to be the gist of Kant’s
reasoning in §104.
Now Kant introduces a distinction between dialectical and analytical opposi-

tions. He reasons as follows:
Quotation 62.6. N. K. Smith [1929], p. 447; translating Kant [1787b] (B), p. 532.
[[O]]f two dialectically opposed judgments both may be false; for the one
is not a mere contradictory of the other, but says something more than is
required for a simple contradiction.
Comment. As for quotation 62.5.
28 Cf. quotation 62.2 above.
The point now is to detect such an inadmissible condition. This is the topic of
the next section. Before I conclude this section, however, I quote a bit from Kant’s
considerations in the Prolegomena, which, at a cursory glance, at least, may appear
to be quite different from the one just discussed.
p. 341.
Contradictory propositions cannot both be false, unless the concept lying
at the ground of both of them is self-contradictory; for example, the propo-
sitions, “A square circle is round,” and “A square circle is not round,” are
both false. For, as to the former, it is false that the circle is round be-
cause it is quadrangular; and it is likewise false that it is not round, i.e.,
angular, because it is a circle. For the logical criterion of the impossibility
of a concept consists in this, that if we presuppose it, two contradictory
propositions both become false[[.]]
Comment. Readers who have gone through the school of Frege or in general some
modern logic will probably be able to make more sense of this quotation if they
replace “concept” by “description”. The next quotation indicates that Kant did see
the need for a wider notion of concept (“problematically”).
The supreme concept with which it is customary to begin a transcendental
philosophy is the division into the possible and the impossible. But since all
division presupposes a concept to be divided, a still higher one is required,
and this is the concept of an object in general, taken problematically, with-
out its having been decided whether it is something or nothing.
Comment. Cf. Frege [1884], p. 105:29 concepts, unlike description, need not subsume
anything under them.
62c. The critical decision regarding the conflict of reason with itself.
I shall now concentrate on the inadmissible condition. This is one of the most
important issues in the whole Critique!of Pure Reason.30
Remark 62.8. N. K. Smith translates “Entscheidung” in the title of section 7 as

“solution”. In view of the subtitle of section 9, “Auflösung der kosmologischen Idee”
(in different respects) which N. K. Smith translates as “solution” too, I prefer to
use the word “decision” to translate “Entscheidung” since it differs from what Kant
deals with in the “Auflösung” which I refer to as “resolution”.
Quotation 62.9. N. K. Smith [1929], p. 434; translating Kant [1781], p. 481 f.

If from our own concepts we are unable to assert and determine anything
certain, we must not throw the blame upon the object as concealing itself
from us. Since such an object is nowhere to be met with outside our idea,
it is not possible for it to be given. The cause of failure we must seek in
29 See quotation 71.16 (3) in these materials.
30 Cf. also section 68b, in particular quotation 68.24 (2).
our idea itself. For so long as we obstinately persist in assuming that there
is an actual object corresponding to the idea, the problem, as thus viewed,
allows of no solution. A clear exposition of the dialectic which lies within
our concept itself would soon yield us complete certainty how we ought to
judge in reference to such a question.
Comment. Note the claim that it is not possible for such an object to be given.
(B 525).
The whole antinomy of pure reason rests upon the dialectical argument:
If the conditioned is given, the entire series of all its conditions is likewise
given; objects of the senses are given as conditioned; therefore, etc.
Kant calls this a “pseudo-rational argument” (ibid.) and sets out to show why.
Since this is a fairly crucial part, I shall follow him closely.
pp. 497 f (B 526).
In the first place, [[. . .]] if the conditioned is given, a regress in the series
of all conditions is set us as a task. For it is involved in the very concept
of the conditioned that something is referred to a condition, and if this
condition is again itself conditioned, to a more remote condition, and so
through all the members of the series. [[. . .]]
Further, if the conditioned as well as its condition are things in them-
selves, then upon the former being given, the regress to the latter is not only
set as a task, but therewith already really given. And since this holds of all
members of the series, the complete series of the conditions, and therefore
the unconditioned, is given therewith[[.]] [[. . .]] If, however, what we are deal-
ing with are appearances—as mere representations appearances cannot be
given save in so far as I attain knowledge of them, or rather attain them
in themselves, for they are nothing but empirical modes of knowledge—I
cannot say, in the same sense of the terms, that if the conditioned is given,
all its conditions (as appearances) are likewise given, and therefore cannot
in any way infer the totality of the series of its conditions. The appearances
are in their apprehension themselves nothing but an empirical synthesis in
space and time, and are given only in this synthesis. It does not, therefore,
follow, that if the conditioned in the [field of] appearance, is given, the
synthesis which constitutes its empirical condition is given therewith and
is presupposed. This synthesis first occurs in the regress, and never exists
without it.
(2) Carus and Ellington [1977], pp. 82 f; translating Kant [1783], p. 342.
If the objects of the world of sense are taken for things in themselves
and the [[. . .]] laws of nature for the laws of things in themselves, the con-
tradiction would be unavoidable.
Quotation 62.12. N. K. Smith [1929], pp. 447 f; translating Kant [1781] (A),
pp. 504 f (B 532 f).
If we regard the two propositions, that the world is infinite in magni-
tude and that it is finite in magnitude, as contradictory opposites, we are
assuming that the world, the complete series of appearances, is a thing in
itself that remains even if I suspend the infinite or the finite regress in the
series of its appearances. If, however, I reject this assumption, or rather this
accompanying transcendental illusion and deny that the world is a thing in
itself, the contradictory opposition of the two assertions is converted into
a merely dialectical opposition. Since the world does not exist in itself, in-
dependently of the regressive series of my representations, it exists in itself
neither as an infinite whole nor as a finite whole.
Kant is a bit clearer, at least in some way, about the kind of independence: the
world does not exist “independently of the regressive series of my representations”.
p. 505 (B 533).
The series of conditions is only to be met with in the regressive synthesis
itself, not in the [field of] appearance viewed as a thing given in and by itself,
prior to all regress. [[. . .]] [[A]]n appearance is not something existing in itself,
and its parts are first given in and through the regress of the decomposing
synthesis, a regress which is never given in absolute completeness[[.]]
Comment. Note the focus on existence as a criterion for meaningfulness.
Thus the antinomy of pure reason in its cosmological ideas vanishes
when it is shown that it is merely dialectical, and that it is a conflict due
to an illusion which arises from our applying to appearances that exist
only in our representations, and therefore, so far as they form a series, not
otherwise than in a successive regress, that idea of absolute totality which
holds only as a condition of things in themselves.
Comment. Let me summarize: The antinomy arises from our applying to appear-
ances an idea of absolute totality which only holds of things in themselves, i.e., an
idea which can only be applied to things insofar they are given independent of our
form of representation.
pp. 506 f (B 534 f).
If the world is a whole existing in itself, it is either finite or infinite. But
both alternatives are false (as shown in the proofs of the antithesis and
thesis respectively). It is therefore also false that the world (the sum of all
appearances) is a whole existing in itself. From this it then follows that
appearances in general are nothing outside our representations—which is
just what is meant by their transcendental ideality.
Comment. As regards the argument of the first part this quotation, cf. section 104d
in the groundworks, in particular p. 1436.
All beginning is in time and all limits of the extended are in space.
But space and time belong only to the world of sense. Accordingly, while
appearances in the world are conditionally limited, the world itself is neither
conditionally nor unconditionally limited.
Comment. The problem in all this is, why should the world be a thing in itself,
in order to be finite or infinite? If this is not clear, read again through quotations
62.11–62.13 above or look up §104 where I shall give my own reading.
p. 342.
Now if I ask about the magnitude of the world, as to space and time, it
is equally impossible, with regard to all my concepts, to declare it infinite
or to declare it finite. For neither assertion can be contained in experience,
because experience either of an infinite space or of an infinite time elapsed,
or again, of the boundary of the world by an empty space or by an an-
tecedent empty time, is impossible; these are mere Ideas. This magnitude
of the world, be it determined in either way, would therefore have to exist in
the world itself apart from all experience. But this contradicts the concept
of a world of sense, which is merely a complex of the appearances whose ex-
istence and connection occur only in our representations, i.e., in experience;
since this latter is not an object in itself but a mere mode of representation.
Hence it follows that, as the concept of an absolutely existing world of sense
is self-contradictory, the solution of the problem concerning its magnitude,
whether attempted affirmatively or negatively, is always false.
Comment. Cf. section 104d in the groundworks regarding the argumentation in
this quotation.
To assume that an appearance, e.g., that of a body, contains in itself before
all experience all the parts which any possible experience can ever reach
is to impute to a mere appearance, which can only exist in experience, an
existence previous to experience. In other words, it would mean that mere
representations exist before they can be found in our faculty of representa-
tion. Such an assertion is self-contradictory[[.]]
pp. 508 f (B 536 f).
The principle of reason is thus properly only a rule, prescribing a regress
in the series of the conditions of given appearances, and forbidding it to
bring the regress to a close by treating anything at which it may arrive as
absolutely unconditioned. It is not a principle of the possibility of experi-
ence and of empirical knowledge of objects of the sense, and therefore not
a principle of the possibility of experience and of empirical knowledge of

objects of the senses, and therefore not a principle of the understanding; for
every experience, in conformity with the given [forms of] intuition, is en-
closed within limits. Nor is it a constitutive principle of reason, enabling us
to extend our concept of the sensible world beyond all possible experience.
It is rather a principle of the greatest possible continuation and extension
of experience, allowing no empirical limit to hold as absolute. Thus it is a
principle of reason which serves as a rule, postulating what we ought to do
in the regress, but not anticipating what is present in the object as it is in
itself, prior to all regress. Accordingly I entitle it a regulative principle of
reason to distinguish it from the principle of the absolute totality of the
series of conditions, viewed as actually present in the object (that is, in the
appearances), which would be a constitutive cosmological principle.
The transcendental ideas [[. . .]] express the peculiar application of reason
as a principle of systematic unity in the use of the understanding. Yet if
we assume this unity of the mode of cognition to pertain to the object of
cognition, if we retard that which merely regulative to be constitutive, and
if we persuade ourselves that we can by means of these ideas enlarge our
cognition transcendently or far beyond all possible experience, while it only
serves to render experience within itself as nearly complete as possible, i.e.,
to limit its progress by nothing that cannot belong to experience—if we do
all this, then we suffer from a mere misunderstanding in our estimate of
the proper application of our reason and of its principles and suffer from a
dialectic which confuses the empirical use of reason and also sets reason at
variance with itself.
Remark 62.17. According to N. K. Smith [1918], p. 506, the general result of the
antinomy of pure reason, the refutation of the existence of the series of conditions
as a whole, is of earlier origin, than the more distinguished treatment in section 9
where different specific solutions for the mathematical and the dynamical ideas are
proposed. At the same time, however, he shows a tendency to run the two together
as the solution of the antinomies.31 It seems worthwhile, therefore, to stress that
Kant is quite careful in his language; this is why I translate him as speaking of the
decision of the conflict and the resolution of the ideas.
The last quotation in this section is meant to confirm that Kant clearly saw
the negative result as applying to all of the cosmological antinomies, not just the
mathematical ones. This is quite important in view of a resolution which only the
dynamical antinomies allow and which I shall discuss in the next section.
31 In particular in the subtitles to section ix, on p. 508 (“solution of the first and second anti-
nomies”), p. 512 (“solution of the third antinomy”), and p. 518 (“solution of the fourth antinomy”).
In his commentary, N. K. Smith still renders “Entscheidung” as “decision”, but no longer in his
translation of the Critique itself, where he renders it as “solution”.
(B 533).
What we have here said of the first cosmological idea, that is, of the
absolute totality of magnitude in the [field of] appearance, applies also to
all the others.
Thus for the negative result of the antinomy of pure reason which is positive
in the sense of transcendental idealism: it confirms the transcendental ideality of
appearances.32 This, however, is not the whole story; it is not yet a resolution of
the ideas in the sense that one would know what to do with them; having found an
inadmissible condition in the antinomies does not invalidate the ideas of reason as
such. Now Kant found a difference between what he called the dynamical and the
mathematical antinomies. This is what the third feature is about.
62d. The different resolutions of the mathematical-transcendental
ideas and the dynamical ones. I begin with the resolution of the mathematical-
transcendental idea which is short and simply negative, as it were.
p. 518 (B 546).
For the solution [[. . .]] of the first cosmological problem we have only
to decide whether in the regress to the unconditioned magnitude of the
universe, in time and space, this never limited ascent can be called a regress
to infinity, or only an indeterminately continued regress (in indefinitum)
Comment. N. K. Smith conveniently neglected in his translation the little German
word “noch”, meaning somewhat “further”: for the resolution nothing more is nec-
essary than to further settle . . . . This wouldn’t fit with N. K. Smith’s line that the
critical decision is already the solution. Note also that Kant speaks of “Auflösung
der ersten kosmologischen Aufgabe”, i.e., “resolution of the first cosmological task”;
I suggest that this be seen in connection with quotation 62.11 (1) above: “a regress
in the series of all conditions is set us as a task ” (“aufgegeben”).
(2) N. K. Smith [1929], p. 456; translating Kant [1781] (A), p. 519 f (B 547 f).
Since the world is not given me, in its totality, through any intuition,
neither is its magnitude given me prior to the regress. We cannot, therefore,
say anything at all in regard to the magnitude of the world, not even that
there is in it a regress in infinitum. All that we can do is to seek for the con-
cept of its magnitude according to the rule which determines the empirical
regress in it. This rule says no more than that, however far we may have
attained in the series of empirical conditions, we should never assume an
absolute limit, but should subordinate every appearance, as conditioned, to
another as its condition, and that we must advance to this condition. This
is the regressus in indefinitum, which, as it determines no magnitude in the
object, is clearly enough distinguishable from the regressus in infinitum.
32 Cf. Bernays [1913] and Kowalewsky [1918] on the antinomies as a foundation of transcen-
dental idealism.
Comment. I find it tempting to try and construct an analogy to the situation of

the undecidability of classical first order logic: there is no general method to de-
cide of any given sequent whether or not it has a terminating LK-reduction (see
lemma 36.33 in conjunction with 48.28 in the tools). Cf. also Wang in quota-
tion 85.21 and Hilbert and Ackermann in quotation 79.8 in these materials.
In a Concluding Note on the Solution of the Mathematical-transcendental Ideas,

and Preliminary Observation on the Solution of the Dynamical-transcendental Ideas
Kant works out a difference between the mathematical and dynamical antinomies
which allows a different resolution of the cosmological ideas concerning causality
and necessity.
p. 529 (B 557).
[[I]]n all this we have been overlooking an essential distinction that obtains
among the objects, that is, among those concepts of understanding which
reason endeavours to raise to ideas.
[[I]]n the mathematical connection of the series of appearances no other than
a sensible condition is admissible, that is to say, none that is not itself a
part of the series. On the other hand, in the dynamical series of sensible
conditions, a heterogeneous condition, not itself a part of the series, but
purely intelligible, and as such outside the series, can be allowed.
The dialectical arguments, which in one or other way sought unconditioned
totality in mere appearances, fall to the ground, and the propositions of
reason, [[in the thus corrected]] interpretation, may both alike be true.
Comment. N. K. Smith translated: “when thus given the more correct interpreta-
tion”. This may suggest that this positive interpretation is more correct than the
negative one; a suggestion which, however, is not contained in the German original:
“in der auf solche Weise berichtigten”.
Remark 62.21. It seems quite important to emphasize the point that the neg-
ative result applies, as a “result” (German: “Erfolg”) to all antinomies (cf. quota-
tion 62.18).33 In his translation, N. K. Smith followed the reading of Hartenstein who
suggested “mathematische Antinomie” for “Antinomie” and thus created a wrong
impression of Kant’s statement. The section in question is the following:
Dadurch nun, daß die dynamischen Ideen eine Bedingung der Erschei-
nungen außer der Reihe derselben, d. i. eine solche, die selbst nicht Er-
scheinung ist, zulassen, geschieht etwas, was von dem Erfolg der Antinomie
gänzlich unterschieden ist. (Kant [1781] (A), p. 531 (B 559).)
33 “Erfolg” literally translated would be “success”. The meaning in the present context, how-
ever, is more like that of consequence, “Folge” which is translated “result” in Carus and Ellington
[1977], p. 88.
Inasmuch as the dynamical ideas allow of a condition of appearances

outside the series of the appearances, that is, a condition which is not
itself appearance, we arrive at a conclusion altogether different from any
that was possible in the case of the mathematical antinomy. (N. K. Smith
[1929], p. 463)
My translation of this section:
Because of this now, that the dynamical ideas allow of a condition of
appearances outside the series of the appearances, that is, a condition which
is not itself appearance, something happens that is completely different from
the result of the antinomy.
My reading of this passage: the antinomy has a result, namely to “curb the preten-
sions” of thought in a way similar to how the concept of a notion of the thing in
itself curbs the pretensions of intuition. This result is not a resolution in the sense of
telling us what to do with the ideas. The resolution with regard to the mathematical
ideas is pretty bleak, i.e., there isn’t much to do, except to determine whether the
series goes on in infinitum or just in indefinitum.34 The series which are subject to
the dynamical ideas, however, allow something that is different from the result of
the antinomy altogether, not just the mathematical antinomy, viz. an object outside
of the series which essentially changes one of the grammatical subjects of the two
conflicting statements.
The passage in question continues as follows:
Diese nämlich verursachte, daß beide dialektischen Gegenbehauptungen für
falsch erklärt werden mußten. Dagegen das Durchgängigbedingte der dyna-
mischen Reihen, welches von ihnen als Erscheinungen unzertrennlich ist, mit
der zwar empirischunbedingten, aber auch n i c h t s i n n l i c h e n Bedingung
verknüpft, dem V e r s t a n d e einerseits und der V e r n u n f t andererseits
Genüge leisten[[.]] (Kant [1781] (A), p. 531 (B 559).)
which is translated as
In it we were obliged to denounce both the opposed dialectical assertions
as false. In the dynamical series, on the other hand, the completely con-
ditioned, which is inseparable from the series considered as appearances is
bound up with a condition which, while indeed empirically unconditioned, is
also non-sensible. We are thus able to obtain satisfaction for understanding
on the one hand and for reason on the other. (N. K. Smith [1929], p. 463)
I should like to suggest the following translation (which is grammatically clumsy,
to say the least, but matches the original German, which isn’t particularly elegant
either):
This namely caused that both the opposed dialectical assertions had to be
declared false. Against that, the completely conditioned of the dynamical
series, which is inseparable from them as appearances, tied to the, actually
empirically unconditioned, but also non-sensible condition, satisfies the un-
derstanding on the one hand and reason on the other.
34 Cf. quotations 62.19 above.
In other words, there is an option to bind up the unconditioned of the dynamical

series with an empirically unconditioned not belonging to the field of appearances;
the unconditioned of the dynamical series does not come with such a condition as
the translation of N. K. Smith seems to suggest. The point seems to be that N. K.
Smith has not done justice to the participle “verknüpft” used in this situation as
expressing a condition, like, “read in this way, the whole thing looks different”.
p. 343.
In the first (the mathematical) class of antinomies the falsehood of
the presupposition consists in representing in one concept something self-
contradictory as if it were compatible (i.e., an appearance as an object
in itself). But as to the second (the dynamical) class of antinomies, the
falsehood of the presupposition consists in representing as contradictory
what is compatible. Consequently, whereas in the first case the opposed
assertions were both false, in this case, on the other hand, where they are
opposed to one another by mere misunderstanding, they may both be true.
(2) Abbott [1873], p. 87; translating Kant [1788], p. 3.
Speculative reason could only exhibit this concept (of freedom) problem-
atically as not impossible to thought, without assuring it any objective
reality[[.]]
(3) Abbott [1873], p. 144; translating Kant [1788], p. 64.
But how is it with the application of this category of causality (and
all the others; for without them there can be no knowledge of anything
existing) to things which are not objects of possible experience, but lie
beyond its bounds? For I was able to deduce the objective reality of these
concepts only with regard to objects of possible experience. But even this
very fact, that I have saved them, only in case I have proved that objects
may by means of them be thought, though not determined à priori; this it
is that gives them a place in the understanding, by which they are referred
to objects in general (sensible or not sensible). If anything is still wanting
it is that which is the condition of the application of these categories, and
especially that of causality, to objects, namely intuition; for where this is
not given, the application with a view to theoretic knowledge of the object,
as a noumenon, is impossible; and therefore if anyone ventures on it, is
(as in the critique of pure reason) absolutely forbidden. Still, the objective
reality of the concept (of causality) remains, and it can be used even of
noumena, but without our being able in the least to define the concept
theoretically so as to produce knowledge. For that this concept, even in
reference to an object, contains nothing impossible, was shown by this, that
even while applied to objects of sense, its seat was certainly fixed in the pure
understanding; and although, when referred to things in themselves (which
cannot be objects of experience), it is not capable of being determined so as
to represent a definite object for the purpose of theoretic knowledge; yet for
any other purpose (for instance, a practical) it might be capable of being
determined so as to have such application. This could not be the case if, as
Hume maintained, this concept of causality contained something absolutely
impossible to be thought.
Remark 62.23. Kant’s treatment of the antinomies is of considerable interest to

my endeavour. What it amounts to for me is this: in the antinomies, something
is expressed which adheres to the form (under which things appear) but not the
genuine content (i.e., the thing as it is in itself). Note that this is on the same line
as Aristoteles’ distinction mentioned in quotation 67.4 and quite in accordance with
the latter’s view of dialectic as being concerned with opinion.
CHAPTER XVI
HEGEL’S SPECULATIVE PHILOSOPHY
While Kant may have a good chance of passing as rational, Hegel is commonly
considered a paradigm of irrationality. Even by many of his devotees he is presented
as the champion of an obscure metaphysics of some absolute spirit. This is certainly
not the view I take, and the quotations of the present chapter are chosen accordingly.
My emphasis lies on illuminating what I take to be the aim of Hegel’s speculative
philosophy. In this vein the first paragraph of this chapter is devoted to the shift
from Kant to Hegel, to give an impression of how Hegel’s speculative philosophy may
be seen as growing out of a consistent critique of Kant’s transcendental philosophy.
In Hegel’s speculative philosophy I shall concentrate on the following issues:
(HG1) the logical turn;
(HG2) the opposition of form and content;
(HG3) truth and the content of thought determinations;
(HG4) contradictions (negative dialectic);
(HG4) Aufhebung (positive dialectic).
Main sources: Hegel [1812] and [1816], Science of Logic I and II, in the trans-
lations of Miller [1969], and Johnston and Struthers [1929],1 Hegel [1830], Encyclo-
pedia I, Science of Logic in the Wallace [1873] translation of the revised edition of
1830; I have marked the Zusätze as such. I shall also make use of [1801] (“Differen-
zschrift”), [1802] (Glauben und Wissen), [1807] (Phenomenology).
It might seem odd that I do not take up Hegel where he explicitly refers to
mathematics, not even in so crucial a topic as that of the infinitesimals. The reasons
for this have little to do with Hegel’s own writings; the reasons are to be found in
the development of the foundations of mathematics since the days of the notion
of the infinitesimals. When mathematicians finally got around to tidying up the
conceptual mess surrounding the notion of the infinitesimals, they stumbled into
another trap, a trap which they set themselves by means of concepts much more
subtle and sophisticated than those of the calculus ever were. It is this result of the
mathematicians’ conceptual struggle which sets the target for the present study, not
so much what Hegel has said explicitly, be that about dialectic, or the infinitesimals.
In other words, it is the spirit of Hegel, not so much the words, but it is the words
of mathematics, not so much its spirit, that I appropriate for my study.
1 The use of different translations is actually due to constraints on access at a certain time;
but it may equally well be seen as a reluctance to give priority to one or the other translation. I
find them all pretty bad with the exception of Behler [1990].
830
§ 63. FROM KANT TO HEGEL 831
§ 63. From Kant to Hegel
Kant’s Critique of Pure Reason initiated an extremely fertile, though controversial,

period of philosophy in Germany which is usually referred to as “German Idealism”.
With regard to the core of transcendental philosophy, viz., the transcendental de-
duction of the categories, it may be characterized as the shift from transcendental
logic to speculative philosophy. The lesson of transcendental philosophy, the proof of
transcendental idealism via a reductio ad absurdum, had to give way to a new assess-
ment of contradictions in Hegel. To some philosophers, particularly the champions
of common sense, this shift was simply a lapse into absurdity.2
The salient features of the shift from Kant’s transcendental philosophy to Hegel’s
speculative philosophy can be collected under three headings:
(i) the metaphysics of knowledge (idealism and thing in itself);
(ii) the sources of knowledge (I, self-consciousness);
(iii) the emergence of contradictions in knowledge (antinomies).
As regards the metaphysics of knowledge, the development from Kant to Hegel is
characterized by a stronger emphasis of the idealistic aspect. The assault on the
concept of the thing in itself may be viewed as its main aspect. As regards the
sources of knowledge, it is characterized by abandoning sense perception as original
source, an empiricist leftover with no real place in the new design, and a shift
to understanding, and finally, with Hegel, to language (logic). This comes with a
concentration on a faculty of synthesis (such as the I) as a foundation of knowledge.
As regards the emergence of contradictions, it is characterized by a re-evaluation of
the phenomenon of antinomies: number of antinomies; “meaning” (solution) of the
phenomenon of antinomies: no longer viewed as limitations of knowledge.
Remarks 63.1. (1) The assault on the concept of the thing in itself may be reminis-
cent of Berkeley’s critique of the “absolute existence of sensible objects in themselves,
or without the mind ”,3 in particular in the formulation of Hegel in quotation 65.2 (1)
below. As for the rest, however, the situation is quite new.
(2) The reception of Kant’s transcendental philosophy by Maimon, Reinhold, and
Schulze hasn’t made it into these materials.
(3) It would have been nice to have a bit of Schelling included in this chapter —
but I’m afraid I just didn’t make it beyond that isolated quotation below.
63a. Fichte’s reception of Kant’s project. Fichte turned his attention to

two points, firstly the notion of the thing in itself, and secondly the transcendental
deduction of the categories, i.e., to topics from the first two groups, the metaphysics
of knowledge and the sources of knowledge. He did not engage with the third group,
the phenomenon of contradictions in the realm of reason.
In presenting Fichte’s ideas I shall use his Science of Knowledge and the two
introductions.
2 Popper can claim to be the paradigm case; cf. quotation 68.4 (1) in these materials. Also:
Reichenbach in quotation 68.5 (2) in these materials.

3 Cf Berkeley [1710], p. 160, and the argumentation of Philonous in Berkeley [1713], pp. 245 f;
see quotations 58.12 (3), 58.13, respectively.

832 XVI. HEGEL’S SPECULATIVE PHILOSOPHY
Fichte’s starting point was Kant’s deduction of the categories. His point was that
the principle underlying the deduction of the categories is not sufficient. In other
words, Fichte concentrated on what Kant in his Critique of Pure Reason called
the transcendental unity of apperception, an objective synthesis which constitutes
knowledge, the unconditioned in knowledge.
Quotations 63.2. (1) Heath and Lachs [1982], p. 4; translating Fichte [1797/98],
p. 2.
I have long asserted, and repeat once more, that my system is nothing other
than the Kantian; this means that it contains the same view of things, but
is in method quite independent of the Kantian presentation.
(2) Heath and Lachs [1982], pp. 51 f; translating Fichte [1797/98], pp. 58 ff.
Now I am very well aware that Kant by no means established a system of the
aforementioned kind; for in that case the present author would have saved
himself the trouble and chosen some other branch of human knowledge
as the scene of his labors. I am aware that he by no means proved the
categories he set up to be conditions of self-consciousness, but merely said
that they were so: that still less did he derive space and time as conditions
thereof, or that which is inseparable from them in the original consciousness
and fills them both; in that he never once says of them, as he expressly
does of the categories, that they are such conditions, but merely implies
it by way of the argument given above. However, I think I also know with
equal certainty that Kant envisaged such a system; that everything that he
actually propounds consists of fragments and consequences of such a system,
and that his claims have sense and coherence only on this assumption.
Whether he had not thought out this system for himself to a pitch of clarity
and precision where he could also have expounded it to others, or whether
he had in fact done so and simply did not want to expound it, as certain
passages seem to indicate, might well, as it seems to me, remain wholly
unexplored, or if it is to be looked into, someone else may do it; for on this
point I have never expressed any view. However such an inquiry might turn
out, the eminent author still retains unique credit for this achievement, of
having first knowingly diverted philosophy away from external objects and
directed it into ourselves. This is the spirit and inmost heart of his whole
philosophy, as it is also the spirit and heart of the Science of Knowledge.
Comments. (α) Cf. Kant [1781] (A), pp. 82 f (B108); quotation 61.15 (1) in these
materials.
(β) It is in a footnote to the first occurrence of the word “spirit” in the last sentence
that Fichte makes the distinction of reading the spirit and reading the letter which
Hegel was to take up in his “Differenzschrift” 4 .

Quotation 63.3. Heath and Lachs [1982], p. 93; translating Fichte [1794/95],
p. 91.
Our task is to discover the primordial, absolutely unconditioned first
principle of all human knowledge. This can be neither proved nor defined,
if it is to be an absolutely primary principle.
It is intended to express that Act which does not and cannot appear
among the empirical states of our consciousness and alone makes it possible.
Quotation 63.4. Zweig [1967], p. 253; translating Kant; quoted after Pereboom
[1990], p. 33.
I hereby declare that I regard Fichte’s Theory of Science [Wissenschaft-
slehre] as a totally indefensible system. For the pure theory of science is
nothing more or less than mere logic, and the principles of logic cannot
lead to any material knowledge. Since logic, that is to say, pure logic, ab-
stracts from the content of knowledge, the attempt to cull a real object out
of logic is a vain effort and therefore a thing that no one has ever done. If the
transcendental philosophy is correct, such a task would involve metaphysics
rather than logic.
Comment. This quotation shows beautifully what I regard as Kant’s basic mis-
judgment — albeit common amongst philosophers — of the nature of logic: because
logic abstracts from the content of knowledge, it can’t have any content. This is on
a similar line as that claim of Kant’s that “mathematical definition can never be in
error” 5 On a different level this is repeated by Putnam: “We interpret our languages
or nothing does.” 6 Furthermore: are space, time, causality, etc. “real objects”?
63b. The assault on the concept of the thing-in-itself. Fichte did the
first step to do away with the notion of the thing in itself and he concentrated
his efforts on actually deducing the sources of knowledge from a single principle;
an enterprise which Kant had envisaged but never accomplished. This is what he
called “Wissenschaftslehre” (Science of Knowledge).
Quotation 63.5. Heath and Lachs [1982], pp. 10 f; translating Fichte [1797/98],
pp. 10 f.
I can freely determine myself to think this or that; for example, the
thing-in-itself of the dogmatic philosophers. If I now abstract from what
is thought and observe only myself, I become to myself in this object the
content of a specific presentation. That I appear to myself to be deter-
mined precisely so and not otherwise, as thinking, and as thinking, of all
possible thoughts, the thing-in-itself, should in my opinion depend on my
self-determination: I have freely made myself into such an object. But I
have not made myself as it is in itself; on the contrary, I am compelled to
presuppose myself as that which is to be determined by self-determination.
I myself, however, am an object for myself whose nature depends, under
5 Kant [1781] (A), p. 731 (B759); see quotation 59.6 (3) in these textscmaterials.
6 Putnam [1980], p. 482; see quotation 95.28 (5) in the materials; cf. also confrontation 113.12
in the groundworks.
certain conditions, on the intellect alone, but whose existence must always
be presupposed.
Now the object of idealism is precisely this self-in-itself. The object of
this system, therefore, actually occurs as something real in consciousness,
not as a thing-in-itself, whereby idealism would cease to be what it is and
would transform itself into dogmatism, but as a self-in-itself; not as an
object of experience, for it is not determined but will only be determined
by me, and without this determination is nothing, and does not even exist;
but as something that is raised above all experience.
By contrast, the object of dogmatism belongs to those [[. . .]], which are
produced solely by free thought; the thing-in-itself is a pure invention and
has no reality whatever. [[. . .]]
Thus the object of idealism has this advantage over the object of dog-
matism, that it may be demonstrated, not as the ground of the explanation
of experience, which would be contradictory and would turn this system it-
self into a part of experience, but still in general in consciousness; whereas
the latter object cannot be looked upon as anything other than a pure in-
vention, which expects its conversion into reality only from the success of
the system.
Comment. Note Fichte’s formulation “not as the ground of the explanation of ex-
perience”; this is diametrically opposed to N. K. Smith’s formulation cited by Werk-
meister in quotation 68.17 (2) in these materials.
Quotations 63.6. (1) Heath and Lachs [1982], p. 30; translating Fichte [1797/98],
p. 34.
Its chosen topic of consideration is not a lifeless concept, passively exposed
to its inquiry merely, of which it makes something only by its own thought,
but a living and active thing which engenders insights from and through
itself, and which the philosopher merely contemplates.
(2) Heath and Lachs [1982], p. 50; translating Fichte [1797/98], p. 57.
According to Kant, all consciousness is merely conditioned by self-
consciousness, that is, its content can be founded upon something outside
self-consciousness; now the results of this foundation are simply not sup-
posed to contradict the conditions of self-consciousness; simply not to elim-
inate the possibility thereof; but they are not required actually to emerge
from it.
According to the Science of Knowledge, all consciousness is determined
by self-consciousness, that is, everything that occurs in consciousness is
founded, given and introduced by the conditions of self-consciousness; and
there is simply no ground whatever for it outside self-consciousness. —I
must explain that in our case the determinacy follows directly from the
fact of being conditioned, and that here, therefore, the distinction question
does not operate at all, and says nothing.
Kant’s conception of a reductio ad absurdum to prove transcendental idealism
seems abandoned; so does the concept of empirical realism. The existence of an
extramental world is not denied nor doubted; there just isn’t any need for it in
Fichte’s Science of Knowledge. In some way Fichte started where Kant halted —
convinced that he had indicated the direction sufficiently — with the intent to carry
out the business of a deduction of the categories in full detail.
Quotations 63.7. (1) Schelling [1800], p. 24.

Da jedes wahre System (wie z.B. das des Weltbaues) den Grund seines
Bestehens in sich selbst haben muß, so muß, wenn es ein System des Wissens
gibt, das Prinzip desselben innerhalb des Wissens selbst liegen.
(2) Schelling [1800], pp. 29 f.
Das Prinzip der Philosophie muß [[. . .]] ein solches sein, in welchem
der Inhalt durch die Form, und hinwiederum die Form durch den Inhalt
bedingt ist und nicht eines das andere, sondern beide wechselseitig sich vor-
aussetzen. — Gegen ein erstes Prinzip der Philosophie ist unter anderem
auch auf folgende Weise argumentiert worden: Das Prinzip der Philosophie
muß sich in einem Grundsatz ausdrücken lassen; dieser Grundsatz soll ohne
Zweifel kein bloß formeller, sondern ein materieller sein. Nun steht aber
jeder Satz, sein Inhalt sei, welcher er wolle, unter den Gesetzen der Lo-
gik. Also setzt jeder materielle Grundsatz bloß dadurch, daß er ein solcher
ist, höhere Grundsätze, die der Logik, voraus. — Es fehlt zu dieser Argu-
mentation nichts, als daß man sie umkehre. Man denke sich irgend einen
formellen Satz, z. B. A = A, als den höchsten; was an diesem Satze logisch
ist, ist bloß die Form der Identität zwischen A und A; aber woher kommt
mir denn A selbst? Wenn A ist, so ist es gleich sich selbst; aber woher ist
es denn? Diese Frage kann ohne Zweifel nicht aus dem Satz selbst, sondern
nur aus einem höheren beantwortet werden. Die Analysis A = A setzt die
Synthesis A voraus. Also ist offenbar, daß kein formelles Prinzip gedacht
werden kann, ohne ein materielles, noch ein materielles, ohne ein formelles
vorauszusetzen.
Aus diesem Zirkel, daß jede Form einen Inhalt, jeder Inhalt eine Form
voraussetzt, ist gar nicht herauszukommen, wenn nicht irgend ein Satz ge-
funden wird, in welchem wechselseitig Form durch Inhalt und Inhalt durch
Form bedingt und möglich gemacht ist.
(3) Schelling [1800], pp. 33 f.
In jedem Satz werden zwei Begriffe miteinander verglichen, d.h. sie wer-
den einander entweder gleich oder ungleich gesetzt. Im identischen Satze
nun wird bloß das Denken mit sich selbst verglichen. — Der synthetische
Satz hingegen geht hinaus über das bloße Denken; dadurch, daß ich das
Subjekt des Satzes denke, denke ich nicht auch das Prädikat — das Prä-
dikat kommt zum Subjekt hinzu. Der Gegenstand ist also hier nicht bloß
bestimmt durch sein Denken; er wird als reell betrachtet; denn reell ist
eben, was durch das bloße Denken nicht erschaffen werden kann.
Wenn nun ein identischer Satz der ist, wo der Begriff nur mit dem
Begriff, ein synthetischer der, wo der Begriff mit dem von ihm verschiedenen
Gegenstand verglichen wird, so heißt die Aufgabe, einen Punkt zu finden,
wo das identische Wissen zugleich synthetisch ist, so viel als: einen Punkt
finden, in welchem das Objekt und sein Begriff, der Gegenstand und seine
Vorstellung ursprünglich, schlechthin und ohne alle Vermittlung eins sind.
Comment. A bit clumsy, the whole thing, it seems to me; but take into account
Schelling’s age: not yet 25. Still, I want to recommend to keep this task in mind.
63c. Some of Hegel’s roots in ancient Greek philosophy. While Kant

did not seem to have a lot to say about any roots of his philosophy in ancient
Greek thought, Hegel certainly does, albeit predominantly in notes collected by his
students.
I begin with a quotation regarding the origin of dialectic, followed by one re-
garding Herakleitos.
Quotations 63.8. (1) Hegel [1833a], p. 296 (cf. Haldane [1892], p. 239).
[[I]]n der eleatischen Schule [[. . .]] sehen wir den Gedanken sich selbst frei
für sich selbst werden; in dem, was die Eleaten als das absolute Wissen
aussprechen, den Gedanken sich selbst rein ergreifen, und die Bewegung
des Gedankens in Begriffen. Wir finden hier den Anfang der Dialektik, d. h.
eben der reinen Bewegung des Denkens in Begriffen; damit den Gegen-
satz des Denkens gegen die Erscheinung oder das sinnliche Seyn, — dessen
was an sich ist gegen das Für-ein-Anderes-Seyn dieses Ansich: und an dem
gegenständlichen Wesen den Widerspruch, den es an ihm selbst hat (die
eigentliche Dialektik).
(2) Hegel [1833a], pp. 347 f, (cf. Haldane [1892], p. 281).
Cicero hat einen schlechten Einfall, wie es ihm oft geht; er meint, [[Herak-
leitos]] habe absichtlich so dunkel geschrieben. Es ist dieß aber sehr platt
gesagt, seine eigene Plattheit, die er zur Plattheit Heraklits macht[[.]] [[. . .]]
Das Dunkle dieser Philosophie liegt aber hauptsächlich darin, daß ein tiefer,
spekulativer Gedanke in ihr ausgedrückt ist; dieser ist immer schwer, dunkel
für den Verstand: die Mathematik dagegen ist ganz leicht. Der Begriff, die
Idee ist dem Verstand zuwider, kann nicht von ihm gefaßt werden.
Comment. To me this sounds like Hegel is speaking here about his own philosophy
and the way it is being received.
Quotation 63.9. Hegel [1833a], pp. 312 f.

Mit Parmenides hat das eigentliche Philosophieren angefangen, die Erhe-
bung in das Reich des Ideellen ist hierin zu sehen. Ein Mensch macht sich
frei von allen Vorstellungen und Meinungen, spricht ihnen alle Wahrheit ab
und sagt, nur die Nothwendigkeit, das Seyn ist das Wahre. Dieser Anfang
ist freilich noch trübe und unbestimmt; es ist nicht weiter zu erklären, was
darin liegt; aber gerade dieß Erklären ist die Ausbildung der Philosophie
selbst, die hier noch nicht vorhanden ist. Damit verband sich die Dialektik,
daß das Veränderliche keine Wahrheit habe; denn wenn man diese Bestim-
mungen nimmt, wie sie gelten: so kommt man auf Widersprüche.
Quotations 63.10. (1) Haldane [1892], p. 252; translating Hegel [1833a], p. 310.
The nothing, in fact, turns into something, since it is thought or is said: we
say something, think something, if we wish to think and say the nothing.
Comment. This is Hegel’s comment to Parmenides fr.2, v.7, quoted in 57.1 in these
materials together with various translations.
(2) Haldane [1892], p. 252; translating Hegel [1833a], pp. 310 f.
Parmenides says, whatever form the negative may take, it does not exist at
all.
Comment. This is Hegel’s comment to Parmenides fr.6.
(3) Hegel [1833a], p. 312.
Das Denken producirt sich, was producirt wird, ist ein Gedanke; Denken ist
also mit seinem Seyn identisch, denn es ist nichts au"ser dem Seyn, dieser
gro"sen Affirmation.
Comment. This is Hegel’s comment to Parmenides fr.8, vv.34 ff.
(4) Haldane [1892], p. 253; translating Hegel [1833a], p. 312.
Thought produces itself, and what is produced is a Thought. Thought is
thus identical with Being, for there is nothing beside Being, this great af-
firmation.
Comment. Notice: Haldane uses “Thought” to translate both, “Denken” (thinking)
as well as “Gedanke” (thought). Also: Haldane did not translate “seinem” in the
second sentence. So, in my reading, it should be:“Thinking is thus identical with its
Being”.
(5) Miller [1969], pp. 94 f; translating Hegel [1812], p. 104.
Parmenides held fast to being and was most consistent in affirming
at the same time that nothing absolutely is not; only being is. As thus
taken, entirely on its own, being is indeterminate, and has therefore no
relation to an other; consequently, it seems that from this beginning no
further progress can be made—that is, from this beginning itself—and that
progress can only be achieved by linking it on to something extraneous,
something outside it. Hence the progress made in affirming that being is
the same as nothing appears as a second, absolute beginning—a transition
which is independent of being and added to it from outside. If being had
a determinateness, then it would not be the absolute beginning at all; it
would then depend on an other and would not be immediate, would not
be the beginning. But if it is indeterminate and hence a genuine beginning,
then, too, it has nothing with which it could bridge the gap between itself
and an other; it is at the same time the end. It is just as impossible for
anything to break forth from it as to break into it; with Parmenides as
with Spinoza, there is no progress from being or absolute substance to the
negative, to the finite. If, nevertheless, there is progress— which, as has
been remarked, in the case of relationless, and so progress-less being can be
accomplished only in an external manner—then this progress is a second,
a fresh beginning.
(6) Miller [1969], p. 100; translating Hegel [1812], p. 112.

The dialectic employed by Plato in treating of the One in the Par-
menides is also to be regarded rather as a dialectic of external reflection.
Being and the One are both Eleatic forms which are the same thing. But
they are also to be distinguished; and it is thus that Plato takes them in
that dialogue. After removing from the One the various determinations of
whole and parts, of being-within-itself, of being-in-another, etc., of shape,
time, etc., he reaches the result that being does not belong to the One, for
being belongs to any particular something only in one of these modes.
Comment. Hegel’s discussion continues, of course, but the point was only to get a
glimpse of his attitude towards Platon’s Parmenides.
Quotations 63.11. (1) Hegel [1833a], pp. 264 f; a translation based on Hegel [1840]
can be found in Haldane [1892], pp. 215 f.
[[D]]ie Aufnahme des Gegensatzes als eines wesentlichen Moments des Abso-
luten hat aber überhaupt bei den Pythagoräern ihren Ursprung. Sie haben
früh, wie später Aristoteles, eine Tafel von Kategorien aufgestellt (daher
man dem Letzteren den Vorwurf machte, von ihnen seine Denkbestimmun-
gen entlehnt zu haben), — die abstrakten und einfachen Begriffe weiter be-
stimmt, obzwar freilich auf eine unangemessene Art; — eine Vermischung
von Gegensätzen der Vorstellung und des Begriffs, ohne weitere Deduktion
oder System der Bewegung. [[. . .]] Es ist dieß Versuch einer weiteren Ausbil-
dung der Idee der spekulativen Philosophie in ihr selbst, in Begriffen. Aber
weiter als auf diese α) vermischte Auflösung, β) bloße Aufzählung scheint
dieser Versuch nicht gegangen zu seyn. Es ist sehr wichtig, daß zunächst
nur Sammlung gemacht werde (wie Aristoteles that) von den allgemeinen
Denkbestimmungen. Es ist ein roher Anfang von näherer Bestimmung der
Gegensätze; ohne Ordnung, ohne Sinnigkeit[[.]]
(2) Haldane and Simson [1894], p. 136 and me; translating Hegel [1833b], p. 317.
Aristoteles always seems to have philosophized only respecting the individ-
ual and particular, and [[he doesn’t seem to say what the absolute, the uni-
versal, what God is]]; he always goes from the individual to the individual.
[[He takes on the whole bulk of the world of ready made ideas (“Vorstel-
lungswelt”) and works his way through it: soul, motion, sensation, memory,
thinking, — his day’s work is (concerns), what is — just like a professor
in a half year’s course; and]] yet appears only to have recognized the truth
in the particular, or only a succession of particular truths[[, — he doesn’t
point out the universal]].
Comment. Haldane’s translation is based on a different edition of Hegel’s Geschichte

der Philosophie; so my “quotation” is a brutal mixture and should be taken cum
grano salis. Not to speak of the fact that it wasn’t Hegel who wrote it in the first
place.
(3) Miller [1969], pp. 831 f; translating Hegel [1816], pp. 337 f.
The older Eleatic school directed its dialectic chiefly against motion, Plato
frequently against the general ideas and notions of his time, especially those
of the Sophists, but also against the pure categories and the determinations
of reflection; the more cultivated scepticism of a later period extended it
not only to the immediate so-called facts of consciousness and maxims of
common life, but also to all the notions of science. Now the conclusion
drawn from dialectic of this kind is in general the contradiction and nullity
of the assertions made. But this conclusion can be drawn in either of two
senses—either in the objective sense, that the subject matter [[Gegenstand ]]
which in such a manner contradicts itself cancels itself out and is null and
void—this was, for example, the conclusion of the Eleatics, according to
which truth was denied, for example, to the world, to motion, to the point;
or in the subjective, that cognition is defective. One way of understanding
the latter sense of the conclusion is that it is only this dialectic that imposes
on us the trick of an illusion. This is the common view of so-called sound
common sense which takes its stand on the evidence of the senses and on
customary conceptions and judgements. [[. . .]]
The fundamental prejudice in this matter is that dialectic has only a
negative result [[.]]
63d. Hegel’s logical turn: logic as an organon. Hegel’s idea of a specula-

tive philosophy as a resumption of metaphysics has roots in Kant’s Critique of Pure
Reason. Right with the first sentence of the first preface to his Science of Logic, it
becomes clear in which direction Hegel is going to push the Kantian approach.
Quotations 63.12. (1) Miller [1969], p. 25; translating Hegel [1812], p. 13.
The complete transformation which philosophical thought in Germany has
undergone in the last twenty-five years and the higher standpoint reached
by spirit in its awareness of itself, have had but little influence as yet on
the structure of logic.
Comments. (α) As regards the phrase “complete transformation . . . ”, German, “Die

völlige Umänderung, welche die philosophische Denkweise . . .”: is this a reminiscence
of Fichte: “die völlige Umkehrung der Denkart”;7 a variation on Kant’s allusion to
Copernicus.
(β) The mention of logic may sound surprising since it was Kant who created a
transcendental logic, thus apparently giving to logic a new dimension. But Hegel
seems to be aiming at something different, something which comes much closer to
logic in the sense of tautology than Kant’s transcendental logic. It seems to me that
what Hegel is preparing here is an assault on Kant’s doctrine of the impotence of
logic as documented in quotations 59.5 (3) and (4), and 63.4.
7 Fichte [1797/98], p. 3.

Recently Kant has opposed to what has usually been called logic an-
other, namely, a transcendental logic. What has here been called objective
logic would correspond in part to what with him is transcendental logic.
The critical philosophy had, it is true, already turned metaphysics into logic
but it, like the later idealism [[. . .]] was overawed by the object, and so the
logical determinations were given an essentially subjective significance with
the result that these philosophies remained burdened with the object they
had avoided and were left with the residue of a thing-in-itself, an infinite
[[impetus]], as a beyond.
Comment. The word “impetus” in double square brackets replaces “obstacle”. The
German word is “Anstoß”.
Quotations 63.13. (1) U. Petersen [2002], p. 840; translating Hegel [1807],

pp. 33 f.8
The activity of separation is the force and labor of the Understanding, that
most amazing (intriguing) and greatest, or rather absolute power. The circle
which rests closed within itself and qua substance contains its moments, is
the immediate and for that reason not amazing (intriguing) relation. But
that the Accidental as such, detached from its circumference (that which
embraces it), the Bound and only Actual in its connection (correlation)
with some other, gains an own existence (Daseyn) and separate freedom, is
the tremendous power of the Negative; it is the energy of the Thought, the
pure I.
Comment. I use capital initial letters to distinguish nouns from adjectives, as is
done in German.
The ‘I’, or becoming in general, this mediation, on account of its simple
nature, is just immediacy in the process of becoming, and is the immediate
itself. Reason is, therefore, misunderstood when reflection is excluded from
the True, and is not grasped as a positive moment of the Absolute. It
is reflection that makes the True a result, but it is equally reflection that
overcomes the antithesis between the process of its becoming and the result,
for this becoming is also simple, and therefore not different from the form of
the True which shows itself as simple in its result; the process of becoming
is rather just this return into simplicity.
Philosophy has to take a new direction. “With what must the science begin?”
According to Kant: transcendental aesthetic, i.e., intuition. According to Hegel:
logic.
8 Miller’s translation in Miller [1977], pp. 18 f, does not seem to catch the meaning that I
gathered from the German.

It is only in recent times that thinkers have become aware of the difficulty of
finding a beginning in philosophy, and the reason for this difficulty and also
the possibility of resolving it has been much discussed. What philosophy
begins with must be either mediated or immediate, and it is easy to show
that it can be neither the one nor the other; thus either way of beginning
is refuted.
(2) Wallace [1873], p. 3; translating Hegel [1830], p. 41.
Philosophy misses an advantage enjoyed by the other sciences. It cannot
like them rest the existence of its objects on the natural admissions of
consciousness, nor can it assume that its method of cognition, either for
starting or for continuing, is one already accepted.
(3) Taubeneck [1990], p. 48; translating Hegel [1817], p. 19.
All sciences other than philosophy deal with issues that are assumed
to be immediate to representation. Such issues are thus presupposed from
the beginning of the science and, in the course of its further development,
determinations considered necessary are also derived from representation.
Such a science does not have to justify the necessity of the issues
it treats. Mathematics [[in general, geometry, arithmetic]], jurisprudence,
medicine, zoology, botany, and so on, can presuppose the existence of mag-
nitude, space, number, law, diseases, animals, plants, and so one. These are
assumed to be ready at hand for representation. It does not occur to us to
doubt the being of such issued[[.]]
Comment. After the mention of mathematics the German original has “überhaupt,
der Geometrie, der Arithmetik” which explains my addition in double square brack-
ets. I seriously doubt, however, that the translator’s reason for dropping geometry
and arithmetic has anything to do with my opinion that after Frege arithmetic no
longer belongs in the list.
(4) Taubeneck [1990], pp. 48 f; translating Hegel [1817], p. 20.
By contrast, the beginning of philosophy involves the awkward problem
that its object immediately and necessarily provokes doubt and controversy.
1) There is a problem regarding content: in order to be seen as not merely a
representation, but as the very object of philosophy, the content must not be
found in the representation. Indeed, the cognitive procedure in philosophy is
actually opposed to representation, and the faculty of representation should
be brought beyond itself through philosophy.
Comment. In the first one of these quotations, Taubeneck translates the German
“Gegenstand” as “issue”. In the second one he has changed to “object”.
(5) Taubeneck [1990], p. 49; translating Hegel [1817], pp. 20 f.
The beginning of philosophy faces the same embarrassment regarding form,
for the beginning as beginning is immediate, but represents itself as medi-
ated. The concept must on the one hand be recognized as necessary and
at the same time the cognitive method cannot be presupposed, since its
derivation occurs within philosophy itself.
The demand made fashionable by Kantian philosophy, that prior to
actual cognition the faculty of cognition should be investigated critically,
offers itself on first glance as a plausible alternative. But this investigation
is itself cognitive, and to claim that it could be initiated without cognition
is meaningless. Furthermore, even the assumption of a faculty of cognition
before actual cognition is a presupposition, both of the unjustified category
or determination of a faculty or power, and of a subjective cognition;—
a presupposition which belongs to the former one. Logic, by the way, is
also part of that required investigation, and more truly than the critical
method, which would have to proceed in a limited way on the basis of its
own presuppositions and according to the nature of its own ability.
Quotation 63.15. Hegel [1812], p. 80.

[[D]]er Anfang soll nicht selbst schon ein Erstes u n d ein Anderes seyn;
ein solches das ein Erstes u n d ein Anderes in sich ist, enthält bereits
ein Fortgegangenseyn. Was den Anfang macht, der Anfang selbst, ist daher
als ein Nichtanalysierbares, in seiner einfachen unerfüllten Unmittelbarkeit,
also a l s S e y n , als das ganz Leere zu nehmen.
[[. . .]]
[[. . .]] Diejenigen, welche mit diesem Anfange unzufrieden bleiben, mö-
gen sich zu der Aufgabe auffordern, es anders anzufangen[[.]]
Comment. Cf. Frege [1893], p. XXVI.9 It is not so much that I am dissatisfied with
this beginning, as with the actual “Fortgang”.
Quotations 63.16. (1) U. Petersen [2002], p. 842; translating Hegel [1809/11],

p. 101.
In philosophy the determinations of the knowledge are not just regarded
as determinations of the things, but at once with the knowledge to which
they belong at least jointly with the objects; or they are taken not only as
objective, but also as subjective determinations, or rather as certain kinds
of relation of the object and the subject to each other.
Comment. This quotation is taken from Hegel’s teaching papers at the Gymnasium
in Nuremberg. Note the conceptual struggle: determinations of the knowledge —
determinations of the things. Hegel makes a distinction which he immediately has
to cancel again.
(2) U. Petersen [2002], p. 842; translating Hegel [1809/11], p. 101.
Our ordinary knowledge imagines the object only which it knows but not
at the same time itself, namely the knowledge.


The vulgar believe that the objects of perception which confront them, such
as an individual animal, or a single star, are independent and permanent
existences, compared with which thoughts are unsubstantial and dependent
on something else. In fact however the perceptions of sense are the properly
dependent and secondary feature, while the thoughts are really independent
and primary. This being so, Kant gave the title objective to the intellectual
factor, to the universal and necessary: and he was quite justified in so doing.
Comment. This is an extremely important point and I am afraid I haven’t given it
sufficient space. On top of it, the translation is bad.
[[C]]ognition is determining and determinate thinking[[.]]
The next couple of quotations revolves around a certain petrification of con-

cepts.
Quotations 63.17. (1) Miller [1977], pp. 19 f; translating Hegel [1807], pp. 34 f.
The manner of study in ancient times differed from that of the modern age
in that the former was the proper and complete formation of the natural
consciousness. Putting itself to the test at every point of its existence, and
philosophizing about everything it came across, it made itself into a uni-
versality that was active through and through. In modern times, however,
the individual finds the abstract form ready-made; the effort to grasp and
appropriate it is more the direct driving-forth of what is within and the
truncated generation of the universal than it is the emergence of the latter
from the concrete variety of existence. Hence the task nowadays consists not
so much in purging the individual of an immediate, sensuous mode of ap-
prehension, and making him into a substance that is an object of thought
and that thinks, but rather in just the opposite, in freeing determinate
thoughts from their fixity so as to give actuality to the universal, and im-
part to it spiritual life. But it is far harder to bring fixed thoughts into a
fluid state that to do so with sensuous existence. [[. . .]] Thoughts become
fluid when pure thinking, this immediacy, recognizes itself as a moment, or
when the pure certainty of self abstracts from itself—not by leaving itself
out, or setting itself aside, but by giving up the fixity of its self-positing[[.]]
(2) Wallace [1873], p. 53; translating Hegel [1830], p. 107 (Zusatz).
The battle of reason is the struggle to break up the rigidity to which the
understanding has reduced everything.
Quotations 63.18. (1) Wallace [1873], p. 113; translating Hegel [1830], p. 184.
In point of form Logical doctrine has three sides: (α) the Abstract side, or
that of understanding; (β) the Dialectical, or that of negative reason; (γ)
the Speculative, or that of positive reason.

(α) Thought, as Understanding, sticks to fixity of characters and their dis-
tinctness from one another: every such limited abstract it treats as having
a subsistence and being of its own.
(β) In the Dialectical stage these finite characterizations or formulae super-
sede themselves, and pass into their opposites.
(γ) The Speculative or Positive-Reasonable grasps the unity of the determi-
nations in their opposition, the Positive, which is contained in their solution
[Auflösung] and transition [Übergang].
Comment. This is taken from the first (Heidelberg) encyclopedia from 1817 which
is translated in Taubeneck [1990], p. 58. The German custom of turning an adjec-
tive into a noun by writing it with a capital initial and providing a definite article
is copied into the English translation. The passage appears unchanged in Hegel
[1830], p. 195, translated in Wallace [1873], p. 119. “Auflösung” is translated as “dis-
solution” by Taubeneck and “disintegration” by Wallace. “Übergang” is translated
as “devolution” by Taubeneck and “transition” by Wallace. “Auflösung” can equally
mean “dissolution” as well as “resolution”, so I settled for “solution” which reflects
this double meaning.
Common Sense, that mixture of sense and understanding, believes the
objects of which it has knowledge to be severally independent and self-
supporting; and when it becomes evident that they tend towards and limit
one another, the interdependence of one another is reckoned something for-
eign to them and to their true nature.
Remark 63.19. Kant mentioned the relation of the subject and the predicate: B98.
In his table of judgments: categorical.
§ 64. Hegel’s speculative philosophy as an extension of Kant’s transcen-

dental philosophy
In this paragraph I shall have a closer look at Hegel’s speculative philosophy. In

view of the picture that Popper paints of Hegel’s attitude to Kant’s philosophy in
quotation 68.4 (1), it seems worth emphasizing that I see his speculative philoso-
phy still in the spirit of transcendental philosophy. My choice of quotations in this
paragraph aims at showing how Hegel’s speculative philosophy grew out of Kant’s
transcendental philosophy; hopefully this may even help making Hegel a bit more
intelligible,
64a. Hegel on Kant and Fichte as regards the categories. First of all,
I find it important to point out that Hegel did indeed acknowledge the step made
by Kant.
§ 64. HEGEL’S PHILOSOPHY AS AN EXTENSION OF KANT’S 845
Quotations 64.1. (1) Wallace [1873], p. 66; translating Hegel [1830], p. 124 (Zu-
satz).
A very important step was undoubtedly made, when the terms of the old
metaphysic were subjected to scrutiny. The plain thinker pursued his un-
suspecting way in those categories which had offered themselves naturally.
It never occurred to him to ask to what extent these categories had a value
and authority of their own. If, as has been said, it is a characteristic of free
thought to allow no assumptions to pass unquestioned, the old metaphysi-
cians were not free thinkers. They accepted their categories as they were,
without further trouble as an a apriori datum, not yet tested by reflection.
The Critical philosophy reversed this. Kant undertook to examine how far
the forms of thought were capable of leading to the knowledge of truth.
(2) Miller [1969], p. 61, footnote 1; translating Hegel [1812], p. 62, footnote *.
I would mention that in this work [[i.e., the Science of Logic]] I frequently
refer to the Kantian philosophy (which to many may seem superfluous)
because whatever may be said, both in this work and elsewhere, about
the precise character of this philosophy and about particular parts of its
exposition, it constitutes the base and the starting-point of recent German
philosophy and this its merit remains unaffected by whatever faults may
be found in it. The reason, too, why reference must often be made to it in
the objective logic is that it enters into detailed consideration of important,
more specific aspects of logic[[.]]
(3) Miller [1969], p. 62; translating Hegel [1812], pp. 63 f.
[[Kant’s]] chief thought is to vindicate the categories for self-consciousness as
the subjective ego. By virtue of this determination the point of view remains
confined within consciousness and its opposition; and besides the empirical
element of feeling and intuition it has something else left over which is
not posited and determined by thinking self-consciousness, thing-in-itself,
something alien and external to thought—although it is easy to perceive
[[einzusehen]] that such an abstraction as the thing-in-itself is itself only a
product of thought, and of merely abstractive thought at that.
(4) Haldane and Simson [1896], p. 428; translating Hegel [1840] (not [1833c], p. 557;
but second edition thereof).
[[I]]f universality and necessity do not exist in external things, the ques-
tion arises “Where are they to be found?” To this Kant, as against Hume,
maintains that they must be a priori, i.e. that they must rest on reason
itself, and on thought as self-conscious reason; their source is the subject,
“I” in my self-consciousness. This, simply expressed, is the main point in
the Kantian philosophy.
Hegel’s criticism of Kant’s treatment of the categories is partly directed towards

psychological notions in the transcendental deduction of categories.
Quotations 64.2. (1) Miller [1969], p. 584; translating Hegel [1816], pp. 14 f.
When one speaks in the ordinary way of the understanding possessed by the
I, one understands thereby a faculty or property which stands in the same
relation to the I as the property of a thing does to the thing itself, that is, to
an indeterminate substrate that is not the genuine ground and the determi-
nant of its property. According to this conception, I possess notions and the
Notion, just as I also possess a coat, complexion, and other external prop-
erties. Now Kant went beyond this external relation of the understanding,
as the faculty of notions and of the Notion itself, to the I. It is one of the
profoundest and truest insights to be found in the Critique of Pure Reason
that the unity which constitutes the nature of the Notion is recognized as
the original synthetic unity of apperception, as unity of the I think, or of
self-consciousness. This proposition constitutes the so-called transcendental
deduction of the categories; but this has always been regarded as one of the
most difficult parts of the Kantian philosophy, doubtless for no other reason
than that it demands that we should go beyond the mere representation of
the relation in which the I stands to the understanding, or notions stand
to a thing and its properties and accidents, and advance to the thought of
that relation.
This original synthesis of apperception is one of the most profound prin-
ciples for speculative development; it contains the beginning of a true ap-
prehension of the nature of the Notion and is completely opposed to that
empty identity or abstract universality which is not within itself a synthe-
sis. The further development, however, does not fulfil the promise of the
beginning. The very expression synthesis easily recalls the conception of an
external unity and a mere combination of entities that intrinsically separate.
Then, again, the Kantian philosophy has not got beyond the psychological
reflex of the Notion and has reverted once more to the assertion that the
Notion is permanently conditioned by a manifold of intuition. It has de-
clared intellectual cognition and experience to be phenomenal content, not
because the categories themselves are only finite but, on the ground of a
psychological idealism, because they are merely determinations originating
in self-consciousness.
[[T]]he stage of the understanding is supposed to be preceded by the stages
of feeling and intuition, and it is an essential proposition of the Kantian
transcendental philosophy that without intuitions notions are empty and
are valid solely as relations of the manifold given by intuition.
(4) Miller [1969], p. 586; translating Hegel [1816], p. 17 f.
[[A]]s regards the relation of the understanding or the Notion to the stages
presupposed by it, the form of these stages is determined by the particular
science under consideration. In our science, that of pure logic, these stages
are being and essence. In psychology the antecedent stages are feeling and
intuition, and then ideation generally. In the phenomenology of spirit, which

is the doctrine of consciousness, the ascent to the understanding is through
the stages of sensuous consciousness and then perception. Kant presupposes
only feeling and intuition. How incomplete to begin with this scale of stages
is is revealed by the fact that he himself adds as an appendix to the tran-
scendental logic or doctrine of the understanding a treatise on the concepts
of reflection—a sphere lying between intuition and the understanding or
being and the Notion
About these stages themselves it must be remarked, first of all, that
the forms of intuition, ideation and the like belong to the self-conscious
spirit which, as such, does not fall to be considered in the science of logic.
It is true that the pure determinations of being, essence and the Notion
constitute the ground plan and the inner simple framework of the forms of
the spirit; spirit as intuiting and also as sensuous consciousness is in the
form of immediate being; and, similarly, spirit as ideating and as perceiving
has risen from being to the stage of essence or reflection. But these concrete
forms as little concern the science of logic as do the concrete forms assumed
by the logical categories in nature, which would be space and time, then
space and time self-filled with a content as inorganic nature, and lastly,
organic nature.
Now Hegel claimed that Kant did not yet manage to realize what he had set
forth as an idea and he acknowledged Fichte for having recognized that. This is
what the next set of quotations is about.
satz).
Kant’s examination of the categories suffers from the grave defect of viewing
them, not absolutely and for their own sake, but in order to see whether
they are subjective or objective.
But after all, objectivity of thought, in Kant’s sense, is again to a certain
extent subjective. Thoughts, according to Kant, although universal and
necessary categories, are only our thoughts—separated by an impassable
gulf from the thing, as it exists apart from our knowledge. But the true
objectivity of thinking means that the thoughts, far from being merely ours,
must at the same time be the real essence of the things, and of whatever is
an object to us.
Comment. Cf. quotation 65.1 (1) below, for some of Hegel’s expositions on subjec-
tive and objective.
(3) Wallace [1873], pp. 68 f; translating Hegel [1830], p. 128.
Kant, it is well known, did not put himself to much trouble in discover-
ing the categories. ‘I’, the unity of self-consciousness, being quite abstract
and completely indeterminate, the question arises, how are we to get at
the specialized forms of the ‘I’, the empirical classification of the kinds of
judgement. Now the various modes of judgement, as enumerated to our

hand, provide us with the several categories of thought. To the philosophy
of Fichte belongs the great merit of having called attention to the need of
exhibiting the necessity of these categories and giving a genuine deduction
of them.
Comment. Hegel does not speak in the German original of “Deduktion”, but about
“abgeleitet”, i.e., “derived”. The term deduction may give rise to a confusion with
Kant’s use of it as a justification (cf. quotations 61.20 in these materials).
(4) U. Petersen [2002], p. 848; translating Hegel [1801], pp. 33 f.
Kantian philosophy needed to have its spirit separated from the letter
and the purely speculative principle lifted out of the rest which belonged
to the reasoning reflection or could be used for it. In the principle of the
deduction of the categories this philosophy is genuine idealism; and it is
this principle which Fichte brought out in pure and strict form and called
it the spirit of Kantian philosophy. That the things-in-themselves (whereby
nothing is expressed objectively than the empty form of the opposition)
are hypostatized again and posited as absolute objectivity like the objects
of the dogmatist: that the categories themselves have been turned partly
into static dead drawers of intelligence, partly into the highest principles,
by means of which the expressions in which the absolute itself is spelled
out, as for instance the substance in Spinoza, are destroyed; and thus the
negative reasoning could posit itself in the place of philosophizing — these
circumstances lie at the most in the form of the Kantian deduction of the
categories, not in their principle or spirit. And had we had by Kant no
other piece of his philosophy than this, the metamorphosis would almost
be incomprehensible. In this deduction of the forms of understanding the
principle of speculation, the identity of the subject and the object, is spelled
out most determinately.
The identity of the subject and the object confines itself to twelve or rather
nine pure thought activities; for the modality does not provide a true ob-
jective determination, in it persists essentially the non-identity of subject
and object.
Comment. Notice: “modality does not provide a true objective determination”.
[[T]]he categories, apart from the fact that they are incomplete if taken
directly from the description are derived empirically from ordinary logic,
without showing how the “transcendental unity of self-consciousness” de-
termines itself in the first place. Similarly, it is not shown how this unity
achieves the multiplicity of the determinations of the categories. Nor is it
shown how the categories are to be deduced in terms of their determinacy.
Comment. This about sums it all up — not just for Kant, but also for Hegel himself.
I take it as a variation on the theme ‘you can’t get from here to there’.

After all it was only formally that the Kantian system established the prin-
ciple that thought is spontaneous and self-determining. Into details of the
manner and the extent of this self-determination of thought, Kant never
went. It was Fichte who first noticed the omission; and who, after he had
called attention to the want of a deduction for the categories, endeavoured
really to supply something of the kind.
Comment. I find something confusing in this way of putting things; Kant was ex-
plicitly concerned with a deduction of the categories, essentially in the sense of a
justification. What I take Hegel to be speaking of is a definition of the categories
which Kant explicitly mentioned that he purposely omitted (cf. quotation 61.15 (1)
in these materials).
Still, even Fichte, did not go far enough, in the eyes of Hegel.
Transcendental idealism in its more consistent development, recognized the
nothingness of the spectral thing-in-itself left over by the Kantian philoso-
phy, this abstract shadow divorced from all content, and intended to destroy
it completely. This philosophy also made a start at letting reason itself ex-
hibit its own determinations. But this attempt, because it proceeded from
a subjective standpoint, could not be brought to a successful conclusion.
(2) Wallace [1873], p. 94; translating Hegel [1830], pp. 162 f (Zusatz).
With Fichte, the ‘Ego’ is the starting-point in the philosophical develop-
ment: and the outcome of its action is supposed to be visible in the cate-
gories. But in Fichte the ‘Ego’ is not really presented as a free, spontaneous
energy; it is supposed to receive its first excitation by a shock or impulse
from without. Against this shock the ‘Ego’ will, it is assumed, react, and
only through this reaction does it first become conscious of itself. Mean-
while, the nature of the impulse remains a stranger beyond our pale: and
the ‘Ego’, with something else always confronting it, is weighted with a con-
dition. Fichte, in consequence, never advanced beyond Kant’s conclusion,
that the finite only is knowable, while the infinite transcends the range
of thought. What Kant calls the thing-by-itself, Fichte calls the impulse
from without—that abstraction of something else that ‘I’, not otherwise
describable or definable than as the negative or non-Ego in general.
(3) Wallace [1873], p. 69; translating Hegel [1830], pp. 128 f.
Fichte ought to have produced at least one effect on the method of logic.
One might have expected that the general laws of thought, the usual stock-
in-trade of logicians, or the classification of notions, judgements, and syl-
logisms, would be no longer taken merely from observation and so only
empirically treated, but be [[derived]] from thought itself. If thought is to be
capable of proving anything at all, if logic must insist upon the necessity
of proofs, and if it proposes to teach the theory of demonstration, its first
care should be to give a reason for its own subject-matter, and to see that
it is necessary.
Comment. As regards the double square brackets: I replaced “deduced” by “derived”
to translate “abgeleitet” into English, mainly to avoid a confusion with the term
“deduction” (“Deduktion”) in Kant’s philosophy. To my mind, Hegel is not concerned
with a question of justification here, but one of definition.
The pure thinking of itself, the identity of the subject and the object,
in the form I = I is the principle of the Fichtean system; and if one sticks
immediately and solely to this principle, as in the Kantian philosophy to
the transcendental principle which lies at the bottom of the deduction of
the categories, then one obtains the boldly expressed true principle of spec-
ulation. So soon, however, as speculation steps out of the concept which it
sets up of itself, and shapes itself into the system, it leaves itself and its
principle and does not return into the same.
Comment. Note the phrase “so soon”; cf. also the motto to the present part C: the
spirit tries to reach its concept.
(5) Nisbet [1991], p. 40; translating Hegel [1821], p. 57.
[[I]]n Fichte [[. . .]] ‘I’, as the unbounded (in the first proposition of his Theory
of Knowledge [Wissenschaftslehre]), is taken purely and simply as some-
thing positive (and thus as the universality and identity of the understand-
ing). Consequently, this abstract ‘I’ for itself is supposed to be the truth;
and limitation — i.e. the negative in general, whether as a given external
limit or as an activity of the ‘I’ itself — is therefore something added to it
(in the second proposition). — The further step which speculative philoso-
phy had to take was to apprehend the negativity which is immanent within
the universal or the identical, as in the ‘I’[[.]]
64b. Hegel on Kant as regards the antinomies, their solution, and

dialectic. The issue of the antinomies is a central one in the relation between He-
gel and Kant. It marks as much a dividing line as it marks their common concern.
In this section I shall focus on some of Hegel’s remarks concerning the treatment
of the antinomies in Kant. In general, Hegel acknowledged the basic idea of Kant’s
transcendental idealism, in particular the emergence of antinomies as a necessary
consequence of reason, but he was dissatisfied with Kant’s evaluation of the antin-
omies.
satz).
The principles of the metaphysical philosophy gave rise to the belief that,
when cognition lapsed into contradictions, it was a mere accidental aberra-
tion, due to some subjective mistake in argument and inference. According
to Kant, however, thought has a natural tendency to issue in contradictions
or antinomies, whenever it seeks to apprehend the infinite.
Comment. Note a certain infidelity in Hegel’s account of Kant’s position: the point
for Kant is not the struggle to apprehend the infinite, but absolute totality or ab-
solute completeness. The point is also made by Zermelo against Cantor’s criticism
of Kant’s antinomies. It is worthwhile having a look at quotations 68.27 in this
context.
Kant was the first definitely to signalize the distinction between Reason
and Understanding. The object of the former, as he applied the term, was
the infinite and unconditioned, of the latter the finite and conditioned. Kant
did valuable service when he enforced the finite character of the cognitions
of the understanding founded merely upon experience, and stamped their
contents with the name of appearance. But his mistake was to stop at the
purely negative point of view, and to limit the unconditionality of Reason
to an abstract self-sameness, without any shade of distinction. It degrades
Reason to a finite and conditioned thing, to identify it with a mere step-
ping beyond the finite and conditioned range of understanding. The real
infinite, far from being a mere transcendence of the finite, always involves
the absorption of the finite into its own fuller nature.
Kant rated dialectic higher—and this is among his greatest merits—
for he freed it from the seeming arbitrariness which it possesses from the
standpoint or ordinary thought and exhibited it as a necessary function of
reason. Because dialectic was held to be merely the art of practising decep-
tions and producing illusions, the assumption was made forthwith that it is
only a spurious game, the whole of its power resting solely on concealment
of the deceit and that its results are obtained only surreptitiously and are
a subjective illusion. True, Kant’s expositions in the antinomies of pure
reason, when closely examined [[. . .]] do not indeed deserve any great praise;
but the general idea on which he based his expositions and which he vindi-
cated, is the objectivity of the illusion and the necessity of the contradiction
which belongs to the nature of thought determinations: primarily, it is true,
with the significance that these determinations applied by reason to things
in themselves; but their nature is precisely what they are in reason and
with reference to what is intrinsic or in itself. This result, grasped in its
positive aspect, is nothing else but the inner negativity of the determina-
tions as their self-moving soul, the principle of all natural and spiritual life.
But if no advance is made beyond the abstract negative aspect of dialectic,
the result is only the familiar one that reason is incapable of knowing the
infinite[[.]]
Comment. Cf. also Hegel [1816], p. 337, for a similar formulation.
It is an infinite merit of the Kantian philosophy [[. . .]] to have given the
impetus to the restoration of logic and dialectic in the sense of the exami-
nation of the determinations of thought in and for themselves. The [[object]],
as it is without thought and the concept, is a [[Vorstellung]] or perhaps a

name; it is in the determinations of thought and the Notion that it is
what it is. [[Therefore, they are, indeed, all that counts]]; they are the true
[[object]] and content of reason, and [[that which one understands otherwise
by object]] and content in distinction from them [[holds]] only through them
and in them. It must not therefore be considered the fault of an [[object]] or
of cognition that these determinations, through their constitution and an
external connexion, show themselves dialectical.
(5) Taubeneck [1990], pp. 64 f; translating Hegel [1817], p. 45.
[[T]]he thought that the contradiction which is posited in reason by the cat-
egories of the understanding is essential and necessary should be regarded
as one of the deepest, most important, and progressive achievements of
modern philosophy[[; although]] this is represented in the Critique of Pure
Reason as if the contradiction were not in the concepts themselves, but
emerges only in their application to the unconditional.
Hegel holds against Kant that his approach is far from being either conclusive
or exhaustive. In particular, he criticizes his evaluation of the antinomies in three
respects. The first concerns the reason of the antinomies.
The explanation offered by Kant alleges that the contradiction does not
affect the object in its own proper essence, but attaches only to the Reason
which seeks to comprehend it.
In this way the suggestion was broached that the contradiction is oc-
casioned by the subject-matter itself, or by the intrinsic quality of the cat-
egories. And to offer the idea that the contradiction introduced into the
world of Reason by the categories of Understanding is inevitable and essen-
tial was to make one of the most important steps in the progress of Modern
Philosophy. But the more important the issue thus raised the more trivial
was the solution. Its only motive was an access of tenderness to the things
of the world. The blemish of contradiction, it seems, could not be allowed
to mar the essence of the world; but there could be no objection to attach
it to the thinking Reason, to the essence of mind. Probably nobody will feel
disposed to deny that the phenomenal world presents contradictions to the
observing mind; meaning by ‘phenomenal’ the world as it presents itself to
the senses and understanding, to the subjective mind. But if a comparison
is instituted between the essence of the world and the essence of the mind,
it does seem strange to hear how calmly and confidently the modest dogma
has been advanced by one, and repeated by others, that thought or Reason,
and not the World, is the seat of contradiction.
Comment. Note that Hegel does not say that it is the object as it is in itself which is
contradictory. Such a claim would not be compatible with his other statements about
things in themselves. Apparently, he is at pains to find the right words without being
in any way definitive: “object in its own proper essence”, German: “Gegenstand an
und für sich”; “subject-matter”, German: “Inhalt selbst”; “the essence of world” and
“the World”, German: “das weltliche Wesen” (by translating “das weltliche Wesen”
differently at different points, Wallace is likely to add confusion to the already
difficult text). Compare quotation 66.7 (3): “thought in its very nature is dialectical”.
(2) Johnston and Struthers [1929], p. 208; translating Hegel [1812], pp. 231 f.
[[I]]n the apagogic detour we see that the very assertion which is to be
its result already occurs. Thus the proof might be more briefly be put as
follows:—
Let it be assumed that substance do not consist of simple parts, but are
composite. Now all composition can be thought away (since it is a merely
contingent relation); it being removed, therefore, no substances remain un-
less they consist of simple parts. But we must have substances, since we
assumed their existence; everything must not vanish, something must re-
main; for we have presupposed that some such persistent entity (which we
called substance) exists. Therefore this something must be simple.
The Thesis and the Antithesis [[of Kant’s antinomy regarding the spa-
tial infinity of the world]] and their proofs [[. . .]] represent nothing but the
contrary assertions that there is a limit, and that this limit equally is can-
celled, that the limit has a beyond, but that it is in relation with the latter
and must pass over to it; which again gives rise to such a limit which is no
limit.
The solution of these antinomies [[. . .]] is transcendental, that is, it con-
sists in the assertion of the ideality of space and time as forms of intuition,—
in this sense, that the world in itself is not in contradiction with itself and
does not transcend itself, but that consciousness in its intuition and in
the relation of intuition to understanding and reason is a self-contradictory
essence. But it is an excessive tenderness towards the world to remove from
it the contradictions and to transfer it into Spirit or Reason, allowing it to
remain there unresolved. In point of fact it is Spirit which is strong enough
to support the contradiction, but it is also Spirit which knows how to re-
solve it. As for the so-called world (whether it be called objective or real
world, or, according to the transcendental idealism, subjective intuition and
sensuousness determined by the category of understanding), it never and
nowhere is without the contradiction, but, since it cannot support it, is
subject to arising and passing away.
Comment. There is an interesting aside regarding this passage: Derrida [1972], p. ix,
quotes the first paragraph of it, without mentioning that Hegel refers to Kant’s an-
tinomies. His10 quotation ends in the middle of the first sentence of the second para
with a full stop: “The solution . . . is transcendental, that is.” For readers who are
prone to view this as a particular subtlety of Derrida’s strategy of “deconstruction”,
I wish to recall quotation 56.2 (2) in these materials, where Kaufmann deals with
“quilt quotations”.
10 It might still be the poetry of Derrida’s translator; I haven’t checked it.
(4) U. Petersen [2002], p. 854, translating Hegel [1830], p. 138; cf. Wallace [1873],
p. 76.
If thought and appearance do not perfectly correspond to one another,
one has first of all the choice to regard the one or the other as the deficient
one. In the Kantian idealism, in so far as it concerns the reasonable, the
deficiency is pushed on the thoughts, so that, therefore, they are supposed
to be insufficient, because they are not adequate to the perceived and to
a consciousness which limits itself to the extent of perception, in which as
such the thoughts are not to be met with. The content of the thoughts for
itself is not addressed in this.
It is no escape to turn round and explain that Reason falls into contradiction
only by applying the categories. For this application of the categories is
maintained to be necessary, and Reason is not supposed to be equipped
with any other forms but the categories for the purpose of cognition.
Comment. This is a very central point: how serious does Kant take the exclusiveness
and constitutiveness of the categories, if he is not prepared to accept consequences
such as contradictions.
(6) Taubeneck [1990], p. 65; translating Hegel [1817], pp. 46 f.
It is the greatest inconsistency to admit, on the one hand, that the
understanding knows only appearances, and to claim on the other hand
that this knowledge is absolute, by such statements as: “Cognition can go
no further”; “Here is the natural and absolute limit of human knowledge.”
For any limit or lack is only recognized through comparison with the idea
of the whole and the complete. It is thoughtless, therefore, not to see that
[[this very]] designation of something as finite or limited contains the proof
of the actuality and presence of the infinite and unlimited.
Comment. I replaced “precisely” by the double bracketed this very which I prefer
strongly to render “eben” into English. Cf. my remarks regarding preferred ways of
proficient speakers of a natural language towards the end of §56.
Remark 64.7. Hegel’s use of the phrase “an und für sich” or “selbst” or combined
as “an und für sich selbst” is notoriously obscure and I shall not attempt any in-
terpretation but only point to a passage in the phenomenology of spirit regarding
the “Sache selbst”.11 Miller [1977], pp. 246 ff, translates it as “heart of the matter”
and then as “matter in hand”, apparently trying to work out a difference that is not
accounted for in Hegel’s original choice of words.
The second respect as to which Kant’s evaluation of the antinomies is not

correct, according to Hegel, concerns the number of antinomies. Hegel claims that
not only the concepts of pure reason lead to antinomies but all concepts.
11 Hegel [1807], p. 315.

[[T]]he Antinomies are not confined to the four special objects taken from
Cosmology: they appear in all objects of every kind, in all conceptions,
notions, and Ideas. To be aware of this and to know objects in this property
of theirs makes a vital part in a philosophical theory. For the property thus
indicated is what [[furthermore determines itself as the dialectical moment
of]] logic.
Comment. The question of the realm and the number of antinomies is fairly cru-
cial for deciding between Kant and Hegel.12 Main problem: to make sense of the
view that antinomies appear in all objects et cetera. As regards the double square
brackets: Wallace translated “was weiterhin sich als das d i a l e k t i s¸h e Moment
des Logischen bestimmt” as “what we shall afterwards describe as the Dialectical
influence in Logic”. Sounds cruel to me.
Kant wanted to give his four cosmological antinomies a show of complete-
ness by the principle of classification which he took from his schema of the
categories. But profounder insight into the antinomial, or more truly into
the dialectical nature of reason demonstrates any Notion whatever to be
a unity of opposed moments to which, therefore, the form of antinomial
assertions could be given.
Comment. Note the phrase “antinomial, or more truly . . . dialectical nature of rea-
son” which I take as an indication that my linking together antinomies and dialectic
in later part of this treatise is indeed in the spirit of Hegel.
Further, Kant did not take up the antinomy in the Notions themselves, but
in the already concrete form of cosmological determinations. In order to
possess the antinomy in its purity and to deal with it in its simple Notion,
the determinations of thought must not be taken in their application to and
entanglement in the general idea of the world, of space, time, matter, etc;
this concrete material must be omitted from consideration of these deter-
minations which it is powerless to influence and which must be considered
purely on their own account since they alone constitute the essence and the
ground of the antinomies.
The Kantian antinomies on closer inspection contain nothing more than
the quite simple categorical assertion of each of the two opposed moments
of a determination, each being taken on its own in isolation from the other.
But at the same time this simple categorical, or strictly speaking assertoric
statement is intended to produce a semblance of proof and to conceal and
disguise the merely assertoric character of the statement[[.]]
12 Cf. quotation 62.4 in these materials as regards Kant’s assertion.

§ 65. Logic, truth, and the content of thought determinations
Hegel’s move to extend the transcendental approach to logic proper, as presented

by my choice of quotations in the foregoing paragraph, constitutes an extremely
radical step and meets with serious difficulties when it comes to a realization. Even
sympathetic readers of Hegel’s Logic will probably admit that the bite Hegel took
was too big for him to chew. Hegel was seriously disadvantaged by the lack of
a powerful logical theory at his time. So I have to be content with ideas being
formulated in a conceptual framework which in no way can stand the test of modern
standards of logic.
In the present paragraph I shall concentrate on what Hegel says about truth,
the role of logic and the dialectic of form and content. I regard this as the crucial
part of Hegel’s philosophy which has to be understood before it is possible to make
sense of the concept of dialectic as a conflict of form and content.
65a. The thing in itself as regards logic. Provided that Hegel was right
in principle, as far as the number of antinomies is concerned, i.e., every category is
antinomical, this would mean that Kant’s argumentation breaks down at a crucial
point. According to Kant, the antinomies are possible because no empirical object
correlates to the (conceptual) object in question. If, however, antinomies occur
already in as basic a concept as being, we are forced to concluding that even being
is a speculative concept, i.e., reaches beyond all possible experience.
But what is left, if we extend the notion of the thing in itself to every concept,
i.e., a thing in itself beyond any determination? Not even nothing! Since nothing is
still committed to the notion of “things” (no-thing), i.e., to a shadow of determina-
tion.
I begin with a quotation concerning subjective and objective to illustrate Hegel’s
position on the basic gnoseological situation.
satz).
Objective and subjective are convenient expressions in current use, the
employment of which may easily lead to confusion. Up to this point, the
discussion has shown three meanings of objectivity. First, it means what has
external existence, in distinction from which the subjective is what is only
supposed, dreamed, & c. Secondly, it has the meaning, attached to it by
Kant, of the universal and necessary, as distinguished from the particular,
subjective, and occasional element which belongs to our sensations. Thirdly,
[[. . .]] it means the thought-apprehended essence of the existing thing, in
contradistinction from what is merely our thought, and what consequently
is still separated from the thing itself, as it exists in independent essence.
Comment. This strikes me as one of the most difficult points in Hegel. It should be
seen in connection with what Hegel says about logic in the next section which might
give an idea of how to make sense of the idea that the notion can be something like
the essential reality of things.
§ 65. LOGIC, TRUTH, AND THE CONTENT OF THOUGHT DETERMINATIONS 857

[[P]]ure science presupposes liberation from the opposition of consciousness.
It contains thought in so far as this is just as much the object in its own
self, or the object in its own self in so far as it is equally pure thought.
Comment. Miller translated “die Sache an sich selbst” as “the object in its own self”.
[[W]]hat at first to ordinary reflection is, as content, divorced from form,
cannot in fact be formless, cannot be devoid of inner determination; if it
were, then it would be only vacuity, the abstraction of the thing-in-itself;
that on the contrary, the content in its own self [[in ihm selbst]] possesses
form, in fact it is through form alone that it has soul and meaning, and that
it is form itself which is transformed only into the semblance of a content,
hence into the semblance of something external to this semblance. With this
introduction of the content into the logical treatment, the subject matter
is not things but their import, the Notion of them.
Comment. The German original would be quite suitable to help understanding
Hegel’s notion of “der Sache selbst”: what Miller translates as “their import ” is in
German “die Sache”. The last sentence in my translation: With this introduction
of the content into the logical consideration, it is not the things, but the affair
(“Sache”), the notion of things, which becomes the object.
Things are called ‘in themselves’ in so far as abstraction is made from all
being-for-other, which means simply, in so far as they are thought devoid
of all determination, as nothings. In this sense, it is of course impossible
to know what the thing-in-itself is. For the question: what ? demands that
determinations be assigned; but since the things of which they are to be
assigned are at the same time supposed to be things in-themselves, which
means, in effect, to be without any determination, the question is thought-
lessly made impossible to answer, or else only an absurd answer is given.
The thing-in-itself is the same as that absolute of which we know nothing
except that in it all is one. What is in these thing-in-themselves, there-
fore, we know quite well; they are as such nothing but truthless, empty
abstractions.
Comment. Compare this quotation with quotation 61.34 (1) where Kant speaks of
the noumenon as abstract from our mode of intuiting it. Hegel puts determination
where Kant talks about intuition. Note the similarity to Berkeley’s “conceiving a
thing which is unconceived ”; it is impossible to describe the undetermined.
The Thing-in-itself [[. . .]] expresses the object when we leave out of
sight all that consciousness makes of it, all its emotional aspects, and all
specific thoughts of it. It is easy to see what is left—utter abstraction, total
emptiness, only described still as an ‘outer world’—the negative of every
image, feeling, and definite thought. Nor does it require much penetration
to see that this caput mortuum is still only a product of thought, such as
accrues when thought is carried on to abstraction unalloyed[[.]] [[. . .]] [[O]]ne
can only read with surprise the perpetual remark that we do not know the
Thing-in-itself. On the contrary there is nothing we can know so easily.
Comment. Note Hegel’s formulation “only a product of thought”.
So much for the negative sense of the thing in itself. With regard to its positive
sense consider the following quotation.
Quotation 65.3. Miller [1969], p. 121; translating Hegel [1812], pp. 137 f.
What, however, the thing-in-itself is in truth, what truly is in itself, of this
logic is the exposition, in which however something better than an abstrac-
tion is understood by ‘in-itself’, namely, what something is in its Notion;
but the Notion is concrete within itself, is comprehensible simply as Notion,
and as determined within itself and connected whole of its determinations,
is cognizable.
Comment. This is a pretty tricky thing.
65b. Hegel’s conception of logic and metaphysics. In view of Hegel’s

basic line of extending the transcendental approach to logic proper it seems only
consistent that logic is going to take the place of metaphysics. This means, against
Kant, that logic is to become an organon. In the present section I shall illustrate
Hegel’s view of logic and metaphysics.
Quotation 65.4. Wallace [1873], p. 25; translating Hegel [1830], p. 66.

Logic is the science of the pure Idea; pure, that is, because the Idea is in
the abstract medium of Thought.
So far, this can still be seen in the tradition of Kant’s idea of a transcendental
logic: the categories, as not being applied to empirical content of the senses, but
pure. It will be clear, however, that Hegel goes beyond Kant.13
Logic might have been defined as the science of thought, and of its laws and
characteristic forms. But thought, as thought, constitutes only the general
medium, or qualifying circumstance, which renders the Idea distinctively
logical. If we identify the Idea with thought, thought must not be taken in
the sense of method or form, but in the sense of the self-developing totality
of its laws and peculiar terms. These laws are the work of thought itself,
and not a fact which it finds and must submit to.
Logic [[. . .]] coincides with Metaphysics, the science of things set and held in
thoughts—thought accredited able to express the essential reality of things.
13 Although this would qualify Hegel as an early representative of Beyondism, he was sensible
enough not to actually write anything entitled “Beyond Kant”.


What we are dealing with in logic is not a thinking about something which
exists independently as a base for our thinking and apart from it, nor forms
which are supposed to provide mere signs or distinguishing marks of truth;
on the contrary, the necessary forms and self-determinations of thought are
the content and the ultimate truth itself.
The question, why the self-determinations of thought are the ultimate truth
itself, is something that Hegel never answers convincingly.
Quotation 65.6. Miller [1969], p. 50; translating Hegel [1812], pp. 45 f.

[[L]]ogic is to be understood as the system of pure reason, as the realm
of pure thought. This realm is truth as it is without veil and in its own
absolute nature. It can therefore be said that this content is the exposition
of God as he is in his eternal essence before the creation of nature and a
finite mind.
Comment. Hegel merges reason, logic, and pure thought. In view of Kant’s state-
ment that “reason does not really generate concepts”,14 his emphasis of the constitu-
tive role of logic attracts my attention. But formulations in terms of the “exposition
of God . . . before the creation” are likely to contribute to the view of Hegel’s Logic
as a philosophy of the absolute spirit, which is certainly not in the spirit of my
enterprise.
Quotations 65.7. (1) Wallace [1873], p. 40; translating Hegel [1830], pp. 88 f (Zu-
satz).
Logic is usually said to be concerned with forms only and to derive the
material for them from elsewhere. But this ‘only’, which assumes that the
logical thoughts are nothing in comparison with the rest of the contents, is
not the word to use about forms which are the absolutely real ground of ev-
erything. Everything else rather is an ‘only’ compared with these thoughts.
Comment. Compare this with Kant: “nothing but forms of thought”, “merely logical
faculty”.15
When logic is taken as the science of thinking in general, it is under-
stood that this thinking constitutes the mere form of a cognition, that logic
abstracts from all contents and that the so-called second constituent belong-
ing to cognition, namely matter, must come from elsewhere; and that since
this matter is absolutely independent of logic, this latter can provide only
the formal conditions of genuine cognition and cannot in its own self [[selbst]]
contain any real truth, nor even be the pathway to real truth because just
that which is essential in truth, its content, lies outside logic.
14 Kant [1781] (A), p. 409; cf. quotation 61.17 in these materials.
15 Kant [1787] (B), p. 305; cf. quotation 61.31 in these materials.

It is declared to be an abuse when logic, which is supposed to be merely
a canon of judgement, is regarded as an organon for the production of
objective insights. The notions of reason in which we could not but have an
intimation of a higher power and a profounder significance, no longer possess
a constitutive character as do the categories, they are mere Ideas; certainly,
we are quite at liberty to use them, but by these intelligible entities in which
all truth should be completely revealed, we are to understand nothing more
than hypotheses, and to ascribe absolute truth to them would be the height
of caprice and foolhardiness, for they—do not occur in any experience.
Comment. This quotation represents well Hegel’s opposition against Kant’s view of
logic as expressed, for example, in quotations 59.5 (3) and (4) in these materials.

The Logic of the Notion is usually treated as a science of form only, and
understood to deal with the form of notion, judgement, and syllogism as
form, without in the least touching the question whether anything is true.
The answer to that question is supposed to depend on the content only.
If the logical forms of the notion were really dead and inert receptacles of
conceptions and thoughts, careless of what they contained, knowledge about
them would be an idle curiosity which the truth might dispense with. On
the contrary they really are, as forms of the notion, the vital spirit of the
actual world. That only is true of the actual which is true in virtue of these
forms, through them and in them. As yet, however, the truth of these forms
has never been considered or examined on their own account any more than
their necessary interconnections.
Comment. This is of particular relevance for me. Note also: “necessary interconnec-
tions”.

Logic, in the essential meaning of speculative philosophy, emerges in
place of what is often dismissed as a separate science and is otherwise called
metaphysics. The nature of logic and the standpoint on which scientific
cognition posit[[ed]] itself [[receives its closer provisional clarification from]]
the nature of metaphysics and then by [[the]] critical philosophy, through
which metaphysics reached its final stage. [[. . .]] Metaphysics, after all, is
only outmoded in relation to the history of philosophy. For itself, it is in
general as it has become, particularly in recent times: the view from the
understanding of the objects of reason.
Comment. The word “metaphysics” in the third line is emphasized in the German
original. As regards the double square brackets: the German “nähere vorläufige“, i.e.,
closer provisional, turns up in Taubeneck’s translation as “more precisely” which
does in no way justice to the “vorläufig”.
(7) U. Petersen [2002], p. 861; translating Hegel [1817], p. 36; [1830], pp. 195 f.
The mere logic of the understanding is contained in the speculative logic
and can be made out of that one immediately; it needs nothing for that,
but to omit from it the Dialectical and Reasonable; thus it turns into what
the ordinary logic is, a story of variously compiled thought determinations
which in their finitude count for something infinite.
Comment. Other translations can be found in Taubeneck [1990], p. 58 and Wallace
[1873], p. 120.
Speculative Logic contains all previous Logic and Metaphysics: it preserves
the same forms of thought, the same laws and objects—while at the same
time remodelling and expanding with wider categories.
Quotation 65.8. Petersen [2002], p. 861; translating Hegel [1830], p. 118.16

The basic illusion in scientific empiricism is always this, that it employs the
metaphysical categories of matter, force, not to mention One, Many, Uni-
versal, and Infinite and so on, furthermore it continues to reason along the
line of these categories, thereby assumes and applies the forms of reasoning,
and doesn’t know in all this, that in this way it itself contains and carries
on metaphysics, and employs those categories and their connections in a
completely uncritical and unconscious way.
65c. Truth, logic, form, and content. The question that has evolved: in
which way can logic be more but purely formal? How can form acquire content?
Hegel’s idea of the dialectic of form and content is probably the most difficult topic
that I have do deal with.
Quotations 65.9. (1) Miller [1969], pp. 44 f; translating Hegel [1812], p. 38.
Hitherto, the Notion of logic has rested on the separation, presupposed
once and for all in the ordinary consciousness, of the content of cognition
and its form, or of truth and certainty. First, it is assumed that the material
of knowing is present on its own account as a ready-made world apart from
thought, that thinking on its own is empty and comes as an external form
to the said material, fills itself with it and only thus acquires a content and
so becomes real knowing.
Further, these two constituents—for they are supposed to be related
to each other as constituents, and cognition is compounded from them in
a mechanical or at best chemical fashion—are appraised as follows: the ob-
ject is regarded as something complete and finished on its own account,
something which can entirely dispense with thought for its actuality, while
thought on the other hand is regarded as defective because it has to com-
plete itself with a material and moreover, as a pliable indeterminate form,
has to adapt itself to its material. Truth is the agreement of thought with
the object, and in order to bring about this agreement—for it does not exist
16 Cf. Wallace [1873], p. 62, for a different translation.
on its own account— thinking is supposed to adapt and accommodate itself

to the object.
Thirdly, when the difference of matter and form, of object and thought
is not left in that nebulous indeterminateness but is taken more definitely,
then each is regarded as a sphere divorced from the other. Thinking there-
fore in its reception and formation of material does not go outside itself; its
reception of the material and the conforming of itself to it remains a modi-
fication of its own self[[seiner selbst]], it does not result in thought becoming
the other of itself; and self-conscious determining moreover belongs only to
thinking. In its relation to the object, therefore, thinking does not go out
of itself to the object; this, as a thing-in-itself, remains a sheer beyond of
thought.
The essential point to keep in mind about the opposition of Form and
Content is that the content is not formless, but has the form in its own self
[[in ihm selbst ]], quite as much as the form is external to it. There is thus a
doubling of form. At one time it is reflected into itself; and then is identical
with the content. At another time it is not reflected into itself, and then is
the external existence, which does not at all affect the content.
The form, when thus thought out in its purity, will have within itself the
capacity to determine itself, that is, to give itself a content, and that a
necessarily explicated content—in the form of a system of determinations
of thought.
The objective logic, then, takes the place 17 rather of former metaphysics
which was intended to be the scientific construction of the world in terms
of thought alone. If we have regard to the final shape in the elaboration of
this science, then it is first and immediately ontology whose place is taken
by objective logic—that part of this metaphysics which was supposed to
investigate the nature of ens in general; ens comprises both being and
essence, a distinction for which the German language has fortunately pre-
served different terms. But further, objective logic also comprises the rest
of metaphysics in so far as this attempted to comprehend with the forms of
pure thought particular substrata taken primarily from figurate conception
[[Vorstellung]], namely the soul, the world and God; and the determinations
of thought constituted what was essential in the mode of consideration.
Logic, however, considers these forms free from those substrata, from the
subjects of figurate conception; it considers them, their nature and worth,
in their own proper character.
Logic being the science of the absolute form, this formal science, in order
to be true, must possess in its own self [[an ihm selbst]] a content adequate to
its form; and all the more, since the formal element of logic is the pure form,
17 Emphasized in German original but not in Miller’s translation.
and therefore the truth of logic must be the pure truth itself. Consequently
this formal science must be regarded as possessing richer determinations
and a richer content and as being infinitely more potent in its influence on
the concrete than is usually supposed.
[[T]]he form of the positive judgement is accepted as something perfectly
correct in itself, the question whether such a judgement is true depending
solely on the content. Whether this form is in its own self [[an und für sich]]
a form of truth, whether the proposition it enunciates, the individual is a
universal is not inherently dialectical, is a question that no one thinks of
investigating. It is straightway assumed that this judgement is, on its own
account, capable of containing truth and that the proposition enunciated
by any positive judgement is true, although it is directly evident that it
lacks what is required by the definition of truth, namely, the agreement of
the Notion and its object; if the predicate, which here is the universal, is
taken as the Notion, and the subject, which is the individual, is taken as
the object, then the one does not agree with the other. But if the abstract
universal which is the predicate falls short of constituting a Notion, for a
Notion certainly implies something more, and if, too, a subject of this kind
is not yet much more than a grammatical one, how should the judgement
possibly contain truth seeing that either its Notion and object do not agree,
or it lacks both Notion and object?
Comment. My problem with this as well as the other quotations in this lot is that
Hegel does not seem to go beyond characterizing his goal in negative contrast to
some form of understanding; or he remains metaphorical; what does it mean to
investigate whether a “form is in its own self (“an und für sich”) a form of truth”?
It is only ordinary abstract understanding which takes the terms of me-
diation and immediacy each by itself absolutely, to represent an inflexible
line of distinction, and thus draws upon its own head the hopeless task of
reconciling them.
(7) Hegel [1830], p. 115; translated in Wallace [1873], p. 59.
In der spekulativen Philosophie ist der Verstand zwar ein Moment, aber ein
Moment, bei welchem nicht stehen geblieben wird.
satz).
Truth is at first taken to mean that I know something is. This is truth,
however, only in reference to consciousness; it is formal truth, bare correct-
ness. Truth in the deeper sense consists in the identity between objectivity
and the notion. It is in this deeper sense of truth that we speak of a true
state, or of a true work of art. These objects are true, if they are what they
ought to be, i.e. if their reality corresponds to their notion.
(2) Miller [1977], p. 22; translating Hegel [1807], p. 38

‘True’ and ‘false’ belong among those determinate notions which are
held to be inert and wholly separate essences, one here and one there,
each standing fixed and isolated from the other, with which it has nothing
in common. Against this view it must be maintained that truth is not a
minted coin that can be given and pocketed ready-made.
Dogmatism as a way of thinking, whether in ordinary knowing or in the
study of philosophy, is nothing else but the opinion that the True consists
in a proposition which is a fixed result, or which is immediately known. To
such questions as, When was Caesar born?, or How many feet were there in
a stadium?, etc. a clear-cut answer ought to be given, just as it is definitely
true that the square on the hypotenuse is equal to the sum of the squares on
the other two sides of a right-angled triangle. But the nature of a so-called
truth of that kind is different from the nature of philosophical truths.
It is one of the fundamental assumptions of dogmatic Logic that Qual-
itative judgments such as ‘The rose is red’ or ‘is not red’ can contain truth.
Correct they may be, i.e. in the limited circle of perception, of finite con-
ception and thought: that depends on the content, which likewise is finite,
and, on its own merits, untrue. Truth, however, as opposed to correctness,
depends solely on the form, viz. on the notion as it is put and the reality
corresponding to it.
Comment. This strikes me as pretty unclear, so I go to the Zusatz.
In common life the terms truth and correctness are often treated as
synonymous: we speak of the truth of a content, when we are only thinking
of its correctness. Correctness, generally speaking, concerns only the formal
coincidence between our conception and its content, whatever the constitu-
tion of this content may be. Truth, on the contrary, lies in the coincidence
of the object with itself, that is, with its notion. That a person is sick, or
that some one has committed a theft, may certainly be correct. But the
content is untrue. A sick body is not in harmony with the notion of body,
and there is a want of congruity between theft and the notion of human
conduct. These instances may show that an immediate judgement in which
an abstract quality is predicated of an immediately individual thing, how-
ever correct it may be, cannot contain truth. The subject and predicate of
it do not stand to each other in the relation of reality and notion.
We may add that the untruth of the immediate judgement lies in the
incongruity between its form and content. To say ‘This rose is red’ involves
(in virtue of the copula ‘is’) the coincidence of subject and predicate. The
rose however is a concrete thing, and so is not red only: it has also an odour,
a specific form, and many other features not implied in the predicate red.
The predicate on its part is an abstract universal, and does not apply to
the rose alone. There are other flowers and other objects which are red too.
The subject and predicate in the immediate judgement touch, as it were,
only in a single point, but do not cover each other.
Comment. What Hegel seems to assume is that the copula expresses an identity
relation which is not satisfied in the immediate judgement. According to Russell
[1918], p. 245, note *, “The confusion of the [[‘is’ of predication and that of identity]]
is essential to the Hegelian conception of identity-in-difference.”
(6) Miller [1977], pp. 24 f; translating Hegel [1807], pp. 40 ff.

42. As for mathematical truths, we should be even less inclined to regard
anyone as a geometer who knew Euclid’s theorems outwardly by rote, with-
out knowing their proofs, without, as we might say, to point the contrast,
knowing them inwardly. [[. . .]] Yet, even in mathematical cognition, the es-
sentiality of the proof does not have the significance and nature of being
a moment of the result itself; when the latter is reached, the demonstra-
tion is over and has disappeared. [[. . .]] The movement of the mathematical
proof does not belong to the object, but rather is an activity external to
the matter in hand. [[. . .]]
43. In mathematical cognition, insight is an activity external to the
thing; it follows that the true thing is altered by it. The means employed,
construction and proof, no doubt contain true propositions, but it must
[[just as much]] be said that the content is false. [[For]] example the triangle
is dismembered, and its parts consigned to other figures, whose origin is
allowed by the construction upon the triangle. Only at the end is the triangle
we are actually dealing with reinstated. During the procedure it was lost
to view, appearing only in fragments belonging to other figures. [[. . .]]
44. But what is really defective in this kind of cognition concerns the
cognitive process itself, as well as its material. As regards the former, we do
not, in the first place, see any necessity in the construction. Such necessity
does not arise from the notion of the theorem; it is rather imposed, and the
instruction to draw precisely these lines when infinitely many others could
be drawn must be blindly obeyed without our knowing anything beyond
except that we believe that this will be to the purpose in carrying out the
proof. In retrospect, this expediency also becomes evident, but it is only an
external expediency, because it becomes evident only after the proof. This
proof, in addition, follows a path that begins somewhere or other without
indicating as yet what relation such a beginning will have to the result that
will emerge. In its progress it rakes up these particular determinations and
relations, and lets others alone, without its being immediately clear what
the controlling necessity is; an external purpose governs this procedure.
45. The evident character of this defective cognition of which math-
ematics is proud, and on which it plumes itself before philosophy, rests
solely on the poverty of its purpose and the defectiveness of its stuff, and
is therefore of a kind that philosophy must spurn.
Comments. (α) This sounds like Hegel was talking about deduction in a Hilbert
style calculus.18 But Gentzen’s work has shown, at least for a certain part of math-
ematics, that this is not necessarily so. Consider the possibility of reduction meth-
ods in symmetric sequential calculi, as presented in section 21c in the tools, for
example, for the case of sentential logic. Cf. also Wittgenstein’s comment on math-
ematics in quotation 86.5 (6) in these materials, in particular regarding equation,
but beware of his misjudgment as regards the nature of logic in the preceding quo-
tation 86.5 (5).
(β) I replaced “none the less” in Miller’s translation by “just as much“ (in double
square brackets) to render “eben so sehr” into English.
I finish this section with a quotation which comes up to the expectations of

Hegel as the apostle of the Absolute.
Quotation 65.11. Wallace [1873], p. 41; translating Hegel [1830], p. 90 (Zusatz).

God alone is the thorough harmony of notion and reality. All finite things
involve an untruth: they have a notion and an existence, but their existence
does not meet the requirements of the notion. For this reason they must
perish and [[this is how]] the incompatibility between their notion and their
existence becomes manifest.
Comment. It may sound pretty wild, this piece, but if you look closer . . . But I’m
not going to interpret or even try to defend Hegel here. Anyway, it’s only a Zusatz.
65d. Content of thought determinations on their own. In this section

my attention will be focused on Hegel’s notion of Denkbestimmungen (thought deter-
minations). I begin with Hegel’s delineation of his notion of thought determinations
from a view which sees them as some sort of boxes which can be filled or not with
objects, but which remain indifferent towards them, a certain common sense view.
Quotations 65.12. (1) Wallace [1873], p. 37; translating Hegel [1830], pp. 83 f (Zu-
satz).
To speak of thought or objective thought as the heart and soul of the world,
may seem to be ascribing consciousness to the things of nature. We feel
repugnance against making thought the inward function of things, especially
as we speak of thought as marking the divergence of man from nature.
I would seem necessary, therefore, if we use the term thought at all, to
speak of nature as the system of unconscious thought, or, to use Schelling’s
expression, a petrified intelligence. And in order to prevent misconception,
‘thought-form’ or ‘thought-type’ should be substituted for the ambiguous
term thought.
Comment. Wallace translates the German “Denkbestimmung” as “thought-form” or
“thought-type”. Literally translated it is “thought determination”. Since this is how
Miller translates it too, I prefer the latter.
18 Cf. the comment by Hodges, for instance, on Hilbert style calculi: barbarously unintuitive;
quotation 96.18 in these materials.


It will now be understood that Logic is the all-animating spirit of all
the sciences, and its categories the spiritual hierarchy. They are the heart
and centre of things: and yet at the same time they are always on our lips,
and apparently at least, perfectly familiar objects. But things thus familiar
are usually the greatest strangers. Being, for example, is a category of pure
thought: but to make ‘is’ an object of investigation never occurs to us.
Comment. Recall “thought as the heart and soul of the world” of the foregoing
quotation. Here is an indication of what Hegel might have had in mind: it is indeed
the copula, bare predication that Hegel wants to make an object of research (not
something like ‘existence’); and this ‘is’ is universal; recall Parmenides and interpret
him as expressing that thinking and predicating is the same.

The finite [[. . .]] subsists in reference to its other, which is its negation and
presents itself as its limit. Now thought is always in its own sphere; its
relations are with itself, and it is its own object. In having a thought for
object, I am at home with myself. The thinking power, the ‘I’, is therefore
infinite, because, when it thinks, it is in relation to an object which is
itself. Generally speaking, an object means a something else, a negative
confronting me. But in the case where thought thinks itself, it has an object
which is at the same time no object[[.]]
[[In so far, then, as]] subjective thought is our very own, innermost, act,
and the objective notion of things constitutes [[the subject matter itself]],
we cannot go outside this our act, we cannot stand above it, and just as
little can we go beyond the nature of things. We can, however, disregard
the latter determination; in so far as it coincides with the first it would
yield a relation of our thoughts to the object [Sache], but this would be a
valueless result because it would imply that the thing, the object, would
be set up as a criterion for our notions and yet for us the object can be
nothing else but our notions of it. The way in which the critical philoso-
phy understands the relationship of these three terms is that we place our
thoughts as a medium between ourselves and the objects, and that this
medium instead of connecting us with the objects rather cuts us off from
them. But this very view can be countered by the simple observation that
these very things which are supposed to stand beyond us and, at the other
extreme, beyond the thought referring to them, are themselves figments
of subjective thought, and as wholly indeterminate they are only a single
thought-thing—the so-called thing-in-itself of empty abstraction.
Comment. The words in double brackets replace “Since, therefore,” and “their es-
sential import”, respectively; the German words are “Insofern also” and “die Sache
selbst”.
Remark 65.13. It seems appropriate at this point to issue a warning, in particular,

as this is written in English and Hegel wrote in German. The English language uses
“is” in “there is” in the sense of “there exists”. This use of “is” is not necessarily
shared by other languages. In French, for instance, one says “there it has” (“il y
a”), in German one says “it gives” (“es gibt”). It would be a gross misinterpretation
to read being in the present context of Hegel’s logic!Hegel’s as existing. Observe,
however, that the function of “is” as copula may be quite a different story.19
The forms of thought are, in the first instance, displayed and stored
in human language. [[. . .]] In all that becomes something inward for men,
an image or conception as such, into all that he makes his own, language
has penetrated, and everything that he has transformed into language and
expresses in it contains a category—concealed, mixed with other forms or
clearly determined as such, so much is logic his natural element, indeed his
own peculiar nature.
It is an advantage when a language possesses an abundance of logical ex-
pressions, that is, specific and separate expressions for the thought deter-
minations themselves; many prepositions and articles denote relationships
based on thought[[.]] [[. . .]] These particles, however, play quite a subordi-
nate part having only a slightly more independent form than the prefixes
and suffixes, inflections and the like. It is much more important that in
a language the categories should appear in the form of substantives and
verbs and thus be stamped with the form of objectivity. In this respect
German has many advantages over other modern languages; some of its
words even possess the further peculiarity of having not only different by
opposite meanings so that one cannot fail to recognize a speculative spirit
of the language in them: it can delight a thinker to come across such words
and to find the union of opposites naïvely shown in the dictionary as one
word with opposite meanings, although this result of speculative thinking
is nonsensical to the understanding.
In life, the categories are used ; from the honour of being contemplated for
their own sakes they are degraded to the position where they serve in the
creation and exchange of ideas involved in intellectual exercise on a living
content. [[. . .]] [[I]]n this process the import and purpose, the correctness and
truth of the thought involved, are made to depend entirely on the subject
matter itself and the thought determinations are not themselves credited
with any active part in determining the content.
19 Cf. also Heidegger [1953], pp. 54 f, quotation 67.9 (2) in these materials, regarding the
etymology of the word “sein”.


[[O]]n this basis, the determinations of thought have the significance of forms
which are only attached to the content, but are not the content itself.
[[T]]he notion is what is mediated through itself and with itself. It is a
mistake to imagine that the objects which form the content of our mental
ideas come first and that our subjective agency then supervenes, and by
the [[. . .]] operation of abstraction, and by colligating the points possessed
in common by the objects, frames notions of them. Rather the notion is
the genuine first; and things are what they are through the action of the
notion, immanent in them, and revealing itself in them.
Quotations 65.15. (1) Johnston and Struthers [1929], pp. 39 f; translating Hegel
[1812], p. 511.
In its positive expression A=A this law is nothing more than empty
tautology. It has therefore rightly been observed that this Law of Thought
is without content and leads no further. Those therefore are stranded upon
empty Identity who take it to be a truth in itself, and are in the habit of re-
peating that Identity is not [[Difference]], but that Identity and [[Difference]]
are different. They do not see that they are themselves here saying that
Identity is different, for they say that Identity is different from [[Difference]];
and since this must at the same time be admitted to be the nature of Iden-
tity, their assertion implies that Identity has the quality of being different
not externally but in its very nature.
Comment. Maybe it is asking for too much; but perhaps the reader can see a faint
similarity to diagonalization in this weird piece of argumentation.
The form of the proposition which expresses Identity contains more
[[. . .]] than Identity simple and abstract: it contains this pure movement
of Reflection, in which the Other figures only as Show and as immediate
disappearance. A is a beginning which imagines a different term that is to
be reached; but this term is never reached; A is—A; the difference is only
a disappearance, and the movement withdraws into herself.
satz).
The forms of thought must be studied in their essential nature and complete
development: they are at once the object of research and the action of
that object. Hence they examine themselves: in their own action they must
determine their limits, and point out their defects. This is that action of
thought, which will hereafter be specially considered under the name of
Dialectic[[.]]
(2) Wallace [1873], pp. 40 f; translating Hegel [1830], p. 89 (Zusatz).

To ask if a category is true or not, must sound strange to the ordinary
mind: for a category apparently becomes true only when it is applied to a
given object, and apart from this application it would seem meaningless to
inquire into the truth. But this is the very question on which everything
turns. We must however in the first place understand clearly what we mean
by Truth. In common life truth means the agreement of an object with our
conception of it. We thus presuppose an object to which our conception
must conform. In the philosophical sense of the word, on the other hand,
truth may be described, in general abstract terms, as the agreement of a
thought-content with itself. This meaning is quite different from the one
given above. At the same time the deeper and philosophical meaning of
truth can be partially traced even in the ordinary usage of language. Thus
we speak of a true friend; by which we mean a friend whose manner of
conduct accords with the notion of friendship. In the same way we speak
of a true work of Art. Untrue in this sense means the same as bad, or
self-discordant. In this sense a bad state is an untrue state; and evil and
untruth may be said to consist in the contradiction subsisting between the
function or notion and the existence of the object. Of such a bad object we
may form a correct representation, but the import of such representation
is inherently false.
The study of truth, or, as it is here explained to mean, consistency,
constitutes the proper problem of logic. In our everyday mind we are never
troubled with questions about the truth of the forms of thought. We may
also express the problem of logic by saying that it examines the forms of
thought touching their capability to hold truth.
To ask if being, existence, finitude, simplicity, complexity, etc. are no-
tions intrinsically and independently true, must surprise those who believe
that a question about truth can only concern propositions (as to whether
a notion is or is not with truth to be attributed, as the phrase is, to a
subject), and that falsehood lies in the contradiction existing between the
subject in our ideas, and the notion to be predicated of it. Now as the
notion is concrete, it and every character of it in general is essentially a
self-contained unity of distinct characteristics. If truth then were nothing
more than the absence of contradiction, it would be first of all necessary in
the case of every notion to examine whether it, taken individually, did not
contain this sort of intrinsic contradiction.
Comment. In other words: concepts can also occur as (grammatical) subjects and
if truth is the absence of contradiction between subject and predicate, a concept
would have to be first of all predicated of itself to see if there is no contradiction.
(Notice that this is what a type theoretical approach is trying to avoid.)
§ 66. CONTRADICTIONS AND THEIR AUFHEBUNG 871
§ 66. Contradictions and their Aufhebung
As I mentioned before, Kant’s employment of the antinomies for a reductio ad

absurdum to prove the transcendental ideality of space and time had to give way to
a new assessment of contradictions in Hegel’s speculative philosophy. The present
paragraph attempts to give an impression of Hegel’s approach; weird as it may look,
it has some very important aspects for my enterprise.
66a. Hegel on contradictions. What Hegel says explicitly about contradic-
tions is notoriously obscure. In this section I shall present a variety of quotations
to give an impression of Hegel’s ideas. I begin with a formulation of what may be
called the ordinary view.
Quotation 66.1. Johnston and Struthers [1929], p. 67; translating Hegel [1812],
p. 546.
Ordinarily Contradiction is removed, first of all from things, from the
existent and the true in general; and it is asserted that there is nothing
contradictory. Next it is shifted into subjective reflection, which alone is
said to posit it when it relates and compares. But really—it is said—it
does not exist even in this reflection, for it is impossible to imagine or to
think anything contradictory. Indeed, Contradiction, both in actuality and
in thinking reflection, is considered an accident, a kind of abnormality or
paroxysms of sickness which will soon pass away.

Die Leerheit des Gegensatzes von sogenannten kontradiktorischen Begriffen
hatte ihre volle Darstellung in dem so zu sagen grandiosen Ausdruck eines
allgemeinen Gesetzes, daß j e d e m Dinge von a l l e n so entgegengesetzten
Prädikaten das eine zukomme und das andere nicht, so daß der Geist sey
entweder weiß oder nicht weiß, gelb oder nicht gelb u. s. f. ins Unendliche.
So far, this may not sound too hairy. The problem lies with what Hegel appar-
ently considers a rejection of this view.
Quotation 66.3. Johnston and Struthers [1929], p. 67; translating Hegel [1812],
pp. 546 f.
With regard to the assertion that Contradiction does not exist, that it is
non-existent, we may disregard this statement. In every experience there
must be an absolute determination of Essence—in every actuality as well
as in every concept. [[. . .]] But ordinary experience itself declares that at
least there are a number of contradictory things about, contradictory ar-
rangements and so forth, the contradiction being present in them and not
merely in an external reflection. But it must further not be taken only as
an abnormality which occurs just here and there: it is the Negative in its
essential determination, the principle of all self-movement, which consists of
nothing else but an exhibition of Contradiction. External, sensible motion
is itself its immediate existence.
Quotations 66.4. (1) Wallace [1873], p. 167; translating Hegel [1830], pp. 268 f.
[[T]]he maxim of identity, reads: Everything is identical with itself, A=A:
and, negatively, A cannot at the same time be A and not A. This maxim,
instead of being a true law of thought, is nothing but the law of abstract
understanding. The propositional form itself contradicts it: for a proposi-
tion always promises a distinction between subject and predicate; while the
present one does not fulfil what its form requires.
The Maxim of Excluded Middle is the maxim of the definite understanding,
which would fain avoid contradiction, but in doing so falls into it.
Contradiction is the very moving principle of the world: and it is ridiculous
to say that contradiction is unthinkable. The only thing correct in that
statement is that contradiction is not the end of the matter, but cancels
itself.
Motion is a paradigmatic form in which contradictions exist. Here are some

more quotations.
Something moves, not because at one moment it is here and at another
there, but because at one and the same moment it is here and not here,
because in this ‘here’, it at once is and is not. The ancient dialecticians
must be granted the contradictions that they pointed out in motion; but it
does not follow that there is no motion, but on the contrary, that motion
is existent contradiction itself.
(2) Johnston and Struthers [1929], p. 68; translating Hegel [1812], p. 547.
[[I]]it is only in so far as it contains a Contradiction that anything moves
and has impulse and activity.
Something [[. . .]] has life only in so far as it contains Contradiction. But if
an existent something cannot in its positive determination also encroach on
its negative, cannot hold fast the one in the other and contain Contradic-
tion with in itself, then it is not living unity, or Ground, but perished in
Contradiction.
At this moment the planet stands in this spot, but implicitly it is the
possibility of being in another spot; and that possibility of being otherwise
the planet brings into existence by moving.
In motion, time posits itself spatially as place, but this indifferent spa-
tiality becomes just as immediately temporal: the place becomes another
[[. . .]]. This difference of time and place is, as the difference of their absolute
unity and their indifferent content, a difference of bodies, which hold them-
selves apart from each other yet equally seek their unity through gravity;—
general gravitation.
Comment. In the context of this quotation it may be worthwhile looking at quota-
tion 93.9 (3) in these materials and its comment.
Das, wodurch sich der Begriff selbst weiter leitet, ist das [[. . .]] N e-
g a t i v e, das er in sich selbst hat; dieß macht das wahrhaft Dialektische
aus. Die D i a l e k t i k, die als ein abgesonderter Theil der Logik betrachtet
und in Ansehung ihres Zwecks und Standpunkts, man kann sagen, gänzlich
verkannt worden, erhält dadurch eine ganz andere Stellung. — Auch die
p l a t o n i s c h e Dialektik hat selbst im Parmenides, und anderswo ohnehin
noch direkter, Theils nur die Absicht, beschränkte Behauptungen durch
sich selbst aufzulösen und zu widerlegen, Theils aber überhaupt das Nichts
zum Resultate.
66b. Negative dialectic. In quotations 63.18, Hegel can be found to work
out a distinction between a negative and a positive reason, or the dialectical and
the speculative; in other place he also speaks of negative and positive dialectic.
satz).
To illustrate the presence of Dialectic in the spiritual world, especially in
the provinces of law and morality, we have only to recollect how general
experience shows us the extreme of one state or action suddenly shifting
into its opposite: a Dialectic which is recognized in many ways in common
proverbs. Thus summum jus summa injuria, which means that to drive an
abstract right to its extremity is to do a wrong. In political life, as every
one knows, extreme anarchy and extreme despotism naturally lead to one
another. The perception of Dialectic in the province of individual Ethics
is seen in the well-known adages: Pride comes before a fall; Too much wit
outwits itself. Even feeling, bodily as well as mental, has its Dialectic. Every
one knows how the extremes of pain and pleasure pass into each other: the
heart overflowing with joy seeks relief in tears, and the deepest melancholy
will at times betray its presence by a smile.
(2) Wallace [1873], pp. 15 f; translating Hegel [1830], pp. 55 f.
To see that thought in its very nature is dialectical, and that, as under-
standing, it must fall into contradiction—the negative of itself—will form
one of the main lessons of logic. When thought grows hopeless of ever
achieving, by its own means, the solution of the contradiction which it has
by its own action brought upon itself, it turns back to those solutions of
the question with which the mind had learned to pacify itself in some of
its other modes and forms. Unfortunately, however, the retreat of thought
has led it, as Plato noticed even in his time, to a very uncalled-for hatred
of reason (misology); and it then takes up against its own endeavours that
hostile attitude of which an example is seen in the doctrine that ‘immedi-

ate’ knowledge, as it is called, is the exclusive form in which we become
cognizant of truth.
Comment. This is a statement explicit enough: “thought in its very nature is di-
alectical”. Cf. also Marx in quotation 67.29 (9) in the next chapter.
(3) Hegel [1809/11], p. 214.
Die Vernunft ist negative oder dialektische, insofern sie das Uebergehen
einer Verstandesbestimmung des Seins in ihre entgegengesetzte aufzeigt.
Gewöhnlich erscheint das Dialektische so, daß von Einem Subject zwei ent-
gegengesetzte Prädicate behauptet werden. Das reinere Dialektische besteht
darin daß von einem Prädicat eine Verstandesbestimmung aufgezeigt wird,
wie sie an ihr selbst eben so sehr das Entgegengesetzte ihrer selbst ist, sie
sich also in sich aufhebt.
(4) Hegel [1807], p. 78.20
Diese d i a l e k t i s c h e Bewegung, welche das Bewußtseyn an ihm selbst,
sowohl an seinem Wissen, als an seinem Gegenstande ausübt, i n s o f e r n
i h m d e r n e u e w a h r e G e g e n s t a n d daraus e n t s p r i n g t, ist ei-
gentlich dasjenige, was E r f a h r u n g genannt wird.
Compared with the simple layout of Kant’s presentation of the antinomies,

Hegel’s argumentation is likely to evoke repulsion. I shall try to illuminate his con-
ception of contradiction with the help of the first contradiction of the Science of
Logic, the identity and non-identity of Being and Nothing. According to Hegel, Be-
ing and Nothing are one and the same with respect to their contents. However,
there is also a difference, but this has not yet come to manifestation.
The most general concept is that of being. Whatever we may say about some-
thing, we say — implicitly — that it is.21 Hegel maintained that the concept of
being leads to contradictions; his reasoning, however, is in general not considered
conclusive, to put it politely. I quote the relevant passages fully so the reader can
get an impression.
Being, pure being without any further determination. In its indeterminate
immediacy it is equal only to itself. It is also not unequal relatively to an
other; it has no diversity within itself nor any with a reference outwards.
It would not be held fast in its purity if it contained any determination or
content which could be distinguished in it or by which it could be distin-
guished from an other. It is pure indeterminateness and emptiness. There is
nothing to be intuited in it, if one can speak here of intuiting; or, it is only
this pure intuiting itself. Just as little is anything to be thought in it, or it
is equally only this empty thinking. Being, the indeterminate immediate, is
in fact nothing, and neither more nor less than nothing.
20 A translation can be found in Miller [1977], p. 55.
21 Watch out! To be does not mean to exist. See also remark 65.13 in these materials.
Comment. Note the way nothing is introduced here. Keep in mind, however, that
Hegel wrote German and the correspondent to the first nothing in the above text
had a small initial “n” whereas the following two a capital one. What would a
German translation of the “nothing” in the first sentence of this comment look like:
small or capital initial letter?
Nothing, pure nothing: it is simply equality with itself, complete emptiness,
absence of all determination and content—undifferentiatedness in itself. In
so far as intuiting or thinking can be mentioned here, it counts as a dis-
tinction whether something or nothing is intuited or thought. To intuit or
think nothing has, therefore, a meaning; both are distinguished and thus
nothing is (exists) in our intuiting thinking; or rather it is empty intui-
tion and thought itself, and the same empty intuition or thought as pure
being. Nothing is, therefore, the same determination or rather absence of
determination, and thus altogether the same as, pure being.
Comment. Notice the implicit criterion of identity: “the same determination”, where,
however, determination does not seem to be taken intensionally but rather exten-
sionally, i.e., what falls under it.
Pure being and pure nothing are, therefore, the same. What is the truth
is neither being nor nothing, but that being—does not pass over but has
passed over—into nothing, and nothing into being. But it is equally true
that they are not undistinguished from each other, that, on the contrary,
they are not the same, that they are absolutely distinct, and yet that they
are unseparated and inseparable and that each immediately vanishes in its
opposite. Their truth is, therefore, this movement of the immediate van-
ishing of the one in the other: becoming, a movement in which both are
distinguished, but a difference which has equally immediately resolved it-
self.
Comments. (α) I do not understand Hegel as saying that the identity of being
and nothingness already constitutes a contradiction. Common sense may well feel
uneasy, but a contradiction only arises when we realize that being and nothingness
are distinct as well.
(β) I take this as an example of Hegel’s notion of the truth of a category which was
only abstractly described, i.e., without an example, in quotation 65.16 (2).
(1) The proposition that Being and Nothing is the same seems so para-
doxical to the imagination or understanding, that it is perhaps taken for
a joke. And indeed it is one of the hardest things thought expects itself
to do: for Being and Nothing exhibit the fundamental contrast in all its
immediacy—that is, without the one term being invested with any attribute
which would involve its connection with the other. The attribute, however,
[[. . .]] is implicit in them—the attribute which is just the same in both. So
far the deduction of their unity is completely analytical: indeed the whole
progress of philosophizing in every case, if it be a methodical, that is to say

a necessary, progress, merely renders explicit what is implicit in a notion.
Comment. As regards the last subclause, compare Findlay [1981], p. 132, quotation
69.25 (2) in these materials. The German phrase is: “das Setzen desjenigen, was
in seinem Begriff schon enthalten ist.” The problem is, of course, how this is to be
done; in particular given Hegel’s dislike of mathematical methods and the fact that
logic has turned essentially mathematical since Hegel’s time.
It may be helpful to add a few of Hegel’s comments.
It is the common opinion that being is rather the sheer other of nothing
and that nothing is clearer than their absolute difference, and nothing seems
easier than to be able to state it. But it is equally easy to convince oneself
that this is impossible, that it is unsayable. Let those who insist that being
and nothing are different tackle the problem of stating in what the difference
consists. If being and nothing had any determinateness by which they were
distinguished from each other then, as has been observed, they would be
determinate being and determinate nothing, not the pure being and pure
nothing that here they still are. Their difference is therefore completely
empty, each of them is in the same way indeterminate; the difference, then,
exists not in themselves but [[only]] in a third, in subjective opinion. Opinion,
however, is a form of subjectivity which is not proper to an exposition of
this kind.
Comment. In view of McKeon’s remarks on Aristoteles’ conception of dialectic in
quotation 67.4 in these materials, observe the notion of opinion coming in here.
The “subjective” before “opinion” is a generous addition by the translator, the
German original only has “Meinen”. I do not at all agree with this addition, even
though Hegel goes on to call opinion a form of subjectivity; it seems to me to
preempt the possibility that this form of subjectivity may have another side, one
which eventually becomes objective. After all, this is what the present study is all
about.
The distinction between Being and Nought is, in the first place, only
implicit, and not yet actually made: they only ought to be distinguished.
A distinction of course implies two things, and that one of them possesses
an attribute which is not found in the other. Being however is an absolute
absence of attributes, and so is Nought. Hence the distinction between the
two is only meant to be; it is a quite nominal distinction, which is at the
same time no distinction.
Ex nihilo nihil fit —is one of those propositions to which great impor-
tance was ascribed in metaphysics. In it is to be seen either only the empty
tautology: nothing is nothing; or, if becoming is supposed to possess an
actual meaning in it, then, since from nothing only nothing becomes, the
proposition does not in fact contain becoming, for in it nothing remains

nothing. Becoming implies that nothing does not remain nothing but passes
into its other, into being.
Comment. I do not follow Hegel in this particular point; my emphasis is on the
difference of the nothing which comes out of nothing.
We cannot be expected to meet on all sides the perplexities which such
a logical proposition produces in the ordinary consciousness, for they are
inexhaustible. Only a few of them can be mentioned. One source among
others of such perplexity is that the ordinary consciousness brings with it
to such an abstract logical proposition, conceptions of something concrete,
forgetting that what is in question is not such concrete something but only
the pure abstractions of being and nothing and that these alone are to be
held firmly in mind.
This is what I called the exceedingly delicate affair. Dealing with this affair,
Hegel warns us to be careful. Our normal conception of language does not apply to
it without certain restrictions, e.g., we are not allowed to claim the identity of being
and nothingness, without adding that they are non-identical at the same time.22
On the other hand, we think we know that being and nothing is not the same,
according to Hegel. But this is the problem.
It may also be worthwhile in the present context to go back to what Hegel said
regarding Parmenides in quotations 63.10.
Quotations 66.10. (1) Hegel [1802], p. 297.

Wie sind synthetische Urtheile a priori möglich? Dieses Problem drückt
nichts Anderes aus, als die Idee, daß in dem synthetischen Urtheil Subject
und Prädikat, jenes das Besondere, dieses das Allgemeine, jenes in der Form
des Seyns, dieß in der Form des Denkens, — dieses Ungleichartige zugleich
a priori, d. h. absolut identisch ist.
Comment. Note the mention of Being and Thinking in this context.
(2) Hegel [1809/11], pp. 148 f.
§. 13.
In der Logik wird das Urtheil seiner reinen Form nach betrachtet, ohne
Rücksicht auf irgend einen bestimmten, empirischen Inhalt. Die Urtheile
unterscheiden sich durch das Verhalten, welches das Subject und das Prä-
dicat in der Rücksicht zu einander hat, in wiefern ihre Beziehung durch und
in dem Begriff oder eine Beziehung der Gegenständlichkeit auf den Begriff
ist. Von der Art dieser Beziehung hängt die höhere oder absolute Wahrheit
des Urtheils ab. Die Wahrheit ist Uebereinstimmung des Begriffs mit sei-
ner Gegenständlichkeit. Im Urtheil fängt diese Darstellung des Begriffs und
seiner Gegenständlichkeit, somit das Gebiet der Wahrheit, an.
22 Cf. quotations 66.15 below.
§. 14.
Indem das Urtheil die Darstellung eines Gegenstandes in den verschie-
denen Momenten des Begriffs ist, so ist es umgekehrt die Darstellung des
Begriffs in seinem Dasein, nicht sowohl wegen des bestimmten Inhalts, den
die Begriffsmomente haben, als weil sie im Urtheil aus ihrer Einheit treten.
Wie das ganze Urtheil den Begriff in seinem Dasein darstellt, so wird dieser
Unterschied auch wieder zur Form des Urtheils selbst. Das Subject ist der
Gegenstand und das Prädikat die Allgemeinheit desselben, welches ihn als
Begriff ausdrücken soll. Die Bewegung des Urtheils durch seine verschiede-
nen Arten hindurch erhebt diese Allgemeinheit in die höhere Stufe, worin
sie dem Begriff so entsprechend wird, als sie überhaupt sein kann, insofern
sie überhaupt Prädicat ist.
66c. Aufhebung, positive dialectic, and the speculative step. One place
where Hegel explicitly talks about dialectic is § 81 in the Encyclopedia (Hegel [1830]),
partly represented in quotation 66.11 (4) below.
Quotation 66.11. (1) Miller [1969], p. 73; translating Hegel [1812], p. 78.
The beginning is not pure nothing, but a nothing from which something
is to proceed; therefore being, too, is already contained in the beginning.
The beginning, therefore, contains both, being and nothing, is the unity of
being and nothing; or is non-being which is at the same time being, and
being which is at the same time non-being.
The analysis of the Beginning [[. . .]] yields the concept of the unity of
Being and Not-being, or (in a more reflected form) the unity of the state of
being differentiated and of being undifferentiated, or the identity of identity
and non-identity. This concept might be considered as the first or purest
(that is, most abstract) definition of the Absolute; which in fact it would
be were we concerned with the forms of definitions and the name of the
Absolute. In this sense, this abstract concept would be the first definition
of the Absolute, and all further determinations and developments would be
richer and more closely determinate definitions of it.
Comment. Note the grammatical form (subjunctive): “it would be were we con-
cerned”.

Being itself and the special sub-categories of it which follow, as well as
those of logic in general, may be looked upon as the definitions of the
Absolute, or metaphysical definitions of God: at least the first and third
category in every [[sphere]] may—the first, where the thought-form of the
[[sphere]] is formulated in its simplicity, and the third, being the return from
differentiation to a simple self-reference.
Comment. Wallace translates “Sphäre” as “triad”. I am not prepared to participate

in that asinine talk of a ‘dialectical triad’; in particular not, since Hegel never used
it himself, not even as an Eselsbrücke.23
[[I]]n its true and proper character, Dialectic is the very nature and essence of
everything predicated by mere understanding—the law of things and of the
finite as a whole. [[. . .]] In the first instance, Reflection is that movement out
beyond the isolated predicate of a thing which gives it some reference, and
brings out its relativity, while still in other respects leaving it its isolated
validity. But by Dialectic is meant the indwelling tendency outwards by
which the one-sidedness and limitation of the predicates of understanding
is seen in its true light, and shown to the negation of them. For anything
to be finite is just to suppress itself and put itself aside. Thus understood
the Dialectical principle constitutes the life and soul of scientific progress,
the dynamic which alone gives immanent connection and necessity to the
body of science[[.]]
Comment. The double square brackets indicate the dropping of the sentence “Di-
alectic is different from ‘Reflection” ’, which has no counterpart in the German
original. Altogether the translation is very free. I add my own — very literal —
translation of the part before the above quotation:
Dialectic is commonly regarded as an external art which produces through
capriciousness a confusion in certain determinate concepts and a mere ap-
pearance of contradiction, so that not these determinations, but this ap-
pearance be a right one and the one to be understood in comparison rather
the true.
This last part of the last sentence is actually quite unclear in the original German.
When those determinations of thought which are only external forms are
truly considered in themselves, this can only result in demonstrating their
finitude and the untruth of their supposed independent self-subsistence,
that their truth is the Notion. Consequently, the science of logic in dealing
with the thought determinations which in general run through our mind
instinctively and unconsciously—and even when they become part of the
language do not become objects of our attention—will also be a recon-
struction of those which are singled out by reflection and are fixed by it as
subjective forms external to the matter and import of the determinations
of thought.
No subject matter is so absolutely capable of being expounded with a
strictly immanent plasticity as is thought in its own necessary development;
no other brings with it this demand in such a degree; in this respect the
Science of Logic must surpass even mathematics[[.]]
23 Literally: pons asinorum, bridge of asses; used in German in the sense of aide-mémoire.
Gives rise to a nice variation of Murphy’s law: provide an ‘Eselsbrücke’ and all the asses will take
it beastly seriously.
Comment. The problem is, of course, the kind of “immanent plasticity” (“imma-
nent plastisch”) in question. I suspect it will be in Hegel’s case that of a German
philosophy professor at the beginning of the nineteenth century, in other words a
thoroughly subjective notion, bordering on idiosyncrasy.
It is [[. . .]] in the grasping of opposites in their unity or of the positive in the
negative, that speculative thought consists. It is the most important aspect
of dialectic, but for thinking which is as yet unpractised and unfree it is the
most difficult.
Comment. The “of dialectic” in the last sentence is an addition of the translator,
but it does make the sentence more suitable for quotation.
(7) Hegel [1817], p. 36; Hegel [1830], p. 195 (unchanged)
The dialectic has a positive result, because it has a determinate content, or
because its result is truly not the empty, abstract nothing, but the negation
of determinate determinations, which are contained in the result for the
very reason that this is not an immediate nothing, but a result.
Comment. Other translations can be found in Taubeneck [1990], p. 58, and Wallace
[1873], p. 119.
Dialectic is one of those ancient sciences that have been most misunderstood
in the metaphysics of the moderns, as well as by popular philosophy in
general, ancient and modern alike. [[. . .]] Dialectic has often been regarded
as an art, as though it rested on a subjective talent and did not belong to
the objectivity of the Notion. [[. . .]] It must be regarded as a step of infinite
importance that dialectic is once more recognized as necessary to reason,
although the result to be drawn from it must be the opposite of that arrived
at by Kant.
Quotations 66.12. (1) U. Petersen [2002], p. 880; translating Hegel [1812], p. 119;
Miller, p. 106.
The balance into which coming-to-be and ceasing-to-be posit them-
selves, is in the first place becoming itself. But this equally goes together in
calm unity. In it, being and nothing are only as something vanishing; but
becoming as such is only through their distinguishedness. Their vanishing,
therefore, is the vanishing of becoming, or the vanishing of the vanishing
itself. Becoming is a posture-less unrest which collapses into a calm result.
This could also be expressed thus: becoming is the vanishing of being
in nothing, and of nothing in being, and the vanishing of being and nothing
generally; but at the same time it is based on their difference. It therefore
contradicts itself in itself, because it unites something that is opposed to
itself; such a unity, however, destroys itself.
This result is the having24 -vanished, but not as nothing; as that it
would only be a relapse into one of the already sublated determinations,
24 To do justice to the German, this should be “being”; but grammar has its own demands.
not a result of the nothing and the being. It is the into calm simplicity born
unity of being and nothing.
Comment. The last bit must be a translator’s nightmare. The notorious camel of
a grammatical construction that won’t go through the eye of the translator.25 Ger-
man: “Es ist die zur ruhigen Einfachheit gewordene Einheit des Seyns und Nichts.”
Miller translates: “It is the unity of being and nothing which has settled into a stable
oneness.”
In Becoming the Being which is one with Nothing, and the Nothing which
is one with Being, are only vanishing factors; [[. . .]]. Thus by its inherent
contradiction Becoming collapses into the unity in which the two elements
are absorbed. This result is accordingly Being Determinate (Being there
and so).
Comment. The translation is hopeless. To make things worse, Wallace added a
sentence, here indicated by the double brackets, which has no counterpart at all in
the German original: “they are and they are not.” This may please dialetheists, but
Hegel is not its author; it is a product of the poetry of the translator — or, perhaps
more so, of the pampering-the-English-reader-attitude.
(3) Miller [1969], pp. 107; translating Hegel [1812], p. 121.
The more precise meaning and expression which being and nothing
receive, now that they are moments, is to be ascertained from the consider-
ation of determinate being as the unity in which they are preserved. Being
is being, and nothing is nothing, only in their contradistinction from each
other; but in their truth, in their unity, they have vanished as these deter-
minations and are now something else. Being and nothing are the same; but
just because they are the same they are no longer being and nothing, but
now have a different significance. In becoming they were coming-to-be and
ceasing-to-be; in determinate being, a differently determined unity, they are
again differently determined moments. This unity now remains their base
from which they do not again emerge in the abstract significance of being
and nothing.
Something is the negation of the negation in the form of being; for this
second negation is the restoring of the simple relation to self; but with this,
something is equally the mediation of itself with itself. Even in the simple
form of something, then still more specifically in being-for-self, subject, and
so on, self-mediation is present; it is present even in becoming, only the
mediation is quite abstract. In something, mediation with itself is posited,
in so far as something is determined as a simple identity. [[. . .]]
This mediation with itself which something is in itself, taken only as
negation of the negation, has no concrete determinations for its sides; it thus
collapses into the simple oneness which is being. Something is, and is, then,
25 My apologies to Austin.
also a determinate being [[Daseiendes]]; further, it is in itself also becoming,

which, however, no longer has only being and nothing for its moments. One
of these, being [[Sein]], is now determinate being [[Dasein]], and, further,
a determinate being [[Daseiendes]]. The second is equally a determinate
being [[Daseiendes]], but determined as a negative of the something—an
other [[Anderes]]. Something as a becoming is a transition [[Uebergehen]],
the moments of which are themselves somethings, so that the transition is
alteration [[Veränderung]]—a becoming which has already become concrete.
Comment. The translation doesn’t help; I added the German expression in some
cases in the hope that this might help to get a rough idea of what is going on. But I
do want to emphasize that I can’t see this kind of conceptual stew ever doing more
than give a few esoteric minds a feeling of sublime thrill.
Quotation 66.13. Miller [1969], p. 389; translating Hegel [1812], p. 481.

The truth of being is essence.
Being is the immediate. Since knowing has for its goal knowledge of
the true, knowledge of what being is in and for itself, it does not stop at
the immediate and its determinations, but penetrates it on the supposition
that at the back of this being there is something else, something other
than being itself, that this background constitutes the truth of being. This
knowledge is a mediated knowing for it is not found immediately with and
in essence, but starts from an other, from being, and has a preliminary
path to tread, that of going beyond being or rather of penetrating into it.
Not until knowing inwardizes, recollects [erinnert ] itself out of immediate
being, does it through this mediation find essence. The German language
has preserved essence in the past participle [gewesen] of the verb to be; for
essence is past—but timelessly past—being.
It is in this context that the notion of Aufhebung is introduced.
satz).
[[W]]e should note the double meaning of the German word aufheben (to
put by, or set aside). We mean by it (1) to clear away, or annul: thus, we
say, a law or a regulation is set aside; (2) to keep, or to preserve: in which
sense we use it when we say: something is well put by. This double usage of
language, which gives to the same word a positive and negative meaning,
is not an accident, and gives no ground for reproaching language as a cause
of confusion. We should rather recognize in it the speculative spirit of our
language rising above the mere ‘either—or’ of understanding.
Comment. The best translation of aufheben I can think of myself is to lift ; it reflects
a bit of the ambiguity of the German word; a ban, for instance, is lifted, i.e.,
removed; someone lifts her head (not to be confused with someone having his face
lifted); someone offers you a lift; a baby is lifted out of the pram, etc. In this it is
not inferior to tollere (cf. next quotation) or ‘sublate’, even though its ambiguity
does not coincide exactly with that of the German word “aufheben”; but there is an
ambiguity, one which, I think, can be extended without too much difficulty to include
the meaning of lifting a contradiction as removing it in one sense and raising it to
and preserving it on a higher level at the same time.26 The point for me here is to get
rid of the halo that surrounds the words “aufheben” and “Aufhebung”, in particular,
amongst non-German speaking philosophers, who seem to get a particular thrill out
of saying them with an air of sublime importance — and the stress on the wrong
syllable; to my mind they are just words which conceal the lack of a proper concept.

To sublate [[Aufheben]], and the sublated [[das Aufgehobene]] (that which ex-
ists ideally as a moment [[das Ideelle]]), constitute one of the most important
notions in philosophy. It is a fundamental determination which repeatedly
occurs throughout the whole of philosophy, the meaning of which is to be
clearly grasped and especially distinguished from nothing. What is sublated
[[Was sich aufhebt]] is not thereby reduced to nothing. Nothing is immedi-
ate; what is sublated, on the other hand, is the result of mediation; it is
a non-being but as a result which had its origin in a being. It still has,
therefore, in itself the determinateness from which it originates.
‘To sublate’ has a twofold meaning in the language: on the one hand it
means to preserve, to maintain, and equally it also means to cause to cease,
to put an end to. Even ‘to preserve’ includes a negative element, namely,
that something is removed from its immediacy and so from an existence
which is open to external influences, in order to preserve it. Thus what is
sublated is at the same time preserved; it has only lost its immediacy but is
not on that account annihilated, The two definitions of ‘to sublate’ which
we have given can be quoted as two dictionary meanings of this word. But it
is certainly remarkable to find that a language has come to use one and the
same word for two opposite meanings. It is a delight to speculative thought
to find in the language words which have in themselves a speculative mean-
ing; the German language has a number of such. The double meaning of
the Latin tollere (which has become famous through the Ciceronian pun:
tollendum est 27 Octavium) does not go so far; its affirmative determina-
tion signifies only a lifting-up. Something is sublated only in so far as it has
entered into unity with its opposite; in this more particular signification as
something reflected, it may fittingly be called a moment.
Comment. Admittedly, this is a difficult piece to translate, but I can’t say I am

impressed by what Miller has done.
I close this section with two sets of quotations which are meant to give an
impression of what Hegel said regarding speculative thought and the speculative
sentence, and the mathematical infinite.
26 According to my Concise Oxford Dictionary from 1976, p. 626: “Raise to higher position”;
this comes pretty close to “aufheben”.

27 The German edition has “esse”.
Since the Notion is the object’s own self, which presents itself as the coming-
to-be of the object, it is not a passive Subject inertly supporting the Acci-
dents; it is, on the contrary, the self-moving Notion which takes its de-
terminations back into itself. In this movement the passive Subject itself
perishes; it enters into the differences and the content, and constitutes the
determinateness, i.e. the differentiated content and its movement, instead
of remaining inertly over against it. The solid ground which argumentation
has in the passive Subject is therefore shaken, and only this movement itself
becomes the object. The Subject that fills its content ceases to go beyond it,
and cannot have any further Predicates or accidental properties. Conversely,
the dispersion of the content is thereby bound together under the self; it
is not the universal which, free from the Subject, could belong to several
others. Thus the content is, in fact, no longer a Predicate of the Subject,
but is the Substance, the essence and the Notion of what is under discus-
sion. Picture-thinking, whose nature it is to run through the Accidents or
Predicates and which, because they are nothing more than Predicates and
Accidents, rightly goes beyond them, is checked in its progress, since that
which has the form of a Predicate in a proposition is the Substance itself. It
suffers, as we might put it, a counter-thrust. Starting from the Subject as
though this were a permanent ground, it finds that, since the Predicate is
really the Substance, the Subject has passed over into the Predicate, and,
by this very fact, has been sublated; and, since in this way what seems to
be the Predicate has become the whole and the independent mass, thinking
cannot roam at will, but is impeded by this weight.
[[T]]he general nature of the judgement or proposition, which involves the
distinction of Subject and Predicate, is destroyed by the speculative propo-
sition, and the proposition of identity which the former becomes contains
the counter-thrust against that subject-predicate relationship.
(3) U. Petersen [2002], p. 884; translating Hegel [1812], pp. 98 f; based on Miller
[1969], p. 90.
[[I]]n so far as the proposition: being and nothing is the same expresses the
identity of these determinations, but, in fact, equally contains them both
as distinguished, the proposition contradicts itself in itself and dissolves
itself. If we record this more closely, then a proposition is posited here
which, considered more closely, has the movement of vanishing through
itself. Thereby, however, happens to its own self [[an ihm selbst]] that which
is to constitute its own peculiar content, namely, becoming.
Comment. My translation might not pamper the English reader, but I realize un-
dertones in Hegel’s German which I find missing in Miller’s translation. I don’t
know, if they come across in my very literal rendering of Hegel’s words into English
(“Hegelenglish”).
(4) Miller [1969], p. 90 (immediately following the preceding quotation); translating

Hegel [1812], p. 99.
The proposition thus contains the result, it is this in its own self [[an
sich selbst]]. But the fact to which we must pay attention here is the defect
that the result is not itself expressed in the proposition; it is an external
reflection which discerns it therein. In this connection we must at the outset,
make this general observation, namely that the proposition in the form of
a judgement is not suited to express speculative truths.
It is the form of simple judgement, when it is used to express specula-
tive results, which is very often responsible for the paradoxical and bizarre
light in which much of the recent philosophy appears to those who are not
familiar with speculative thought.
To help express the speculative truth, the deficiency is made good in the
first place by adding the contrary proposition: being and nothing are not the
same, which is also enunciated as above. But thus there arises the further
defect that these propositions are not connected, and therefore exhibit their
content only in the form of an antinomy whereas their content refers to one
and the same thing, and the determinations which are expressed in the two
propositions are supposed to be in complete union—a union which can only
be stated as an unrest of incompatibles, as a movement. The commonest
injustice done to a speculative content is to make it one-sided, that is, to give
prominence only to one of the propositions into which it can be resolved.
It cannot be denied that this proposition is asserted; but the statement is
just as false as it is true, for once one of the propositions is taken out of
the speculative content, the other must at least be equally considered and
stated.
Comment. It remains mysterious to me how, according to Hegel, a speculative con-
tent can be expressed adequately.
64. One difficulty which should be avoided comes from mixing up the
speculative with the ratiocinative methods, so that what is said of the
Subject at one time signifies its Notion, at another time merely its Predicate
or accidental property. The one method interferes with the other, and only
a philosophical exposition that rigidly excludes the usual way of relating
the parts of a proposition could achieve the goal of plasticity.
Speculative thought consists only in this, that thought holds fast Contra-
diction, and, in Contradiction, itself[[.]]
65. As a matter of fact, non-speculative thinking also has its valid rights
which are disregarded in the speculative way of stating a proposition. The
sublation of the form of the proposition must not happen only in an immedi-
ate manner, through the mere content of the proposition. On the contrary,
this opposite movement must find explicit expression; it must not just be
the inward inhibition mentioned above. This return of the Notion into itself
must be set forth. This movement which constitutes what formerly the proof
was supposed to accomplish, is the dialectical movement of the proposition
itself. This alone is the speculative in act, and only the expression of this
movement is a speculative exposition.
The mathematical infinite is interesting, first, because its introduction
has widened the scope of mathematics and has led to important results
therein; next it is remarkable because this science has not yet succeeded
in vindicating this use of it conceptually (“concept” being here taken in its
proper meaning).
[[M]]etaphysics, though it opposes them, cannot deny or invalidate the bril-
liant results of the employment of the mathematical infinite; while mathe-
matics cannot achieve clearness about the metaphysics of its own concept,
nor, therefore, about the derivation of the methods which the employment
of the infinite necessitates.
[[T]]he calculation of the infinite admits of, and demands, modes of proce-
dure which mathematics, when it operates with finite magnitudes, must
altogether reject, and at the same time it treats these infinite magnitudes
as finite Quanta, seeking to apply to the former those same methods which
are valid for the latter. The most important step in the development of
this science is to have imposed on transcendent determinations and the
treatment of these the form of ordinary calculation.
Comment. Cf. Poincaré’s question in quotation 74.2 (2).
In mathematics a magnitude is defined as that which can be increased or
diminished; in general as an indifferent limit. Now since the infinitely great
or small is that which cannot be increased or diminished, it is in fact no
longer a quantum as such.
Comment. With hindsight a paradigmatic point of failure, at least as regards the
infinitely great, and that according to book-lore.
66d. Dialectic, method, and encyclopedia. The present section is meant

to provide material regarding questions of Hegel’s method; if he had one at all, more
specifically, a dialectical method; if dialectic is a method at all, and the role of the
encyclopedic structure of Hegel’s system.
I begin with a quotation which may well serve as a key-note.
Quotation 66.17. Miller [1977], p. 43; translating Hegel [1807], p. 43.

70. Should anyone ask for a royal road to Science, there is no more
easy-going way than to rely on sound common sense; and for the rest, in
order to keep up with the times, and with advances in philosophy, to read
reviews of philosophical works, perhaps even to read their prefaces and first
paragraphs. For these preliminary pages give the general principles on which
everything turns, and the reviews, as well as providing historical accounts,
also provide the critical appraisal which, being a judgement, stands high
above the work judged. This common road can be taken in casual dress;
but [[in the robes of the high priest strides along]] the high sense for the
Eternal, the Holy, the Infinite [[ — ]] on a road that is [[much rather itself
the immediate being in the centre, the ingenuity of deeply original ideas
and high brainwaves]]. But just as profundity of this kind still does not
reveal the source of essential being, so, too, these sky-rockets of inspiration
are not yet the empyrean. True thoughts and scientific insight are only to
be won through the labour of the Notion. Only the Notion can produce
the universality of knowledge which is neither common vagueness nor the
inadequacy of ordinary common sense, but a fully developed, perfected
cognition; not the uncommon universality of a reason whose talents have
been ruined by indolence and the conceit of genius, but a truth ripened to
its properly matured form so as to be capable of being the property of all
self-conscious Reason.
Comment. I changed the part in the double brackets because Miller’s translation
didn’t seem to me to do justice to Hegel’s sarcasm.
Quotation 66.18. Hegel [1801], pp. 71 f.

Das Philosophiren, das sich nicht zum System konstruirt, ist eine beständige
Flucht vor den Beschränkungen, — mehr ein Ringen der Vernunft nach
Freiheit, als reines Selbsterkennen derselben, das seiner sicher, und über
sich klar geworden ist. Die freie Vernunft und ihre That ist Eins, und ihre
Thätigkeit ein reines Darstellen ihrer selbst.
In dieser Selbstproduktion der Vernunft gestaltet sich das Absolute in
eine objektive Totalität, die ein in sich selbst getragenes und vollendetes
Ganze ist, — keinen Grund außer sich hat, sondern durch sich selbst in
ihrem Anfang, Mittel und Ende begründet ist. Ein solches Ganzes erscheint
als eine Organisation von Sätzen und Anschauungen. Jede Synthese der
Vernunft, und die ihr korrespondirende Anschauung (die beide in der Spe-
kulation vereinigt sind) ist, als Identität des Bewußten und Bewußtlosen,
für sich im Absoluten und unendlich; zugleich aber ist sie endlich und be-
schränkt, insofern sie in der objektiven Totalität gesetzt ist, und andere
außer sich hat. Die unentzweiteste Identität objektiv — die Materie, sub-
jektiv — das Fühlen (Selbstbewußtseyn) ist zugleich eine unendlich entge-
gengesetzte, eine durchaus relative Identität. Die Vernunft, das Vermögen
(insofern) der objektiven Totalität vervollständigt sie durch ihr Entgegen-
gesetztes; und produzirt durch die Synthese beider eine neue Identität, die
selbst wieder vor der Vernunft eine mangelhafte ist, die ebenso sich wieder
ergänzt. Am Reinsten giebt sich die weder synthetisch noch analytisch zu
nennende Methode des Systems, wenn sie als eine Entwicklung der Ver-
nunft selbst erscheint; welche die Emanation ihrer Erscheinung, als eine
Duplicität, nicht in sich immer wieder zurückruft, — (hiermit vernichte-
te sie dieselbe nur), — sondern sich in ihr zu einer durch jene Duplicität
bedingten Identität konstruirt, diese relative Identität wieder sich entgegen-
setzt: so daß das System bis zur vollendeten Totalität fortgeht, sie mit der
entgegenstehenden subjektiven zur unendlichen Weltanschauung vereinigt,
deren Expansion sich damit zugleich in die reichste und einfachste Identität
kontrahirt hat.
Comment. Notice the quick step: “vollendete Totalität”. There is nothing in Hegel’s
argumentation that would indicate how the procedure described would ever be
capable of completion — quite to the contrary.
The next set of quotations is meant to give an impression of Hegel’s conception

of encyclopedia with a special emphasis on the status of the logic.
Quotations 66.19. (1) Taubeneck [1990], p. 54; translating Hegel [1817], p. 29

(§. 11.).
The whole of science is the presentation of the idea; its division, there-
fore, can only be conceptualized on this basis. Now since the idea is reason
identical to itself, which, in order to be for itself stands in opposition to
itself and is itself an other, but in this other is identical to itself, science
falls into three parts: (1) logic, the science of the idea in and for itself; (2)
the philosophy of nature, as the science of the idea in its otherness; (3) the
philosophy of spirit, the science of the idea as it returns to itself from its
otherness.
(2) Taubeneck [1990], p. 67; translating Hegel [1817], p. 50 (§. 37.).
Pure science or logic divides into three parts: into the logic of being, the
logic of essence, and the logic of the concept or the idea.—That is, it divides
into the logic of the immediacy of thought, of the reflection of thought, and
of the thought which returns from reflection into itself and abides by itself
in its reality.
(3) Taubeneck [1990], p. 263 (#474.); translating Hegel [1817], p. 50 (§. 474.).
This concept of philosophy is the self-thinking idea, truth aware of itself
[[. . .]], or logic with the significance that it is generality preserved in concrete
content. In this way science returns to its beginning, with logic as the result.
The presupposition of its concept, or the immediacy of its beginning and
the aspect of its appearance at that moment, are suspended.
Remark 66.20. As a recommendation for interested readers: the very last chapter
of Hegel’s Science of Logic (Miller [1969], pp. 824–844, translating Hegel [1816],
pp. 327–353) contains most valuable reading regarding method and — indirectly, at
least — the role of dialectic.
Hitherto philosophy had not found its method; it regarded with envy the
systematic structure of mathematics and, as we28 have said, borrowed it or
had recourse to the method of sciences which are only amalgams of given
material, empirical propositions and thoughts—or even resorted to a crude
rejection of all method. However, the exposition of what alone can be the
true method of philosophical science falls within the treatment of logic itself;
for the method is the consciousness of the form of the inner self-movement
of the content of logic.
[[The only thing]] to achieve scientific progress[[, and the quite elementary
understanding of which is essentially to be striven for]]—is the recognition of
the logical principle that the negative is just as much positive, or that what
is self-contradictory does not resolve itself into a nullity, into abstract noth-
ingness, but essentially only into the negation of its particular content[[.]]
Comment. Miller translates: “this quite simple insight”, which renders “simple” an
adjective to “insight”. My reading is that the object of the insight, or understanding,
is the logical principle, and that the understanding of this has to be basic, i.e., in
character more an adverb than an adjective: to understand in quite elementary a
way. I am afraid my translation is not elegant; but then, as I said before, I am not
prepared to pamper the English reader, in particular not at the cost of distorting
Hegel’s sentence construction. Admittedly, my rendering of “ganz einfache Einsicht”
as “quite elementary understanding” is an interpretation too.
(3) Wallace [1873], pp. 133 f; translating Hegel [1830], pp. 215 f.
[[T]]he only way to secure any growth and progress in knowledge is to hold
results fast in their truth. There is absolutely nothing whatever in which we
cannot and must not point to contradictions or opposite attributes; and the
abstraction made by understanding therefore means a forcible insistence on
a single aspect, and a real effort to obscure and remove all consciousness of
the other attribute which is involved. Whenever such contradiction, then,
is discovered in any object or notion, the usual inference is, Hence this
object is nothing. Thus Zeno, who first showed the contradiction native to
motion, concluded that there is no motion: and the ancients, who recognized
origin and decease, the two species of Becoming, as untrue categories, made
use of the expression that the One or Absolute neither arises nor perishes.
Such a style of dialectic looks only at the negative aspect of its result,
and fails to notice, what is at the same time really present, the definite
result, in the [[case of Being Determinate (“Dasein”)]] a pure nothing, but a
Nothing which includes Being, and, in like manner, a Being which includes
Nothing. Hence Being Determinate is (1) the unity of Being and Nothing,
in which we get rid of the immediacy in these determinations, and their
contradiction vanishes in their mutual connection—the unity in which they
28 Hegel doesn’t use the plural when he is referring to himself; this “we” is a product of the
translator; original: “wie gesagt”.

are only constituent elements. And (2) since the result is the abolition of
the contradiction, it comes in the shape of a simple unity with itself: that
is to say, it also is Being, but Being with negation or determinateness: it is
Becoming expressly put in the form of one if its elements, viz. Being.
I could not pretend that the method which I follow in this system of
logic—or rather which this system in its own self [[an ihm selbst]] follows—
is not capable of greater completeness, of much elaboration in detail; but
at the same time I know that it is the only true method. This is self-
evident simply from the fact that it is not something distinct from its object
and content; for it is the inwardness of the content, the dialectic which it
possesses within itself, which is the mainspring of its advance.
Comment. Hegel’s system of logic follows a method, “an ihm selbst”. So does Hegel
have a method? I think it is clear enough from this quotation that the question
asked in this way is rather misleading. Nevertheless it provides a wonderful toy for
the academic playground.29
Method may appear at first as the mere manner peculiar to the process
of cognition, and as a matter of fact it has the nature of such. But the
peculiar manner, as method, is not merely a modality of being determined
in and for itself ; it is a modality of cognition and as such is posited as
determined by the Notion and as form in so far as the form is the soul
of all objectivity and all otherwise determined content has its truth in the
form alone. If the content again is assumed as given to the method and
of a peculiar nature of its own, then in such a determination method, as
with the logical element in general, is a merely external form. Against this
however we can appeal not only to the fundamental Notion of the science
of logic; its entire course, in which all possible shapes of a given content
and of objects came up for consideration, has demonstrated their transition
and untruth; also that not merely was it impossible for a given object to
be the foundation to which the absolute form stood in a merely external
and contingent relationship but that, on the contrary, the absolute form
has proved itself to be the absolute foundation and ultimate truth. From
this course the method has emerged as the self-knowing Notion that has
itself, as the absolute, both subjective and objective, for its subject matter,
consequently as the pure correspondence of the Notion and its reality, as a
concrete existence that is the Notion itself.
Accordingly, what is to be considered here as a method is only the
movement of the Notion itself, the nature of which movement has already
been cognized; but first, there is now the added significance that the No-
tion is everything, and its movement is the universal absolute activity, the
self-determining and self-realizing movement. The method is therefore to
be recognized as the unrestrictedly universal, internal and external mode;
29 Cf. Forster [1993], p. 131, for instance.
and as the absolutely infinite force, to which no object, presenting itself

as something external, remote from and independent of reason, could offer
resistance or be of a particular nature in opposition to it, or could not be
penetrated by it. It is therefore soul and substance, and anything what-
ever is comprehended and known in its truth only when it is completely
subjugated to the method ; it is the method proper to every subject matter
because its activity is the Notion.
The method of absolute cognition is to this extent analytic. That it finds
the further determination of its initial universal simply and solely in that
universal, is the absolute objectivity of the Notion, of which objectivity the
method is the certainty. But the method is no less synthetic, since its subject
matter, determined immediately as a simple universal, by virtue of the
determinateness which it possesses in its very immediacy and universality,
exhibits itself as an other. This relation of differentiated elements which the
subject matter thus is within itself, is however no longer the same thing as is
meant by synthesis in finite cognition; the mere fact of the subject matter’s
no less analytic determination in general, that the relation is relation within
the Notion, completely distinguishes it from the latter synthesis.
This is no less synthetic than analytic moment of the judgement, by
which the universal of the beginning of its own accord determines itself as
the other of itself, is to be named the dialectical moment.
(3) Taubeneck [1990], p. 138 (#189.); translating Hegel [1817], p. 143 (§. 189.).
The method is in this way not an extraneous form, but the soul and
the concept of the content itself, and is only different from it insofar as the
determinations of the concept are as content also in themselves the totality
of the concept. This concept, however, shows itself to be inappropriate to
the elements and contents as long as it is without form.
Quotations 66.23. (1) Behler [1990], p. 300; translating Hegel [1828], p. 177.30
Most, however, indeed all controversies and contradictions must allow them-
selves to be harmonized through the apparently simple means of taking up
only what expresses itself in asserting, simply to observe and compare it
to the additional that one likewise asserts. To know what one says is much
rarer than one thinks, and it is with the greatest injustice that the accusa-
tion of not knowing what one says is considered the harshest.
Comment. This was written by the mature Hegel, about three years before his
death; it leaves me in no doubt that he adhered to a certain form of consistency.
The philosophical proposition, since it is a proposition, leads one to believe
that the usual subject-predicate relation obtains, as well as the usual atti-
tude towards knowing. But the philosophical content destroys this attitude
and this opinion. We learn by experience that we meant something other
30 This quotation was instigated by a quotation in Bubner [1976], p. 36.
than we meant to mean; and this correction of our meaning compels our
knowing to go back to the proposition, and understand it in some other
way.
Comment. I should like to see this being read in conjunction with quotation 66.9 (1)
regarding the difference of being and nothing: an opinion. Unfortunately, Miller’s
translation is not very suitable to bring out the meaning that I understand the
German original to have, so I should also like to see it being read in conjunction
with variation 120.1.
[[I]]t is an important consideration—one which will be found in more detail
in the logic itself—that the advance is a retreat into the ground, to what is
primary and true, on which depends and, in fact, from which originated,
that with which the beginnings is made. Thus consciousness on its onward
path from the immediacy with which it began is led back to absolute knowl-
edge as its innermost truth. This last, the ground, is then also that from
which the first proceeds, that which at first appeared as an immediacy. [[. . .]]
The essential requirement31 for the science of logic is not so much that the
beginning be a pure immediacy, but rather that the whole of the science be
within itself a circle in which the first is also the last and the last also the
first.
We see therefore that, on the other hand, it is equally necessary to
consider as result that into which the movement returns as into its ground.
In this respect the first is equally the ground, and the last a derivative[[.]]
Comment. It is in passages like this that I see the wisdom of Hegel’s philosophy —
in conjunction with the impotence of language.
By virtue of the nature of the method [[. . .]], the science exhibits itself as
a circle returning upon itself, the end being wound back into the beginning,
the simple ground, by the mediation; this circle is moreover a circle of
circles, for each individual member as ensouled by the method is reflected
into itself, so that in returning into the beginning it is at the same time the
beginning of a new member.
Comment. Miller translated “dabei” as “moreover” which doesn’t appeal to me but
it is hard to find a better way; what I understand is more at the same time.
The movement of the Notion is development : by which that only is [[made]]
explicit which is already implicitly present. In the world of nature it is
organic life that corresponds to the grade of the notion. Thus e.g. the plant
is developed from its germ. The germ virtually involves the whole plant,
but does so only ideally or in thought: and it would therefore be a mistake
to regard the development of the root, stem, leaves, and other different
31 The word “requirement” is added in the English translation. The German original has here
only: “Das Wesentliche”, i.e., “The essential”.

parts of the plant, as meaning that they were realiter present, but in a very
minute form, in the germ.
Comment. I don’t find Wallace’s translation “by which that only is explicit” satis-
factory; this is why I added “made”; alternatively: “by which that only is posited”.
(3) Miller [1969], pp. 842; translating Hegel [1917], p. 352.
The method is the pure Notion that relates itself only to itself; it is therefore
the simple self-relation that is being. But now it is also fulfilled being, the
Notion that comprehends itself, being as the concrete and also absolutely
intensive totality. In conclusion, there remains only this to be said about
this Idea, that in it, first, the science of logic has grasped its own Notion.
In the sphere of being, the beginning of its content, its Notion appears as a
knowing in a subjective reflection external to that content. But in the Idea
of absolute cognition the Notion has become the Idea’s own content. The
Idea is itself the pure Notion that has itself for subject matter and which, in
running itself as subject matter through the totality of its determinations,
develops itself into the whole of its reality, into the system of the science
[of logic], and concludes by apprehending this process of comprehending
itself, thereby superseding its standing as content and subject matter and
cognizing the Notion of the science.
[[T]]he Idea freely releases itself in its absolute self-assurance and inner poise.
By reason of this freedom, the form of its determinateness is also utterly
free—the externality of space and time existing absolutely on its own ac-
count without the moment of subjectivity.
Comment. Now it looks to me like Hegel has arrived where Kant started: the forms
of intuition, space and time.32 Hegel has done his thing: logic precedes the tran-
scendental aesthetic and has as its result the very elements of the transcendental
aesthetic. (On the other hand: phenomenology precedes logic as an introduction.)
32 In the Encyclopedia, the logic is followed by the philosophy of nature, the first chapter of
which is entitled “mathematical mechanics”, the first two sections of which deal with space and
time, respectively.
CHAPTER XVII
AFTER THE GREAT MASTERS
Es hört gar nicht auf, daß der Eine daher, der Andere dorther
einen Fall und Instanz beibringt, nach der auch noch etwas mehr und anderes
bei diesem und jenem Ausdruck zu verstehen,
in dessen Definition also noch eine nähere oder allgemeinere Bestimmung
aufzunehmen und darnach die Wissenschaft einzurichten sey.1
Kant is a sort of highway

with lots and lots of milestones.
Then all the little dogs come
and each deposits his bits at the milestones.2
Artistotle had spoken,

and it was the part of humbler men
merely to repeat the lesson after him.3
To the extent that philosophical thinking is still attempted,

it manages only to attain an epigonal renaissance
and variations of that renaissance.4
The way I presented different stages of transcendental idealism in the foregoing

chapters might suggest that the matter were fairly straightforward. This, however,
is far from being the case. Philosophers of all persuasions, not just those in the
tradition of Kant and Hegel, have struggled with more or less — or more less —
success to arrive at something that might be called a common denominator as re-
gards transcendental and speculative philosophy. The present chapter is devoted to
a cursory glance at a variety of attitudes — mainly within the tradition of Kant,
Fichte, and Hegel — with the intent to reflect a bit of the forlornness of the επίγονοι
of a philosophical revolution that got half way stuck; other discussions of relevant
topics such as the synthetic-analytic distinction in a non-Kantian/Hegelian (non-
transcendental) tradition can be found in part E in these materials.
1 Hegel [1812], p. 44.
2 Einstein, quoted after Rosenthal-Schneider [1980], p. 90, pinched from Wang [1986], p. 74.
3 Russell [1914], p. 33.
4 Heidegger [1966], p. 433. Speaking for himself, I suppose, and, of course, anticipating Der-
rida’s marginalism.
894
§ 67. GENERAL CONSIDERATIONS 895
It will be clear that the point of quotations in this chapter can no longer be the
characterization of a relatively coherent philosophical position. The aim is, in fact,
modest; it is mainly to give an impression of a mostly pathetic situation.5
Although I do not think that Kant’s heirs manage to do more than “den bela-
chenswerten Anblick zu geben, daß einer (wie die Alten sagen) den Bock melkt, der
andere ein Sieb unterhält”,6 there is something to the way in which they do this,
that deserves attention.
§ 67. General considerations
Transcendental idealism is not a common sense philosophy. In fact, it is demanding

on one’s capacity to reflect. So it will not come as a surprise that it has given rise to
many misinterpretations and misunderstandings. Generations of philosophers have
tried to read sense — or non-sense, according to taste — into Kant’s and Hegel’s
philosophy. On top of an inherent difficulty in their subject, they are often found to
have had a difficult writing style, and that — even worse — in a language which is
notoriously difficult anyway, and unfamiliar to the majority of philosophers today.
Before I try to give an impression of these efforts I try to locate some sources of the
problems in question.
Remark 67.1. Neither Kant’s transcendental philosophy nor Hegel’s speculative

philosophy provides a satisfactory framework for my enterprise. Still, it is their
ideas that figure prominently in it. A natural reaction in this situation seems to
be to consult the secondary literature. Natural as it may look, it is the source of
even more confusion and most of what follows is meant to serve as a warning. Sec-
ondary literature is overwhelming, more, it is actually defeating. The sheer amount
presents a problem: there might be a pearl hidden, but by the time you worked your
way through the effusions of logical incompetence and verbal incontinence, you’re
incapable of recognizing it.
So, for the most part, at least, I consider secondary literature a waste of time
and the reader may find that this attitude reveals itself in the way I deal with
it. Admittedly, I’m very likely to have done injustice to some authors, just by
overlooking or not recognizing their work. But, in view of what I’ve said above,
I can’t be bothered spending my time digging in book mountains. In particular not,
because I have found one pearl at least: Allison’s [1983] is not just by far the best
access to Kant’s transcendental idealism I’ve ever come across, I find it simply good,
which is better than the best in a bad lot.
But then, readers might find it curious, to find a whole chapter dedicated to,
essentially, secondary literature. The point is simple: everything has its virtue; and
the virtue of secondary literature is that it provides good bad examples. In the third
book I shall take advantage of various possible — and impossible — positions by
5 The lack of structure of the present chapter is partly a reflection of the lack of structure
of its subject, but also a reflection of my cursory reading in a field that I do not take all that
seriously, at least not as a source of original ideas.
6 Kant [1781], p. 58.
896 XVII. AFTER THE GREAT MASTERS
way of paraphrasing formulations actually put forward. To a considerable extent,

the present chapter provides the original formulations for that purpose.
67a. Aspects of the historical reception of ancient dialectic. Among

the various aspects of the ancient notion of dialectic, the following features are most
commonly mentioned.
(i) dialogue/rhetoric,
(ii) ideas/universals,
(iii) knowledge and opinion, and
(iv) logic.
I take them as moments which contribute to the modern notion of dialectic. None
of them, however, can account for the speculative power with which dialectic was
endowed by Hegel.
Quotation 67.2. Hall [1967], p. 386.

If [[. . .]] the lost work of Protagoras did begin, as several subsequent
writers attest, with the claim that on every subject two opposite state-
ments (λóγoι) could be made, and if the book continued with a content of
statement and counterstatement, then Protagoras deserves to be considered
the ancestor of the medieval or of the Hegelian dialectic [[.]]
Comment. Note: statement — counterstatement. But is this enough to speak of

dialectic?
The next group of quotations revolves around dialectic and dialogue.
Quotations 67.3. (1) Russell [1946], pp. 109 f.

Dialectic, that is to say, the method of seeking knowledge by question
and answer, was not invented by Socrates. It seems to have been first prac-
ticed systematically by Zeno, the disciple of Parmenides; in Plato’s dialogue
Parmenides, Zeno subjects Socrates to the same kind of treatment to which,
elsewhere in Plato, Socrates subjects others. But there is every reason to
suppose that Socrates practised and developed the method.
(2) McKeon [1975], p. 1.
The origins of dialectic and dialogue are closely connected in Greek thought,
both etymologically and logically, by logos. Diogenes Laertius begins his ac-
count of Plato’s philosophy with a borrowed and doubtless familiar account
of the affinity of dialogue and dialectic. “A dialogue is a discourse (logos)
consisting of question and answer on some philosophical or political ques-
tion, with due regard to the characters introduced and to the style (lexis)
employed. Dialectic is the art of discourse (logoi) by which we either refute
or establish some propositions by means of question and answer.”
(3) McKeon [1975], p. 3.

The invention of dialectic and the invention of dialogue tend to be attributed
to a single thinker, Zeno of Elea is among those credited, on the alleged
authority of Aristotle, with the invention of dialectic, and he is also said
to have been the first to write dialogues. Plato brought the dialogue to its
perfection, however, and he is therefore judged its inventor. It was Plato
who, in like fashion gave dialectic its philosophic scope and inclusiveness.
(4) McKeon [1975], pp. 11 f.
Diogenes Laertius [[. . .]] informs us in his brief account of Euclides of Me-
gara,7 that Euclides was a student of Parmenides and that his disciples were
called Megarians, then Eristics, and then Dialecticians, “because they put
their arguments (logoi) in the form of question and answer.” The illustration
of this dialectic is taken from ethics and physics, with Eleatic overtones:
the supreme good is really one, and all that is contradictory of the good has
no existence, but its development is taken from logic. Euclides impugned
a demonstration by attacking the conclusion, not the premisses, and his
disciples centered their attention on self-refuting paradoxes, like The Liar
and The Veiled Figure, and wrote about propositions, predications, and
inferences.
Comment. Note the mention of self-refuting paradoxes.
The next set of quotations revolves around knowledge and opinion. The emer-
gence of a negative association in the term “dialectic”.
Quotation 67.4. McKeon [1975], pp. 12 f.

With Aristotle the conception of dialectic changed, and the history of
dialectic ceased to be the whole of the history of philosophy. Aristotle made
explicit the range of the meanings of logos — as statement, thought, essence,
and action — and used them to differentiate kinds of syllogisms: apodeic-
tic based on the causes of things, dialectic based on the opinions of men,
and sophistic or eristic based on errors or tricks of statement or thought.
Dialectical argument is not scientific argument. It is concerned with the
opinions of men, not the nature of things. Knowledge of things and their
properties, however, is expressed as opinions, and a dialectical examination
of the opinions of men may be used as a propaedeutic to scientific inquiry
or to test the principles of scientific demonstration.
Comment. This strikes me as an interesting point: dialectic is “concerned with the
opinions of men, not the nature of things.” It reminds of Hegel’s remarks concerning
being and nothing “the difference . . . exists not in themselves but in a third, in
subjective opinion.” 8
7 Not to be confused with the mathematician, who lived about 100 years later. Cf. quota-
tion 70.4 (4) on p. 985 in these materials.

8 Miller [1969], p. 92; cf. quotation 66.9 (1) in these materials. Also relevant in this context:
the distinction of form and content.

Quotations 67.5. (1) McKeon [1975], p. 13.

Dialectic is the counterpart or antistrophe of rhetoric. Both are con-
cerned with opinions: dialectic with the opinions of all men or of experts to
provide techniques to establish or refute definitions, genera, properties, or
accidents as predicables of a subject; rhetoric with the opinions of particular
audiences to provide techniques of persuasion dependent on the character of
the speaker. the emotions of the audience, and the style and organization of
the speech. The two related arts not concerned with essences or things, and
the subjects of their propositions are not defined univocal terms, but places
or topoi, empty of meanings and references, used as sources of arguments
and discoveries or inventions.
(2) McKeon [1975], p. 16.
The changing character of dialectic as an art is indicated in its chang-
ing relation to rhetoric. For Plato dialectic is the art of demonstration,
and rhetoric is a sham art if it is separated from dialectic; and he there-
fore refutes rhetorical speeches or transforms them into dialectical proofs.
For Aristotle dialectic and rhetoric are counterparts, antistrophes, and he
therefore examines the differences of argument which distinguish dialectical
syllogisms and inductions from rhetorical enthymemes and examples.
Finally, I come to the connection between dialectic and logic which seems to
have developed first with the Stoics, and then survived into the Middle Ages.
Quotations 67.6. (1) Hall [1967], p. 387.

By “dialectic” the Stoics primarily meant formal logic, in which they par-
ticularly developed forms of inference belonging to what we now call the
propositional calculus. But they applied the term “dialectic” widely: for
them it also included the study of grammatical theory and the considera-
tion of meaning-relations and truth. This widened scope [[. . .]] was accepted
by Cicero and perhaps overemphasized by Seneca, who wrote that dialectic
“fell into two parts, meanings and words, that is, things said and expressions
by which they are said”[[.]]
Comment. In the context of dialectic, this distinction is worth noting here; it may
well indicate a glimpse of the dialectic of form and content as found in Hegel’s Logic.
(2) Hall [1967], p. 387.
In the Middle Ages “dialectic” continued to be the ordinary name for
logic: for example, the first medieval logical treatise was the Dialectic of
Alcuin.
67b. The challenge of the ancient Greeks. Not surprisingly, the fragments
of Parmenides cause consternation, at least in certain sections. It may be helpful
to read the following quotations alongside with quotations 63.10 from the foregoing
chapter.
Quotations 67.7. (1) Cornford [1939], p. 34, footnote 1.

I follow Zeller and Burnet in reading ἔστιν, ‘it is possible’. [[. . .]] I cannot
believe that Parmenides meant: ‘To think is the same thing as to be.’ He
nowhere suggests that his One Being thinks, and no Greek of his date or for
long afterwards would have seen anything but nonsense in the statement
that ‘A exists’ means the same thing as ‘A thinks’.
Comment. Perhaps this can serve as a warning against reading ‘to think’ as a ‘thing’
and ‘to be’ as ‘existence’ ? Cf. quotation 67.8 (6) below.
(2) Cornford [1939], p. 34, footnote 2.
Parmenides certainly held that there can be no thought without an object
which is; but nothing in the poem supports the interpretation that thinking
is the same thing as its object. Burnet’s translation: ‘the thing that can be
(ἔστι) thought and that for the sake of which the thought exists is the same’
is rather tautologous: it amounts to ‘what can be thought is the object of
thought’.
Comment. What is interesting for me, in view of what comes later, is the apparent
low regard for tautology. Tautologies, or rather the ‘playing up’ of certain tautolo-
gies, is an extremely challenging phenomenon in my study.
Quotations 67.8. (1) Burnet [1914], p. 67.

[[T]]he rise of mathematics in [[the]] Pythagorean school had revealed for the
first time the power of thought. To the mathematician of all man it is the
same thing that can be thought (ἔστι νοεῖν) and that can be (ἔστιν εἶναι),
and this is the principle from which Parmenides starts. It is impossible to
think what is not, and it is impossible for what cannot be thought to be.
The great question, Is it or is it not? is therefore equivalent to the question,
Can it be thought or not?
(2) A. E. Taylor [1926], p. 359.
Parmenides [[. . .]] wrote the words ταυτὸ γὰρ ἔστι νοεῖν τε καὶ εἶναι (“it is
the same thing which can be thought of and can be”)[[.]]
Comment. This seems to be a new variation on the theme; at least I know of no
fragment in precisely this form. As regards the translation, Taylor seems to follow
Burnet and Zeller as characterized by Cornford in quotation 67.7 (1) above.
(3) Cornford [1939], p. 43.
Since Being is ‘whole’ and complete, there can be no other being left
outside it, no second object of thought. And it is unchangeable, since there
is nothing that it ‘is not’ and could come to be by changing.
(4) Heidegger [1953], p. 110.
Was der Spruch des Parmenides ausspricht, ist eine Bestimmung des
Wesens des Menschen aus dem Wesen des Seins selbst.
(5) Derrida [1971a], p. 182.

“Thought”—that which lives under this name in the West—could never
emerge or announce itself except on the basis of a certain configuration of
noein, legein einai and of the strange sameness of noein and einai spoken
of in Parmenides’ poem.
Comment. A “certain configuration”; I suppose one can’t go wrong with a specifi-
cation of this kind.
(6) Bass (translator), in Derrida [1971a], p. 197, footnote 29.
[[T]]he reader should keep in mind that in Greek, German, and French in-
finitives can be used as substantives, while they cannot in English. Thus
the point in Parmenides’ poem is the sameness of “the ‘to think’ ” (noein,
denken, penser ) and “the ‘to be’ ” (einai, sein, être).
Quotations 67.9. (1) A. E. Taylor [1926], p. 369.

[[T]]he ultimate source of our perplexities is the ambiguity of the word “is.”
We get contradictory results according as “is” is taken to be the symbol
of predication (Peano’s ǫ), or that of existence (Peano’s ∃). Many of the
inferences turn simply on this confusion of a predication with what we
now call an “existential proposition.” It is legitimate parody to employ this
fallacy, because, as we can see from the remains of the poem of Parmenides,
the whole of Eleaticism lies in ignoring the distinction.
Comment. How can you see from remains something that applies to the whole?
As regards the content, see remark 103.1 (1); the point is crucial for my notion of
dialectic.
(2) Aune [1986], p. 4.
To understand why a philosopher as astute as Aristotle might undertake
a study as peculiar as “the science of being qua being”—and to decide
whether such a study really makes sense—we must first appreciate the wide
variety of things that can be said to “be” in some sense or other. If these
various “beings” are described in a sufficiently general way, we shall find
that they can be arranged into fundamentally different kinds or categories
and that they are sometimes so different from one another that we can
reasonably wonder how they all fit together in the same world or universe.
The wonder we might feel here will provide some basis, at least, for the sort
of investigation Aristotle had in mind.
Perhaps the least problematic group of things that can be said to be (or
exist) includes the familiar objects of everyday experience: people, animals,
rocks, pieces of furniture.
Comment. Charmingly helpless attempts like these to read sense into a translation
from a different language, culture, and time may well evoke sympathetic feelings,
but the incredible arrogance that lies in the phrase “to decide whether such a study
really makes sense” should serve as a warning. Here is a cultivation of ignorance
and lack of understanding that is about to establish itself as an academic standard,
a kind of McDonald’s standard in philosophy, as it were.
Quotations 67.10. (1) Heidegger [1953], p. 18.

Wer vom Nichts redet, weiß nicht was er tut. Wer vom Nichts spricht,
macht es durch solches Tun zu einem Etwas. Sprechend spricht er so gegen
das, was meint. Er wider-spricht sich selbst. [[. . .]] Ein solches Reden über
das Nichts besteht aus lauter sinnlosen Sätzen.
(2) Krell [1978], pp. 96 f; translating Heidegger [1929], pp. 27 f.
Interrogating the nothing—asking what and how it, the nothing is—turns
what is interrogated into its opposite. The question deprives itself of its own
object. Accordingly, every answer to this question is also impossible from
the start. For it necessarily assumes: the nothing “is” this or that. With
regard to the nothing, question and answer alike are inherently absurd.
[[. . .]] The commonly cited ground rule of all thinking, the proposition
that contradiction is to be avoided, universal “logic” itself, lays low this
question. For thinking, which is always essentially thinking about some-
thing, must act in a way contrary to its own essence when it thinks the
nothing.
Since it remains wholly impossible for us to make the nothing into
an object, have we not already come to the end of our inquiry into the
nothing—assuming that in this question “logic” is of supreme importance,
that the intellect is the means, and thought the way, to conceive the nothing
originally and to decide about its possible exposure?
But are we allowed to tamper with the rule of “logic”?
Comment. This quotation contains quite a remarkable collection of commonplace
prejudices about logic. The most interesting aspect for me here is revealed in the last
question; Heidegger considers the possibility of tampering with the “rule” of logic,
but apparently not with the content of its claims, like, for instance, “the proposition
that contradiction is to be avoided”. Nowhere in his lecture on the question of “what
is metaphysics? ” does he consider the possibility that there is something wrong —
in simple technical-logical terms — with the reasoning that leads to his conclusion
that “every answer . . . is impossible from the start”.9 In fact, this is an option
which is closed to him, if he is to promote “Angst” as an “original revelation of the
nothing” (ibid. p. 103). Heidegger deprives the nothing of an objective character in
order to be able to launch his subjective foundation of the nothing in a “fundamental
experience” (ibid. p. 99).
(3) Krell [1978], p. 97; translating Heidegger [1929], pp. 28.
[[T]]he nothing is the negation of the totality of beings; it is nonbeing pure
and simple. But with that we bring the nothing under the higher determi-
nation of the negative, viewing it as the negated. However, according to the
reigning and never-challenged doctrine of “logic,” negation is a specific act
of the intellect. How then can we in our question of the nothing, indeed in
the question of its questionability, wish to brush the intellect aside? Are we
9 For those who want to know what is wrong in technical-logical respect: the reasoning employs
contraction. Cf. section 41a in the tools for technical results concerning the incompatibility of
contraction and (unrestricted) abstraction.
altogether sure about what we are presupposing in this matter? Do not the
“not,” negatedness, and thereby negation too represent the higher determi-
nation under which the nothing falls as a particular kind of negated matter?
Is the nothing given only because the “not,” i.e., negation, is given? Or is it
the other way around? Are negation and the “not” given only because the
nothing is given? That has not been decided; it has not even been raised
expressly as a question. We assert that the nothing is more original than
the “not” and negation.
Comment. Is this more than empty rhetoric? Anyway, I am happy to leave it to
the masters of ceremonies and their servers. In the groundworks, negation etc.
will be defined on the basis of unrestricted abstraction. What Heidegger seems to
leave unquestioned in his starting point is the “totality of beings”; he only focuses on
“negation”. But negation can be defined in terms of generalization and implication,
in the sense of saying, ‘not A’ means ‘A implies everything’. This I take to be close
in spirit to Hegel. It may be worthwhile, in the present context, to have a look at
quotations 69.29 in this chapter (Henrich).
Quotations 67.11. (1) R. Robinson [1941], pp. 74 f.

Although dialectic can be used to advantage in many and various
spheres, its subject-matter is in one sense always the same, and was so
considered throughout Plato’s life. It is always the search for “what each
thing is” (Rp. 533B). That is to say, it seeks the “essence” of each thing
(oυσία, Rp. 534B), the formal and abiding element in the thing. It regards
“what neither comes into being nor passes away, but is always identically
the same” (Phlb. 61E). Thus it presupposes that things have unchanging
essences; and if anyone denies this he absolutely destroys the power of di-
alectic (Pa. 135BC).
(2) Hall [1967], p. 386.
[[D]]ialectic always had the same subject matter: it sought the unchanging
essence of each thing.
(3) A. E. Taylor [1926], p. 293
[[W]]e may say that what the Republic calls “dialectic” is, in principle, simply
the rigorous and unremitting task of steady scrutiny of the indefinables
and indemonstrables of the sciences, and that, in particular, his ideal, so
far as the sciences with which he is directly concerned goes, is just that
reduction of mathematics to rigorous deduction from expressly formulated
logical premisses by exactly specified logical methods of which the work of
Peano, Frege, Whitehead, and Russell has given us a magnificent example.
Quotations 67.12. (1) R. Robinson [1941], p. 75.

Plato did not separate dialectic from philosophy as we tend to separate
say logic or methodology from metaphysics. Dialectic was not a propaedeu-
tic to philosophy. It was not a tool that you might or might not choose
to use in philosophizing. It was philosophy itself, the very search for the
essences, only considered in its methodical aspect. The method occurred

only in the search, and the search only by means of the method.
Comment. I find this remark interesting in view of Hegel’s ‘identification’ of meta-
physics, logic, and speculative philosophy.
(2) R. Robinson [1941], pp. 75 f.
Plato’s dialectic is often of such a nature that to our minds it ought to
be separated from philosophy. To us he often seems to be discussing neither
physical nor metaphysical reality, but only the human logical apparatus of
conceptions and terms. But still, in a manner very strange and unnatural to
us, he regards himself as talking not logic but ontology. [[. . .]] [[H]]is theory
of communion or κοινωνία has to us such a strongly logical air that most
interpreters represent it as a theory of the copula or of predication or the
like, while yet Plato himself regards it as a theory of reality.
Comment. This is a charmingly unvarnished manifestation of a complete lack of
understanding. If you drop the “human” in the “human logical apparatus of con-
ceptions and terms” you have the topic of the present treatise with the study of
predication providing the philosophical backbone for the unity of logic and ontol-
ogy. Differently put, it seems to me that Robinson is saying: Platon’s dialectic seems
logical in character and should be separated from metaphysics. I hope it is clear
that the whole point of the present study is to approach metaphysics through the
dialectic inherent in logic.
(2) Cornford [1939], p. 92.
Plato’s Parmenides repudiates the doctrine which some critics ascribe to
the real Parmenides that ‘to think is the same as to be’[[.]]
Comment. This is Cornford’s conclusion from Plato [Parmenides], 132c. Cf. quota-
tion 67.7 (1) regarding Cornford’s reading of Parmenides’ fragment.
(3) Forster [1993], p. 154.10
[[A]] case could be made for seeing Parmenides’s argument for the inco-
herence of the notion of not-being as a successful exposure of deep self-
contradictions in the thought of his contemporaries. [[. . .]] Hegel, in his ar-
gument for the incoherence of the category Nothing in the Logic, attempts
to revive a form of the Parmenidean paradox.
Comment. And, I want to add, my attempt is to employ Cantor’s diagonal argument
for this very purpose.
Quotations 67.13. (1) Russell [1946], p. 210.

What, exactly is meant by the word ‘category’, whether in Aristotle or in
Kant and Hegel, I must confess that I have never been able to understand.
I do not myself believe that the term ‘category’ is in any way useful in
philosophy, as representing any clear idea. There are, in Aristotle, ten cat-
egories: substance, quantity, quality, relation, place, time, position, state,
action, and affection. The only definition offered of the term ‘category’ is:
10 Cf. also Pinkard in quotation 69.31 (6) below.
‘expressions which are in no way composite signify’ — and then follows the
above list. This seems to mean that every word of which the meaning is
not compounded of the meanings of other words signifies a substance or a
quantity or etc. There is no suggestion of any principle on which the list of
ten categories has been compiled.
Comment. This is also Kant’s objection to Aristoteles’ list of ten categories although
I can’t see that he did any better. The problem here may be similar to that of a
concept such as “elements”. Ancient Greeks: fire, water, earth, air.11 Still, the idea
seems to have made sense to a lot of people and today we have a pretty impressive
table of elements.
(2) Cooke [1938], p. 2.

What is the subject of the Categories? In ordinary usage κατ ηγoρία, ren-
dered in English as ‘category,’ meant nothing more than ‘a predicate.’ This
meaning it seems highly probable that it retains in [[Aristoteles’ categories]].
The ten categories, then, are ten predicates.
(3) Ryle [1960/89], p. 59.
Aristotle borrowed “categoria” from legal parlance, where it meant “accu-
sation”, and stretched it to mean anything that could be asserted truly or
falsely of anything.
Comment. I’ve got a feeling somehow that I’ve read something of this kind some-
where in Heidegger, broadly laid out of course. Modern Greek usage of the corre-
sponding verb κατηγορώ, according to my Pocket Oxford Greek Dictionary is in the
sense of “to accuse”, “to charge”, but also “to criticize”.
Quotation 67.14. Russell [1946], p. 61.

Thales [[. . .]] thought everything was made of water; Anaximenes thought
air was the primitive element; Heraclitus preferred fire. At last Empedocles
suggested a statesmanlike compromise by allowing four elements, earth, air,
fire and water.
I close this section with a quotation from the entry “category” in the Cambridge
Dictionary of Philosophy.
Quotation 67.15. Meiland, Cambridge Dictionary of Philosophy, p. 108, first col-

umn.
category, an ultimate class. Categories are the highest genera of entities
in the world.
Comment. Should I have warned you not to get your hopes too high?

67c. Aspects of metaphysics and epistemology. Metaphysics is an easy

target for criticism of all sorts. When Kant came forward with his first Critique
he was apparently quite confident that he was about to settle the problems. Since
then, it seems, his contribution can be regarded as part of them and there is no lack
of high flying ambitions unaware of his approach.
As regards the assault on metaphysics by Carnap and others in the tradition
of so-called logical positivism, reader will find bits in chapter XXII. In the present
section I confine myself to aspects of subsequent and independent developments.

Empiricism and idealism alike are faced with a problem to which, so
far, philosophy has found no satisfactory solution. This is the problem of
showing how we have knowledge of other things than ourself and the opera-
tions of our own mind. Locke considers this problem [[in quotation 58.8 (1)]].
[[. . .]] From this it would seem to follow immediately that we cannot know
of the existence of other people, or of the physical world, for these, if they
exist, are not merely ideas in my mind. Each one of us, accordingly, must,
so far as knowledge is concerned, be shut up in himself, and cut off from
all contact with the outer world.
This, however, is a paradox, and Locke will have nothing to do with
paradoxes.
(2) Russell [1946], pp. 591 f.
Locke assumes it known that certain mental occurrences, which he calls
sensations, have causes outside themselves, and that these causes, at least to
some extent and in certain respects, resemble the sensations which are their
effects. But how, consistently with the principles of empiricism, is this to be
known? We experience the sensations, but not their causes; our experience
will be exactly the same if our sensations arise spontaneously. The belief
that sensations have causes, and still more the belief that they resemble
their causes, is one which, if maintained, must be maintained on grounds
wholly independent of experience. The view ‘knowledge is the perception of
the agreement or disagreement of two ideas’ is the one that Locke is entitled
to, and his escape from the paradoxes that it entails is effected by means of
an inconsistency so gross that only his resolute adherence to common sense
could have made him blind to it.
This difficulty has troubled empiricism down to the present day. Hume
got rid of it by dropping the assumption that sensations have external
causes, but even he retained this assumption whenever he forgot his own
principles, which was very often. His fundamental maxim, ‘no idea without
an antecedent impression’, which he takes over from Locke, is only plausible
so long as we think of impressions as having outside causes, which the very
word ‘impression’ irresistibly suggests. And at the moments when Hume
achieves some degree of consistency he is wildly paradoxical.
No one has yet succeeded in inventing a philosophy at once credible and
self-consistent. Locke aimed at credibility, and achieved it at the expense
of consistency. Most of the great philosophers have done the opposite. A

philosophy which is not self-consistent cannot be wholly true, but a philoso-
phy which is self-consistent can very well be wholly false. The most fruitful
philosophies have contained glaring inconsistencies, but for that very reason
have been partially true. There is no reason to suppose that a self-consistent
system contains more truth than one which like Locke’s is obviously more
or less wrong.
Comment. I’d love to know the notion of credibility in question. I suspect it is a
paragon of ethnocentricity, but the point of ethnocentricity is that it works by not
being made explicit.
(3) Russell [1959], pp. 42; pinched from Hylton [1993], p. 449.
I was at this time [[1890s]] a full-fledged Hegelian, and I aimed at construct-
ing a complete dialectic of the sciences. [[. . .]] I accepted the Hegelian view
that none of the sciences is quite true, since all depend upon some abstrac-
tion, and every abstraction leads, sooner or later, to contradiction.
In view of the central theme of Kant’s Critique, transcendental idealism and

empirical realism, let me continue with the problem of realism.
Quotation 67.17. Popper [1982], p. 3.

[[S]]ome [[. . .]] theories are so bold that they can clash with reality: they are
the testable theories of science. And when they clash, then we know that
there is reality: something that can inform us that our ideas are mistaken.
And this is why the realist is right.
Comment. That’s how simple philosophy can be — if you are just simple-minded
enough. When theories clash, it is because of reality; don’t bother about the con-
ceptual framework. Compare confrontation 100.5 in the groundworks.
Quotations 67.18. (1) Sayers [1985], p. x.

The objective world can be known to consciousness, Lenin insists, only
because consciousness is a reflection of reality. This idea, in some form or
other, is basic to all versions of realism and materialism in epistemology.
(2) Sayers [1985], pp. xiv f.
[[Materialism is]] the theory that consciousness does not exist independent of
matter and that all reality is ultimately material in nature. Idealism is the
opposite of this: it is the view that reality is ultimately ideal in character,
and can be seen as a construct or creation of mere ideas, interpretations,
or whatever.
Comment. The point of including such a quotation in these materials is not so
much that it represents a fairly unrefined mind-matter dualism, but to give an
example of what is useless for my approach to dialectic, despite being explicitly
concerned with “dialectic and the theory of knowledge”.
If this was bad, things get worse. The attempt to overcome naive realism and to
describe an alternative way of looking at things seems to fall prey to naive realism
again, since it is, if done in an unconscious way, a simple claim about an independent
state of affair, hence a claim in the tradition of naive realism.
Quotations 67.19. (1) Adorno [1966/67], p. 30.

Die Anstrengung, die im Begriff des Denkens selbst, als Widerpart zur passi-
vischen Anschauung, impliziert wird, ist bereits negativ, Auflehnung gegen
die Zumutung jedes Unmittelbaren, ihm sich zu beugen. Urteil und Schluß,
die Denkformen, deren auch Kritik des Denkens nicht entraten kann, ent-
halten in sich kritische Keime; ihre Bestimmtheit ist allemal zugleich Aus-
schluß des von ihnen nicht Erreichten, und die Wahrheit, die sie organisieren
wollen, verneint, wenngleich mit fragwürdigem Recht, das nicht von ihnen
Geprägte. Das Urteil, etwas sei so, wehrt potentiell ab, die Relation seines
Subjekts und seines Prädikats sei anders als im Urteil ausgedrückt. Die
Denkformen wollen weiter als das, was bloß vorhanden, »gegeben« ist.
Comment. The battle of words. The quotation marks around the gegeben (given)
can’t conceal the resilience of the naive realistic pattern of thought. Even that which
can only be assumed hypothetically as the beyond is pressed into the category of
the ‘being at hand’ (Vorhandenen). Adorno does not escape the forms of thought
he is criticizing, he just tries to obscure their presence. After all, what is it that is
forced into the form of a judgment? It is naive-realistic metaphysics which suggests
that there is anything that succumbs and thus distracts from what that actually is
which is beyond: thought as act. Adorno’s linguistic acrobatics cannot conceal the
lack of insight into the nature of the dialectic of the forms of thought. Note also
the phrase “passivische Anschauung”, which puts Adorno in the company of Locke
rather than Kant.12
(2) Adorno [1969], p. 45.
Was den Positivisten wie Musik in den Ohren mißtönt, ist das nicht
ganz in Sachverhalten Vorhandene, das der Form der Sprache bedarf. Je
strikter diese den Sachverhalten sich anschmiegt, desto höher entragt sie der
bloßen Signifikation und nimmt etwas wie Ausdruck an. [[. . .]] Wörtlichkeit
und Präzision sind nicht dasselbe, eher tritt beides auseinander. Ohne ein
Gebrochenes, Uneigentliches gibt es keine Erkenntnis, die mehr wäre als
einordnende Wiederholung.
(3) Adorno [1969], p. 126.
[[D]]as Ideal der einstimmigen, möglichst einfachen mathematisch eleganten
Erklärung versagt, wo die Sache selbst [[. . .]] nicht einstimmig, nicht einfach
ist, auch nicht neutral dem Belieben kategorialer Formung anheimgegeben,
sondern anders, als das Kategoriensystem der diskursiven Logik von seinen
Objekten vorweg erwartet.
12 Cf. quotation 58.7 (1) in these materials. Kant, in quotations 61.1, speaks — in translation
— of receptivity as a ‘capacity’ or ‘power’. Cf. also Solomon in quotation 68.6 below.

Comment. In view of what I am aiming at: Adorno, just like those who he criticizes,
imposes a framework, he doesn’t derive it. Different as this framework may be, the
fact of the simple imposition (as “vorweg”) makes it just as ‘short’ vis à vis its
object, as that of the “Kategoriensystem der diskursiven Logik”. (Note: Adorno is
speaking of society as the “Sache” here.)
(4) Adorno [1966/67], p. 28.

Denken [[. . .]], das frisch-fröhlich von vorn anfängt, unbekümmert um die
geschichtliche Gestalt seiner Probleme, wird erst recht deren Beute.
(5) Adorno [1966/67], p. 163.
Denken, das an der Identität irre ward, kapituliert leicht vor dem Unauf-
löslichen und bereitet aus der Unauflöslichkeit des Objekts ein Tabu fürs
Subjekt, das irrationalistisch oder szientifisch sich bescheiden, nicht an das
rühren soll, was ihm nicht gleicht, vorm gängigen Erkenntnisideal die Waf-
fen streckend, dem es dadurch noch Respekt bekundet.
Quotation 67.20. Heidegger [1953], p. 1.

Warum ist überhaupt Seiendes und nicht vielmehr Nichts? Das ist die
Frage. [[. . .]]
[[. . .]] Jeder wird einmal, vielleicht sogar dann und wann, von der verbor-
genen Macht dieser Frage gestreift, ohne recht zu fassen, was ihm geschieht.
In einer großen Verzweiflung z.B., wo alles Gewicht aus den Dingen schwin-
den will und jeder Sinn sich verdunkelt, steht die Frage auf. Vielleicht nur
einmal angeschlagen wie ein dumpfer Glockenschlag, der in das Dasein her-
eintönt und mählich wieder verklingt.
Comment. Does Derrida’s Glas bear some relation to this “Glockenschlag”?

[[G]]ehört es zum Wesensgeschick der Metaphysik, daß sich ihr der eigene
Grund entzieht, weil im Aufgehen der Unverborgenheit überall das Wesende
in dieser, nämlich die Verborgenheit, ausbleibt, und zwar zugunsten des
Unverborgenen, das als das Seiende erscheint?
Quotations 67.22. (1) Aune [1986], p. 10.

Although [[. . .]] logical positivists have insisted that metaphysical theo-
ries of reality are inherently meaningless because they are baseless, or “un-
verifiable,” this extreme view is no longer in vogue. In the past few decades
the trend has been to defend a scientific metaphysics, one according to
which reality is substantially what the physical sciences say it is.
Comment. I don’t care whether a view is in ‘Vogue’ or ‘Women’s Weekly’, as long

as I can keep the superstition that comes with the label “scientific” at a healthy
distance.
(2) Aune [1986], p. 11

As it now exists, the subject of metaphysics can be described by a
distinction that became standard in the seventeenth and eighteenth cen-
turies. According to this distinction, metaphysics has two principal divi-
sions: general metaphysics and special metaphysics. General metaphysics
include ontology and most of what has been called universal science; it is
concerned, on the whole, with the general nature of reality: with problems
about abstract and concrete being, the nature of particulars, the distinc-
tion between appearances and reality, and the universal principles holding
true of what has fundamental being. Special metaphysics is concerned with
certain problems about particular kinds or aspect of being. These special
problems are associated with the distinction between the mental and the
physical, the possibility of human freedom, the nature of personal identity,
the possibility of survival after death, and the existence of God.
Comment. Personally, I’m more interested in survival before death than after; but
I don’t think this counts in the face of the unfathomable lack of depth of the issues
addressed.
(3) Harré [1972], p. 100.
Metaphysics is the study of the most general categories within which we
think.
Philosophy in the dialectical tradition is often characterised by a lack of rel-
evance. The following quotation is meant to give an example of what I have in
mind.
Quotations 67.23. (1) Bhaskar [1975], p. 21.
[[M]]en in their social activity produce knowledge which is a social product
much like any other, which is no more independent of its production and
the men who produce it than motor cars, armchairs or books, which has
its own craftsmen, technicians, publicists, standards and skills and which
is no less subject to change than any other commodity. This is one side
of ‘knowledge’. The other is that knowledge is ‘of ’ things which are not
produced by men at all: the specific gravity of mercury, the process of
electrolysis, the mechanism of light propagation. None of these ‘objects of
knowledge’ depend upon human activity. If men ceased to exist sound would
continue to travel and heavy bodies fall to the earth in exactly the same
way, though ex hypothesi there would be no-one to know it.
Comment. Of men and things.13 There isn’t much left of the delicacy of the prob-
lem of conceptual cognition. Where would knowledge about money, market, lit-
erature, etc. range in Bhaskar’s view? To me Bhaskar’s idea of knowledge looks
extremely narrow: it is doubtful that even mathematics would classify, without
adhering strictly to Platonism. Certainly linguistics can’t count as knowledge, if
“knowledge is ‘of ’ things which are not produced by men at all“; or, perhaps, all
13 How about this: women in their private passivity happen to conceive men which poses a
threat to the security of men and their social activity once women turn away from them.
that is created by God? I rather suspect that what manifests itself here is an extreme
form of ‘hard science’ idea of knowledge, as rigid as it is unaware.
(2) Bhaskar [1975], p. 38.
For the transcendental realist it is not a necessary condition for the existence
of the world that science occurs. But it is a necessary condition for the
occurrence of science that the world exists and is of a certain type.
Comment. Poor Kant! Is it necessary to remind readers of Kant’s characterization of
transcendental realism like, for instance, in quotation 60.5 (3) in these materials?
(3) Rée [1989], p. 315.
Transcendental Arguments move from the premise that a certain kind of
knowledge is possible (say, arithmetic), to the conclusion that a priori “con-
ditions of its possibility” must be fulfilled. The view that such arguments
are crucial to philosophy is due to Kant’s proposal for a “transcendental”
philosophy[[.]]
Comment. I suspect this is somewhat based on what Kant says in quotations
59.4 (2) and (3), but it still hurts.
Quotation 67.24. Ryan [1982], p. 9.

In very broad terms, deconstruction consists of a critique of meta-
physics, that branch of philosophy, from Plato to Edmund Husserl to Paul
Ricoeur, which posits first and final causes or grounds, such as transcen-
dental ideality, material substance, subjective identity, conscious intuition,
prehistorical nature, and being conceived as presence, from which the mul-
tiplicity of existence can be deduced and through which it can be accounted
for and given meaning.
Comment. These terms are very broad indeed; so broad that I consider them thor-
oughly useless, except as an example of what I want to avoid in this study. More
specifically: in so far as deconstruction indulges in the fancy illusion that critique is
possible without taking position, it is irrelevant; in so far as it does not, it has not
yet come into existence.
67d. The concept of dialectic. With another ambitious concept coming up,
things can’t really be expected to get better.
Quotations 67.25. (1) McKeon [1975], p. 25.

[[L]]ater philosophers have found themselves in historical situations in which
they faced a choice between skeptical dialectic and dogmatic dialectic, func-
tional phenomenal dialectic and transcendental idealistic dialectic, human-
istic cultural dialectic and materialistic historical dialectic.
(2) Hall [1967], p. 385.
[[A]]mong the more important meanings of the term have been (1) the
method of refutation by examining logical consequences, (2) sophistical
reasoning, (3) the method of division or repeated logical analysis of genera
into species, (4) and investigation of the supremely general abstract notions
by some process of reasoning leading up to them from particular cases or

hypotheses, (5) logical reasoning or debate using premises that are merely
probable or generally accepted, (6) formal logic, (7) the criticism of the
logic of illusion, showing the contradictions into which reason falls in trying
to go beyond experience to deal with transcendental objects, and (8) the
logical development of thought or reality through thesis and antithesis to a
synthesis of these opposites.
Quotations 67.26. (1) Krings [1964], p. 58; my translation

[[T]]he dialectic is a method by means of which the knowing can be under-
stood, but not contentually increased.
(2) Bubner [1973], p. 129; my translation.
If dialectic is anything at all, then it is a method. If it is a method, then
it is a procedure to gain knowledge. Thus it is not primarily a content or
a particularly distinguished complex of contents. There are no reserves of
specific objects where dialectic were to dwell, so to speak, and other ways
of thinking excluded.
(3) Kojève [1947], p. 176.
If you please, the Hegelian “method” is purely “empirical” or “positivist”:
Hegel looks at the Real and describes what he sees, and nothing but what
he sees. In other words, he has the “experience” (Erfahrung) of dialectical
Being and the Real, and thus he makes their “movement” pass into his
discourse which describes them.
(4) Kojève [1947], p. 259.
Hegelian Dialectic is not a method of research or of philosophical ex-
position, but the adequate description of the structure of Being, and of the
realization and appearance of Being as well.
(5) Hintikka [1981], p. 213.
[[T]]he logic of dialectic can be expected to be nothing but the logic of
dialogue. Moreover, the kind of dialogue from which dialectic originated
is easily identified. It is the Socratic dialogue which consists in questions
whose answers are hoped to elicit information from the answerer, including
information which he knew earlier only tacitly. Hence the logic of dialogue
relevant here is that of sequences of information-seeking questions and an-
swers.
Comment. There seems no trace of awareness of the notion of dialectic that emerged
with transcendental philosophy. Note the phrase “dialectic can be expected”.
Quotations 67.27. (1) Ollman [1993], p. 10.

Dialectics is not a rock-ribbed triad of thesis-antithesis-synthesis that serves
as an all-purpose explanation; nor does it provide a formula that enables
us to prove or predict anything; nor is it the motor force of history. The
dialectic, as such, explains nothing, proves nothing, predicts nothing, and
causes nothing to happen. Rather, dialectics is a way of thinking that brings
into focus the full range of changes and interactions that occur in the world.
As part of this, it includes how to organize a reality viewed in this manner
for purposes of study and how to present the results of what one finds to
others, most of whom do not think dialectically.
(2) Ollman [1993], p. 11.
Dialectics restructures our thinking about reality by replacing the com-
mon sense notion of “thing,” as something that has a history and has exter-
nal connections with other things, with notions of “process,” which contains
its history and possible futures, and “relation,” which contains as part of
what it is its ties with other relations. Nothing that didn’t already exist has
been added here.
(3) Ollman [1993], p. 23.
First and foremost, and stripped of all qualifications added by this
or that dialectician, the subject of dialectics is change, all change, and
interaction, all kinds and degrees of interaction.
Comment. This is dialectic stripped of Hegel’s qualifications too: ‘change’ is taken
from representation (“Vorstellung”); but Hegel’s dialectic does not primarily deal
with representations; contrast quotation 65.16 (1) in these materials.
Dialectic, as an isolated feature of Hegel’s philosophy, separated from the ide-

alistic and transcendental background, seems to have become a target for some
progressive-dynamic philosophers.
Quotations 67.28. (1) Bunge [1975], p. 63

We construe dialectics as an ontological theory. And we submit that di-
alectical ontology has a plausible kernel surrounded by a mystical fog. The
plausible kernel of dialectics is constituted by the hypotheses (i) that every
thing is in some process of change or other, and (ii) that at certain points
in any process new qualities emerge[[.]]
Comment. The dashing venture of commonplace philosophy.
(2) da Costa and Wolf [1980], p. 191.
McGill and Parry’s elucidation of the central dialectical position on the
unity of opposites is a model of clarity and balanced elucidation. Their
treatment makes use of techniques and terminology of contemporary ana-
lytic philosophy and does much to make the principle of the unity of oppo-
sites intelligible to philosophers trained in the analytic tradition. For this
reason, their formulation of the principle is more immediately amendable
to the sort of formalization we are interested in than a treatment couched
in more traditional Hegelian or Marxist language.
[[. . .]]
The three interpretations [[. . .]] with which we are concerned are:
“4. A concrete system or process is simultaneously determined by op-
positely directed forces, movements, tendencies, i.e., directed toward A and
-A.
5. In any concrete continuum, whether temporal or non-temporal, there

is a middle ground between two contiguous opposite properties A and -A,
i.e., a stretch of the continuum where it is not true that everything is either
A or -A.
6. In any concrete continuum, there is a stretch where something is
both A and -A.” (McGill and Parry 1948, p. 422)
Comment. Do I need to mention that I want to keep a healthy distance to this
brand of arrogant naivety?
(3) Barth [1981], p. 46.
A 1. Reconstruction: principles and components
Dialectical philosophers have, again and again, requested of the critics
of the Hegelian system of logic that their criticism satisfy two conditions.
The first condition is that it should be “immanent”, the second that it
should not be too piecemeal but rather holistic. With these requests I am
in full sympathy. It follows that all attempts at formalizing Hegel’s logic in
the language of any modern system of logic whatsoever should be rejected,
whether higher-order, four-valued or what not. Such attempts are ad hoc,
unhistorical and unphilosophical.
No modern system of logic has a model even vaguely resembling Hegel’s.
As against those who have tried to formalize certain aspects of Hegelian
logic in some modern system, I hold that a complete description of an
exotic logic should contain at least the following elements:
a. a survey of the categories in which practitioners of that logic think
(logical and other categories) and speak (syntactic categories);
b. a description of the model or model structure characteristic of that
logic and composed of the said categories in a certain array;
c. a survey of the technical terms employed in the object system(s);
d. a survey of the syntactic principles assumed in or underlying the use
of language by which the system is traditionally described;
e. a semantics, i.e. the relationship according to the object system(s)
between model structure or ontology on the other hand and terminology
and syntax on the other; this semantics should issue in the truth con-
cept — if any — in that system, and in its definitions of “deduction” and
[[“]]induction”;
f . a survey of the principles of inference (the set of supposedly “good”
inference rules), such that the connection — or lack of connection — with
the model structure, with the syntax and with the semantics becomes as
clear as possible;
g. a discussion of the consequence of the answers to the above questions
for the possibility of discussion (criticism and defense) of statements of the
kind that are generated by the system;
Comment. The list goes on to i; but I don’t think it is necessary to go through the
pain of quoting it further. The point is clear: here is good advice for the field worker
who is going out to investigate some largely unexplored and/or misunderstood tribe
with an “exotic logic”. What is particularly curious in this piece is the similarity
to a certain kind of — not altogether obsolete — anthropology: the naivety with

which alien categories are employed (such as model structure, semantics, definition
of induction, etc. in this particular case) in order to make sense of the activities
of “practitioners of that logic”. It may be worth emphasizing at this point, as a
member of the strange tribe concerned, that I intend to somewhat reverse the role
model and investigate the strategies of people who adhere to something like the
classical “truth concept” to respond to the challenge of the various forms of Cantor’s
diagonal method. But I want to avoid the arrogance of a position that approaches
other strategies as “exotic” ones.14
Quotations 67.29. (1) Marx [1887], p. XXX; translation of Marx [1873], p. 27.
My dialectic method is [[regarding its foundation]] not only different
from the Hegelian, but is its direct opposite. To Hegel, the life-process of
the human brain, i.e., the process of thinking, which, under the name of
“the Idea,” he transforms into an independent subject, is the demiurgos of
the real world, and the real world is only the external, phenomenal form of
the “the Idea.” With me, on the contrary, the ideal is nothing else than the
material world reflected by the human mind, and translated into forms of
thought.
[[. . .]] The mystification which dialectic suffers in Hegel’s hands, by no
means prevents him from being the first to present its general form of
working in a comprehensive and conscious manner. With him it is standing
on its head. It must be turned right side up again, if you would discover
the rational kernel with the mystical shell.
(2) Marx [1887], pp. 41 f; translation of Marx [1867], p. 85.
A commodity appears, at first sight a very trivial thing, and easily
understood. Its analysis shows that it is, in reality, a very queer thing,
abounding in metaphysical subtleties and theological niceties. So far as it
is a value in use, there is nothing mysterious about it, whether we consider
it from the point of view that by its properties it is capable of satisfying
human wants, or from the point that those properties are the product of
human labour. It is as clear as noon-day, that man, by his industry, changes
the forms of the materials furnished by nature, in such a way as to make
them useful to him. The form of wood, for instance, is altered, by making
a table out of it. Yet, for all that, the table continues to be that common,
every-day thing, wood. But, so soon as it steps forth as a commodity, it is
changed into something transcendent. It not only stands with its feet on
the ground, but, in relation to all other commodities, it stands on its head,
and evolves out of its wooden brain grotesque ideas, far more wonderful
than “table-turning” ever was.
Comment. The translation is terrible; “wunderlich”, for instance, is translated as
“wonderful”.
14 May I say it? I’d rather be exotic than idiotic.

(3) Marx [1887], pp. 42 f; translation of Marx [1867], p. 86.

A commodity is therefore a mysterious thing, simply because in it the
social character of men’s labour appears to them as an objective character
stamped upon the product of that labour, because the relation of the pro-
ducers to the sum total of their own labour is presented to them as a social
relation, existing not between themselves, but between the products of their
labour. [[. . .]] In order, therefore, to find an analogy, we must have recourse
to the mist-enveloped regions of the religious world. In that world the pro-
ductions of the human brain appear as independent beings endowed with
life, and entering into relation both with one another and the human race.
So it is in the world of commodities with the products of men’s hands. This
I call the Fetishism which attaches itself to the products of labour, so soon
as they are produced as commodities, and which is therefore inseparable
from the production of commodities.
This Fetishism of commodities has its origin, as the foregoing analy-
sis has already shown, in the peculiar social character of the labour that
produces them.
(4) Marx [1887], p. 45; translation of Marx [1867], p. 88.
[[W]]hen we bring the products of our labour into relation with each other as
values, it is not because we see in these articles the material receptacles of
homogeneous human labour. Quite the contrary: whenever, by an exchange,
we equate as values our different products, by that very act, we also equate,
as human labour, the different kinds of labour expended upon them. We
are not aware of this, nevertheless we do it. Value, therefore, does not stalk
about with a label describing what it is. It is value, rather, that converts
every product into a social hieroglyphic. Later on, we try to decipher the
hieroglyphic, to get behind the secret of our own social products; for to
stamp an object of utility as a value, is just as much a social product as
language.
(5) Marx [1887], p. 56; translation of Marx [1867], p. 99.
It is plain that commodities cannot go to market and make exchanges of
their own account. We must, therefore, have recourse to their guardians[[.]]
[[. . .]] In order that these objects may enter into relation with each other
as commodities, their guardians must place themselves in relation to one
another, as persons whose will resides in those objects and must behave in
such a way that each does not appropriate the commodity of the other, and
part with his own, except by means of an act done by mutual consent.
(6) Marx [1887], p. 26, footnote 1; translation of Marx [1867], p. 72, footnote 21.
[[O]]ne man is king only because other men [[relate as]] subjects to him. They,
on the contrary, imagine that they are subjects because he is king.
Comment. The original translation in the bracketed part is “stand in relation of”.
This does not sufficiently bring out that it is the men themselves who relate (German
original: “sich verhalten”) as subjects to the king; hence my change.
(7) Marx [1845], p. 5, (zweite Feuerbach These).

Die Frage, ob dem menschlichen Denken gegenständliche Wahrheit zu-
komme — ist keine Frage der Theorie, sondern eine praktische Frage. In der
Praxis muß der Mensch die Wahrheit, i. e. Wirklichkeit und Macht, Dies-
seitigkeit seines Denkens beweisen. Der Streit über die Wirklichkeit oder
Nichtwirklichkeit des Denkens — das von der Praxis isoliert ist — ist eine
rein scholastische Frage.
Comment. Note the phrase “Wahrheit, i. e. Wirklichkeit und Macht”. Cf. Lenin in
quotation 69.2 (2) in these materials, and my comment on what I call “Hume’s
fallacy” in remark 118.12 (1) in the groundworks.
(8) Marx [1845], p. 5, (elfte Feuerbach These).
Die Philosophen haben die Welt nur verschieden interpretiert, es kömmt
darauf an, sie zu verändern.
(9) Marx [1852], p. 146.
Men make their own history, but not of their own free will; not under cir-
cumstances they themselves have chosen but under the given and inherited
circumstances with which they are directly confronted. The tradition of the
dead generations weighs like a nightmare on the minds of the living. And,
just when they appear to be engaged in the revolutionary transformation
of themselves and their material surroundings, in the creation of something
which does not yet exist, precisely in such epochs of revolutionary crisis
they timidly conjure up the spirits of the past to help them; they borrow
their names, slogans and costumes so as to stage the new world-historical
scene in this venerable disguise and borrowed language.
Comment. Cf. Hegel in quotation 66.7 (2) in the previous chapter.

[[D]]ialectic has played a very unfortunate rôle not only in the development
of philosophy, but also in the development of political theory. A full un-
derstanding of this unfortunate rôle will be easier if we try to see how
Marx originally came to develop such a theory. We have to consider the
whole situation: Marx, a young man who was progressive, evolutionary and
even revolutionary in his thought, listening to the lectures of Hegel, then a
famous professor in Berlin.
Comment. The emphasis here, I suggest, lies on the phrase “consider the whole
situation”. The particular situation to consider would be a young Marx at the tender
age of twelve or thirteen at the Berlin University listening to Hegel. Marx was
born in 1818, Hegel died in 1831. In the German translation of this paper, Popper
seems to have distanced himself from this little bit of factual travesty and he only
talks about Marx getting under the influence of Hegel.15 Anyway, what Popper is
concerned with is the whole situation and a full understanding, not with silly little
facts and their insubordinate contingency.
15 Popper [1965], p. 287.
Quotations 67.31. (1) Fogarasi [1954], p. 66.

Zwei einander widersprechende Sätze können zugleich wahr sein, wenn in
ihnen die auf verschiedenen (zeitlichen, geschichtlichen usw.) Beziehungen
beruhenden Widersprüche zum Ausdruck kommen.
(2) Narski [1973], p. 299.
Die ursprüngliche Problem-Antinomie für die erkenntnistheoretische Ana-
lyse der Bedeutung ist die Antinomie von Zeichen und Bedeutung. Wir
formulieren sie folgendermaßen: „Die Bedeutung unterscheidet sich vom Zei-
chen und unterscheidet sich nicht von ihm.“ Das heißt mit anderen Worten,
sie existiert als etwas relativ Selbständiges im Unterschied zum Zeichen
„existiert zur gleichen Zeit und nicht zur gleichen Zeit“.
(3) Narsky [1975], p. 80.
[[D]]er objektive dialektische Widerspruch ist der innerliche Widerspruch
von zwei Hauptaspekten oder von zwei entgegengesetzten Tendenzen; und
diese Aspekte und Tendenzen sind wechselseitig bedingt, zusammengebun-
den und voneinander in ihrer Entwicklung abhängig.
(4) Krajewski [1981], p. 183.
Das Problem der Existenz der dialektischen Logik und ihres Verhältnisses
zur formalen Logik erregte in der UdSSR, in Polen und der DDR viele
Auseinandersetzungen, die in den fünfziger Jahren besonders heftig waren.
Allmählich überwog der folgende Standpunkt: wenn man von einer dialek-
tischen Logik spricht, darf man dies nur im Sinne einer Theorie der Ent-
wicklung von Begriffen und des ganzen Wissens. (Ich will bemerken, daß
eine solche Theorie noch nicht ausgearbeitet worden ist.) Dialektische Lo-
gik kann jedoch keine Alternative zur formalen bilden. Die letztere hat eine
universelle Gültigkeit und alle Erwägungen, auch die dialektischen müs-
sen ihren Regeln folgen. Dieser Standpunkt dominiert mittlerweile bei den
Philosophen der erwähnten Länder, jedenfalls unter jenen, die direkt mit
der Wissenschaft zu tun haben. Die Ansicht, es gäbe eine „höhere“ dialek-
tische Logik, wird meistens als dogmatisches Überbleibsel angesehen und
manchmal „Hegelei“genannt.
Jetzt während meines Aufenthalts in der Bundesrepublik, habe ich mit
Erstaunen festgestellt, daß dieses Überbleibsel hier sehr lebendig ist! Viele
westdeutsche Philosophen betrachten, wahrscheinlich der Hegelschen Tra-
dition folgend, Dialektik als eine Alternative zur formalen Logik. Ich bin
überzeugt, daß jede Entwertung der formalen Logik zu Unklarheit und Ver-
wirrung führt.
Comment. Note the alternative: dialectical logic and formal logic; not: classical
logic. Anyway: I hope it is clear that the present study is unashamedly engaged
with the dogmatic remnant (Hegelei), and that from a position of competence in
mathematical logic, a position which I prefer to that of any kind of ideological
position, even if it comes in the guise of anti-dogmatism.
The last quotation in this paragraph concerns logic.

Quotation 67.32. Harris [1987], p. 1.

[[C]]ontemporary logicians have asserted that all logical truths are analytic
and tautological, so that they can never give us new knowledge, even though
they may serve to derive hitherto unnoticed consequences from what is
already known. But this view of logic is a misconception. After all, it is
admitted on all hands that logic is the science that sets out and develops
the principles fundamental to the method of every science and identifies the
norms that justify the claim of any discipline to be a science at all. And
nobody nowadays is likely to deny that science is of practical importance.
If so, surely logic, at least indirectly, must have relevance to the interest of
practical living.
Comment. What is characteristic in this piece is the lack of something to point at
when it comes to reject the tautological nature of logic. Instead we are led through
a jungle of “nobody nowadays is likely to deny”-reasoning to end up with a “surely,
at least indirectly, it must”-conclusion. There is more to the inadequacy of logic
than that.
§ 68. Struggling with transcendental philosophy
Transcendental philosophy today is a sore sight. It may not come as a surprise

that a sophisticated view such as Kant’s is vulnerable to misunderstandings of all
sorts and sometimes terrible mutilation. In particular, basic concepts in Kant’s
transcendental philosophy were subjected to everything from a change of meaning
to an emptying of meaning. In some areas an inflation of “transcendentals” raged,
with a transcendental I-don’t-know-what popping up like mushrooms.16
In this paragraph I pick sporadic examples from attempts to make sense of
Kant’s transcendental philosophy for the sake of illustrating the situation.
68a. How many Kants? I begin with a quotation of a rather general kind
which, however, receives its relevance from the fact that it is by the author of the
almost authoritative translation of Kant’s Critique of Pure Reason into the English
language,17 as well as a Commentary which doesn’t seem to be regarded as equally
authoritative, if my impression from selected reading is correct. The quotation is
taken from his Commentary.
Quotation 68.1. N. K. Smith [1918], p. vii.

The Critique of Pure Reason is more obscure and difficult than even a
metaphysical treatise has any right to be. The difficulties are not merely due
to defects of exposition; they multiply rather than diminish upon detailed
study; and, as I shall endeavour to show in [[my]] Commentary, are traceable
to two main causes, the composite nature of the text, written at various
16 Amongst others I have come across “refutation”, “analysis”, “situation”, “critique”, “argu-
ment”, and “meditation” adorned with the adjective transcendental.

17 At least still at the time that this was written (roughly late eighties of the twentieth
century). New translations only appeared in the mid-nineties.

§ 68. STRUGGLING WITH TRANSCENDENTAL PHILOSOPHY 919
dates throughout the period 1769–1780, and the conflicting tendencies of

Kant’s own thinking.
Comment. There is a third cause for difficulties which N. K. Smith doesn’t consider
and this is a lack of metaphysical understanding, be that on the translator’s or on
the reader’s part. The least I want to say is that I don’t share N. K. Smith’s attitude;
differently put, my difficulties with reading Kant’s Kritik der reinen Vernunft did
not multiply upon detailed study, and I find Kant’s thinking quite consistent, cum
grano salis; but then, I did not read Kant as N. K. Smith translated him. To readers
who want to explore the possibilities of a consistent reading of Kant’s Critique of
Pure Reason, I recommend Allison [1983].
Quotation 68.2. Priest [1995], p. 82.

Interpreting the Critique is a very sensitive issue [[. . .]]. This is so partly
because of the abstract and jargonated way in which Kant expresses himself.
Comment. Here it is, the jargon evoked at the end of quotation 56.3 (1). An Eng-
lish reader who hasn’t been pampered enough by the translator, or rather over-
pampered? For an example of what Priest considers Kant’s jargon, see the end of
quotation 68.23 (2) in the present paragraph.
Although my way of characterizing Kant’s transcendental philosophy mainly by

means of citation in chapter XV may look fairly safe, there is, apparently, still room
for quite a different view of what is suggested there; a view which does not classify
Fichte and Hegel as belonging to the same tradition. At the same time, there is a
view which doesn’t place Kant in the tradition of Hume.

Hume, by his criticism of the concept of causality, awakened him from
his dogmatic slumbers so at least he says, but the awakening was only
temporary, and he soon invented a soporific which enabled him to sleep
again. Hume, for Kant was an adversary to be refuted[[.]]
Comment. The phrase “dogmatic slumber” can be found not only in regard to Hume
but also in a letter to Garve with regard the Antinomy of Pure Reason.18
(2) Russell [1946], p. 646.
German philosophers, from Kant to Hegel, had not assimilated Hume’s ar-
guments. I say this deliberately, in spite of the belief which many philoso-
phers share with Kant, that his Critique of Pure Reason answered Hume. In
fact, these philosophers—at least Kant and Hegel—represent a pre-Humian
type of rationalism, and can be refuted by Humian arguments.
Comment. I have to admit that I haven’t assimilated Hume’s arguments either, and
I want to add that this is not just due to the fact that I took him too seriously in
his invitation to throw a book of metaphysical illusion into the fire.19 An “assimi-
lation” of Hume’s arguments can only mean, in my understanding, to follow Hume
18 Cf. Allison [1983], p. 35.
19 Cf. the end of chapter XIV.
in excluding practice from epistemology; in other words, to subscribe to what I call

“Hume’s fallacy” in remark 118.12 (1) in the groundworks.
Quotations 68.4. (1) Popper [1963], p. 176.

Kant believed in the Enlightenment. He was its last great defender. I realize
that this is not the usual view. While I see Kant as the defender of the
Enlightenment, he is more often taken as the founder of the school which
destroyed it — of the Romantic School of Fichte, Schelling, and Hegel. I
contend that these two interpretations are incompatible.
Fichte, and later Hegel, tried to appropriate Kant as the founder of
their school. But Kant lived long enough to reject the persistent advances
of Fichte, who proclaimed himself Kant’s successor and heir. In A Pub-
lic Declaration Concerning Fichte which is too little known, Kant wrote:
‘May God protect us from our friends. . .. For they are fraudulent and per-
fidious so-called friends who are scheming for our ruin while speaking the
language of good-will.’ It was only after Kant’s death, when he could no
longer protest, that this world-citizen was successfully pressed into the ser-
vice of the nationalistic Romantic School, in spite of all his warnings against
romanticism, sentimental enthusiasm, and Schwärmerei.
Comment. The expression Schwärmerei with its adjective schwärmerisch seems to
have been quite popular with Kant. It can be found in Kant [1787] (B), p. xxxiv,
where it is translated as fanaticism by N. K. Smith (p. 32), and p. 128, where it is
translated as enthusiasm (p. 128), and further in Kant [1783], p. 293, to characterize
Berkeley’s idealism and there it is translated as visionary by Carus and Ellington,20
and appears in the latter’s German-English list of terms as vagaries.21 An eminent
use of the word in modern German language would be to characterize a very positive
attitude of a 14 year old school girl towards her teacher (not necessarily male); or
that of Popper here to Kant.
(2) Popper [1940], p. 414.
The struggle of the earlier rationalists and empiricists was thoroughly dis-
cussed by Kant, who tried to offer some synthesis—a compromise, or rather,
a modified form of empiricism. His main interest was to reject pure ratio-
nalism. In his Critique of Pure Reason, he maintained that the scope of
our knowledge is limited to the field of possible experiences and that pure
reasoning beyond this field—the attempt to build a metaphysical system
out of pure reason—has no justification whatsoever.
Comment. The claim that Kant “thoroughly discussed” the earlier rationalists and
empiricists seems to me a bit along the line of factual travesty already mentioned
in quotation 67.30. What I also find interesting is that Popper is apparently preoc-
cupied with knowledge about an outer world; not the Socratic reflexive knowledge,
the spirit of the Delphic Apollon: know thyself, γνῶθη σεαυτόν.22
20 Quotation 60.5 (8) in these materials.

21 Carus and Ellington [1977], p. 125.
22 To do justice to contemporary academic reality, read: Sell thyself.
A last couple of quotations somewhat relating to the shift from Hume to Kant
and from Kant to Hegel.
[[F]]ar from being content with Kant’s refutation of metaphysics, philoso-
phers busied themselves with building up new metaphysical systems based
on pure speculation[[.]]
Comment. Kant didn’t seem all that content with his own refutation of metaphysics
either; or why would he have thought of busying himself with building a “system of
pure (speculative) reason”?23
(2) Reichenbach [1951], p. 72.
Hegel has been called the successor of Kant; that is a serious misun-
derstanding of Kant and an unjustified elevation of Hegel. Kant’s system,
though proved untenable by later developments, was the attempt of a great
mind to establish rationalism on a scientific basis. Hegel’s system is the
poor construction of a fanatic who has seen one empirical truth and at-
tempts to make it a logical law within the most unscientific of all logics.
Whereas Kant’s system marks the peak of the historical line of rationalism,
Hegel’s system belongs in the period of decay of speculative philosophy
which characterizes the nineteenth century.
Comment. Speaking of fanatics . . . . Compare also section 86e.
Quotation 68.6. Solomon [1974], p. 277.
Kant’s great anti-sceptical move in the ‘Transcendental Analytic’ is the mu-
tual rejection of the Cartesian claim to the epistemological priority of the
mental and the empiricist model of knowledge as passive-receptive ‘repre-
sentation’. In place of both, Kant supplies the revolutionary notion of a
priori synthesis, the idea that objects are not simply given in experience
but rather constituted or synthesized as a necessary connection between
events, e.g. causal connection, is not given in experience (thus agreeing
with Hume) but is supplied by the Understanding as a necessary condition
of every experience. But Kant, like most great revolutionaries, is not quite
bold enough to step surefootedly on the new intellectual ground he had
liberated. Even as he attacks the Cartesian epistemological priority of the
mental, he still finds problematic the idea that we could know things-in-
themselves. And even as he attacks the ‘myth of the given’ of empiricism
and the ‘blindness’ of non-conceptualized experience, he hangs on to the
empiricist notion of ‘impressions’ (Empfindung) which are given in atom-
istic bits and then synthesized to give us objects. And as he argues for the
active role of the Understanding and the Imagination in perception, Kant
retains the conservative belief that there is but one set of categories and
consequently but one possible conception of the world. The move from the
idea that we supply the categories by which objects are synthesized to the
idea that we might supply other categories doesn’t entice him.
23 Cf. Kant [1781] (A), p. xxi; quotation 60.3 (6) in these materials.
Comment. As regards the last part: Kant explicitly makes the restriction “to us
humans at least”. Interesting also: “the idea that we might supply other categories”,
as if this were a matter of free will. Have you ever tried to think four-dimensionally?
In this context: cf. ‘Folk’ in confrontations 113.12 in the groundworks.
68b. Transcendental idealism, the analytic-synthetic distinction, and

universally valid knowledge. I begin with a few quotations which I take to
reflect a bit of the poor reception of Kant’s attempt to allow for what he labeled
empirical realism within his transcendental idealism.
Quotations 68.7. (1) Sayers [1985], p. xiv.

By realism I mean the belief that there is an objective, material world,
which exists independently of consciousness, and which is knowable by con-
sciousness.
Comment. This characterization involves knowledge. Merging, however, the meta-
physical claim with the epistemological claim makes short work of Kant’s distinction
and, as Sayers then realizes, excludes Kant’s empirical realism as realism.
(2) Sayers [1985], p. xiv.
If even Kant’s empirical realism is allowed to count as a form of realism, it
is difficult to think of any philosopher who is not a realist. Even Berkeley
would qualify; and the term becomes useless.
Comment. Roughly, the question comes down to whether Kant is right in his claim
that transcendental realism entails dogmatic idealism.24
(3) N. K. Smith [1918/23], p. 612.
In the Critique of Pure Reason, when arguing on phenomenalist lines, Kant
has maintained that on the basis of transcendental idealism an empirical
realism can be established; but nowhere does he face the difficulties which
such a position involves. In one set of passages he refers us to things in
themselves as the primary causes of our sensations; in yet another set of
passages it is physical stimuli which are cited as the causes of our sensations.
Nowhere does he explain how, if objects be appearances, conditioned by
mental processes, they can also be causes, initiating and yielding material
for these processes.
Comment. I don’t know how difficult it is to realize that N. K. Smith mistakes Kan-
t’s notion of empirical realism for some kind of transcendental realism, or rather
dogmatic or common sense realism, i.e., he keeps asking for what is behind our
experiences, or causing our experiences, when Kant speaks of the reality of expe-
rience. Reader are encouraged to look up quotation 60.5 (4) in these materials.
I shall make a similar move with regard to mathematical reality, and I expect a
similar lack of comprehension on the part of common sense realists.
24 Cf. quotations 60.5 and 60.7 as regards Kant’s assessment of his own idealism versus that
of Berkeley.
How does knowledge relate to the world and how is it possible to have knowledge
which is necessary and universally valid?
Popper’s account of Kant’s answer to this question is hardly more than a per-
siflage, but it is still worth looking at it for the sake of getting aware of what can
be done to a philosophical idea — given a minimum of philosophical education: the
confusion of what the world is like and what our knowledge of the world is like.

[[Kant’s]] reasoning was somewhat as follows: The mind can grasp the
world or, rather, the world as it appears to us, because this world is not
utterly different from our mind—because it is mind-like. And it is so, be-
cause in the process of obtaining knowledge, of grasping the world, our
mind is, so to speak, actively digesting all that material which enters it
by our senses. It is forming, moulding this material; it impresses on it its
own intrinsic forms—the forms of our thought. What we call “nature”—the
world in which we live, the world as it appears to us—is already a world
digested, is a world formed by our mind. And being thus assimilated by the
mind, it is mind-like.
Comment. Popper wavers between “the world” and “the world as it appears to us”. It
is not just that a common sense champion struggles with the intricate argumentation
of a speculative philosopher. Kant himself did not clearly distinguish between the
world itself and the world as it appears to us. It may be worthwhile mentioning
that Kant never asked the question “How can our mind grasp the world?”
(2) Popper [1940], p. 420.
[[T]]he idealistic answer, which has been varied by different idealistic philoso-
phers but remains fundamentally the same, namely “Because the world is
mind-like”, has only the appearance of an answer. But we shall see that it
has not the slightest justification, if we only consider some analogous argu-
ment like: “How can this mirror reflect my face?”—“Because it is face-like.”
Comment. Do you feel that there is something wrong with this analogy? The point
is simple: Popper has swapped the grammatical subject (nominative) and object
(accusative) in his analogy: the mind grasps the world, because the world (as it
appears) is mind-like. Now, if you take the analogy seriously, it runs: the mirror
can reflect my face, because my face (as it appears) is mirror-like. What this means
is that my face is such that a mirror can reflect it; unlike my thoughts, which cannot
be reflected by a mirror. Whether Popper did this out of foolishness or mischief is
a question which I shall not engage with, simply because I don’t think they are
separable in Popper’s thought.
Quotations 68.9. (1) Hamlyn [1967], p. 29.

Kant’s criterion of objectivity is the criterion of intersubjectivity—validity
for all men; he is pointing out that from the point of view of the critical
philosophy something may be objective in this sense without being a feature
of something independent of the mind.
(2) Hamlyn [1967], pp. 30 f.

The problem is how we can distinguish between what is contributed by the
mind and what is not, between the self and not-self, as Fichte put it. In
Kant’s view objectivity was equivalent to validity for all men, but that it was
at all possible to distinguish between what was due to the mind and what
was not seemed guaranteed only by the existence of things-in-themselves.
With the rejection of the latter, experiences and experiencer became only
two sides of the same coin. For this reason the general trend of idealism
was toward the coherence theory of truth—the view that experiences and
judgments are true to the extent that they cohere with one another, forming
a coherent system.
Quotations 68.10. (1) Rescher [1973], p. 3

[[Conceptual idealism]] maintains that the concepts we standardly employ
in constituting our view of reality — even extramental, material reality —
involve an essential (though generally tacit) reference to minds and their
capabilities. [[. . .]] On this view, what the mind ‘makes’ is not nature it-
self but the mode-and-manner-determining categories in terms of which we
conceive it. Accordingly, it is maintained that the mind shapes [[. . .]] not
nature itself, but nature as it is for us.
Comment. The strong emphasis on the mind almost seems to suggest that Rescher
grants the mind a monopoly on concepts.
(2) Putnam [1978], p. 1.
With Kant a new view emerges: the view that truth is radically mind-
dependent.
Comment. Compared with Popper’s crude assessment of Kant this is almost so-
phisticated. Still, it is very vague.
Quotation 68.11. Allison [1983], p. 76 f.

Kant [[suggests]] that the synthetic judgment contains a “determina-
tion,” whereas the analytic judgment contains only a “logical predicate.”
Since Kant maintains both that existential judgments are synthetic and
that ‘existence’ is not a real predicate, this account of synthetic judgments
obviously cannot be accepted as it stands. In other words, it cannot be
maintained that the possession of a logical predicate that is also a real
predicate is criterial for the syntheticity of a judgment is synthetic, not
because its logical predicate ‘existence’ is a real predicate or determina-
tion, but rather because its logical subject is one, and the judgment simply
asserts the existence of an object corresponding to this subject.
Comment. What Kant says is that ‘being’ (German: “Sein”) is not a real predicate.25
To translate “Sein” as “existence” in a context which explicitly points out its role as
a copula strikes me as curious; existence just doesn’t serve as a copula.
25 Kant [1781], p. 598; see quotation 60.12 in these materials.
68c. The transcendental unity of apperception and the system of cat-

egories. The present section is concerned with the fate of KT6 and KT7, the tran-
scendental deduction of the categories and the transcendental unity of apperception.
The task of a systematic deduction of the categories from some form of transcen-
dental unity, the aspect of Kant’s enterprise which was so strongly emphasized by
Fichte, seems to have widely abandoned though there still seem to be scattered
attempts to find something within knowledge which fills the place of Kant’s tran-
scendental unity of apperception, the core of Kant’s transcendental deduction of
the categories.
I begin with a quotation from Strawson’s Bounds of Sense as a deterrent.
Quotation 68.12. Strawson [1966], p. 15.

There are limits to what we can conceive of, or make intelligible to our-
selves, as a possible general structure of experience. The investigation of
these limits, the investigation of the set of ideas which forms the limiting
framework of all our thought about the world and experience of the world,
is, evidently, an important and interesting philosophical undertaking. No
philosopher has made a more strenuous attempt on it than Kant.
Comment. What Strawson seems to assume here is that the general structure of
experience has a limiting effect on what we can know. This may not sound too bad at
first; but actually it is little more than trivializing the aim of Kant’s transcendental
philosophy, in particular, the project of a transcendental deduction of the categories.
To see why, it may be helpful to recall Kant’s comments on Locke’s enterprise
(“physiology of the human understanding” 26 ), in particular as elaborated in Kant
[1787] (B), pp. 118 f and pp. 127 f.27
Quotations 68.13. (1) Bittner [1973], p. 1535, my translation.

Kant’s Transcendental Philosophy seems to rely here on some introspection
which reveals our original achievements of knowing; such an introspection,
however, does not exist. [[. . .]] [[T]]he problem is [[. . .]] what sort of conclusion
this is supposed to be which leads us to the transcendental functions of
knowing[[.]]
(2) Körner [1967], p. 318.
A transcendental deduction can [[. . .]] be defined quite generally as a
logically sound demonstration of the reasons why a particular categorical
schema is not only in fact, but also necessarily employed, in differentiating
a region of experience. [[. . .]]
Among the most important and interesting examples of attempted tran-
scendental deductions are, of course, those found in Kant’s philosophy[[.]]
26 Kant [1781] (A), p. ix, cf. quotation 59.2 in these materials.

27 See quotations 61.22 in these materials.
(3) Harris [1987], p. 86.

[[T]]ranscendental logic stems as a whole from the identity and self-con-
sciousness of the subject, on which the unity of the object and its coherence
is made to depend. The object is a whole because of the necessary synthetic
activity of the subject, an activity dispensable to cognition through and by
means of a flux of representations. This emphasis on the activity of the
subject, important as it is, as well as the reflection (frequently reiterated
by transcendentalists) that all knowledge comes and can come only through
consciousness, of which the identity and self-awareness of the subject is the
unavoidable condition, commits transcendental logic to an idealism which
brings with it insuperable problems, despite the disavowals by its authors
of subjectivism and their common protestations against the accusations of
their critics.
Quotations 68.14. (1) Strawson [1966], p. 85.

[[W]]as not the Metaphysical Deduction held to prove precisely that “by
their means alone an object can be thought”? Why should it be necessary
to seek another and independent proof of the same conclusion?
(2) Allison [1983], p. 115.
The demonstration of the objective reality of [[the]] categories is the ex-
plicit task of the Transcendental Deduction. Before this can be undertaken,
however, it is first necessary to show that there are such concepts and to
identify them. This is the function of the section of the Critique called the
Clue to the Discovery of all Pure Concepts of the Understanding. In the
Second Edition of the Critique Kant entitles this section the Metaphysical
Deduction[[.]]
(3) Allison [1983], p. 123.
[[The]] Metaphysical Deduction [[. . .]] is intended to establish the agreement
between the table of the logical functions of judgment (§9) and that of
the pure concepts of the understanding (§10). It is embedded in a com-
parison of the respective concerns of general and transcendental logic. The
focal point of this comparison, which is cryptic in the extreme, is a brief
characterization of the transcendental functions of the imagination and the
understanding. The former is said to synthesize our representations, and
the latter to “bring this synthesis to concepts” (A78/B103). This is the first
discussion in the Critique of the transcendental functions of these faculties.
It contains, however, little more than a series of bald assertions, and much
of it is only intelligible in light of the subsequent treatment of this topic in
the Transcendental Deduction.
(4) Allison [1983], p. 133.
A basic exegetical difficulty that is unique to the Second Edition De-
duction stems from the division of the argument into two parts, each of
which presumably establishes the necessity of the categories. The first part
(§§15–21) asserts their necessity with respect to objects of sensible intuition
in general. The claim is that any sensible content, whatever its inherent na-
ture, must be subject to the categories if it is to be brought to the unity
of consciousness, that is, if it is to be thought or conceptualized. Establish-
ing this result is equivalent to demonstrating that the categories are the
necessary rules for any discursive intelligence. The second part (§§24–26)
argues for the necessity of the categories with respect to human sensibility
and its data. This portion of the argument thus presupposes the results of
the Transcendental Aesthetic.
The problem is how one is to understand the connection between these
two parts and their corresponding arguments. Are they intended as two
distinct yet complementary proofs of the categories, or are they rather two
steps in a single proof?
(5) Allison [1983], p. 135.
[[T]]he notion of objective reality has an ontological sense. To claim that a
concept has objective reality is to claim that it refers or is applicable to an
actual object. Thus a fictional concept, such as ‘unicorn,’ would not have
objective reality, although it could very well function as a predicate in an
objectively valid judgment, such as ‘unicorns do not exist.’ In the case of
the categories, which alone concern us here, the claim of objective reality
is equivalent to the claim that they have a reference or applicability to
whatever objects are given to us in intuition (objects of possible experience).
That is why the demonstration of the objective reality (but not the objective
validity) of the categories requires the establishment of their connection
with the forms or conditions of human sensibility. We shall see that this
connection is made in the second part of the Deduction by means of the
conception of the transcendental synthesis of the imagination.
Comment. This is a crucial point. Frege’s considerations on concepts and objects
will help to bring more structure into the problem.
It seems quite likely that Kant obtained his antinomies first, and then con-
structed the system of the sources of knowledge to suit them,28 in particular the
idea that they are due to an inner contradiction in the defining properties of the
subject. This makes it look even more promising to concentrate on the antinomies
for the task of making sense of Kant.
Quotation 68.15. Russell [1903], pp. 456 f.

433. Broadly speaking, the way in which Kant seeks to deduce his
theory of space from mathematics (especially in the Prolegomena) is as
follows. Starting from the question: “How is pure mathematics possible?”
Kant first points out that all the propositions of mathematics are synthetic.
He infers hence that these propositions cannot, as Leibniz hoped, be proved
by means of a logical calculus; on the contrary, they require, he says, certain
synthetic à priori propositions, which may be called axioms, and even then
(it would seem) the reasoning employed in deductions from the axioms
is different from that of pure logic. Now Kant was not willing to admit
that knowledge of the external world could be obtained otherwise than by
experience; hence he concluded that the propositions of mathematics all
deal with something subjective, which he calls a form of intuition. Of these
forms there are two, space and time; time is the source of Arithmetic, space
of Geometry. It is only in the forms of time and space that objects can be
experienced by a subject; and thus pure mathematics must be applicable to
all experience. What is essential, from the logical point of view, is, that the
à priori intuitions supply methods of reasoning and inference which formal
logic does not admit; and these methods, we are told, make the figure (which
may of course be merely imagined) essential to all geometrical proofs. The
opinion that time and space are subjective is reinforced by the antinomies,
where Kant endeavours to prove that, if they be anything more than forms
of experience, they must be definitely self-contradictory.
In the above outline I have omitted everything not relevant to the phi-
losophy of mathematics. The questions of chief importance to us, as regards
the Kantian theory, are two, namely, (1) are the reasonings in mathematics
in any way different from those of Formal Logic? (2) are there any con-
tradictions in the notions of time and space? If these two pillars of the
Kantian edifice can be pulled down, we shall have successfully played the
part of Samson towards his disciples.
434. The question of the nature of mathematical reasoning was ob-
scured in Kant’s day by several causes. In the first place, Kant never
doubted for a moment that the propositions of logic are analytic, whereas
he rightly perceived that those of mathematics are synthetic. It has since
appeared that logic is just as synthetic as all other kinds of truth; but
this is a purely philosophical question, which I shall here pass by. In the
second place, formal logic was, in Kant’s day, in a very much more back-
ward state than at present. It was still possible to hold, as Kant did, that
no great advance had been made since Aristotle, and that none, therefore,
was likely to occur in the future. The syllogism still remained the one type
of formally correct reasoning; and the syllogism was certainly inadequate
for mathematics. [[. . .]] In the third place, in Kant’s day, mathematics itself
was, logically, very inferior to what it is now. [[. . .]] [[T]]here probably did not
exist, in the eighteenth century, any single logically correct piece of math-
ematical reasoning, that is to say, any reasoning which correctly deduced
its result from the explicit premisses laid down by the author. Since the
correctness of the result seemed indubitable, it was natural to suppose that
mathematical proof was something different from logical proof. But the fact
is, that the whole difference lay in the fact that mathematical proofs were
simply unsound.
Comment. As regards Russell’s comment on the synthetic character of logic, Wang
[1986], p. 60, remarks that he was amused reading this. Hylton [1993], p. 482, note 45,
remarks that “Russell makes a similar point nearly ten years later; see Problems of
Philosophy (London: Oxford University Press, 1912), 79.”
Quotation 68.16. Derrida [1971a], pp. 186 f

When Kant proposes a system of categories governed by the “faculty of
judgment,” which is the same as the “faculty of thought,” is grammar still
the guiding thread of the investigation? This is by no means excluded;
but what kind of historical labyrinth are we drawn into then? What kind
of entanglement of linguistic and philosophical structures would have to
be taken into account? In effect, the relation of the Kantian categories
to language is mediated by an entire philosophical stratification (viz., the
entire heritage of Aristotle, which is to say, many things), and by an entire
set of linguistic displacements whose complexity is easily glimpsed. The
enormousness of this task does not reduce its necessity.
Comment. In the face of such enormousness, it is probably not surprising that
Derrida takes refuge in les belles lettres. For those, however, who value systematic
thinking, I wish to point out that there lies a challenging possibility hidden in the
grammar of formalized theories. It has its roots in a principal incompletability of
formalized (mathematical) theories and brings with it the possibility of a progres-
sion of theories on a somewhat logical basis. Differently put, outside the “historical
labyrinth”, the “entanglement of linguistic and philosophical structures”, or the “en-
tire heritage of Aristotle”, there is a phenomenon of pristine beauty, viz., that of
logic’s own limitations. I do not, however, want to deny that it will be difficult to
employ this for a foundation of categories, but taking a notion like ‘undecidability’
by analogy will certainly not help.29
68d. The phenomena-noumena distinction, the notion of the thing

in itself, and the antinomies. Like the conception of transcendental idealism
notions like that of phenomena, noumena and a thing in itself are being given a
hard time under the treatment of straight thinking, i.e., thinking which wants to get
through to the object, without taking notice of itself. Kant’s precautions in empha-
sizing the only negative sense of a noumenon do not seem to have been sufficient.
Admittedly, it is difficult to understand, in particular, I assume, for analytically
inclined thinkers who tend to regard concepts as referring to something. The main
point seems to be to avoid taking certain expressions as simply referring, and thus
open to all sorts of discourse, but as problematic. The classic example is Kant [1787],
p. 306: “if we entitle certain objects . . .”.30
Quotations 68.17. (1) N. K. Smith [1918/23], p. 614.

In Kant’s view the ultimate source of all spontaneity and agency lies in
things in themselves.
Comment. It is not just that I have avoided any quotation in chapter XV that would
support such a claim, I really wouldn’t know where to find support for it. Bluntly,
I think it is one of the weirdest claims about Kant’s view. Cf. the immediately
following quotation.
29 Cf. quotations 95.18 in these materials regarding Derrida’s reception of undecidability.
(2) Werkmeister [1981], p. 302.

Norman Kemp Smith, the translator of Kant’s Critique of Pure Reason
who should have known better, states boldly: “The thing in itself is regarded
as the sole true substance and as a real cause of everything which happens
in the natural world”.
Comment. To do justice to N. K. Smith, I should like to point out, that, in this
particular quotation, I understand him as speaking about Kant’s pre-Critical posi-
tion, albeit in the context of what makes it into the Critique. The first section of
Werkmeister’s [1981] is worth a look for a selection of conflicting interpretations of
Kant’s concept of a thing in itself. Compare also Fichte in quotation 63.5, last part
(see the comment).
(3) Allison [1983], p. 237.
Of all the criticisms that have been raised against Kant’s philosophy, the
most persistent is that he has no right to affirm the existence of things in
themselves, noumena, or a transcendental object, much less to talk about
such things as somehow “affecting” the mind.
(4) Allison [1983], p. 238.
[[A]]lthough Kant does admit noumena in “purely negative sense,” he goes
on to insist that the whole force of this admission is to underscore the
limitation of “our kind of intuition” to objects of our senses, and thus to
allow for the (logical) possibility of “some other kind of intuition, and so
for things as its objects” (A286/B342–43).
(5) Allison [1983], p. 243.
Although Kant denied the possibility of knowledge of noumena on the
grounds that such knowledge would require intellectual intuition, he did
not totally reject the concept of a noumenon. On the contrary, he sought to
reinterpret it in such a way that it could be incorporated into his transcen-
dental account. This is accomplished by giving it the function of a limiting
concept.
[[. . .]]
In his initial treatment of the problem in the Inaugural Dissertation,
Kant used the “limitation of sensibility” brought about by the introduction
of the concept of the noumenon to provide the basis for a positive theory
of the non-sensible. In the Critique, by contrast, he notes that by limiting
sensibility, which is accomplished by “applying the term noumenon to things
in themselves (things not regarded as appearances),” the understanding
also limits itself. It does so because it recognizes “that it cannot know these
noumena through any of the categories, and that it must therefore think
them only under the title of an unknown something” (A256/B312).
(6) Allison [1983], p. 245.
[[T]]he object to which I refer my representations must be described merely
as a transcendental object, not as a noumenon, because I am lacking a
faculty of nonsensible intuition. The underlying assumption is that if I had
such a faculty, the object would be a genuine noumenon, and I would know
it as it is in itself.
There is, indeed, a strange ambiguity in Kant’s argumentation concerning the

thing in itself. Although Kant claims that we shall never have any knowledge about
the thing as it is in itself, but only through our conception (which is a constitutive
part of knowledge), he is not prepared to give up talking about it. The thing in
itself is still present as a ghost throughout Kant’s Critique of Pure Reason.
Quotations 68.18. (1) Rescher [1981], p. 297.

On Kantian principles, there is just no prospect of establishing any cognitive
contact with a “mind-independent reality” in any strict or literal construc-
tion of the term. There is no realm of mind-independent realia that exist
altogether “in themselves;” and even if there were, we could have nothing
to do with them — they would literally, be nothing to us.
(2) Rescher [1981], p. 298.
The conception of “things in themselves” is accordingly a creature of
the human mind, a way that the human mind has of construing its objects.
As such, it has every bit as much (or, rather, as little) ontological weight
as is the case with space and time as creatures of our sensibility or with
causality and actuality as creatures of our understanding (A247). Things
in themselves are not the furnishing of a mind-independent world but the
mind-purported correlates of a mind-constructed conception.
Comment. I find this a beautifully clear and simple account of the notion thing in
itself. Note, however, that Kant is a bit more specific: he doesn’t talk about the
human mind indiscriminately.31 A noumenon is a creature of the understanding, as
regards intuition, to “curb the pretensions of sensibility”. This hierarchical structure
is important for my appropriation of Kant’s solution of the antinomies in §104 in
the groundworks.
(3) Rescher [1981], p. 293.
Our understanding cannot operate outside the supposition of their [[i.e.,
noumena]] existence any more than our sensibility can operate outside the
space-time framework at the perceptual level.
Comment. Such a claim about the concept of things-in-themselves as virtually a
conditio sine qua non for the operating of the understanding (comparable to that
of the cosmological ideas) does not find support in my quotations (cf. section 61f
above) and I don’t know of any other sections in Kant’s Critique which would
support this view. The indication Rescher gives is encapsulated in a subjunctive
clause and hardly sufficient to back up his ambitious claim.
(4) Posy [1981], p. 323.
A model is a pair hM, Φi, where M is a model-structure and Φ interprets
L on M. Let us assume that A is some judgment about empirical objects.
31 In fact, there is no entrance at all under “mind” in the index to N. K. Smith’s translation.
The model hM, Φi says that A is known if A is designated at the α0 of

M . A is known a posteriori if we are sensitive to the possibility (however
remote) that it is subject to revision. The model will reflect this by having
a non-isolated node, α′ , (which is inaccessible to α0 ) at which A is not
designated. This is how we accommodate the possibility of errors. Within
a galaxy our judgments are persistent. We revise beliefs by jumping to
an inaccessible galaxy. If, however, A reflects the very structure of our
empirical a[[p]]paratus—and thus hold a priori — then of course A must be
successful in every empirically possible situation. The model captures this
by designating such A at every non-isolated node.
Comment. This is a curious attempt to use possible world semantics to illustrate
Kant’s philosophy. I do not really know what to think of it myself; but I do not
want to exclude that there is something to it.
Quotations 68.19. (1) Werkmeister [1981], p. 304.

[[I]]n most of the relevant statements Kant uses the terms ‘thing in itself’
and ‘things in themselves’ as abbreviations of the expression ‘things viewed
(or contemplated) without reference to our experiencing them in sensory
intuition’.
Comment. Note Werkmeister’s specification: “sensory intuition”.
(2) Gram [1967], p. 502.
Kant talks of a thing in itself as a kind of object. It is the kind of object that
remains when we abstract from the conditions under which that object is
given to us.
Comment. As opposed to this: Kant [1781] (A), p. 288,32 where Kant explicitly
emphasizes that the concept of a noumenon is not the concept of an object. What
is important here is to specify what sort of abstraction is involved and what sort
of result can be expected. Gram does not specify; he just talks of “conditions”. A
lot of problems regarding Kant’s notion of a thing in itself arise from such a lack of
specification.
(3) Gram [1967], p. 499.
[[T]]he referent of the concept ‘world’ does not exist as a thing in itself.
Comment. This is not only far away from Kant’s actual choice of words; the notion
of referent of a concept is alien to transcendental idealism altogether.
A point which plays a dominant role in Kant’s critical decision of the conflict of
reason with itself and my choice of quotations in section 62c: why we can only say
about the world that it is finite or infinite, if it is a thing in itself. In other words,
what is the connection between things in themselves and tertium non datur.

Quotations 68.20. (1) Strawson [1966], p. 211.

The series of appearances “is bound up with a condition which, while em-
pirically unconditioned, is also non-sensible”; appearances “must themselves
have grounds which are not appearances”, “they must rest upon a transcen-
dental object which determines them as mere representations”.
Comment. There is some strange sort of confusion showing up in this potpourri
of fragmented Kant citations.33 To begin with, Strawson links the question of the
noumenon with the solution of the dynamical antinomies. This requires a bit of
manipulation in his first citation: for a start, the subject “the completely uncon-
ditioned” of the original sentence is replaced by “the series of appearances”, and
then Strawson subscribes to the mistranslation by N. K. Smith which I pointed
out in section 62c. As regards the second citation, Strawson has abridged from “if
[[appearances]] are viewed not as things in themselves, but merely as representa-
tions, connected according to empirical laws, they must themselves have grounds
which are not appearances.” In other words: Strawson cites the ‘then’ part of an
‘if—then’-sentence only. But this has nothing in particular to do with the dynami-
cal antinomies: by virtue of our classifying something as representations, they must
have ground somewhere else. The point is a conceptual one.
(2) Strawson [1966], p. 236.
The supersensible: things as they are in themselves. There exists the
sphere of supersensible reality, of things, neither spatial nor temporal, as
they are in themselves.
Comment. That’s far away from the problematic concept of a thing in itself which
Kant introduced, far away from critique, back to dogmatism.
The first problem concerns the existence of antinomies, the second their solu-
tion. I begin with the antinomies themselves.
Quotation 68.21. N. K. Smith [1918], p. 432.

[[P]]reoccupation with the problem of antinomy was the chief cause of the
revolution which took place in Kant’s views in 1769, and which found ex-
pression in his Dissertation 1770. It was the existence of antinomy which
led Kant to recognise the subjectivity of space and time. That is to say, it
led him to develop that doctrine of transcendental idealism which reappears
in the concluding section of the Aesthetic, and which was recast and de-
veloped in the Analytic. Already in the Dissertation it supplies the key for
the solution of the problems concerning infinity. The impossibility of com-
pleting the space, time, and causal series, and the consequent impossibility
of satisfying the demands of the mind for totality, simplicity and uncondi-
tionedness, do not, it is there maintained, discredit reason, but only serve
to establish the subjectivity of the sensuous forms to which the element of
infinitude is in all cases due.
33 Cf. quotation 56.2 (2) in these materials on “quilt quotations”.
Comment. Sometimes this seems to be taken as an indication that the whole Cri-
tique was artificially constructed around the antinomy. Anyway, I do not take Kant
to assert the “subjectivity of space and time”. True, space and time are “subjective
conditions of intuition” (cf. quotations 61.9 (1) and (2) in these materials), but
Kant also emphasizes their “objective validity”. Kant speaks (in translation) of the
“transcendental ideality” and “empirical reality” of space and time, and he speaks of
space and time as the “subjective conditions of intuition”; but I know of no occasion
where he speaks of the subjectivity of space and time. If it were appropriate to speak
simply of the subjectivity of space and time, I would not be able to make much sense
of Kant’s question “how subjective conditions of thought can have objective validity”.
But then, N. K. Smith makes it very clear that he finds Kant obscure,34 and I think
I can see why.
There is a strong tendency to dismiss Kant’s antinomies as artificial constructs.

The following couple of quotations is meant to give an impression.
Quotations 68.22. (1) Fried [1940], p. 207.

[[N]]either a new metaphysics nor a new psychology is required to escape
from the web of necessary contradiction so neatly woven by Kant. The
contradiction that is supposedly proved, we shall maintain, is grounded,
not as Kant claims, in any assumption of a world in itself, or in any inborn
characteristic of reason, but in an ambiguity of phrase and in a logical
procedure which first plants its contradictions and then simulates surprise
at discovering them. If our contention is correct, the proofs of the antinomies
are nothing more than a philosophical tour de force exhibiting, at their best,
a series of linguistic manœuvres.
Comment. The point about any alleged ambiguity, be that of spheres (types) or
levels of language or what have you,35 is always whether it can be avoided, and if
so at what costs. This will be an issue in the groundworks in the context of what
I call ‘systemic ambiguity’. It is usually not an issue that philosophers find worth a
closer investigation. (They might end up in logic and realize that a notion like that of
a ‘world in itself’ contains an inherent ambiguity of phrase which can be the source
of antinomies.) Notice also the disregard for the linguistic conditions of knowledge
(similar in character to the widespread disregard for tautology). Philosophers want
the real thing, not just complications that arise from an inherent condition, like
that of grammar, for instance.
(2) Hallett [1984], p. 228.
Kant frequently manipulates his proofs by covert appeal to his own special
conception of the world. If my suggestion is correct, then the respective
analyses of the set-theoretic contradictions and Kant’s first Antinomy will
be as follows. Set theory: comprehension principle (or naive realism) plus
logic yields contradictions. Kant: naive realism plus principles inconsistent
34Cf. quotation 68.1 above.
35Ramsey, in quotation 77.3 (2) speaks of an “ambiguity in the psychological or epistemolog-
ical notions of meaning and asserting”.
with naive realism plus logic yields contradictions. Stated crudely, it would
seem that Kant injects the contradictions into his proofs, and what he
claims to be a fundamental and inevitable problem is, in fact, an artificial
one.
One may argue that none of the antinomies presented in the Critique of Pure
Reason stands firm. But there are the antinomies of the foundations of mathematics.
They will be good enough for my purpose later on, so I don’t really need the
antinomies of pure reason. What is still interesting, however, is the role of the
concepts of phenomena and noumena in Kant’s critical decision.
Quotations 68.23. (1) Hallett [1984], p. 228.
To put it shortly: Kant conceives of the physical world as the world
of appearances; this is quite different from the naive view of the world as
independent from us[[.]]
Comment. Quite classic the sloppy language regarding Kant’s view of the physical
world and the world of appearances. As regards the ‘naive view of the world’,
perhaps it has to be pointed out that Kant started explicitly with the intention to
turn this view upside down (‘Copernican revolution’).
(2) Priest [1995], pp. 82 f.
Phenomena are, essentially, those things that are perceivable via the senses.
I use ‘thing’ in a fairly loose way here, to include objects such as buildings,
countries and stars; and events such as the extinction of dinosaurs, plane
journeys and the death of Hegel. Noumena, or at least what we can say
about them, are more problematic, as we shall see. However, essentially,
those things are noumena which are not phenomena. [[. . .]]
The distinction between phenomena and noumena makes perfectly good
sense for a non-Kantian, as much as for a Kantian. And all can agree that
phenomena are in space and time (or just time in the case of internal sensa-
tions). Many would argue, however, that not all things in (space and) time
are phenomena. For there are many physical entities, including those that
are responsible for our perceptions (such as photons and electromagnetic
radiation), which are not themselves perceivable.
It is therefore important to note that Kant has a somewhat idiosyn-
cratic view about what sorts of things phenomena are. For Kant thinks that
objects in themselves cannot be perceived, or intuited in his jargon; what
are perceived are our mental representations of such objects. He calls this
view ‘Transcendental Idealism’.
Comment. I include this quotation as a kind of warning: when there is a clash
between common sense and a theory, and it sounds idiosyncratic, it does not always
come from the theory.
Quotations 68.24. (1) Gram [1967], p. 506.
[[T]]he argument that we cannot apply either of two mutually exclusive
predicates to a collection depends upon certain characteristics possessed by
the collection as a whole.
Comment. Is this a reminiscence to Russell?

(2) Gram [1967], p. 507.
To say that they are things in themselves is to imply that they must
have one of two contradictory predicates. The fact that neither of two con-
tradictory predicates applies to them shows that the opposition between
these predicates is dialectical and that there is a third alternative; namely,
that space and time are just forms of our sensibility.
Comment. In view of my presentation of Kant’s “critical decision of the conflict of
reason with itself” in section 62c, this strikes me as a curious formulation; although
quite compatible with my Kant quotations, I nevertheless find them a bit tilted.
It almost seems to suggest that we don’t have to stick to classical logic when not
dealing with things in themselves. It may be worthwhile, therefore, to recall Kant’s
line as presented in section 62c: ‘the world as a whole’ is a description which makes
an implicit assumption, namely that the idea of absolute totality is applied here to
appearances which exist outside of our representations. It is enough to make this
assumption explicit to see that there is no real opposition.
(3) Priest [1995], p. 107.
Kant claims that if the objects in the series of conditions were noumena
then the limit would exist (for example, A198=B526 f). This appears to
have been so obvious to him that he never bothered to argue for it.
Comment. If that’s how you feel too, have a look at quotation 62.11 (1) in these
materials.
Quotations 68.25. (1) N. K. Smith [1918/23], p. 511.

Kant’s differential treatment of the two sets of antinomies is arbitrary, and
would seem to be due to his having attempted to superimpose, with the
least possible modification, a later solution of the antinomies upon one
previously developed.
Comment. First N. K. Smith “corrects” Kant (following Hartenstein) by translating
“mathematical antinomies” where Kant wrote “Antinomie”,36 and then he states
that “Kant’s differential treatment of the two sets of antinomies is arbitrary”. So let
me emphasize again: there is a subtlety in Kant’s treatment of the antinomies, part
of which is the distinction of a “decision” and a “resolution” which is obliterated
in N. K. Smith’s translation. The “critical decision” regarding the conflict of reason
with itself applies to all antinomies;37 the “resolutions” are different in the sense that
the mathematical antinomies allow of no ‘positive’ reading, as it were, whereas the
dynamical ones do. This is a good opportunity to point out that I shall follow Kant
in his distinction of a “decision” and a “resolution” of the antinomies in the following
simple sense: in the case of the logical paradoxes, the “decision” is contained in the
sacrifice of contraction, and the “resolution” is contained in the introduction of the
Z-inferences.
36 Cf. remark 62.21 in these materials.
(2) Allison [1990], p. 22.

Kant insists that transcendental idealism is the key to the resolution of the
entire antinomical conflict, but he also maintains that this idealism yields
different resolutions to the different antinomies. Thus, whereas he resolves
the first two, or mathematical antinomies by claiming that neither the thesis
nor the antithesis is true, he resolves the last two, or dynamical, antinomies
by claiming that both may be (not that they are) true or, more precisely,
that the competing claims are compatible.
Not surprisingly, however, this distinction has been met with consider-
able skepticism by Kant’s critics. In fact, it has seemed to many like an ad
hoc device, designed to allow Kant to have his cake and eat it too. Typical
of this line of response is Strawson[[.]]
Comment. What then follows is most of my next quotation.
(3) Strawson [1966], p. 209.
It seems obvious what the correct “critical” solution of this conflict
[[of pure reason as regards freedom]] should be. Since things in space and
time are appearances, the series of ever remoter causes should no more be
regarded as existing as a whole than the series of ever remoter temporal
states of the world or the series of ever remoter spatial regions of the world.
Since the series does not exist as a whole, there is no question of its existing
either as an infinite whole or, as is asserted in the thesis, as a finite whole
with a first, uncaused member. Every member of the series which is actually
“met with” in experience, however, may, and must, be taken to have an
antecedent cause. The thesis, then, is false, the antithesis true.
Kant, as I have said, passes over in silence the possibility of this “con-
ventional” critical solution.
Comment. In view of Kant’s distinction between the critical decision of the conflict
of pure reason with itself and the solutions of the cosmological ideas, Strawson
strikes me as just messing up things. But then, he might just be a victim of the
translator’s poetry.
(4) Allison [1990], p. 23, regarding the above quotation by Strawson.
Although he was anticipated by Schopenhauer, it is nonetheless somewhat
curious to find Strawson claiming that the critical solution to the antinomies
is to deny the thesis and affirm the antithesis. Not only does this fly in the
face of Kant’s explicit claim that in the mathematical antinomies neither
the thesis nor the antithesis is true, it also fails to distinguish between the
regulative demand always to seek further conditions and the antithesis’s
dogmatic assertion of the presence of an actual infinity of conditions. In
short, it fails to notice that Kant regards the antithesis as being as dogmatic
in its own way as the thesis.
(5) Loparic [1990], p. 302.
After having established the correct formulation of the first cosmolog-
ical problem, Kant goes on to solve it. He does it by proving the truth of
the antithesis which says that the world has no beginning, and no limits
in space, but is infinite as regards both time and space. [[. . .]] This is, he
says, the first and negative answer to the cosmological problem regarding
the magnitude of the world (B548).
Comment. This does not agree with my presentation of Kant’s solution of the “Cos-
mological idea of the totality of the composition of the appearances of a cosmic
whole” (B545). What Kant says is, in fact, that “the world has no first beginning in
time and no outermost limit in space.” What he doesn’t add is that it is infinite as
regards both time and space; instead, we find a footnote pointing out the difference
to the antithesis which is that there “we inferred the actual infinity of the world”.
(6) Priest [1991], p. 365.

[[T]]he root of the antinomies is, Kant assures us, the confusing of phenom-
ena with noumena, and so the taking one or the other to behave incorrectly
(A498–9, B526–7). [[. . .]] [[W]]hen Kant attempts to work out the details he
gives two conflicting solutions to the Antinomies. The first (A498, B526, ff)
is meant to apply to all the Antinomies, and is to the effect that the limit
does not exist (to suppose so is to take phenomena for noumena). The sec-
ond (A533, B561, ff) is specific to the Third and Fourth Antinomies, and is
to the effect that the limit exists, but the pertinent categories do not apply
to it (to suppose so is to take noumena for phenomena).
Comment. Here is another victim of N. K. Smith’s equitranslation of “Entschei-

dung” and “Auflösung”. On top of that Priest confounds things in themselves and
noumena; but admittedly, Kant’s terminology is confusing. Again, I shall try to
clarify my position in §104.
Quotations 68.26. (1) Lyotard [1991], p. 115.

The infinite as a whole is to the infinite of the sensible world what a “noume-
nal” object (KRV, 257–75; 287–308 ), the object of thought alone, is to a
single phenomenon, the object of knowledge. However, if understanding
can advance to the infinite through the regressive series of the conditions
of phenomena in the world, it is because it is supported by the Idea of the
infinite as a whole, which only “a faculty that is itself supersensible” can
form (103; 99 ). The notion of the serial infinite proceeds from the notion
of the actual infinite. What this means is that even in “the pure intellectual
estimation of magnitude” (in der reinen intellektuellen Grö"senschätzung)
made by understanding, “the infinite of the world of sense,” which is at
the horizon of this estimation, must be “completely comprehended under a
concept” (unter einem Begriffe ganz zusammengefaßt: ibid.). This concept
belongs so little to understanding that mathematical estimation never suc-
ceeds in thinking the infinite of the world of sense “by means of numerical
concepts”—the only ones it has at its deposal (ibid.). The object of the
concept in question, the infinite as totality actually given in thought, does
not belong to the world; it is the “substrate” (Substrat : ibid.) underlying it.
Moreover, the thought that conceives of this object is called reason.
(2) Lyotard [1991], p. 119.

[[C]]ausality or freedom is accepted as is the conditioning of phenomena
because freedom is of a noumenal order, whereas the conditioning of phe-
nomena rules the phenomenal world.
In view of my attempt to establish a link between dialectic and Cantor’s diag-

onal method it seems worthwhile to have a look at what Cantor himself said about
Kant’s antinomies.
Quotations 68.27. (1) Cantor [1886], p. 375.

Ohne ernste kritische Vorerörterung wird der Unendlichkeitsbegriff von
Kant in dessen „Kritik der reinen Vernunft“, in dem Kapitel über die „An-
tinomien der reinen Vernunft“ an vier Fragen behandelt, um den Nachweis
zu liefern, daß sie mit gleicher Strenge bejaht und verneint werden können.
Es dürfte kaum jemals, selbst bei Berücksichtigung der Pyrrhonischen und
Akademischen Skepsis, mit welcher Kant so viele Berührungspunkte hat,
mehr zur Diskreditierung der menschlichen Vernunft und ihrer Fähigkeiten
geschehen sein, als mit diesem Abschnitt der „kritischen Transzendental-
philosophie“. Ich werde gelegentlich zeigen, daß es diesem Autor nur durch
einen vagen, distinktionslosen Gebrauch des Unendlichkeitsbegriffs (wenn
unter solchen Verhältnissen überhaupt noch von Begriffen die Rede sein
kann) gelungen ist, seinen Antinomien Geltung zu verschaffen, und dies
auch nur bei denen, die gleich ihm einer gründlichen mathematischen Be-
handlung solcher Fragen gern ausweichen.
(2) Zermelo [1932], p. 377, regarding the above quotation.
Der Kantischen Lehre von den „Antinomien der reinen Vernunft“ scheint
hier Cantor doch nicht gerecht zu werden. Nicht um eine Widerlegung oder
Ablehnung des Unendlichkeitsbegriffes handelte es sich hier bei Kant, son-
dern um seine Anwendung auf das Weltganze, um die Tatsache, daß die
menschliche Vernunft sich durch ihre innere Natur ebenso gedrängt findet,
die Welt als begrenzt wie als unbegrenzt, als endlich wie als unendlich an-
zunehmen — eine Tatsache, die weder durch mathematische Theorien wie
die Cantorsche Mengenlehre noch durch seine wohl nicht sehr tiefgreifende
Polemik aus der Welt geschafft werden kann. Auch wer wie der Herausgeber
[[Zermelo]] die Kantische Theorie der Mathematik, wonach alle mathemati-
schen Sätze auf „reine Anschauung“ gegründet sein sollen, grundsätzlich ab-
lehnt, wird doch zugeben müssen, daß in dieser Lehre von den „Antinomien“
eine tiefere Einsicht, ein Einblick in die „dialektische“ Natur des mensch-
lichen Denkens zum Ausdruck kommt. Und ein eigentümliches Schicksal
fügte es, daß gerade die „Antinomien der Mengenlehre“, deren mindestens
formale Analogie mit den Kantischen nicht wohl in Abrede gestellt werden
kann, ein ganzes Menschenalter hindurch der Ausbreitung und Anerken-
nung der Cantorschen Leistungen im Wege gestanden haben.
Comment. Note the acknowledgment of an ‘at least formal analogy’ between the
Kantian antinomies and those of set theory.38 Apart from that, I find it immensely
soothing to find a solid understanding of Kant’s doctrine amongst at least some
mathematicians.
I close this paragraph with a quotation to indicate that the emergence of an-
tinomies in the realm of pure reason has also evoked some rather dubious sanction.
Quotation 68.28. Adler, quoted after Fried [1940], p. 204.

To the extent that questions in natural theology concerning God, creation,
the soul, are parts of larger questions which require sacred theology, founded
on revelation, for their adequate solution, Kant was right in regarding a
purely rational approach to the whole question in each case as doomed to
the frustrations of an illusory dialectic, a dialectic which did not properly
recognise its own limitations in philosophical mysteries.
Comment. None.
§ 69. Struggling with speculative philosophy
In this paragraph I shall have a look at how the salient features of Hegel’s speculative
philosophy fare in secondary literature. Among these are most important:
(i) the idealist conception;
(ii) the antinomies (negative dialectic);
(iii) deduction of categories (positive dialectic).
Speculative philosophy seems even worse off than transcendental philosophy. Its es-
oteric appearance has attracted many a verbal acrobat which, in turn, has provoked
many a wise guy to answer. Together they have managed to stink out the place,
sometimes in the name of clearing the air a bit.
The last decade has seen a number of new books on Hegel,39 but I have not
come across anything that I would consider more than academic ruminations. In
particular one of the central themes in Hegel’s speculative outlook on philosophy,
viz., that of the unity of logic, metaphysics, and idealism, fares badly. In fact,
many writers on Hegel do not seem to realize that there is a problem of immense
importance touched upon in Hegel’s speculative philosophy, a problem that is not
created by a peculiar language or some other form of idiosyncrasy, and that despite
the fact that Hegel’s own formulations may be idiosyncratic to the maximum. They
stick to questions of exegesis, the meaning of words, what Hegel really had in mind,
what he said in different versions of his logic, why he changed his mind in later
editions et cetera. All this may well be very scholarly indeed, but it does little to
nothing to help tackling the underlying problem. It is scholastic in the mean sense
38 A similar view is expressed by Hessenberg in quotations 73.18 (1) and (2) in these
materials.
39 Some of which are not even listed in the bibliography. Readers who are interested in the
standard of contemporary Hegel scholarship will find plenty of bad examples in the Cambridge
Companion to Hegel (Beiser [1993]) and Henrich [1986].
§ 69. STRUGGLING WITH SPECULATIVE PHILOSOPHY 941
that conciliation is sought in the writings of the Masters; the strategy is rereading
the Masters, and the test is the rehearsal before the academic mediocracy (members
of which aspire to be Masters themselves). The seeds of the Copernican Revolution
did not fall on fertile ground.
In the present study, I am concerned with a theoretical problem; it is with
regard to this problem, which shall only be formulated in chapter XXVIII, that I
deal with Hegel’s ideas, and, consequently, with secondary literature. It is in this
vein that I here include the following quotation.40
Quotation 69.1. (1) Stepelevich, Preface to The Young Hegelians. An Anthology.
Humanities Press: Atlantic Highlands NJ [1997], p. ix.
A distinction must be made between being a Hegelian philosopher and a
student of Hegelian philosophy, for the practice of this philosophy extends
well beyond the mere scholarly recollection of that thought. To philosophize,
as a Hegelian, is to take up, develop, and apply the dialectical methodology
of Hegel to a point that would extend beyond the limits found in Hegel
himself.
69a. Science of Logic, dialectic, and the idealist conception.
Quotations 69.2. (1) Lenin [1929], p. 114.
Obviously, Hegel takes his self-development of concepts, of categories,
in connection with the entire history of philosophy. This gives still a new
aspect to the whole Logic.
(2) Lenin [1929], p. 191.
Remarkable: Hegel comes to the “Idea” as the coincidence of the Notion
and the object, as t r u t h, t h r o u g h the practical, purposive activity of
man. A very close approach to the view that man by his practice proves
the objective correctness of his ideas, concepts, knowledge, science.
Quotations 69.3. (1) Russell [1914], p. 37 f.
Hegel and his followers widened the scope of logic [[. . .]]. In their writings,
logic is practically identical with metaphysics. In broad outline, the way
this came about is as follows. Hegel believed that, by means of a priori
reasoning, it could be shown that the world must have various important
and interesting characteristics, since any world without these characteristics
would be impossible and self-contradictory. Thus what he calls “logic” is an
investigation of the nature of the universe, in so far as this can be inferred
merely from the principle that the universe must be logically self-consistent.
I do not myself believe that from this principle alone anything of importance
can be inferred as regards the existing universe, But, however that may be,
I should not regard Hegel’s reasoning, even if it were valid, as properly
belonging to logic: it would rather be an application of logic to the actual
world. Logic itself would be concerned rather with such questions as what
self-consistency is, which Hegel, so far as I know, does not discuss. And
40 I am grateful to Valerie Kerruish for having brought this passage to my attention.
though he criticises the traditional logic, and professes to replace it by an

improved logic of his own, there is some sense in which the traditional logic,
with all its faults, is uncritically and unconsciously assumed throughout his
reasoning.
(2) Findlay [1958], p. 81.
Hegel’s main mistake, both in what he says about his Dialectic and in what
he tries to do with it, lies in his assumption that it has a kind of deduc-
tive necessity, different from, but akin to, that of a mathematical system,
whereby we shall find ourselves forced along a single line of reasoning[[.]]
Comment. It will, hopefully, be clear that this is just what I am aiming at: a
deductive theory of dialectic, albeit with steps in between which are not purely
deductive but comparable to adding reflection principles in formalized theories of
arithmetic.
There is a respectable academic tradition which recognizes Hegel’s place in the

tradition of transcendental philosophy without, however, really pushing his line any
further. My first quotation is meant to provide an example.
Quotations 69.4. (1) Gadamer [1971], p. 76.

With his Logic Hegel seeks to bring the transcendental philosophy ini-
tiated by Kant to its conclusion. According to Hegel, Fichte was the first
to grasp the universal systematic implications of Kant’s way of viewing
things from the perspective of transcendental philosophy. At the same time,
however, Hegel was of the opinion that Fichte’s own “Doctrine of Science”
did not really finish the task of developing the entirety of human knowl-
edge out of self-consciousness. To be sure, Fichte’s contention is that his
“Doctrine of Science” had done precisely that. He saw, in the spontaneity
of self-consciousness, the actual, underlying operation, “the active deed”
(Tathandlung), as he calls it.
(2) Gadamer [1971], p. 79.
A glance back at Greek philosophy is necessary, too, if one is to under-
stand Hegel’s conception of the method through which he sought to convert
traditional logic into a genuine philosophical science—the method of dialec-
tic. Dialectic develops from the magnificent boldness of the Eleatics, who, in
opposition to what appears to be the case in sense experience, held strictly
and relentlessly to what thought and thought alone demands. [[. . .]]
[[. . .]] Hegel comments with a measure of justification that Plato’s di-
alectic is deficient in that it is only negative and does not reach any scientific
insight. As a matter of fact, Plato’s dialectic is, properly speaking, not a
method at all and least of all the transcendental method of Fichte or He-
gel. It has no absolute beginning. Nor is it founded on an ideal of absolute
knowledge which could be said to be free from all opposition between know-
ing and what is known and be held to embrace all knowledge in such a way
that the entire content of knowledge would be exhausted in the continuing
determination of the concept in relationship to itself.
(3) C. Taylor [1975], p. 225.

The Logic is the second of the great derivations of Hegel’s vision, and a
crucial one [[. . .]], because it is the only real candidate for the role of strict
dialectical proof. If the real exists and has the structure it has by pure
conceptual necessity, then the task of the Logic is to show this conceptual
structure by pure conceptual argument.
Comment. I like Taylor’s label “vision”; it seems very appropriate to me — with
regard to both, the positive as well as the negative associations that the word
“vision” may evoke. But the phrase “if the real exists . . . ” seems to tell of an ongoing
dominance of a naive realistic framework.
(4) Pinkard [1988], pp. 26 f.
The rather striking beginning of the Science of Logic presents the basic
structure of the whole work. Hegel describes this somewhat obliquely: he
speaks of it as the “ground of the whole science” and says that the progres-
sion in the theory is only a “further determination” of the beginning. Or, as
he also puts it, “Thus the beginning of philosophy is the foundation [Grund-
lage] which is present and preserved throughout the entire subsequent de-
velopment, remaining completely immanent in its further determinations.”
What this means is that the moves (or “logic”) of the conceptions of being
and nothing represent the structure of the work. The logic of other more
developed conceptions of being are constructed according to the logic of the
conception of pure being: each conception of being takes its determinate-
ness, as it were, from its own “nothing,” from what Hegel calls its negation.
It is the assumed necessity to distinguish being from nothing that estab-
lishes the pattern for each successive move in the Science of Logic. Each
new conception is to be introduced as a way to avoid the contradictions of
a lower level.
Hegel proposes that we attempt to construct all our categorial concep-
tions in terms of these kinds of moves that are necessary to avoid contra-
dictions at a certain level of discourse — in more Hegelian terms, between
a categorial conception and what he calls its determinate negation. The de-
terminate negation of a category is in Hegelian parlance its “other.” Used in
this sense, “determinism”, for example, is the Hegelian “other” to “freedom.”
The involvement of a conception with its other leads to a new conception
that is thereby justified as a product of and a solution to the dilemmas
found at the lower level. The higher level conception is one in which the
dilemmas of the lower level conceptions do not appear.
Hegel calls such a movement an Aufhebung. Within the Hegelian theory
the terms means “integration”: the higher level conception integrates the
logic of the lower ones. The determinations that a conception gains by its
references to its other, that is, by the move from it to its negation, are
integrated within the higher conception.
Hegel holds later categories to be implicit in the earlier ones. In what
sense, however, are they implicit? Hegel holds that the later categories
follow with necessity from the earlier ones. This belief in necessity cashes
out in the following way. Hegel holds that the later categories are not only
the explanations of the possibility of the earlier ones but also the only
explanations that are possible. By accepting the validity of the categories
at the lower level, one accepts the dilemmas inherent to them, and since the
dilemmas are unacceptable, one must therefore also accept the resolutions
of the dilemmas. Accepting one set of categories thus necessarily implies
accepting the other categories that resolve the logical dilemmas of the earlier
ones.
This sense of “implicit,” however, rests on accepting Hegel’s belief that
the later categories are the only possible resolutions to the dilemmas found
in the earlier ones. This idea is part of Hegel’s Kantian heritage of the
“transcendental science of reason.” If Hegel’s solutions, however, are shown
not to be the only ones or if the so called logical dilemmas turn out not to be
contradictions but something weaker, then the Kantian, “scientific” element
of Hegel’s theory would be severely undermined. [[. . .]] [[I]]t is precisely that
element of Hegel’s thought that proves the most difficult to sustain.
(5) Duquette [1990], p. 2.
Hegel’s Logic compares with Kant’s conception of a transcendental logic
in that it does not make reference to empirical objects or principles and,
unlike pure general logic, is not devoid of content. However, while, like the
Kantian categories, the Hegelian categories are pure thoughts, it is in the
specification of the nature of the content of these pure thoughts that the
relation of Hegel’s to Kant’s logic becomes problematic. Whereas for Kant
the content of pure thought is objecthood relative to a consideration of the
[[. . .]] sensuous conditions of space and time, for Hegel space and time do
not enter into any such definition of the content of pure thought.
(6) Duquette [1990], p. 3.
Hegel’s Logic is a sort of “transcendental dialectic” but with an epis-
temological significance that Kant could not allow it to have, for Hegel
takes up at least some of what Kant would call “transcendental ideas” and
attempts to constitute them into an organon for knowledge. Moreover, He-
gel characterizes the “movement” of concepts in his logic as a dialectic,
although this now has a positive as well as negative significance. In other
words, speculative dialectic is for Hegel neither a sophistical play of illusion
by pure reason nor a critique of this illusion, as it is for Kant, but rather
involves an intelligible construction of meaning via an immanent develop-
mental sequence of pure thought determinations. This presumption of the
intelligibility and epistemological efficacy of the Hegelian categories, in hav-
ing a thought content independent of reference to the sensuous conditions
of the experience of an object, is where lies the divergence from Kant’s
treatment of the concepts of pure reason.
(7) Hanna [1986], p. 253.
Hegel’s logic, as developed both in the Science of Logic and in the Ency-
clopedia Logic, can be understood only as a criticism of what he calls the
“common logic” (EL, 36/81) and also sometimes “formal logic” or “ordinary
logic.” Common logic is perhaps best exemplified by Kant’s Logic: it deals
with the formal conditions of truth in judgments and includes the theory of
the syllogism and identity. Hegel’s logic, as an ontological logic (EL, 36/81),
manifestly goes far beyond the scope of the common logic; it is by no means
either a bare denial or even a revision of common logic. Hegel’s logic in fact
preserves the entire edifice of common logic while still using the critique
of the latter as a motivation for its own self-development toward a more
comprehensive and radically new sense of logic.
(8) Hanna [1986], pp. 255 f.
By establishing his own logic as a development beyond the common
logic, and as a higher-order activity that consists in the “system of pure rea-
son, as the realm of pure thought” (SL, 50/I, 44), Hegel is saying that the
common logic can be viewed from two quite distinct perspectives. Viewed
on its own terms and at its own level, common logic is simply a discipline
among of “alongside” the other scientific disciplines (SL, 58/I, 54). As such
its procedures and notions have a certain integrity and efficacy that cannot
be denied. As Hegel puts it: “the purpose of the science [of common logic]
is to become acquainted with the procedures of finite thought: and, if it
is adapted to its presupposed object, the science is entitled to be styled
correct” (EL, 22/75). But viewed from a higher viewpoint, namely that of
ontology, the common logic can be seen to rest on certain enabling pre-
suppositions which are also at the same time crippling limitations from an
ontological point of view. These limitations prevent the common logic from
passing directly over into philosophical significance: “they bar the entrance
to philosophy [and must] be discarded at its portals” (SL, 45/I, 38). Only a
transformation or “reconstruction” (SL, 52/I, 46) of the conceptions of the
common logic by means of a thorough critique of it, can provide the basis
of the transition from common logic to Hegelian logic. Thus in order to
become adequately ontological or properly philosophical the common logic
must “disappear.” Again, this does not mean that Hegel is denying the ef-
ficacy and efficiency of common logic at its own level. He is denying only
the implicit and therefore uncriticized claim of common logic to ontological
adequacy.
Comment. What I regard here as the “enabling presuppositions which are also at the
same time crippling limitations” are the rules of contraction, more generally, perhaps
the structural rules. Contraction is the ‘essence’ of truth definite propositions, and
thus of classical logic; cf. also quotations 85.28 in these materials.
(9) Zimmerli [1989], p. 192.

We can, by means of a pragmatic interpretation of Hegel’s Wissenschaft
der Logik, understand this work as a formal logic which, by dealing with
the semantics of metaphysics justifies itself, and as a metaphysics which
legitimizes itself as a result of the self-justification of formal logic itself.
Comment. And best of all, we do not even have to specify any laws of that formal
logic; such are the benefits of a philosophical position uninhibited by formal logic’s
own constraints.
(10) Flay [1990], pp. 155 f.
The dialectical critique of the Logic has its force [[. . .]] in the fact that it
shows that a given comprehension of reality—or some aspect of such a
comprehension—turns on itself or is self-refuting. In dialectical critique the
critique “counts” in the eyes of those criticized because it shows that what
they hold to be true is self-refuting in terms of their own standards and
the standards of reality to which they appeal. This is the one characteristic
that all dialectic shares, from Plato until today.
The first point, then, is that the general content of the Logic involves
(a) traditional positions of the understanding concerning various categories
held to be definitive of the nature of thought and of being, and (b) a di-
alectical analysis of these positions which shows that they are self-refuting
because of their own dialectical nature.
Finding the locus of the force and necessity of the dialectic, then,
involves seeking an answer to the question of how this self-refutation is
brought about. No formal-logical answer to this question will serve our
purpose; for any formal-logical response is bound to a simple theory of im-
plication and to a sense of analytical distinction that the dialectic shatters.
If one argues that there may be some as yet unknown formal logic that
will give us the answer, or that the logic of Hegel’s Logic is to be treated
in a more intuitive way, the response to this is that, in no case does logic
determine the force of an argument.
Comment. This can almost count as a paradigm of dialectical reasoning: a certain
awareness of the role of self-refutation, a rejection of formal logic as a relevant tool
and an inflation of the words “dialectic” and “dialectical”. So far as the rejection of
formal logic is concerned, the reason given displays beautifully the limited horizon:
“any formal-logical response is bound to a simple theory of implication”.41 If it is
formal, it must be simple; no consideration regarding the notion of implication that
is involved in self-refutation. Formal logic, like any other tool, may not hold the
answer, but it is necessary for the analysis of a problem, in this case of a certain
self-refutation. And that is where the logical ignorance of the majority of Hegelians
makes itself most painfully felt: as soon as there is a conceptual challenge, the magic
word “dialectic” makes its appearance, and, of course, it “shatters”.
(11) Flay [1990], p. 168.
The final irony is that Hegel’s system carries within itself its own destructive
ironies, and demands the acceptance of constructive irony if this intention to
escape the prison of the narrow understanding is ultimately to find success.
Hegel’s philosophy, it is true, involves us with a certain kind of closure.
However, it is a closure which is dialectical in nature, which constantly
41 At this point I should like to recommend quotation 79.9 (2) in these materials (Church
[1934], p. 360) to the reader’s attention.

opens up new categories and developments by working on and out of itself.

The ultimate irony is that closure can be truly achieved only by recognizing
the openness installed by dialectic and the irony which makes the dialectic
itself matter.
Comment. To get an idea of the conceptual depth of the term “irony” in this passage
I should like to recommend replacing it by “bingo” throughout. The ultimate irony, in
my view, is the historical development from Hegel to the rubbish of his interpreters.
Can it ever be reversed?
Beyond this, there is the more eye-catching topic of dialectics as the driving
power behind the Logic.
Quotations 69.5. (1) Hall [1967], p. 388.

[[T]]here is indeed a Hegelian dialectic, involving the passing over of thought
or concepts into their opposites and the achievement of a dialectic, involv-
ing the passing over of thought or concepts into their opposites and the
achievement of a higher unity. But if it is a process that arrives at a higher
truth through contradictions, it does not constitute a new conception of
dialectic. Hegel actually showed his awareness of the traditional notion by
paying tribute to “Plato’s Parmenides, probably the greatest masterpiece of
ancient dialectic.” And even the doctrine that dialectic is a world process
— not merely a process of thought but also found in history and in the
universe as a whole — was not wholly new, but goes back to Heraclitus and
the Neoplatonist Proclus. Here again Hegel, with his interest in the history
of philosophy, was aware of his predecessors. What seems to be genuinely
new in Hegel’s view of dialectic is the conception of a necessary movement.
Dialectic was said to be the “scientific application of the regularity found
in the nature of thought.” The “passing over into the opposite” was seen
as a natural consequence of the limited or finite nature of a concept or
thing. The contradiction in thought, nature, and society, even though they
are not contradictions in formal logic but conceptual inadequacies, were
regarded by Hegel as leading by a kind of necessity, to a further phase of
development.
(2) Hamlyn [1967], p. 31.
Dialectical method. Contradictions arise during the application of philo-
sophical categories like those of the One and the many, so that the philoso-
pher finds himself asserting both a thesis and its antithesis, in a manner
similar to that expounded in Kant’s antinomies. There is, Hegel thinks, a
method which reason can pursue in order to resolve any such contradic-
tion. Reason has to find a synthesis, some category which will reconcile
those which produce the apparent contradiction. But the resolution may,
in turn find itself opposed to a further antithesis which demands another
synthesis and so on. This method Hegel calls dialectic. According to him, it
provides the key to understanding how the ideas of reason may be charted.
In the end they will be seen to be dependent on the ultimate, absolute idea
which provides the ground for everything else. Thus, the idea of something
absolute and unconditional which Kant had rejected is restored.
Comment. This is typical school knowledge, a lifeless uninspired skeleton. Its place
is in an encyclopedia, but not the kind that Hegel wrote. Still, some of those logicians
who took to ‘formalizing’ (aspects of) Hegelian dialectic are lacking that kind of a
principal backbone.
None of the conceptions evoked in the first of the foregoing couple of quotations
(e.g., world process, process of thought, passing over, higher unity) is fully worked
out, neither as regards the ancient conception of dialectic nor Hegel’s. Otherwise I
could join those authors who seem to see no principal problem as regards Hegelian
dialectics, and the point would just be to read — and understand, of course —
Hegel.
Quotations 69.6. (1) Sayers [1985], p. 32.

[[I]]n the hands of Hegel and Marx at least, dialectic is no mere magic for-
mula. On the contrary, it is a fully developed, systematically worked out
philosophical theory[[.]]
Comment. Sayers’ remark is characteristic for a strong tendency within dialectical
philosophy to pass off Hegel’s dialectic as a full fledged theory. This is often com-
bined with the esoteric mysticism of a sect in which the words of the Great Master
are treated like a revelation. It should be painfully clear that there is no room for
such an attitude in the present study. If Hegel’s dialectic really is “a fully devel-
oped, systematically worked out philosophical theory”, then the present enterprise
is seriously misguided.
(2) Bubner [1973], p. 167; my translation
It is known that the question what pure ideas have to do with each other,
how they get together to form a unity and how they are still distinguishable
from each other, has led to the development of a methodical dialectic as it
is exposed in a highly speculative form in the State and the later dialogues
of Platon. Hegel has taken over the dialectical-methodical solution of the
problems of structure of a sphere of pure thought and carried through in
his Logic in a precise and exhaustive way.
(3) E. Weil [1973], p. 62.
What positively speaking is the Hegelian dialectic? To one who asks this
question the sole response one can give conscientiously will be: read Hegel.
Comment. With helpful advice like this there shouldn’t be any problem left.
Hegel’s dialectic has indeed two faces, an extremely abstract one as developed
in particular in the first chapters of the Science of Logic, and another very practical
one, as most beautifully shown in the well-known §§ 243–246 of his Philosophy of
Rights. The latter may well have been the cause for some considerable overestimation
of dialectic as a method. Still, there are authors who are quite aware of the problems
as the first of the following quotations indicates.
Quotations 69.7. (1) Findlay [1958], p. 58.

Hegel, it is well known, practised a philosophical technique called ‘Dialectic’,
and his philosophy may be called a dialectical philosophy. Exactly what is
meant by calling his philosophy ‘dialectical’ is, however, far from clear,
nor whether it is a good or a bad manner of philosophizing. The meaning
and worth of the Hegelian Dialectic is, in fact teasingly obscure even to
those who have studied Hegel longest and most sympathetically, who have
brooded deeply over the discrepant accounts that he gives of his method,
and on the Protean tricks through which he operates it. If one starts by
thinking Dialectic easy to characterize, one often ends by doubting whether
it is a method at all, whether any general account can be given of it, whether
it is not simply a name covering any and every of the ways in which Hegel
argues. And if one tries to distinguish between the way in which the method
should be used, and the way in which Hegel actually uses it, one soon
finds that his practice provides no standards by means of which its detailed
working can be tested. Clearness is not helped by the cross-lights thrown by
the use of dialectical ideas and methods by the Marxists, who try to operate
Hegelian machinery with a quite alien and unsuitable fuel. Nor is it helped
by the unbounded prejudice of those who, often without the smallest direct
knowledge of Hegel condemn him for violating elementary logical rules.
(2) Solomon [1983], pp. 21 f (quoted after Forster [1993], p. 131).
Hegel has no method as such . . . Hegel himself argues vehemently against
the very idea of a philosophical method.
(3) Kojève [1947], p. 179.
Hegel’s method [[. . .]] is not at all dialectical, and Dialectic for him is
quite different from a method of thought or exposition. And we can even
say that, in a certain way, Hegel was the first to abandon Dialectic as a
philosophic method. He was, at least, the first to do so voluntarily and with
full knowledge of what he was doing.
The dialectical method was consciously and systematically used for the
first time by Socrates-Plato. But in fact it is as old as philosophy itself. For
the dialectical method is nothing but the method of dialogue—that is, of
discussion.
(4) Flach [1964], p. 55.
Der Hegelsche Gedanke der Dialektik oder, wie wir genauer sagen müssen,
der dialektischen Methode bildet weder einen gesicherten Bestand der Phi-
losophie noch ist er dem Vergangenen zuzurechnen. Obwohl man sich nach
wie vor auf eben diesen Hegelschen Gedanken zu berufen pflegt, hat es die
Hegel-Forschung noch nicht dahin bringen können, verläßlich anzugeben,
was die dialektische Methode bei Hegel ist oder sein soll oder sein muß.
Comment. In view of the focus on the dialectical method “bei Hegel” in this quota-
tion, I wish to seize the opportunity to emphasize that I don’t think that Hegel has
anything like a specific dialectical method, but only some vague idea of some curi-
ous phenomena in conceptual thought which might give a clue to the way in which
thought forms are connected to each other in a somewhat hierarchical structure.
Quotations 69.8. (1) Wood [1993], pp. 415 f.

For Hegel, dialectic or dialectical reason constituted part of an ambitious
program to canonize his system of speculative philosophy both as a replace-
ment for traditional Aristotelian logical theory and as the metaphysical
basis for philosophical thinking in general.
(2) Beiser [1993], p. 20.
[[T]]he Hegelian dialectic makes demands of a tall order, which perhaps can
never be fulfilled. Yet there can be no doubt that the dialectic presented an
original and ingenious solution to the problem facing Hegel: how to legiti-
mate metaphysics in the face of the Kantian critique of knowledge. Even if
Hegel’s dialectic fails, we cannot accuse him of an uncritical indulgence in
metaphysics.
Comment. The implicit consensus is curious: Hegel wanted to do metaphysics. It
was for this purpose that he invented his dialectic.
(3) Burbidge [1993], p. 94.
For Hegel, logic is thought thinking itself. That is, we are to think about
the process of thinking, to identify its distinctive components and the ways
they are related. We are not interested in a casual, psychological dynamic,
but rather the kind of thinking that are universal and binding, the kinds
of thinking most reflective people share. Logic spells out these most-basic
intellectual operations.
That cannot be done all at once. The most sensible way to proceed is
to look at the most-elementary characteristics of thought and then to build
in complications in an orderly way, so that we need only understand what
has already been discussed to do justice to the new item being discussed.
Comment. What I find most-perplexing is the insolence with which people sell off
their silly ideas as Hegel’s own and then propagate the “most sensible way”. It’s the
bumptious style in conjunction with the trivial content that I find so hard to take.
(4) Wood [1990], p. 5.
Of course, the history of philosophy is a history of spectacular failures.
Descartes failed to put the sciences on an absolutely indubitable basis in
his first philosophy. Kant also failed to establish metaphysics as the forever
closed and finished science of the transcendental forms of empirical knowl-
edge. Yet Hegel’s failure was essentially more final and unredeemable than
theirs, since even the problems of Hegel’s logic remain alien and artificial
to us in ways that the problems of Cartesian and Kantian philosophy do
not.
Comment. Admittedly, this is Hegel’s greatest flaw: that the problems of his logic
“remain alien and artificial” to the “us” of the community of mediocre Hegel students.
Probably the same could be said of the problems of modern logic, had the “us” of
that community ever bothered to engage with them. I expect the enterprise of the
present study to remain doubly “alien and artificial” to that community: it tries to
bring together ideas from Hegel’s Logic with problems in modern logic.
Quotations 69.9. (1) Günther [1940/41], p. 139, footnote 4.

Der Verfasser [[Günther]] weiß ganz genau, daß überall in der logistischen
Literatur gegen die klassische Logik polemisiert wird und daß sich der Logi-
zismus als „neue Logik“ der „alten Logik“ des Aristoteles entgegensetzt. (Vgl.
etwa R. Carnap: «Die alte und die neue Logik», Erkenntnis I, S.12–26.) Aber
alle Kritik der Symbolrechner richtet sich n i e gegen die philosophische
Idee dieser Logik, sondern immer nur gegen die mangelnde Differenzierung
und nicht weit genug getriebene Ausführung dieser Theorie des Denkens.
Höchst charakteristisch ist eine Anmerkung Tarskis in seiner «Einführung
in die mathematische Logik» [[. . .]]. Wir lesen dort (S.46): „Die ganze alte
Aristotelische Logik kann fast restlos auf die Lehre von den Grundbeziehun-
gen zwischen Mengen, also auf ein kleines Bruchstück der Klassentheorie,
zurückgeführt werden.“ — Wenn die „alte“ Logik ein Bruchstück der „neu-
en“ ist, dann ist letztere eben eine Fortsetzung der Aristotelischen[[.]]
(2) Wood [1990], pp. 4 f.
Viewed from a late twentieth-century perspective, it is evident that
Hegel totally failed in his attempt to canonize speculative logic as the
only proper form of philosophical thinking. Many of the philosophical para-
doxes Hegel needs in order to make his system work are based on shallow
sophistries; the resolution to paradoxes supplied by his system is often arti-
ficial and unilluminating. When the theory of logic actually was revolution-
ized in the late nineteenth and early twentieth centuries, the new theory was
built upon precisely those features of traditional logic that Hegel thought
most dispensable. In light of it, philosophical sanity now usually judges that
the most promising way to deal with the paradoxes that plague philosophy
is the understanding’s way. Hegel’s system of dialectical logic has never won
any acceptance outside an isolated and dwindling tradition of incorrigible
enthusiasts.
Comment. How do I always manage to end up on the side of insanity? Could it
possibly have something to do with competence in logic?
(3) Zimmerli [1989], p. 191.
In professional philosophical language we define “logic” as either be-
ing the stock of syntactic connection-rules which standardizes what we call
“correct inference,” or the theoretical discussion of these rules. The respec-
tive semantics of the collection of syntactic rules no longer constitute a
characteristic object of logic, at least if we do not refer to the syntax of the
connection-rules themselves.
Comment. This “professional philosophical language” has the distinctly stale taste
of knowledge extracted from a dictionary. It may well be possible to get away with
a “definition” of logic of this kind in philosophical circles, but it is useless when it
comes to tackling the question whether modal logic, for instance, is logic proper.
Suffice it here to say that the problem of a delineation of logic is a central one in
my approach to dialectical logic which will be found in §115 in the groundworks.
(4) Zimmerli [1989], p. 200.
The reason for the truth of propositions (whether so-called “formal” or
“empirical” propositions is irrelevant) [[. . .]] is located in some form of rela-
tion of signs, designated reality, and pragmatics of signs. And this relation
cannot be seen either in the signs, or in the designated reality or in the
pragmatics of signs. Whatever this relation may be and however we may
for its part theoretize it, it is at any rate necessary to presuppose it.
Comment. In other words: this kind of philosophy doesn’t really know what it is
that it is talking about, but whatever it is, it is certainly necessary to presuppose
it. At the same time it is has distanced itself from the tradition of Hegel and moved
closer to that of analytic philosophy. In this respect it may be worthwhile reminding
the reader of Hegel’s attempt at locating truth in thought forms.42
Problems with the notion of dialectic are further increased by cheap slogans
such as the talk about a ‘dialectical triad’, an aspect of dialectic culminating in the
so-called “triad of thesis, antithesis, synthesis”. Some clever salesperson must have
launched it into the school books and now it is repeated by all the people who have
never had a look at Hegel’s writings and sometimes even those who did.43

Dialectic is a theory which maintains that something—for instance human
thought—develops in a way characterised by the so-called dialectic triad:
thesis, anti-thesis, synthesis.
Comment. Here it is, the ‘dialectical triad’. Just how dull Popper’s idea of dialectic
is becomes obvious in the next quotation.
(2) Popper [1940], p. 404.
First, some idea or theory or movement is given, which may be called “the-
sis”. Such a thesis will often produce opposition, because probably it will be,
like most things in the world, of limited value — it will have its weak spots.
This opposing idea or movement is called “anti-thesis” because it is directed
against the first, the thesis. The struggle between the thesis and the anti-
thesis goes on until some solution develops which will, in a certain sense,
go beyond both thesis and anti-thesis by recognising the relative value of
both, i.e., by trying to preserve the merits and to avoid the limitations of
both. This solution, which is the third step, is called “synthesis”.
Comment. I just hope it is clear that such an “explanation” of dialectic is included
in these materials not because of its illuminative value for the simple-minded, but
as a deterrent.
42Cf. quotations 65.16 in these materials.
43Did I already mention Murphy’s law in this context: if there is a complex issue, coin a
phrase that an idiot can use and every idiot will?
(3) Popper [1940], p. 411.

Dialectic, or more precisely, the dialectic triad, maintains that certain
developments, or certain historical processes, occur in a certain typical way.
It is, therefore, an empirical descriptive theory, comparable, for instance,
with a theory which maintains that most living organisms increase their size
during some stage of their development, then remain constant, and lastly
decrease until they die; or with the theory which maintains that opinions
are usually held first in a dogmatic attitude, then in a somewhat sceptical
attitude, and only afterwards, in a third stage, in a scientific, i.e., critical,
attitude. Like such theories, dialectic is not applicable without exceptions—
as long as we are careful not to force the dialectic interpretation. Like those
theories, dialectic is rather vague. And like those theories, dialectic has
nothing particular to do with logic.
Comment. I bet that Popper, without hesitation, would have placed himself in the
third attitude; if he had ever considered this a question at all. As far as the claim
that “dialectic has nothing particular to do with logic” is concerned, Popper is in
the company of logically illiterate Hegelians.
(4) Hall [1967], pp. 387 f.
Hegel is commonly supposed to have presented his doctrines in the form
of the triad or three-step (Dreischritt) of thesis, antithesis, and synthesis.
This view appears to be mistaken insofar as he did not actually use the
terms; and even though he evinced a fondness for triads, neither his dialectic
in general nor particular portions of his work can be reduced simply to a
pattern of thesis, antithesis, and synthesis. The legend of this triad in Hegel
has been bolstered by some English translations which introduce the word
“antithesis” where it is not required.
Comment. Compare my comment to quotation 66.11 (3) regarding the term “triad”
in Wallace’s translation.
(5) Wood [1990], p. 3.
[[T]]he grotesque jargon of “thesis,” “antithesis,” and “synthesis” began in
1837 with Heinrich Moritz Chalybäus, a bowdlerizer of German idealist
philosophy, whose ridiculous expository devices should have been forgotten
along with his name.
Comment. In a note Wood refers to ‘G. E. Mueller, “The Hegel Legend of ‘Thesis–
Antithesis–Synthesis’,” Journal of the History of Ideas 19 (1958), 411–414.’ The
article is reprinted in Stewart [1996], but I think it is as eminently forgettable as is
the jargon it criticizes.
Remarks 69.11. (1) A nice example of what can be done via translation can be
found in the translation of Hegel’s Vorlesungen über die Philosophie der Geschichte
by J. Sibree,44 p. 366, where “Reaction” is translated as “antithesis”.
(2) There exists a certain tendency to “explain” Hegel’s philosophy in terms of some
“absolute spirit”. I don’t want to comment on this; it plays no role for my enterprise.
44 Prometheus Books, Buffalo, New York 1991.
69b. Logic and Hegel’s approach to contradictions and their Aufhe-

bung . Hegel’s idea of dialectic is inseparable form his view of the role of contradic-
tions in knowledge. The latter, however, is also one of those most fiercely contested.
It still seems to lead to an outcry among the champions of rationality. Many an
attempt has been made to discredit his philosophy because of his assessment of
contradictions. In the sequel, impressively foolish arguments were put forward on
both sides, one of their main features being the logical law ex falso quodlibet or
rather ex contradictione quodlibet, which reads in logical terminology A ∧ ¬A → B.
Popper argued that this law cannot be restricted without sacrificing virtually all
of logic, in particular modus ponens. Some philosophers argued against this that
it only proves how powerful dialectics is, if everything can be proved; and some
logicians created paraconsistent logic, today in some parts of the world a flourishing
branch of non-classical logic.45
Neither Hegel nor any of his successors ever managed to illuminate the ori-
gin and necessity of the contradictions; but there are plenty of wild speculations
regarding contradictions.
I begin with a few quotations regarding the nature of the contradictions, in
particular whether they violate logic.
Quotations 69.12. (1) McTaggart [1896], pp. 8 ff.

It is sometimes supposed that the Hegelian logic rests on a defiance of
the law of contradiction. That law says that whatever is A can never at
the same time be not-A. But the dialectic asserts that, when A is any
category, except the Absolute Idea, whatever is A may be, and indeed
must be, not-A also. Now if the law of contradiction is rejected, argument
becomes impossible. It is impossible to refute any proposition without the
help of this law. The refutation can only take place by the establishment
of another proposition incompatible with the first. But if we are to regard
the simultaneous assertion of two contradictories, not as a mark of error,
but as an indication of truth, we shall find it impossible to disprove any
proposition at all. Nothing, however, can ever claim to be considered as
true, which could never be refuted, even if it were false. If then the dialectic
rejected the law of contradiction, it would reduce itself to an absurdity, by
rendering all argument, and even all assertion, unmeaning.
The dialectic, however, does not reject that law. An unresolved con-
tradiction is, for Hegel as for every one else, a sign of error. The relation
of the thesis and antithesis derives its whole meaning from the synthesis,
which follows them, and in which the contradiction ceases to exist as such.
“Contradiction is not the end of the matter, but cancels itself.” [[. . .]]
In fact so far is the dialectic from denying the law of contradiction, that
it is especially based on it. The contradictions are the cause of the dialectic
process. But they can only be this if they are received as marks of error.
Comment. As regards the law of contradiction: a similar point is made by K. R.
Popper [1940].
45 Regarding the basics of its formal treatment, see §26 in the tools.
(2) Bröcker [1962], p. 26; my translation.

The contradiction is not an inevitable stage on the path of truth, but a
sign that one has left it. It is in principle avoidable and has no speculative
power.
The next set of quotations tries to represent another attempt to make sense
of Hegel’s treatment of contradictions, equally futile as all the rest, but almost
burlesque.
Quotations 69.13. (1) Flach [1964], p. 62.
Die gemeinschaftliche Bezogenheit der Glieder einer Bestimmungsreihe auf
den zu bestimmenden Begriff, oder schlicht: die Bestimmungsfunktion, ver-
bietet den Widerspruch zwischen den Gliedern der Bestimmungsreihe. In
dieser Rücksicht lehrt Hegel die Notwendigkeit des Widerspruchsausschlus-
ses.
[[. . .]] Ein jeder bestimmende Begriff hat zur Bedingung seiner letztli-
chen Bestimmtheit das Totum aller übrigen (zur Bestimmung tauglichen)
Begriffe. [[. . .]] Jeder bestimmende Begriff ist bestimmt nur dank seiner Zu-
gehörigkeit zum System der bestimmenden Begriffe überhaupt, dessen Ord-
nung der durchgängige Widerspruch bildet. In dieser Rücksicht lehrt Hegel
die Notwendigkeit des Widerspruchs.
Hegel lehrt also ebenso die Notwendigkeit des Widerspruchs wie die
Notwendigkeit der Vermeidung des Widerspruchs.
Comment. Was den „Widerspruch zwischen den Gliedern der Bestimmungsreihe“
angeht, so wäre Hegel hiernach strenger als die moderne klassische Logik.46
(2) Flach [1964], p. 63.
In der Triade von Thesis, Antithesis und Synthesis begreift Hegel jedweden
Begriff als die Implikation der Totalität der Begriffe nach den zwei Seiten der
absoluten Begründung und der uneingeschränkten Konkretion des Denkens.
Hegel [[. . .]] emphasizes that he is not talking of ‘contradiction’ in some
half-hearted or equivocal manner: he is not saying that X is A in one sense,
but not A in another, that it is A in so far as it is X but not in so far as
it is something else. [[. . .]] Hegel makes it as plain as possible, that it is not
some watered-down, equivocal brand of contradiction, but straightforward,
head-on contradiction, that he believes to exist in thought and the world,
and to be an eliminable component in self-conscious spiritual reality.
(2) Findlay [1958], p. 79.
Ordinary thought steers clear of contradiction by refusing to apply its con-
cepts in unwonted cases, and a deductive system avoids them by the sheer
precision of its abstractions, in which all factors that might lead to hesi-
tation or conflict have been deliberately excluded. Contradiction will not
arise as long as one remains resolutely at a single level of discourse, which
46 Cf. Frege in quotation 71.16 (4) and, perhaps, (5) in these materials.
one does not seek to connect, nor see in relation, with other forms of dis-
course. It arises only when one tires of the deadness and sheer senselessness
of such one-level discourse, and tries to pass on to something deeper: its
point of emergence is not within smoothly functioning patterns of discourse,
so much as between them. [[. . .]]
We may hold, in fact, that Hegel’s notion and use of contradiction,
confusing as it in many ways is, none the less embodies one of the most
important of philosophical discoveries, whose full depth has not even yet
been properly assessed.
Comment. My point: it is not possible to resolutely remain at a single level of dis-
course, at least not as soon as there is a sufficient amount of expressibility in a lan-
guage. What I take to support my point is Gödel’s interpretation of the metatheory
of arithmetic within its own object theory (and thereby constructing an undecidable
sentence). In other words: it is not just “when one tires of the deadness and sheer
senselessness”, but one just can’t avoid having passed beyond it. Differently put
again: it may well be that one refuses to apply one’s concepts in unwonted cases,
but there is no way of recognizing them beforehand.
(3) C. Taylor [1975], p. 80.
Hegel holds that the ordinary viewpoint of identity has to be abandoned
in philosophy in favour of a way of thinking which can be called dialectical
in that it presents us with something which cannot be grasped in a single
proposition or series of propositions, which does not violate the principle
of non-contradiction: ∼(p.∼p). The minimum cluster which can really do
justice to reality is three propositions, that A is A, that A is also ∼A; and
that ∼A shows itself to be after all A.
(4) Dulckeit [1989], p. 119.
The Law of Identity, which Hegel takes up in the “Doctrine of Essence”
says that “everything is identical with itself.” But if we wish to express the
fact that something is self-identical we must put it in the form A=A, a
proposition which clearly uses two terms, not one. So while the object o
is one, to understand and state that fact requires two terms “o=o.” Ac-
cording to Hegel, this is consistent with Leibniz’ Law of the Identity of
Indiscernibles. Implicit in the Law of Identity, then, is something which
simply eludes the logic of the understanding: the idea that identity makes
sense only as self-identity which, in turn, implies self-relation and, there-
fore, contains essentially the idea of difference (LL, §116). Difference is
the necessary condition for grasping (as well as expressing) the notion of
identity. What is more, as is the case for all opposites for Hegel, identity
and difference are interdependent, or mediated, such that either one makes
sense only in terms of the other.
Quotation 69.15. (1) Inwood [1983], p. 102.

The old metaphysics contained [[. . .]] contradictions. Hegel inherited from
Kant the belief that by its procedures valid, or equally valid, proofs could
be given, for example, both of the proposition that there is a free will and
of the proposition that there is not, both of the proposition that the world
is infinite in space and of the proposition that it is not, and so on. The
awareness of such contradictions as these, however, cannot in itself gen-
erate second-order reflection upon metaphysics and its procedures, for it
presupposes it. Full awareness that one’s beliefs or procedures are contra-
dictory already involves the detachment from them that we are attempting
to explain.
Comment. This makes a good deterrent too. It is quite representative of the sort
of drivel that is cultivated in the strongholds of philosophy: full of wishy washy
notions such “full awareness”, “second-order reflection”, “one’s beliefs or procedures”;
this is the opportunity where I should seriously like to recommend a bit of linguistic
analysis.
(2) Harris [1987], pp. 165 f.
Misconception of the function of negation and contradiction is the
root of widespread misunderstanding and faulty interpretation of dialec-
tic, among its defenders as well as among its detractors. The former are
mostly Marxist theorists who base their arguments on the wrong reasons,
or at best, on half-truths. Lenin advocates dialectical logic in preference to
formal on the ground that the latter is based on the Laws of Identity and
Noncontradiction, which hold, he says, only of static and unchanging sub-
jects but are violated by the phenomena of motion and life (hence Zeno’s
paradoxes). Formal logic, he therefore maintains, can be used only in rela-
tion to the immobile and inanimate, while dialectical logic, which rejects
the Laws of Identity and Noncontradiction, recognizing the inherent con-
tradictions in the nature of things, is the true logic of movement and life.
In this there is but half the truth, for [[. . .]] movement and life are not the
precondition of dialectic, but are forms of its manifestation or expression
in spatiotemporal existence. Dialectic is prior to movement, as it is to time
and change, and it establishes the Laws of Identity and Noncontradiction,
so far from rejecting them.
Comment. In a subsequent paragraph Harris deals with Popper [1940] as a partic-
ularly gross example of “misunderstandings and misrepresentations of Kant, Hegel,
and Marx” which I wholeheartedly recommend to the reader’s attention. Notice also
the mention of “static and unchanging subjects” and compare Girard in quotation
85.29 (1) in these materials.
(3) Forster [1993], pp. 143.
Hegel’s true situation is, I think, as follows. On the one hand, he rec-
ognizes with the rest of us that it is unacceptable to make contradictory
claims about reality [[. . .]]. On the other hand, his own philosophical view-
point is inextricably involved in affirming contradictions, but it does not
affirm them of reality and so does not fall foul of his and our proscription
of this. His viewpoint avoids affirming contradictions of reality because it
does not use or recognize the validity of the concept of reality. It renounces
the distinction between reality and thought (being and thought, object and
thought, object and subject, object and concept, etc.). And consequently,
Hegel’s view, it renounces these concepts themselves, since the distinction
is, in his view, an essential part of their very definition. Thus he writes that
“Pure science presupposes liberation from the opposition of consciousness.
It contains thought insofar as this is just as much the object in its own self,
or the object in its own self insofar as it is equally pure thought. However,
strictly, “to talk of the unity of subject and object, . . . of being and thought,
etc. is inept, since object and subject, etc. signify what they are outside of
their unity.” Hegel’s philosophical viewpoint thus officially makes no claims
whatsoever about reality, and a fortiori no contradictory claims about it.
Of what, then, if not reality, does Hegel wish to confirm contradictions?
[[. . .]] [[H]]e does not merely wish to affirm them of thoughts or concepts [[. . .]].
Rather, he wishes to affirm them of whatever is left once the essentially
oppositional concepts of reality or object, on the one hand, and thought or
concept, on the other, have been overcome and synthesized. Hegel variously
calls this Reason, the Logos, the Absolute Idea, the Concept, Absolute
Spirit.
Comment. The “rest of us” does certainly not include dialetheists; but I hasten to
declare that I do not want to be included in that “us” either, and that not just
because of their lousy Hegel reception.
(4) Wood [1990], p. 3.
We might compare Hegel’s treatment of philosophical paradoxes with
the later Wittgenstein’s. Wittgenstein held that contradictions or paradoxes
do not “make our language less usable” because, once we “know our way
about” and become clear about exactly where and why they arise, we can
“seal them off”; we need not view a contradiction as “the local symptom
of a sickness of the whole body.” For Wittgenstein contradictions can be
tolerated because they are marginal and we can keep them sequestered from
the rest of our thinking; for Hegel, they arise systematically in the course
of philosophical thought, but they do no harm.
Comment. One probably has to be a philosopher to enjoy the privilege of being able
to say that contradictions can be tolerated, without feeling the necessity to say how.
Such is the privilege of the logically illiterate. Apart from that, to portray Hegel
as holding that contradictions “arise systematically in the course of philosophical
thought, but they do no harm” sounds like deep irony; a bit like saying that the
mixing of genes is unavoidable in sexual reproduction, but it does no harm.
(5) Solomon [1974], p. 280, footnote 7.
The best definition I know of ‘aufheben’ has been proffered by the
hardly Hegelian philosopher Frank Ramsey in his Foundations of Mathe-
matics (pp. 115–16): ‘the truth lies not in one of the two disputed views but
in some third possibility which has not been thought of, which we can dis-
cover by rejecting something assumed as obvious by both the discussants’.
Comment. Probably good common sense, but to attribute such a view to Hegel is a
bit much, though it may well be common nonsense. At best it sounds a bit like what
Kant offered as a solution to his antinomies: an inadmissible condition assumed as

correct which has to be rejected as a result of a contradiction; but the subjective
character of the process (“a third possibility which has not been thought of”) puts
it at a great distance to Hegel.
Quotation 69.16. (1) Pinkard [1990a], pp. 19 f.

Mr. Duquette argues that we must understand the beginning of the
Logic as involving a principle of dialectical thought that the ‘understand-
ing’ with its insistence on ‘ordinary logic’ cannot grasp. This is a very
important issue, for it involves our whole understanding of Hegel’s project.
First, I want to argue that Hegelian dialectic does not challenge ordinary
logic. Second, I want to suggest at least that Hegel’s Logic should not be
taken strictly as a logic at all but only as an understanding of philosophical
explanation.
[[. . .]]
Mr. Duquette says that ordinary logic cannot make sense of the begin-
ning of the Logic, with its talk of ‘being’ and ‘nothing’ passing over into
‘becoming’. Ordinary logic, he says, cannot understand this beginning nor
how new content is supposed to arise from it. What exactly is this ‘ordinary
logic’ ? Mr. Duquette says that it is the logic of finite things. (What of the
logic of infinite things? We have a good example of such in the Cantorian
mathematics of the transfinite, or, for that matter, in the logic of infinite
sets).
Comment. It may be worthwhile mentioning here that my approach to dialectical
logic is inseparable from my view of Cantor’s theory of transfinite cardinals as a
product of false logic. But in my treatment of the question of the logic of the finite
and the infinite one doesn’t have to go as far as transfinite mathematics. The totality
of natural numbers already marks the point of divergence from a classical position,
and that in the realm of quantification.47
(2) Pinkard [1990a], p. 22.
[[N]]one of Hegel’s dialectic in the Logic is in opposition to ‘ordinary logic’.
Comment. There may be no agreement as to what “Hegel’s dialectic in the Logic”
consists in, but Pinkard can state that it is not “in opposition to ‘ordinary logic’.”
I wish to warn against such pre-emptive claims.
(3) Pinkard [1990a], p. 20.
[[L]]ogic per se does not require me to put things into either/or dichotomies;
just note that the truth table for ‘x or y’ is different from the truth table
for ‘either x or y’.
Comment. According to my understanding, “to put things into either/or dichoto-
mies” means to say of things that they are ‘either x or not x’. It may be worthwhile
noting that the truth table for ‘either x or not x’ is the same as that for ‘x or not
x’. There is, however, another principal point which I regard as crucial: are truth
47 Cf. my treatment of the natural numbers and mathematical induction in chapter XXXIV
in the groundworks.
tables adequate for determining the laws that govern a logical constant? Of course,
they are not; logical reasoning is needed on top of that; but this does not seem to be
a point that would have entered philosophical consciousness. Contrast, for instance,
Girard/Lafont/Taylor [1989], p. 30, regarding the structural rules in Gentzen-style
logic.48 In §129, I show how to introduce logical constants as truth functions in
dialectical logic, and they are exactly the same as in classical logic — as far as their
truth tables go. But, they are handled differently. To understand this point is crucial
for an understanding of my approach to dialectic.
(4) Pinkard [1990a], pp. 24 f.
Perhaps we should conclude that a speculative logic in any strict sense really
is impossible. [[. . .]] If [[. . .]] we mean by logic ‘those rules of inference that
are truth-preserving’ [[. . .]], then Hegel’s speculative logic is not, precisely
speaking, a logic at all. [[. . .]] What would we lose if we[[ ]]decided that,
strictly speaking, there is no speculative logic? We lose nothing really of
Hegel’s theory; but we might clear the air a bit.
Comment. I include this quotation here to make it clear that I don’t think it is up to
anybody to decide this question, least of all logically poorly equipped philosophers;
what one loses by sticking to the classical logic and what alternatives there are, these
are serious questions which require some substantial competence and originality in
logical research. (If this were the proposal of a logician, I could say that what you
would lose by deciding such a question would be your credibility as a logician; but
this doesn’t work with philosophers; they live according to the principle “Und ist
der Ruf erst ruiniert, so lebt’s sich völlig ungeniert”.) As regards the question what
we lose of Hegel’s theory, it should be clear that there is no agreement amongst
philosophers as to what Hegel’s theory is all about and, accordingly, the question
would have to be specified. We would probably not lose anything of Pinkard’s in-
terpretation of Hegel’s theory, but this is hardly a point of Hegel’s theory.
Quotations 69.17. (1) Harris [1983], p. 7.

Formal logic denies the mutual interdependence of form and content
and claims to abstract entirely from the latter without detriment to either.
This, Hegel avers, is because it operates on the reflective level of the un-
derstanding, which makes hard and fast distinction and then elevates them
into separations or dichotomies which a higher reason shows to be unwar-
ranted. Hence the presumed independence of logic from metaphysics, which
the formal logician alleges, and which even in his own practice proves to
be false, results from too absolute a separation between thinking and its
object, between the form and the content of thought, and between the real
and our knowledge of it.
Comment. How does the denial of the mutual interdependence of form and content
manifest itself on the level of its axioms and/or rules of formal logic? Or is it the
mere fact of using axioms and/or rules that does it?

(2) Pinkard [1990a], p. 23.

Hegel was also worried about logic’s formality, since he thought it doubtful
that logic could be ‘true’ if it were purely formal. He could have avoided that
worry altogether if he had been in the position to hold the contemporary
view that logic is not intended to provide truth at all but just to preserve
it.
Comments. (α) What Pinkard seems to be saying is that Hegel could have avoided
worrying about the truth of logic had he abandoned the project of a truth gener-
ating logic altogether, like modern logic — at least tried to do. As if Hegel had
never written a Science of Logic. But even that would not solve the problem of
the foundation of (higher order) logic of being caught between incompleteness and
inconsistency. What I want to say is that modern logic does not escape the problem
of paradoxes in the sense that they prevent the formulation of a truth predicate
within a system containing a certain portion of arithmetic. One would also have
to close one’s eyes against this to stop worrying. In other words, Hegel could have
avoided worrying had he accepted the kind of blinkers promoted by Pinkard.
(β) Regarding ‘truth preserving’: the axiom of unrestricted abstraction is truth
preserving, but incompatible with classical logic.49 The point here is that (following
Tarski) truth definitions for theories based on classical logic depend essentially on an
induction on the length of formulas which means that if logic is intended to capture
reasoning that preserves truth, then it doesn’t live up to this intention. And this
is what brings me to Hegelian dialectic. For those who can see the problem it is
just annoying that important epistemological concepts such as truth are ‘framework
dependent’,50 whatever the ‘contemporary view’ of the day may be.
(γ) Actually, the situation regarding truth preservation is even more critical than
that: the first of the canonical derivability conditions together with the (indirect)
fixed point property (diagonalization) spoils truth preservation in a sense elaborated
in section 114c in the groundworks and commented on in remark 114.3.
(δ) What Pinkard is advocating here is essentially Kant’s view: logic can not be a
science of the speculative understanding.51
(3) Hanna [1986], p. 255.
Hegel’s philosophical use of common logic is a higher-order activity than
the common-logical activity, and does not therefore by any means compete
with the common logic at its own level. Hegel’s higher-order comments
about the common logic are ontological remarks or recommendations, not
common-logical remarks or recommendations.
Comment. This sounds to me like a desperate attempt to save a little niche for
Hegel. I don’t see, however, that it is necessary to exempt Hegel from ‘common
logic’. Hegel can well be interpreted as having a lot to say with regard to ‘common’
logic ‘at its own level’.
49 Cf. Priest [1983b].
50 Cf. Putnam in quotation 68.10 (2) above. On the other hand, take Gödel in quotation
95.24 regarding the notion of computability.
I close this section with some quite paradigmatic specimen of philosophical

drivel.
Quotation 69.18. Burbidge [1993], p. 90.

The pattern of Russell’s argument can be found elsewhere. A thought
is applied to itself. That act of self-reference reveals an incongruity: its op-
eration comes in conflict with what it says. Since any paradox demands
solution, we identify what the problems are, explain why they emerge, and
suggest how to overcome them in a more adequate way. A legitimate so-
lution does not jump to an arbitrary theory that rejects the legitimacy of
anything like the original paradox, but probes into the grounds of the con-
tradiction — why it arises in the first place — and so gets to the heart
of the matter: the central, essential core that is involved in thinking such
thoughts.
This is the process of thinking that Hegel analyzes in the second book
of the Science of Logic: “The Doctrine of Essence.”
Comment. At the end of the first passage, after “thoughts”, there is an endnote
worth quoting: “For those knowledgeable about Hegelian terminology, I am here
talking about der Sache selbst.” May I add for those not knowledgeable about
either Hegelian terminology or the German language, that “Sache” is female, so
nominative as well as accusative is “die Sache”; so at best, i.e., provided it is more
than showing-off, Burbidge is talking about die Sache selbst. Readers might find it
helpful to consult quotation 66.15 (1) where Hegel does not explicitly speak of “der
Sache selbst”,52 but which nevertheless might help to counter-balance the simple-
minded mechanism that Burbidge presents. Also of interest in this context: quota-
tion 65.1 (3), in particular my comment.
69c. Forms of thought and the idealistic conception. In order to do

justice to Hegel’s conception of dialectic it is most important to grasp the idea that
“Denkbestimmungen” are not just empty boxes without any content of their own
but are constitutive for knowledge, i.a.w., they have a content of their own.
Unfortunately, this seems to meet with great difficulties.
Quotations 69.19. (1) Inwood [1983], p. 266.53

[[W]]hat Hegel would need to show if he is to establish that pure thoughts
are contentful on their own is not simply that a contentful narrative must
contain thoughts and no ‘individual events and situations’ at all. [[. . .]]
But why, in any case, is Hegel so anxious to show that pure thoughts
have an independent content? Sometimes [[. . .]] he speaks as if it is to secure
our cognitive access to supersensible objects.
52 I am sorry about this; it is not an attempt to confuse the reader, only this is the geni-
tive case; which might well be the reason for Burbidge’s grammatical blunder in the first place.
Recommendation: if you feel insecure about which form the article takes, just drop the German
article, replace it by the English one and just speak of the “Sache selbst”. You reduce the risk of
coming across as a bumptious idiot.
53 I recommend reading this quotation in the context of quotations 65.1 (2) and 65.9 (3).
Comment. The first step away from a frog perspective of “contentful narrative”
would be to recognize a more general problem, viz., that of how “the empty forms
of logic come to disgorge so rich a content?” 54 such as, for instance, arithmetic
(numbers are supersensible objects of some sort). The point of the present study
is to bring Hegel’s dialectic and the problem of a logical foundation of arithmetic
together, but this is a project that requires some basic capacity of abstract thought.
(2) Inwood [1983], p. 269.

Granted that pure thinking is possible, why should we indulge in it?
Hegel gives various answers to this question[[.]] [[. . .]] [[Some]] answers which
he gives tend to assume that philosophy as a whole is a sort of pure thinking
and attempt to meet the question: Why should we engage in pure thinking
rather than, for example, empirical scientific thinking? [[. . .]].
[[. . .]] Hegel often blurs the distinction between different types of think-
ing. In particular, his answer to the question why we should think purely
rather than at some other level is bedevilled by his failure to distinguish
it from the distinct question: Why should our thinking at any one level
advance once we are engaged in it?
Comment. Nothing has to be granted regarding the possibility of pure thinking: a

form of pure thinking is nicely realized in theories of higher order logic. Unfortu-
nately, this is unlikely to be recognized by philosophers who are caught in some
subjective (psychologistic) idea of ‘thinking’.55 But it could be sufficient to get rid
of a whole lot of “why should we”-asking by simply pointing out that there is no
more reason why anybody should indulge in pure thinking than there is reason why
anybody should indulge in higher order logic. These are activities that are better
left to people who have some sense for it, in the case of pure thinking at least a
grasp of a certain constitutiveness of thought (forms) in the (formulation of) knowl-
edge.56 Research into an unexplored area like that of the progression of theories
in higher order logic, for instance, can’t be left to people who expect to have that
area mapped out and secured beforehand so they could be advised as to why they
should engage in it.
Hegel’s idealism has given rise to numerous comments and, above all, misun-
derstandings.
54 Frege [1884], p. 22, cf. quotation 71.7 (1) in these materials; contrast the passage which
Frege quotes from Mill in that same quotation. Cf. also: Gödel’s struggle with analyticity and
content, as presented in quotation 87.11 (2) in these materials.
55 Hegel [1812], p. 65, for instance, speaks of “reine Denkformen”, i.e., “pure forms of thinking”,
i.e., “pure” is an adjective of “forms”, not of thinking; the forms are pure, as one would expect
of a logic; Miller [1969], p. 63, translates “forms of pure thought”. Sure, Hegel also uses the terms
“reine Gedanken” (“pure thought”) and “reines Denken” (“pure thinking”), but I don’t see that this
justifies reading Hegel’s Logic in terms of some subjective thinking.
56 Personally I find it hard to understand why anybody who has as little affinity to Hegel’s
thought as Inwood seems to have, would undertake to write a full length book on his philosophy.
Nevertheless, it is a good indicator of the poor state of Hegel interpretation.
Quotations 69.20. (1) Westphal [1989], p. 140.

The first point to note about Hegel’s “idealism” is that it is not any
sort of Berkeleian or Kantian idealism, nor is it a phenomenalism. Hegel’s
idealism is an expression of what any non-skeptical view must hold: our
conceptions can capture the way the world itself is. That Kant thought
otherwise is Hegel’s main complaint against Kant. Hegel regards himself as
a defender of the powers of human cognition, not by reducing the objects of
knowledge to a set of subjective (or intersubjective) states, but by arguing
that our conceptions can be adequate to comprehend the world itself.
(2) Westphal [1989], p. 141.
One important point in [[Enz. §41z2 & §42z3]] is Hegel’s insistence that the
contents of our conceptions (when we have true knowledge, at least) and
the structure of the world are the same. [[. . .]]
This is to say, the object of knowledge is the world itself.
Comment. The phrases “true knowledge” and “structure of the world” are curious
here. Nowhere in the passages mentioned is Hegel talking about true knowledge
that would fit this context, nor about a structure of the world. My suspicion is that
Westphal thinks of applications of categories to objects given from outside and truth
as correspondence to how the world itself is, but not the categories as they are “in
and for themselves”.57 This would be a fairly common misinterpretation of Hegel.
(3) Colletti [1969], p. 7.
The central theme of Hegel’s thought is his thesis of the identity of idealism
and philosophy.
Comment. Stated without any further proviso this strikes as misleading. However:
identity of metaphysics, speculative philosophy and logic.
Some interpreters have tried to grasp the concept of dialectic on the basis of an
intimate connection of the human as observer and the world. I just mention here
that I find such attempts unsatisfactory because of their subjective character.
Quotations 69.21. (1) Günther [1962], p. 57.

Unser ganzes wissenschaftliches Denken hat bisher danach gestrebt, das
Sein der Welt ganz objektiv zu erkennen, so wie es “an sich”, d.h. unabhängig
und unentstellt durch mögliche “optische” Verzerrungen unseres Bewußtsein
(in dem es sich notgedrungen spiegelt) wirklich besteht. [[. . .]] Radikale Ob-
jektivation erschien auch als ein Ideal in den Geisteswissenschaften. Der
Historiker Leopold von Ranke sprach es aus, wenn er den sehnsüchtigen
Wunsch äußerte, daß er sein Selbst völlig auslöschen möchte, um die Dinge
so zu sehen, wie sie wirklich seien.
Heute ist dieses Ideal tot. Es existiert nicht einmal mehr in den exak-
testen Naturwissenschaften. Schon im Anfang der dreißiger Jahre sah sich
Werner Heisenberg veranlaßt, in einem philosophischen Rechenschaftsbe-
richt über die moderne Quantenmechanik zu schreiben: “Eine ganz scharfe
57 Enz. §41z2. Translated by Wallace as “absolutely and for their own sake”.
Trennung der Welt in Subjekt und Objekt (ist) nicht mehr möglich”, und
dementsprechend hat der “völlig isolierte Gegenstand [[. . .]] prinzipiell keine
beschreibbare Eigenschaften mehr”.
(2) Titze [1965], p. 296.
Es spielt der Mensch als Denken selbst in diesem Vorgang mit. Wie
er sich zu etwas verhält, wie er es betrachtet, das ist bei der logischen
Entwicklung einer Theorie mit in Betracht zu ziehen.
Comment. Note the emphasis of the role of the human being in this account. It is
by no means clear what relevance the attitude of the human being could have on
the logical development of a theory. Two things are mixed up here: psychology and
logic. It is not the human being, who plays a role, but description. Moreover, this
is far too vague to be of any help. It has to be made precise in which way this plays
a role.
69d. Thought determinations, truth, and logic. The notion of truth is
another dark point in Hegel’s philosophy. There is a certain tendency to try to
cope with Hegel’s remarks as concerns truth by distinguishing a correspondence
theory and a coherence theory of truth and a lot of people seem to be having a jolly
good time juggling with these words. The issue become particularly silly when the
Hegelian tradition is extended to include Blanshard, Bradley etc.
Quotations 69.22. L. J. Cohen [1978], p. 357.
Hegelian idealists [[. . .]] believe that everything is interconnected with ev-
erything else: all true propositions entail one another. As Blanshard put it,
Hegelians believe that
In the real world . . . a change in one fact or event would necessitate that all others
be different. Suppose I climb the hill behind my farm house in Vermont and look
across at Mount Washington. I am wearing a felt hat at the time. It is sensible or
quite sane to argue that if I had worn a straw hat instead, that fact would have
made a difference to Mount Washington? I not only believe it would, but that
the argument for this conclusion is strong almost to demonstration.
Comment. The quotation is identified as Blanshard, “The Nature of Thought, 1939,

Vol.II,” p. 293. My misgivings about a position of this kind is that it reduces the
problem to some kind of New Age slogan, the “interconnectedness of everything”.
The example of Blanshard’s hat and Mount Washington may well attract the atten-
tion of beautiful souls, but when it comes to studying a connection and its impact
on what is generally called ‘facts’, I prefer something like the fixed point property
in arithmetic.
Quotations 69.23. (1) Westphal [1989], p. 113.
What, then, is to be made of Hegel’s notorious rejection of the cor-
respondence theory of truth? Part of Hegel’s notoriety in this regard is
simply mistaken. Hegel certainly does reject correspondence between con-
ception and the object as a criterion of truth but he does not reject it as
the nature of truth (insofar as epistemological issues are concerned).
(2) Westphal [1989], pp. 113 f.

There are many passages in which Hegel does dismiss the notion of truth as
the correspondence of conception and object in favor of a notion of truth as
the correspondence of a thing with its own “concept” or nature. (Recall the
idiom, “a true friend,” as someone whose friendship is genuine and abiding.)
It is important to notice about this second notion of “truth” first, that it
concerns value judgments, and second, that its application presupposed that
“truth” in the ordinary sense of correspondence of conception and object is
unproblematic. [[. . .]]
It is important to stress that Hegel’s second sense of “truth” concerns
value judgments in order to reinforce the point that these two are distinct
roles to play in Hegel’s philosophy. [[. . .]] [[E]]ngaging in the determinations of
value that pervade Hegel’s philosophy presupposes that truth in the sense of
correspondence of conceptions and their objects is available and unproblem-
atic; it does not supplant this more ordinary conception of truth. If Hegel
thinks that determinations of value are the preeminent philosophical con-
cern, and if he accordingly disparages concern with mere “correctness,” it is
only because he thinks that he has shown that having “correct” conceptions
which correspond with their objects is unproblematic.
Comment. Compare quotation 65.10 (5), where Hegel speaks of “correctness” and
“formal coincidence between our conception and its content”.
In view of what I am aiming at I suggest that Hegel’s notion of truth be seen

in connection with his idea of the content of thought determinations in themselves.
The project of a transcendental deduction of categories (positive dialectic) is the
basic problem and the criterion for dialectic in the tradition of Hegel: the possibility
of a deduction of the categories.

The notion of Being is chosen by Hegel as the beginning of his logical
Dialectic because an acknowledgment of Being — that there is something
or other, or that there is this or that — seems to him the simplest and most
fundamental of thinking approaches.
Comment. I regard this is a misinterpretation of Being as Existing.
(2) Sartre [1943], p. 44.

[[W]]e are tempted to consider being and non-being as two complementary
components of the real—like dark and light. In short we would then be
dealing with two strictly contemporary notions which would somehow be
united in the production of existents and which it would be useless to con-
sider in isolation. Pure being and pure non-being would be two abstractions
which could be reunited only on the basis of concrete realities.
Such is certainly the point of view of Hegel.
(3) Sartre [1943], p. 47.

[[W]]hat needs examination here is especially Hegel’s statement that being
and nothingness constitute two opposites, the difference between which on
the level of abstraction under consideration is only a simple “opinion.”
To oppose being to nothingness as thesis and antithesis, as Hegel does,
is to suppose that they are logically contemporary. Thus simultaneously
two opposites arise as the two limiting terms of a logical series. Here we
must note carefully that opposites alone can enjoy this simultaneity because
they are equally positive (or negative). But non-being is not the opposite
of being; it is its contradiction. This implies that logically nothingness is
subsequent to being since it is being, first posited, then denied. It can not
be therefore that being and non-being are concepts with the same content
since on the contrary non-being supposes an irreducible mental act.
(4) Duquette [1990], p. 5.
[[W]]hen we attempt to fix a determination of Being simply as Being we
find that it has none, that like the concept of Nothing its referent is pure
indeterminacy. Moreover, conceptual thought is caught in a dilemma in
which it must, at least according to the rules of ordinary understanding,
affirm exclusively either the identity or the distinctness of these concepts,
and yet it finds itself affirming both. This impasse can be described as the
movement of thought to and fro between the concepts of Being and Nothing,
or as the passing of each concept into the other. The impasse is resolved only
when thought steps back or distances itself, as it were, from this oscillation
and recognizes that there is indeed a moment of determinateness found
in thinking these concepts. Thought recognizes that the very movement of
thought in this conceptual connondrum [[sic!]] thinking, and this movement
is signified in the concept of Becoming. Thus, Becoming, as the first form
of dialectical thought, can be understood as the “conceptual synthesis” of
Being and Nothing.
(5) Pinkard [1990a], p. 22.
The move from being to existence is speculative. There is no deductive
inference here; rather new determinateness is added. That is why it is called
speculative: we add something new that cannot be deduced from the given
content. Speculative inferences therefore do not challenge ordinary logic;
but they are also not ruled out by it. The particular speculative inference is
justified in that only such a move—identifying being with existence—will
allow one to avoid the contradiction.
Comment. This sounds like an effusion of a free floating fantasy, unburdened by
logical nitty-gritty. In view of the speculative inference added in the form of Z-
inferences in §132 it may be worthwhile having a look at paraphrase 120.10 on
p. 1639 in the groundworks. I suppose hardly any two Hegel-interpreters will
agree on what the nature of the move from Sein to Dasein consists in,58 but I dare
say that many will agree that is not one of identification.
58 Pinkard renders “Dasein” as “existence”, not without a preceding remark, however.
Quotations 69.25. (1) C. Taylor [1975], pp. 225 f.

[[W]]e think of our concepts as instruments of our thought which may or
may not apply adequately to reality. We start off, that is, with a notion of
thought as over against the world about which we think. This dualism is
linked to another; since a concept or a category is thought of as a universal
which can apply to many contents, we are tempted to think of it as an
abstract form over against the sensible contents to which it applies.
This double dualism naturally leads us to think that a study of concepts
is quite distinct from a study of reality; and more particularly, that the
necessary relations between concepts which we may discover from such a
study in no way allow us to conclude to necessary relations among the things
to which they apply. Logic, as a study of these relations is thus necessarily
formal, touching our way of thought, and not the contents that we think
about.
But Hegel [[. . .]] does not accept this notion of the concept and the
two dualisms it entails. Thought and the determinations through which it
operates (the Denkbestimmungen, or categories) are not the apanage of a
subject over against the world, but lie at the very root of things. For the
reality which we perceive as finite subjects is the embodiment of Geist or
infinite subject. But the life of Geist is rational thought, a life which is
carried in our own thought, i.e., that of finite subjects, and is adequately
expressed in so far as we think rationally. The rational, truly universal
thought which is expressed in our categories is thus spirit’s knowledge of
itself. Since the external reality to which these categories apply is not only
an embodiment of Geist, but is posited by Geist as its embodiment, and
hence reflects the rational necessity of thought, in grasping the categories
of thought about things, we are also grasping the ground plan or essential
structure to which the world conforms in its unfolding.
(2) Findlay [1981], p. 132.
[[D]]ialectical reasoning looks to what is implied by an assertion though not
at all covered in its basic foundation[[.]] [[. . . ]] Dialectical reasoning, in short
involves that genuine passage beyond premisses that is also involved in
passing from an object-language to a meta-language.
(3) Findlay [1981], p. 140.
I do not, further, despair of the formalization of higher types of diction at
their own level —I have myself long wished to elaborate a purely Platonian
language—but the reduction of higher levels to lower should never be at-
tempted. Dialectical Metabases can as such never be formalized, for they
represent quantum leaps of language, and of the thought behind language.
Dialectic is therefore forever safe from the formal logician.
Comment. It might be worth a mention here that my point in bringing dialectic

and formal logic together is that (higher order) logic is not safe from dialectic, not
vice versa. Notice also that the paradigm of quantum leaps provides no principal
argument against formal logic; quantum leaps in physics can be numerically and
theoretically conceptualized.
(4) Kosok [1966], p. 237.
The formalization of Hegel’s dialectic logic rests upon the contention
that Hegel’s intuitively generated system can be represented as a meta-
language structure in which a given set of elements on one level are capable
of being analyzed from a meta-level which refers to the original elements
from a perspective of reflection, thereby bringing out and expressing prop-
erties about that level not capable of being formulated within the original
level itself.
Comment. This sounds quite interesting, but the actual realization is just schematic
and certainly not suitable to provide a derivation of thought-forms, at least not in
the sense of providing laws like those of modality, for instance.

The talk about the “truth of Being” has a justified meaning in Hegel’s Sci-
ence of Logic, because here truth means the certainty of absolute knowledge.
And yet Hegel, as little as Husserl, as little as all metaphysics, does not ask
about Being as Being, that is, does not raise the question as to how there
can be presence as such. There is presence [[Anwesenheit]] only when clear-
ing [[Lichtung]] holds sway. Clearing is named with alētheia, unconcealment,
but not thought as such.
Quotation 69.27. Bradley [1893], p. 318.

The Absolute, considered as such, has of course no degrees; for it is
perfect, and there can be no more or less in perfection[[.]]
Comment. Compare this with quotation 65.10 (2), where Hegel remarks that being
might be considered as the “first definition of the Absolute, and all further determi-
nations and developments would be richer and more closely determinate definitions
of it.”
Quotations 69.28. (1) Rosen [1982], pp. 117 f.

What really interests Hegel in the Transcendental Deduction is not that
the purely formal nature of self-consciousness makes the ordering operation
of a synthesis of the manifold necessary but that the intrinsic relationship
which self-consciousness has to its contents should give synthesis its princi-
ple. This opens up the possibility of a synthesis which is truly creative —
not just the act of an abstract universal classifying a manifold of content
received from outside. Hegel cannot, of course, attribute such a synthesis
to Kant. But he claims that Kant fails to follow up the breakthrough which
his initial discovery of the intrinsic relationship of synthesis represents and
regresses to a ‘psychological’ picture of the relationship between the concept
and received materials of experience.
(2) Bröcker [1962], p. 35; my translation.

What Kant realizes [[erkennt]] and Hegel fails to realize [[verkennt]] is
this, that the procedure of experience relies on the thinking determination
of intuitions and that thought by itself and apart from this relationship to
intuition is unable to achieve anything at all.
The move from intuition (mind) to predication (logic) makes one question ur-
gent again: the applicability of the forms of thought. Why are the purely logically
derived categories applicable to reality?
The following set of quotations is meant to give a bit of an impression of the
futility of attempts to make sense of Hegel’s Logic without logic.
Quotation 69.29. (1) Henrich [1975], p. 213.

›Negation‹ ist unstreitig eine der bedeutendsten methodischen Grund-
operationen der Logik Hegels. Ohne Bezug auf den Sinn von Negation läßt
sich nicht verdeutlichen, was ihr Verfahren von allen anderen unterscheidet.
(2) Henrich [1975], p. 216.
[[S]]chon der erste Gedanke in Hegels Logik, der offenkundig in die Klasse der
negativen Ausdrücke gehört und der der »Bestimmtheit« noch vorangeht,
— die Kategorie des »Nichts«, die sich sprachlich aus der Substantivierung
des Adverbs herleitet, ist von Hegel so gefaßt worden, daß er sich vielmehr
als rudimentäre Form der von der Aussage und zugleich von irgendeinem
bestimmten negierten Prädikat abgelösten Form der Verneinung verstehen
läßt.
Comment. Henrich does not focus on the “Substantivierung” as a relevant “Grund-
operation” in Hegel’s Logic, but on one of its earliest products. He does so without
giving any consideration to the possibility of the linguistic derivation being in any
way relevant for the logical content in Hegel’s Logic.
(3) Henrich [1975], p. 225.
2. Hegels Logik benutzt verschiedene Typen von Negation. Sie hat die
Zahl der Bedeutungen von Negation im natürlichen Denken noch dadurch
vergrößert, daß sie die Aussageform als solche ontologisierte und daß sie
eine eigentümliche Form von selbstbezüglicher Negation ausbildete und in
ihr schließlich zwei Bedeutungen von Negation konfundierte. Mit diesem
Arsenal von Negationsbegriffen arbeitet die Logik, ohne ausreichend ana-
lysiert zu haben und vor allem ohne anders als rudimentär das Problem zu
erörtern, unter welchen Bedingungen und in welcher Folge diese Negations-
begriffe zu gebrauchen sind.
(4) Henrich [1975], pp. 225 f.
3. Der Gedanke der Selbstbezüglichkeit ist formal die Voraussetzung
dafür, Andersheit an sich, also selbstbezügliche doppelte Negation zu den-
ken. Beziehung nur auf sich, also reine Selbstbeziehung, definiert aber nach
Hegel das, was das Wort ›Sein‹ und was ganz im allgemeinen ›Unmittelbar-
keit‹ heißt. So scheint Selbstbeziehung der erste und ursprüngliche Gedanke
der Logik zu sein. Wirklich erklärt Hegel oft und ausnahmslos. daß die Ne-
gativität nur ein Moment am Absoluten sei. Das könnte bedeuten, daß es
der eigentlich grundlegende Gedanke der Logik Hegels zwar verlange, mit
dem Gedanken der Negativität zusammengebracht zu werden, daß er aber
zunächst unabhängig von ihr zu denken ist und daß dies auch die Voraus-
setzung dafür ist, daß der Gedanke einer Negativität überhaupt eingeführt
werden kann, die absolut und das heißt selbstbezüglich ist. Doch dem ist
entgegenzuhalten, daß der einfachste Gedanke ›Sein‹ eben deswegen als blo-
ße Beziehung auf sich gefaßt sein könnte, weil er nur in dieser Fassung sich
dazu eignet, mit Strukturen der selbstbezüglichen Negation identifiziert zu
werden.
Comment. “Gedanke der Selbstbezüglichkeit”, “Gedanke der Logik”, “Gedanke der
Negativität”, “Gedanke ›Sein‹” — Henrich kommt "uber den abstrakten Gedanken
nicht hinaus; no definition, no concept, no derivation, no realization, no content —
just empty shells. Given unrestricted abstraction, self-referentiality is a result, it
can be derived. A certain self-referentiality (“Selbstbezüglichkeit”) can already be
established in arithmetic (Gödel). This is not a question of how things are to be
thought, but what can be derived from a concept.
(5) Henrich [1975], p. 226.
4. Hat man erkannt, in wie hohem Maße in Hegels Logik Rücksicht ge-
nommen ist auf die Form der negativen Aussage, aber so, daß sie zugleich
als ein Gedanke vom Dasein aufgefaßt ist, und übersieht man die vielfäl-
tigen Konsequenzen dieses Grundzuges seiner Theorie, so wird man die
Selbstinterpretation, die Hegel ihr gegeben hat, nicht mehr ohne weiteres
übernehmen können. Was sich aus übersehbaren Gründen durch die Ver-
schiebung der Bedeutung natürlicher Operationen und Begriffe gewinnen
läßt, das hat seinen Ursprung offenbar in dem konstruktiven Willen eines
Theoretikers. Es kann nicht geradezu als Selbstdarstellung einer objektiven
Vernunft gelten.
Comment. Here we are. Without some understanding of modern logic it is impos-
sible to locate self-referentiality and the emergence of complex forms of negation
in logical operations and concepts, or at least investigate this possibility. Henrich
takes refuge in the strategy of attributing it to the will of a theoretician and a shift
of natural operations and concepts and, as a result, Hegel’s Logic cannot count as
a “Selbstdarstellung einer objektiven Vernunft” — the wisdom of a Hegel interpre-
tation without logic.
69e. The ‘non-metaphysical reading’ of Hegel.59 Recent years have seen

a curious twist in US-American Hegel interpretation, apparently instigated by a
somewhat missionary German Hegelian. The present section is meant to give a kind
of a cross-section through what may be called the “Hartmann school”.
I begin with a few quotations from Hartmann’s key paper entitled Hegel: A
Non-Metaphysical View. I wish to point out, however, that these quotations are by
59 As regards the label ‘non-metaphysical reading’, cf. Rockmore in quotation 69.32 (4) below.
no means meant to give a structured view of Hartmann’s paper, simply because I

don’t think it is possible to give a structured view of a dog’s breakfast.
Quotations 69.30. (1) Hartmann [1972], p. 105.
How could a presuppositionless beginning lead to anything; how could the
absence of determination lead to richness? Thus there must be operative a
contrary consideration, pointing from the ordered richness of granted con-
tent back to its antecedents. The linear progression cannot be deduction,
it can only be reconstruction; what it is heading for is granted. The ques-
tion merely is how to dispose categorial items in a sequence that could be
considered an “explanation” of categories, regressively, and a disclosure, or
“presentation,” of further categories progressively.
Comment. This is a good example of what I subsume under μαλακία ,60 or just
shabbiness. A rhetorical question of the kind “how could . . . ” is being used to intro-
duce a “thus there must be . . . ”-conclusion. With regard to the question “how could
the absence of determination lead to richness?”, compare quotation 71.7 (1) in these
materials, where Frege asks the question “how do the empty forms of logic come
to disgorge so rich a content?”.
(2) Hartmann [1972], pp. 108 f.
A category is the claim that being matches what thought thinks of it. We
could not account for being in terms other than those of thought. Thus
thought, to set up its own genealogy, or the justification of its match with
reality, has to regard its antecedent determinations as stances of grasped
being, or of being grasped in various degrees of coincidence with thought.
There are categories for comparatively incomprehensible things, such as
being; for comparatively comprehensible ones, such as essence; and for al-
together comprehensible ones, such as thought itself. Thus the categorial
claim is that being, to the extent that it is categorized, enters into thought
and makes a difference to thought pro tanto and yet it is thought that has to
provide the means for establishing such ingredience of being into thought.
What is a “match” of being and thought, has to be considered a coincidence,
an identity of being and thought.
The procedure to establish the ingredience of being into thought, or
to establish the rationality of being in a series of categorial determina-
tions, is accordingly to consider the otherness of being with respect to
thought as a negation, and the difference such otherness makes to thought
as “determinate negation.” The negative and the overcoming of the neg-
ative are constitutive of the movement of understanding. In his effort to
point to antecedents of determinacy, Hegel introduces the negative at the
very beginning of the Logic as the account thought gives of being prior to
any determination. All incumbent determinations are developed from this
conflict of being and negation which, in the zero case, in the absence of
any determination, is the indifference of the two. Hegel, in developing de-
terminations from such a conflict of being and negation, is not guilty of
60 Cf. the end of the preramble to part F on p. 1384 of the groundworks.
mistaking negation for otherness or irreducible novelty, as some scholars of

neo-Kantian persuasion have charged. Rather, we should say that, in or-
der to rationalize otherness, Hegel has to transform it into the negation of
what is already appropriated by thought. This replacement of otherness or
novelty by negation in the service of comprehension is what is called the
dialectic. In a dialectical framework, being appears as the complement to
thought, exhaustible in moves of negation and in determinate, or double,
negation. Clearly, this is possible only if the explanatory sequence is not
deduction, but only reconstruction, granting categorical content.
(3) Hartmann [1972], p. 110.
Over against the metaphysical reading, Hegel’s philosophy appears to us as
categorial theory, i.e., as non-metaphysical philosophy, or as a philosophy
devoid of existence claims and innocent of a reductionism opting for certain
existences to the detriment of others. The only claim is that the categories
granted for reconstruction be not empty or without instantiation.
(4) Hartmann [1972], pp. 112.
Categories are not the sum-total of axioms necessary for disciplines such
as geometry or mathematics, and mutatis mutandis for other disciplines,
to get on their feet. Categorial theory answers only the peculiar questions
a philosopher may have as to what it is that a certain discipline is about.
Categorial questions are luxury questions.
Comment. In less polite words, Hartmann’s categorial theory opts for philosophical
wank. I still would appreciate to have a less metaphorical rendering of what it means
for geometry and mathematics “to get on their feet”. Whatever it may mean, I have a
distinct feeling that it is a blessing for geometry and mathematics that Hartmann’s
categories are not necessary for them to get on their feet.
(5) Hartmann [1972], p. 115.

According to [[Hegel’s]] approach, the mind’s reference to being can be dis-
cussed, this side of being, only in thought. We seem to remain entrenched
on the side of the mind and therefore isolated from any referent of men-
tal acts. However, for Hegel, we can accommodate reference in a notion of
thought such that reference to being is already a constitutive feature of its
being thought. Restriction to the conceptual or categorial level will work,
but only if concepts are taken dialectically. This position, however, is not
available to language-oriented philosophy; this philosophy cannot theorize
about the relationship of language and the world on pain of paradox (of
which paradox Wittgenstein’s map theory in the Tractatus is an early state-
ment). The dialectical approach will always be superior where questions of
comprehensibility, i.e., transcendental or speculative questions, are asked.
Answers to such questions can, as in the case of the problem of reference,
be provided to the extent that the question is couched in categorial, and
this means, for Hegel, dialectical and systemic terms.
(6) Hartmann [1972], p. 124.

In ontology in a narrower sense, in that of the Logic, Hegel’s thought can
be assessed as strict and indispensable; in fact, on the present interpreta-
tion, Hegel’s claim appears, contrary to a metaphysical interpretation of
his philosophy, as a very modest one. His achievement is seen to lie in a
hermeneutic of categories.
Quotations 69.31. (1) Pinkard [1979], p. 211.

[[Hegel’s]] aim is to show that many apparent contradictions between basic
categories or between competing categorial systems are only that: apparent
and not real contradictions. The apparent contradiction can be avoided if
one correctly expands and orders one’s categorial scheme.
Comment. Contrast Findlay in quotation 69.14 (1).
(2) Pinkard [1979], p. 213.
[[Hegel’s]] proposal was that the indispensability of the category could be
shown by demonstrating it to be a logical condition for the determinateness
of some other category (this is apparently what is meant by Hegel’s talk
of the ‘immanent development’ of thought). A category would then be a
general determination which can be established by more or less a priori
and not empirical considerations.
(3) Pinkard [1979], p. 224.
The notion of the ‘absolute idea,’ despite its highly metaphysical sound-
ing title, may be taken to be simply the ‘meta-logic,’ so to speak, of the
theory. [[. . .]] The absolute idea, that is, is not a specific concept but the
whole logic of the concepts of being, essence and conceptuality.
(4) Pinkard [1979], p. 226.
This way of reading the absolute idea as simply the method of the WdL
may not, of course, appear satisfactory to more traditional interpreters of
Hegel, for it requires one more or less to excise certain passages as being
best taken as metaphorical. A good case can of course be made for claiming
that Hegel actually took the absolute idea not to be merely the method of
the WdL but to be also some kind of metaphysical entity. That it can be
so taken or that Hegel perhaps wanted it to be so taken is not in dispute
here. What is being claimed is only that it need not be so taken — despite
what Hegel himself might have thought.
(5) Pinkard [1988], pp. 5 f.
[[B]]riefly put, Hegelian dialectic is the attempt to show that some appar-
ently incompatible basic set of categories (or the basic categories of seem-
ingly incompatible philosophies) can actually be shown to be compatible
when they are put in the context of a larger context of categories. Hegel’s
philosophy has the misleading appearance of always occurring in “triplets”
since he always begins with some basic category, shows how it conflicts
with an equally basic category, and then offers a speculative explanation
using some new category or set of categories to show how the conflict is
only apparent. Hegelian dialectic is no mysterious form of logic that tran-

scends or is an alternative to ordinary logic. It is a strategy of explanation
for a philosophical program that attempts to reconcile most of the major
dualisms of the history of philosophy.
I shall also argue that Hegel himself did not completely understand the
implications of the style of philosophy that he was advocating. This can be
understood in terms of the Kantian roots of Hegel’s theory. Of particular
importance is the Kantian idea of a “science of reason.” I will argue that
it was because Hegel took himself to be engaged in something like the
Kantian “science of reason” that he was mistakenly led to see his dialectic
as providing not only explanations of the possibility of categories but also
derivations of the necessity of that set of categories.
Comment. The first paragraph of this quotation comes close to Popper in quotation
69.10 (2) above.
(6) Pinkard [1990b], p. 835.
Hegel thought that he could show that if one began, as Parmenides did, with
the notion of Being itself, one could show that one could not have consistent
thought unless one also supplemented that simple idea with a number of
other categories. In that way, the set of categories found in the Logic can
claim to be more than just one among many alternative sets. By beginning
with the minimal presuppositions of a Parmenidean concept of Being and
showing how the problems with it generate further categories, Hegel thought
he could thereby also claim a kind of priority for his categories that had
eluded other philosophers.
Comment. This is terribly clumsy and vague on top of it; but there seems to be
a faint idea of something relevant happening in that curious thing called “being”.
The point of the present study is to take being in two of its meanings, predication
and inclusion, together with the possibility of turning a propositional form into a
substantive (or: noun; the phrase is Henrich’s: “substantivierte Aussageform”) and
then derive other logical and even more complex theoretical functions. The latter is
certainly too ambitious for Pinkard.
(7) Pinkard [1994], p. 348, note 20.
In the Logic, the subject is how “thought” explains itself, and the movement
there progresses in terms of how each new category of the Logic serves to
let thought explain and justify itself as an inferential system. At each stage,
because of contradictions and incoherences, thought is forced to move on
until at the end of the logical progression (in the “absolute idea”) it is
supposedly capable of giving an account of itself and its inferential method
that it managed to construct in order to reach that point at which it can
reflexively give such an account of itself. But in each case “thought” gives
an account of itself as grasping the object of thought exhaustively and
then “finds” that its account generates a skepticism against itself. (Thus a
Parmenidean account of thought grasping “pure being” generates on its own
terms the denial of the claim it originally makes, that “thought” grasps pure
being as different from “nothing.”) At the end of the dialectical movement

in the Logic, such dialectical “thought” is supposedly to have demonstrated
that it is both self-subsuming (that all the various moves within the system
are moments of itself) and self-explanatory (that it explains the structure of
the system of thought in terms internal to the logical system, not in terms
of its matching up with any kind of metaphysical reality).
(8) Pinkard [1989], p. 8 (quoted from Ameriks [1991], p. 391).
[[I]]n Hegel’s eyes what is important in the Kantian philosophy is not its
attempt to derive everything from the conditions of self-consciousness, but
its attempt to construct a self-subsuming, self-reflexive explanation of the
categories.
(9) Pinkard [1990b], p. 838.
[[W]]e can endorse the idea of speculative dialectic as a method while none-
theless foregoing Hegel’s necessitarian claims for it. We are then left with
the task of constructing categorial claims simply in terms of how well they
do their explanatory job.
Comment. This, I suppose, is how far one can go, if one has no good grasp of
the tools of modern logic. Still, I would have liked to see the job (“of constructing
categorial claims”) done, at least, some beginnings. Mind you, there is always Hart-
mann’s endorsement of luxury61 which doesn’t encourage hopes for what will count
as “explanatory”.
(10) Pinkard [1989], p. 10.
Hegel’s notion of speculation may be understood as an alternative to
deduction in the logical sense. In deduction, nothing new appears; the con-
clusion is already contained in the premises. There is nothing surprising
here. After all, deduction and induction concern those rules of inference
that are truth-preserving (deduction always preserving it, induction usually
preserving it). In speculation, however, something new emerges, something
not contained in the premises. One has a novelty, something not strictly
derivable from that with which one started — let us call this novelty a
posit. Hegel’s system is speculative in this sense.
Comment. This is probably how far you can get if you are caught in the schemes
of school knowledge. Contrast Findlay’s formulation in quotation 69.25 (2) above.
(11) Pinkard [1991], p. 296.
It might be well to just avoid the term, ‘logic’, when speaking of Hegel’s
theory. After all, by no means is it a theory of the rules of truth-preserving
inference. It is a ‘logic’ however in the weaker sense in which we speak of
a ‘logic’ of ordinary language, or the ‘logic’ of investigation and so on: that
is it is a theory of the basic substantive principles and categories we use
in thinking about the world. But it is also supposed to be more than that:
Hegel presents it as a theory about which categories and principles are the
true categories and principles for thinking about the world.
61 Cf. quotation 69.30 (4) above.
Quotations 69.32. (1) Brinkmann [1994], p. 57.

Two things have obscured an understanding of Hegelian philosophy more
than anything else. One is the claim that Hegel’s dialectic constitutes a
violation of the laws of contradiction and excluded middle; the other is the
verdict that Hegel is fundamentally a metaphysician. Klaus Hartmann has
argued succinctly and convincingly that the dialectic is to be viewed as
a procedure for the systematic construal and concatenation of categorial
concepts, for which the principle of avoiding contradiction is absolutely
essential[[.]]
Comment. The problem with statements of this kind is their global notions, in this
specific example the notions of contradictions and metaphysics. In both cases it is
a lack of distinction which renders the remark virtually useless. The first thing to
realize is that the realm of contradictions, like that of the infinite, is not without dis-
tinctions. There are certain contradictions which have to be avoided, others which
don’t. Likewise, there is a certain kind of metaphysics which Hegel is not pursuing,
another form, essentially in the meaning that Kant gave to it, he is. If Hegel’s dialec-
tic involves “Substantivierung von Aussageformen”, and if this is to be understood as
unrestricted abstraction, then Hegel’s dialectic violates classical logic, more to the
point, certain consequences of the so-called ‘principle of (excluded) contradiction’,
e.g., of the form ¬(A ↔ ¬A).
(2) Brinkmann [1994], p. 74, endnote 1.
By a metaphysical reading I understand a view which interprets Hegel’s
categorial claims concerning the logical structure of being or types of being
as claims concerning the existence of entities. A typical Hegelian retort to a
metaphysical interpretation in this sense of metaphysical would be to point
out that the bare existence of something is a mere contingent fact which
as such is of no special philosophical interest. It is only when it comes to
determine what something is in reality, i.e., qua object of thinking, that an
important question is being raised.
(3) Brinkmann [1994], p. 72.
The price to pay for preserving the principle of contradiction consists in the
fact that the new concept is not and cannot be deductively inferred from
its predecessors (because nothing can be deduced from a contradiction). It
is a categorial novelty. Hence Hegel’s insistence that philosophy grants a
familiarity with concepts and that the dialectic is a reconstruction, not a
deduction of categories.
Comments. (α) Notice the claim in brackets: “nothing can be deduced from a con-
tradiction”. Champions of some undergraduate course in propositional calculus will
now emphatically claim: quite to the contrary, everything can be deduced from a
contradiction; witness Popper in quotations 96.16. And those who go on to the next
part, more specifically section 73c, comment to quotation 73.12 (1), will learn that
Cantor used a contradictory (inconsistent) set to prove the well-ordering theorem
(a proof which was considered inconclusive by Zermelo). A glimpse of what I want
to do with contradictions can be found in chapter XXXV in the groundworks;

but this already requires some skill in logic.
(β) The expression “categorial novelty” is credited to Klaus Hartmann. An end-
note to the last sentence refers the reader to Miller [1969], p. 39.62 This reference,
however, only supports the first part of the claim, not the last denying that dialec-
tics is “a deduction of the categories”. Rather what Hegel goes on to say is that
with regard to strictly immanent plasticity “the Science of Logic must surpass even
mathematics”.
(4) Rockmore [1994], p. 45.
Hartmann offers what he calls a non-metaphysical reading of Hegel’s
thought. He holds that we can provide a categorial and systematic inter-
pretation of Hegel’s position which reveals the latter’s achievements as a
hermeneutic of categories. And he suggests the interest of exploring ques-
tions of foundational philosophy in general in the light of a categorial and
systematic interpretation.
I close this chapter with a critical voice which I take to indicate that even
a modest attempt like that of the Hartmann school to bring thought and being,
perhaps not yet into a Parmenidean unity, but at least a bit closer, fails vis à vis
the stubborn prejudice of the duality of thought and the thing.
Quotation 69.33. Pippin [1989], p. 187.

[[R]]eading the Logic as thoughts reflective determination of itself, where
that self-determination involves progressively more self-consciousness about
what it can think “on its own” and about its own self-determining activity,
seems not to take into account the ontological dimension of the Logic itself
(as opposed to the ontological implications of the Logic of thought). On the
face of it, there are several places where Hegel, in discussing the limitations
of a Notion or its development, slips frequently from a “logical” to a material
mode, going far beyond a claim about thought or thinkability, and making
a direct claim about the necessary nature of things, direct in the sense that
no reference is made to a “deduced” relation between thought and thing.
Comment. The point in view of my enterprise is, of course, that the phrase ‘self-
consciousness about what it can think “on its own” ’ has to be filled with content. As
it here stands, the opposition between “logical” and “material” mode is unresolved.
62 See quotation 66.11 (5) in these materials; translation of Hegel [1812], p. 31.
Part D
THE PHILOSOPHY OF
MATHEMATICS AND LOGIC
One of the endearing things
about mathematicians
is the extent to which they will go
to avoid doing any real work.
Matthew Pordage1
1 From: H. Eves, Return to Mathematical Circles, Boston: Prindle, Weber and Schmidt, 1988.
PART D
THE PHILOSOPHY OF MATHEMATICS AND LOGIC
Were it only for the emergence of antinomies within metaphysics, the situation
would appear pretty hopeless for dialectic and speculative philosophy and never
would I have started writing the present study.2 The concepts of philosophy are
just too blunt to let hope for a fruitful treatment of this delicate affair. This is why
I regard the emergence of antinomies in the foundations of so exact a science as
mathematics a real blessing for the project of a logical foundation of dialectic. With
their help I can hope to make some progress in my concern with the antinomies of
metaphysics.
Paradoxes have been part of the philosophical folklore since the ancient Greeks.
They may be impressive or entertaining in themselves; they are, in my view, but a
symptom.
For centuries, or millennia even, they have been dismissed with a wave of schol-
arly arrogance, until they emerged in the foundations of mathematics. Even then
they tended to be attributed to false logic, but at least they were taken seriously.
The present part is devoted to a historical sketch of the various directions
research took in this field from the emergence of antinomies to the establishment
of the fixed-point property. Its concern is not so much the precise form of the ideas
involved as their genesis and mutual connection. It is the fixed-point property, which
is of truly gnoseological relevance. The antinomies are just one side of it. And it will
be the fixed-point property which, eventually, will make my idea of dialectic work.
The foundations of mathematics provide a paradigmatic field of investigation
for transcendental philosophy. Indeed, it was Kant’s claim that arithmetics is syn-
thetic a priori which Frege challenged. Unfortunately, however, philosophers in the
transcendental tradition seem to be fast running out of their depth when it comes
to basic ideas and concepts in modern mathematics. The present part will not be
able to change that fundamentally; it only aims at providing a sort of map for
someone who wants to know where I understand my theory of dialectic to have its
touchstone: it is in the foundation of arithmetic that my theory of dialectical logic
will be put to the test — Hic Rhodus, hic saltus.
Sporadically, a certain familiarity with the tools provided in book one will
now be assumed.
2 And certainly would have saved me, and perhaps also you, at lot of time and effort.
981
CHAPTER XVIII
THE EMERGENCE OF ANTINOMIES IN THE

FOUNDATIONS OF MATHEMATICS
In this chapter I shall present a rather compressed sketch of the history of ideas
that led to the emergence of antinomies within the foundations of mathematics. As
in the case of transcendental idealism I do not try to give a representative survey
but a selected choice of ideas which play a role for my approach to dialectical logic.
Special emphasis is laid on some of Frege’s ideas.
The highlighting of the following four points in the foundations of mathematics
seems somewhat classic:1
(i) axiomatics,
(ii) the elimination of the concept of indivisibles,
(iii) the arithmetization of analysis, and
(iv) the set theoretical foundation of the concept of number.
Of these only the first and certain aspects of the fourth are relevant for me. What
I am interested in is located in an area delineated by axiomatics, in particular
Euclidean- vs. non-Euclidean geometry, aspects of set theory, such as the axiom of
choice, and — above all — Frege’s logical foundation of arithmetic and the emer-
gence of paradoxes. The first paragraph of the present chapter is dedicated to ax-
iomatics, the second to topics in set theory and the logical foundation of arithmetic
and the third paragraph focuses on the paradoxes. Again, I shall not try to give
an historically just account of the foundations of mathematics, but pick the points
that are relevant for my project.
§ 70. Ways of axiomatics
So far I have said that mathematics is of interest for me because of the emergence
of antinomies. But mathematics has always been interesting for philosophers, long
before the emergence of antinomies. In chapter XIV, I quoted Berkeley and Hume
regarding mathematics and in chapter XV, I quoted Kant. As in the case of the
philosophical materials, I start with the ancient Greeks who recognized the unique
gnoseological position of mathematics.
70a. Eukleides’ elements, the idea of axiomatics, and the unique case
of mathematical knowledge. The security of mathematical knowledge challenges
philosophers. So does the apparently abstract nature of mathematical objects.
1 Cf. Hilbert and Bernays [1934/68], p. 1, for a similar list of points.
982
§ 70. WAYS OF AXIOMATICS 983
Quotations 70.1. (1) Lee [1955], pp. 313 f; translating Plato’s Republic, 510c. The
dialogue is between Sokrates (first speaker) and Glaukon.2
‘[[. . .]] I think you know that students of geometry and calculation and the
like begin by assuming there are odd and even numbers, geometrical figures
and the three forms of angle, and other kindred items in their respective
subjects; these they regard as known, having put them forward as basic as-
sumptions which it is quite unnecessary to explain to themselves or anyone
else on the grounds that they are obvious to everyone. Starting from them,
they proceed through a series of consistent steps to the conclusion which
they set out to find.’
‘Yes, I certainly know that.’
‘You know too that they make use of and argue about visible figures,
though they are not really thinking about them, but about the originals
which they resemble; it is not about the square or diagonal which they
have drawn that they are arguing, but about the square itself or diagonal
itself, or whatever the figure may be. The actual figures they draw or model,
which themselves cast their shadows and reflections in water — these they
treat as images only the real objects of their investigation being invisible
except to the eye of reason.’
‘That is quite true.’
‘This type of thing I called intelligible, but said that the mind was
forced to use assumptions in investigating it, and did not proceed to a first
principle, being unable to depart from and rise above its assumptions; but
it used as illustrations the very things [[. . .]] which in turn have their images
and shadows on the lower level [[. . .]], in comparison with which they are
themselves respected and valued for their clarity.’
Comment. Notice: “the mind was forced to use assumptions”.
(2) Lee [1955], p. 332; translating Plato’s Republic, 525. The dialogue is still between
Sokrates (narrator) and Glaukon.
‘You know,’ I said, ‘now that we have mentioned the study of arith-
metic, it occurs to me what subtle and widely useful instrument it is for our
purpose, if one studies it for the sake of knowledge and not for commercial
ends.’
‘How is that?’ he asked.
‘As we have just said, it draws the mind upwards and forces it to argue
about numbers in themselves, and will not be put off by attempts to confine
the argument to collections of visible or tangible objects. You must know
how the experts in the subject, if one tries to argue that the unit itself is
divisible, won’t have it, but make you look absurd by multiplying it if you
try to divide it, to make sure that their unit is never shown to contain a
multiplicity of parts.’
‘Yes, that’s quite true.’
2 This quotation was instigated by Becker [1964], pp. 95 f.

984 XVIII. THE EMERGENCE OF ANTINOMIES
‘What do you think they would say, Glaucon, if one were to say to them,
“This is very extraordinary — what are these numbers you are arguing
about, whose constituent units are, so you claim, all precisely equal to each
other, and at the same time not divisible into parts?” What do you think
their answer would be to that?’
‘I suppose they would say that the numbers they mean can be appre-
hended by reason, but that there is no other way of handling them.’
‘You see therefore,’ I pointed out to him, ‘that this study looks as if it
were really necessary to us, since it so obviously compels the mind to use
pure thought in order to get at the truth.’
Quotation 70.2. Tredennick [1938], p. 321; translating Aristoteles Prior Analytics

41a 23–30.
Everyone who carries out a proof per impossible proves the false conclu-
sion by syllogism and demonstrates the point at issue ex hypothesi when
an impossible conclusion follows from the assumption of the contradictory
proposition. E.g., one proves that the diagonal of a square is incommensu-
rable with the sides by showing that if it is assumed to be commensurable,
odd become equal to even numbers. Thus he argues to the conclusion that
odd becomes equal to even, and proves ex hypothesi that the diagonal is
incommensurable, since the contradictory proposition produces a false re-
sult.
Comment. There is a footnote: “For the proof see Euclid, Elements, x. app. 27
(Heiberg and Menge).”
Remark 70.3. Eukleides seems to be commonly regarded as the “father” of ax-

iomatics. According to Struik [1967], p. 40, however, Elements (τα στοιχεῖα ) was
the title of all Greek axiomatic treatises.
Quotation 70.4. (1) T. L. Heath [1908/25], p. 1.

As in the case of the other great mathematicians of Greece, so in Eu-
clid’s case, we have only the most meagre particulars of the life and per-
sonality of the man.
Most of what we have is contained in the passage of Proclus’ summary
relating to him, which is as follows:
“Not much younger than these (sc. Hermotimus of Colophon, and
Philippus of Medma) is Euclid, who put together the Elements, collect-
ing many of Eudoxos’ theorems, perfecting many of Theaetetus’, and also
bringing to irrefragable demonstration the things which were only some-
what loosely proved by his predecessors. This man lived in the time of
the first Ptolemy. For Archimedes, who came immediately after the first
(Ptolemy), makes mention of Euclid: and, further they say that Ptolemy
once asked him if there was in geometry any shorter way than that of the
elements, and he answered that there was no royal road to geometry. [[. . .]]”
Comment. This seems to be the passage where they have all got their information
from.3 The source is given as ‘Proclus, ed. Friedlein, p. 68, 6–20’.
(2) T. L. Heath [1908/25], p. 2.
We may infer then from Proclus that Euclid was intermediate be-
tween the first pupils of Plato and Archimedes. Now Plato died in 347/6,
Archimedes lived 287–212, Eratosthenes c. 284–204 b.c. Thus Euclid must
have flourished c. 300 b.c., which date agrees well with the fact that Ptolemy
reigned from 306 to 283 b.c.
(3) T. L. Heath [1908/25], p. 2.
It is most probable that Euclid received his mathematical training in
Athens from the pupils of Plato; for most of the geometers who could have
taught him were of that school, and it was in Athens that the older writers of
elements, and the other mathematicians on whose works Euclid’ Elements
depend, had lived and taught. [[. . .]]
One thing is however certain, namely that Euclid taught, and founded
a school Alexandria.
(4) T. L. Heath [1908/25], p. 3.
In the middle ages most translators and editors spoke of Euclid as
Euclid of Megara. This description arose out of a confusion between our
Euclid and the philosopher Euclid of Megara who lived about 400 b.c.
Quotations 70.5. (1) T. L. Heath [1908/25], p. 153; translating Eukleides.
1. A point is that which has no part.
2. A line is breadthless length.
3. The extremities of a line are points.
4. A straight line is a line which lies evenly with the points on itself.
5. A surface is that which has length and breadth only.
6. The extremities of a surface are lines.
7. A plane surface is a surface which lies evenly with the straight lines
on itself.
Comment. These are the first 7 out of 23 definitions. Compare these definitions
with the modern axiomatic approach by Hilbert as presented in quotations 70.22
below.
(2) T. L. Heath [1908/25], pp. 154 f; translating Eukleides.
Let the following be postulated:
1. To draw a straight line from any point to any point.
2. To produce a finite straight line continuously in a straight line.
3. To describe a circle with any centre and distance.
4. That all right angles are equal to one another.
5. That, if a straight line falling on two straight lines make the interior
angles on the same side less than two right angles, the two straight
lines, if produced indefinitely, meet on that side on which are the
angles less than the two right angles.
3 For instance Struik [1967], p. 50.
One
√ rumour has it that Eukleides’ Elements were a reaction to the discovery
that 2 is not rational. The following quotation suggests a different background.
Quotations 70.6. (1) A. E. Taylor [1926], p. 290.

To save mathematical science in the face of Zeno’s arguments it became
necessary in the fourth century to reconstruct the whole system, and the
reconstruction is preserved for us in the Elements of Euclid. The men by
whom the actual reconstruction was done, Eudoxos, Theaetetus, and their
companions, so far as they are known to us, were all associates of Plato
himself in the Academy, and it is quite certain that this revision of the
accepted first principles of mathematics was one of the chief problems to
which the school devoted itself.
Comment. I do not feel competent to judge such a claim, but it strikes me as quite
bold. Compare the following quotation.
(2) E. T. Bell [1940/45], p. 73.
As Plato (430–349 b.c.) preceded Euclid (365?–275? b.c.) by about
twenty years, it is possible but improbable that the geometer was influenced
by the philosopher. It may be regrettable, but it appears to be true, that
creative mathematicians pay little attention to philosophers whose mathe-
matical education has not gone much beyond the elementary vocabulary.
Comment. This raises, above all, three questions; firstly, who counts as a “creative
mathematician”? (Eukleides? or did he only compile?) Secondly, who counts as a
philosopher “whose mathematical education has not gone much beyond the elemen-
tary education”? Thirdly, how is one to count the different answers to the foregoing
two questions? Cantor, it seems from quotation 68.27 (1), would have been happy
to place Kant in the category of “philosophers whose mathematical education has
not gone much beyond the elementary vocabulary”; but there is Zermelo’s defence
of Kant in quotation 68.27 (2). Frege acknowledged Kant’s distinction of analytic
and synthetic judgments as quotation 71.26 (1) shows. Hilbert related in his idea
of proof theory to Kant,4 and Bernays declared himself to having been influenced
by the Neo-Kantian tradition of Leonard Nelson and being a member of the philo-
sophical circle around Ferdinand Gonseth.5 Compare also Hessenberg in quotations
73.18. All that points back to Kant’s philosophical ideas about mathematics and
metaphysics. These misgivings notwithstanding, I do think there is something to
Bell’s point — in principal; it is just badly phrased and, to make things worse, as
far as Eukleides is concerned it is purely based on speculation.6
Quotations 70.7. (1) Eukleides, quoted after Frege [1884], p. 39, footnote *.
Μονάς ἐστι, καθ΄ ἣν ἕκαστον τῶν ὄντων ἓν λέγεται. ᾿Αριθμὸς δὲ τὸ ἐκ μονάδ-
ων συνκείμενον πλῆθως.
4 Cf quotation 76.6 (3), for instance.

5 See the back cover of Bernays [1976].
6 Contrast the cautious formulation of Heath in quotation 70.5 (3) above.
(2) Austin [1950], p. 39e , footnote 1; translating Eukleides as quoted above.

A unit is that by virtue of which each of the things that exist is called one.
A number is a multitude composed of units.
70b. Leibniz’ characteristica universalis. Leibniz’ idea of assigning every

expression a number is of great historical interest, albeit technically irrelevant for
what I want to do.7
I begin with a set of quotations from Leibniz [1677], Dialogus de connexione
inter res et verba, which is said to engage with Hobbes’ view.8
Quotations 70.8. (1) Ariew and Garber [1989], p. 269; translating Leibniz [1677],
p. 26.
A. [[. . . Y]]ou think that truth and falsity are in things, and not in
thoughts?
B. Yes, indeed.
A. Then a thing is false?
B. Not a thing, I think, but a thought or proposition about a thing.
A. And falsity pertains to thoughts, not to things.
B. I am forced to admit it.
A. Therefore truth as well?
B. So it seems.
Comment. B has still doubts about the correctness of the last consequence which I
have cut off.
(2) Leibniz [1677], p. 28.
A. Sed quoniam causam esse necesse est, cur cogitatio aliqua vera aut
falsa futura sit, hanc ubi quaeso quaeremus?
B. In natura rerum puto.
A. quidsi ea oriatur ex natura tua.
B. Certe non ex sola. Nam necesse est meam et rerum de quibus cogito
naturam talem esse, ut quando methodo legitima procedo, propositionem,
de qua agitur, concludam seu veram reperiam.
(3) Ariew and Garber [1989], p. 270; translating Leibniz [1677], p. 28.
A. But since there must be some reason [causa] why a given thought is
going to be true or false, where, I ask, shall we look for it?
B. In the nature of things, I think.
A. What if it arises from your own nature.
B. Certainly not from there alone. For it is necessary that my nature
and the nature of the things I’m thinking about are such that, when I
proceed using a legitimate method, I infer a proposition of the sort that is
at issue, that is I find a true proposition.9
7 Cf. Hegel’s comment in quotation 95.1 (2) in these materials.
8 Ariew and Garber [1989], p. 268 f.
9 My reading of the last sentence: For it is necessary that my nature and that of the things
about which I think are of such a kind, that when I proceed according to right method, I recognize
as conclusive or true, the proposition to which I come.
B. Can anyone depart so far from good sense [bona mens] as to persuade
himself that truth is arbitrary and depends on names, when it is agreed that
the Greeks, the Latins, and the Germans all have the same geometry?
Quotations 70.9. (1) Leibniz [1677], pp. 32 & 34.

Nam etsi characteres sint arbitrarii, eorum tamen usus et connexio habet
quiddam quod non est arbitrarium, scilicet proportionem quandam inter
characteres et res, et diversorum characterum easdem res exprimentium
relationes inter se. Et haec proportio sive relatio est fundamentum veritatis.
Efficit enim, ut sive hos sive alios characteres adhibeamus, idem semper
sive aequivalens seu proportione respondens prodeat. Tametsi forte aliquos
semper characteres adhiberi necesse sit ad cogitandum.
(2) Ariew and Garber [1989], p. 271; translating Leibniz [1677], pp. 32 & 34.
For though the characters are arbitrary, their use and connection have some-
thing that is not arbitrary, namely, a certain [[proportion]] between charac-
ters and things, and certain relations among different characters expressing
the same thing. And this [[proportion]] or this relation is the ground of truth.
For it brings it about that whether we use these characters or others, the
same thing always results, or at least something equivalent, that is, some-
thing corresponding in proportion always results. This is true even if, as it
happens, it is always necessary to use some characters for thinking.10
Comment. This is a central passage in view of my endeavour. It should be read
in the light of Gödel’s incompleteness results, more specifically in the light of the
(indirect) fixed-point property. What it means then is that there is something non-
arbitrary in any symbolic representation of thought (given a certain substitution
function is available, etc.), but it does not quite seem to be what Leibniz had in
mind.11
(3) Leibniz [1677], p. 36.
Quanquam ergo veritates necessario supponant aliquos characteres, imo ali-
quando de ipsis characteribus loquantur [[. . .]] non tamen in eo quod in iis
est arbitrarium, sed in eo quod est perpetuum, relatione nempe ad res con-
sistunt semperque verum est sine ullo arbitrio nostro, quod positis talibus
characteribus talis ratiocinatio sit proventura, et positis aliis, quorum nota
ad priores relatio sit alia quidem, sed etiam relationem servans ad priores
ex characterum relatione resultantem, quae substituendo vel comparando
apparet.
Comment. This is an extremely important point! Roughly, my approach to dialectic
is based on the idea that it is not always possible to avoid speaking — at least
implicitly — about the symbols while using them, and that this is the source of a
10 The double square brackets replace “correspondence [proportio]” in the first case, and just
“correspondence” in the second.

11 Cf. Odifreddi [1989], p. 87; quotation 94.4 (4) in these materials.
confusion (ambiguity) which spoils the picture of continuity (reliability) in relation

which Leibniz evokes at the end of the above quotation.
Therefore, although truths necessarily presuppose some characters, indeed,
sometimes they deal with the characters themselves [[. . .]] truths don’t con-
sist in what is arbitrary in the characters, but in what is invariant [per-
petuus] in them, namely, in the relation they have to things. And it is al-
ways true, independent of any decision of ours, that if, given such and such
characters, such and such a reasoning will succeed, then it will likewise
succeed given others whose relation to the former ones is known; however,
the reasoning preserves a relation in the former ones that results from the
relation among the characters. This is something obvious from substituting
and comparing.
(5) Leibniz [1678] (quid sit idea?), pp. 62 & 64.
Exprimere aliquam rem dicitur illud, in quo habentur habitudines, quae
habitudinibus rei experimendae respondent. Sed eae expressiones variae
sunt; exempli causa, modulus Machinae exprimit machinam ipsam, sceno-
graphica rei in plano delineatio exprimit solidum, oratio exprimit cogita-
tiones et veritates, characteres exprimunt numeros, aequatio Algebraica ex-
primit circulum aliamve figuram: et quod expressionibus istis commune est,
ex sola contemplatione habitudinum exprimentis possumus venire in cogni-
tionem proprietarium respondentium rei exprimendae. Unde patet non esse
necessarium, ut id quod exprimit simile sit rei expressae, modo habitudinum
quaedem analogia servetur.
Quotations 70.10. (1) Leibniz [1685–7], p. 156.

Defin. 1. Eadem sunt quorum unum potest substitui alteri salva veri-
tate. Si sint A et B et A ingrediatur aliquam propositionem veram, et ibi
in aliquo loco ipsius A pro ipso substituendo B fiat nova propositio eaque
itidem vera, idque semper succedat in quacunque tali propositione, A et B
dicuntur esse Eadem; et contra si Eadem sint A et B, procedet substitutio
quam dixi. Eadem etiam vocantur coincidentia; aliquando tamen A quidem
et A vocantur idem, A vero et B si sint eadem vocantur coincidentia.
Defin. 2. Diversa sunt quae sunt non eadem, seu in quibus substitutio
aliquando non succedit.
Coroll. Unde etiam, quae non sunt diversa, sunt eadem.
Charact. 1. A ∞ B significat A et B esse eadem vel coincidentia.
Charact. 2. A non ∞ B, vel B non ∞ A significat A et B esse diversa.
Comment. Cf. Quine in quotation 90.20. I use ≡ for ∞. Cf. also Frege in quotation
71.14 (1) below.
(2) Parkinson [1966], p. 123; translating Leibniz [1685–7], p. 156.
Definition 1. Those terms are ‘the same’ of which one can be substituted
for the other without loss of truth. Thus, suppose that there are A and B;
that A is an ingredient of some true proposition, and that on substituting
B for A in some occurrence of A there a new proposition is formed, which

is also true. If this always holds good in the case of any such proposition,
then A and B are said to be ‘the same’; conversely, if A and B are the same,
the substitution which I have mentioned will hold good. The same terms
are also called ‘coincident’; sometimes, however, A and A are called ‘the
same’, whereas A and B, if they are the same, are called ‘coincident’.
Definition 2. Those terms are ‘different’ which are not the same, i.e. in
which a substitution sometimes does not hold good.
Corollary. Consequently, those terms which are not different are the same.
Symbol 1. ‘A = B’ means that A and B are the same, or coincident.
Symbol 2. ‘A 6= B’, or ‘B 6= A’, means that A and B are different.
Comment. Notice that “the same” is introduced as a definition; not a law or a
principle, as which it is commonly referred to in dictionaries such as Flew [1979] or
Audi [1995] and by many authors.

Charact. 3. A + B ∞ L significat A inesse ipsi L vel contineri.
Schol. Etsi A et B habeant aliquid commune, ita ut ambo simul sint
majora ipso L, nihilominus locum habebunt quae hoc loco diximus aut
dicemus.
(2) Parkinson p. 123; translating Leibniz [1685–7], p. 158.
Symbol 3. ‘A + B = L’ means that A is in L, or is contained in it.
Note. Even if A and B have something in common, so that both taken
together are greater than L, what we have said or shall say here will still
hold.
Comment. In the terminology of my tools this may be formulated as: M ⊆ N iff
M ∪ N1 = N, which strikes me as somewhat ambiguous; to be correct it would have
to run: M ⊆ N iff there is N1 such that M ∪ N1 = N, which is proposition 1.11 in
the tools.

Axioma 1. Si idem secum ipso sumatur, nihil constituitur novum, seu
A + A∞A.
Schol. Equidem in numeris 2 + 2 facit 4, seu bini nummi binis additi
faciunt quatuor nummos, sed tunc bini additi sunt alii a prioribus; si iidem
essent, nihil novi prodiret, et perinde esset ac si joco ex tribus ovis facere
vellemus sex, numerando primum 3 ova, deinde uno sublato residua 2, ac
deinde uno rursus sublato residuum 1.
(2) Parkinson [1966], p. 124; translating Leibniz [1685–7], p. 162.
Axiom i. If the same term is taken with itself, nothing new is constituted;
i.e. A + A = A.
Note. It is true that, in the case of numbers, 2 + 2 makes 4, or two coins
added to two coins make four coins; but then the two which are added are
other than the previous ones. If they were the same, nothing new would
emerge, and it would be as if we wished for a joke to make six eggs out
of three, by first counting three eggs, then removing one and counting the
remainder, two, and then removing one again and counting the remainder,
one.
Comment. This is a clear statement of contraction for propositions. Two notions
of “or” (and “and”) have to be distinguished; Leibniz’ claim holds for one of them,
but not the other. This is a central point in the present study. Cf. also Girard in
Quotations 70.13. (1) Leibniz [1685–7], pp. 170 & 172.

Theor. X.
Detractum et Residuum sunt incommunicantia.
Si L − A∞N, dico A et N nihil habere commune. Nam ex definitione
detracti et Residui omnia quae sunt in L manent in N praeter ea quae sunt
in A, quorum nihil manet in N.
Theor. XI.
In duobus communicantibus id cui inest quicquid utrique commune est,
et duo propria sunt tria incommunicantia inter se.
Sint A et B communicantia et A∞P + M, et B∞N + M, sic ut quidquid
est in A et B, sit in M, nihil vero ejus in P et N, dico P, M, N esse incom-
municantia. Nam tam P quam N sunt incommunicantia cum M, quia quod
inest in M, est in A et B simul, et nihil tale est in P aut N. Deinde P et
N sunt incommunicantia inter se, alioqui itidem quod ipsis commune est,
foret in A et B.
(2) Parkinson [1966], p. 128; translating Leibniz [1685–7], pp. 170 & 172.
theorem x
What has been subtracted and the remainder are uncommunicating.
If L — A = N, I assert that A and N have nothing in common. For
from the definition of ‘what has been subtracted’ and of ‘remainder’, all
the terms which are in L remain in N except those which are in A, of which
nothing remains in N.
theorem xi
In two communicating terms, that in which there is whatever is common
to each, and the two exclusive parts, are three terms which do not commu-
nicate with each other.
Let A and B be communicating, and A = P + M and B = N + M, in
such a way that whatever is in A and B is in M, but nothing of M is in
P and N; I assert that P, M and N are uncommunicating. For both P and
N are uncommunicating with M, since what is in M is at the same time
in A and B, and nothing of this kind is in P or N. Consequently P and N
are uncommunicating with each other, otherwise that which is common to
them would be in A and B.
Comment. In the set theoretical notation of my tools theorem X runs: M ∩ ∁M = ∅,
which is proposition 1.14x. In dialectical logic this depends on the kind of “and”
involved in the definition of the intersection: accumulating, or not. Cf. proposition
128.15i in the elements for a counter-example within dialectical logic and unre-
stricted abstraction.
Quotations 70.14. (1) Leibniz [1679], p. 70 (Couturat [1903], p. 49).

(1) Terminus est subjectum vel prædicatum propositionis categoricæ.
[[. . .]]
(3) Cuilibet Termino assignetur suus numerus characteristicus, qui ad-
hibeatur in calculando, ut terminus ipse adhibetur in ratiocinando. [[. . .]]
(4) Regula inveniendi numeros characteristicos aptos hæc unica est,
ut quando Termini dati conceptus componitur in casu recto ex conceptibus
duorum pluriumve aliorum terminorum, tunc numerus termini dati Charac-
teristicus producator ex terminorum termini dati conceptum componentium
numeris characteristicis invicem multiplicatis.
(2) Parkinson [1966], p. 17; translating Leibniz [1679], p. 70.
(1) A ‘term’ is the subject or predicate of a categorical proposition. [[. . .]]
[[. . .]]
(3) Let there be assigned to any term its symbolic number [numerus charac-
teristicus], to be used in calculation as the term itself is used in reasoning.
[[. . .]]
(4) The one rule of discovering suitable symbolic numbers is this: that when
the concept of a given term is composed directly of the concepts of two or
more other terms, then the symbolic number of the given term should be
produced by multiplying together the symbolic numbers of the terms which
compose the concept of the given term.
(3) Becker [1964], p. 359; quoting Leibniz.
Einen Charakter nenne ich ein anschauliches Merkzeichen, das Gedan-
ken darstellt. Die Kunst der „Charakteristik“ ist die Kunst, Charaktere so
zu bilden und anzuordnen, daß sie die Gedanken wiedergeben, d.h. daß sie
unter sich dieselbe Beziehung haben wie die Gedanken unter sich.
Comment. This is a crucial point for me: it enables me to avoid the mystique
surrounding the notion of reference. It strikes me as clearer than Frege’s way of
saying “expressing a thought”.
(4) Burks [1986], p. 2; translation of Leibniz in: Gerhardt [1875-90], vol. VII, p. 31.
All our reasoning is nothing but the joining and substituting of char-
acters, whether these characters be words or symbols or pictures.
Quotations 70.15. (1) Ariew and Garber [1989], pp. 6 f; translating Leibniz [1678-
79], p. 46 & 48.
[[O]]ne can devise a certain alphabet of human thoughts and [[. . .]] through
the combination of the letters of this alphabet and through the analysis of
words produced from them, all things can both be discovered and judged.
[[. . .]]
[[. . .]] I often wondered why, as far as the recorded history of mankind
extends, no mortal had approached such a project, for meditations of this
kind ought to be among the first to occur to those reasoning in proper

order [[. . .]] The real reason why people have missed the doorway [into this
discovery] is, I think, because principles are, for the most part, dry and
insufficiently agreeable to people, and so, barely tasted, they are dismissed.
(2) Ariew and Garber [1989], p. 8; translating Leibniz [1678-79], p. 52.
Once the characteristic numbers of most notions are determined, the
human race will have a new kind of tool, a tool that will increase the power
of the mind much more than optical lenses helped our eyes, a tool that will
be as far superior to microscopes or telescopes as reason is to vision.
(3) Ariew and Garber [1989], p. 9; translating Leibniz [1678-79], p. 54.
[[T]]wo people who argue look to me almost like two merchants who owe
money to one another from numerous transactions, but who never want to
reckon up the accounts, while meanwhile each in different ways exaggerates
what he himself is owed by the other and exaggerates the validity and size of
certain particular claims. Thus, the controversy will never end. We should
not be surprised that this happens in a large proportion of the controversies
where the matter is unclear, that is, where the dispute cannot be reduced
to numerical terms. But now our characteristic will reduce them all to
numerical terms, so that even reasons can be weighed, just as if we had a
special kind of balance.
Quotation 70.16. Parkinson [1966], p. 93; translating Leibniz [1690] (C 421)

(1) ‘A=B’ is the same as “ ‘A=B” is a true proposition’.
(2) ‘A6=B’ is the same as “ ‘A=B” is a false proposition’.
(3) A=AA; i.e. the multiplication of a letter by itself is here without effect.
(4) AB=BA, i.e. transposition makes no difference.
(5) ‘A=B’ means that one be can be substituted for the other, B for A or
A for B, i.e. that they are equivalent.
(6) ‘Not’ immediately repeated destroys itself.
(7) Therefore A=not-not-A.
Comment. I focus on the first seven of a list of all together 20 points because every
single one of these first seven has been an issue in modern logic and sometimes
subjected to restrictions: (1) truth predicate (Tarski), (2) falsity predicate, (3) con-
traction (dialectical logic, linear logic), (4) exchange, (5) extensionality, (6) and (7)
double negation (Intuitionism).
I close this section on Leibniz with a few mixed quotations.
Quotations 70.17. (1) Parkinson [1966], pp. 5 f; translating Leibniz [1666], p. 199.
[[W]]arning must be given that the whole of this art of complications is
directed to theorems, or, to propositions which are eternal truths, i.e. which
exist, not by the will of God, but by their own nature. But as for all singular
propositions which might be called historical (e.g. ‘Augustus was emperor
of Rome’), or as for observations (i.e. propositions such as ‘All European
adults have a knowledge of God’—propositions which are universal, but
whose truth has its basis in existence, not in essence, and which are true
as if by chance, i.e. but the will of God)—of these propositions there is no
demonstration, but only induction; except that sometimes an observation
can be demonstrated through an observation by the mediation of a theorem.
(2) Leibniz [1690/91], pp. 210 & 212.
[[L]]’art d’inventer est peu connu hors des Mathematiques, car les Topiques
ne servent ordinairement que de lieux memoriaux pour ranger passable-
ment nos pensées, ne contenant qu’un catalogue des Termes vagues et des
Maximes apparentes communement receues. J’avoue que leur usage est tres
grand dans la rhetorique et dans les questions qu’on traite populairement,
mais lorsqu’il s’agit de venir à la certitude, et de trouver des verités cachées
dans la theorie et par consequent des avantages nouveaux pour la prac-
tique, il faut bien d’autres artifices. Et une longue experience de reflexions
sur toute sorte de matieres accompagnée d’un succes considerable dans les
inventions et dans les decouvertes m’a fait connoistre qu’il y a des secrets
dans l’Art de penser, comme dans les autres Arts. Et c’est là l’objet de la
Science Generale que j’entreprend de traiter.
Comment. The idea of a “Science Generale”, scientia generalis, in the context of
l’art d’inventer, strikes me as very congenial to what I want to do. According to my
view, les autres artifices have been provided with modern logic and, in particular,
the arithmetization of metatheory which can take the place of Leibniz’ idea of a
characteristica universalis, with les secrets dans l’Art de penser being located in
the form of fixed point properties.
(3) Ariew and Garber [1989], p. 321; translating Leibniz (second letter to Samuel
Clarke, late November 1715).
The great foundation of mathematics is the principle of contradiction or
identity, that is that a proposition cannot be true and false at the same
time, and that therefore A is A and cannot be not A. This single principle
is sufficient to demonstrate every part of arithmetic and geometry, that is,
all mathematical principles.
Comment. In other words, Leibniz saw mathematics as analytic. On its own, how-
ever, the principle of contradiction is certainly not sufficient to prove all of arith-
metic. What is needed, in addition, is some provision to more, however, the direct
fixed point property of unrestricted abstraction effectively refutes this principle:
there is a fixed point for negation, i.e., neg(f ) = f for some f .
(4) Ariew and Garber [1989], pp. 337 f.12
47. I will here show how men come to form to themselves the notion of
12 This quotation was instigated by Jammer [1954/93], p. 117, where the source is given as A
Collection of Papers which passed between the late learned Mr. Leibnitz and Dr. Clarke London
1717, p. 195. The quotation here is, according to Ariew and Garber [1989], p. 320, a “slightly
updated version of the translation Clarke published”. Changes can be seen by comparison with
the quotation in Jammer: capitals of nouns have been turned into small letters and a number of
semicolons have been replaced by commas.
space. They consider that many things exist at once, and they observe in
them a certain order of coexistence, according to which the relation of one
thing to another is more or less simple. This order is their situation or
distance. When it happens that one of those coexistent things changes its
relation to a multitude of others which do not change their relations among
themselves, and that another thing, newly come, acquires the same relation
to the others as the former had, we then say it is come into the place of
the former; and this change we call a motion in that body wherein is the
immediate cause of the change. And though many, or even all, the coexisting
things should change according to certain known rules of direction and
swiftness, yet one may always determine the relation of situation which
every coexistent acquires with respect to every other coexistent, and even
that relation which any other coexistent would have to this, or which this
would have to any other, if it had not changed or if it had changed any
otherwise. And supposing or feigning that among those coexistents there is
a sufficient number of them which have undergone no change, then we may
say that those which have such a relation to those fixed existents as others
had to them before, have now the same place which those others had. And
that which comprehends all those places is called space.
(3) Ariew and Garber [1989], pp. 292 f; Preface to the New Essays (intended as a
response to Locke, 1703-05).
Although the senses are necessary for all our actual knowledge, they are
not sufficient to give us all of it, since the senses never give us anything
but instances, that is, particular or individual truths. Now all the instances
confirming a general truth, however numerous they may be, are not suffi-
cient to establish the universal necessity of that same truth, for it does not
follow that what has happened before will always happen in the same way.
[[. . .]] As a result it appears that necessary truth, such as we find in pure
mathematics and particularly in arithmetic and geometry, must have prin-
ciples whose proof does not depend on instances nor, consequently, on the
testimony of the senses, although without the senses it would never occur
to us to think of them. This is a distinction that should be noted carefully,
and it is one Euclid understood so well that he proves by reason things that
are sufficiently evident through experience and sensible images.
70c. The development of non-Euclidean geometries. Apart from the

paradigmatic role that the axiomatic approach of Eukleides’ Elements played for
mathematics in general, there is the problem of Eukleides’ fifth postulate which led
to the development of non-Euclidean geometries and which serves today somewhat
as a paradigm case for the study of the relative consistency and independence of
some particular axiom within an axiom system.
Quotation 70.18. T. L. Heath [1908/25], p. 211.

The book Euclides ab omni naevo vindicatus (1733) by Gerolamo Sac-
cheri (1667–1733), a Jesuit, and professor at the University of Pavia [[. . .]]
is of much greater importance than all the earlier attempts to prove Post. 5
because Saccheri was the first to contemplate the possibility of hypothe-

ses other than that of Euclid, and to work out a number of consequences
of those hypotheses. [[. . .]] Saccheri, however, was the victim of the precon-
ceived notion of his time that the sole possible geometry was the Euclidean,
and he presents the curious spectacle of a man laboriously erecting a struc-
ture upon new foundations for the very purpose of demolishing it after-
wards; he sought for contradictions in the heart of the systems which he
constructed, in order to prove thereby the falsity of this hypotheses.
Main point: the sum of the three angles of a triangle; equal to, greater than, or
less than two right angles; the first one is equivalent to Eukleides’ fifth postulate.
Quotations 70.19. (1) Becker [1964], p. 173.

Lambert ist sich wohl darüber im klaren gewesen, daß sämtliche Hypothe-
sen [[ Saccheri’s]] zu einer in sich konsequenten zweidimensionalen Geome-
trie führen, wie die im folgenden [[fourth and fifth para of the following
quotation]] wiedergegebene Bemerkung zeigt. Damit hat er im Grunde als
erster die Widerspruchsfreiheit der beiden nicht-euklidischen Geometrien
erwiesen.
(2) T. L. Heath [1908/25], p. 213; partly translating Lambert [1786], § 82.13
The third part of Lambert’s tract is devoted to the discussion of the
same three hypotheses as Saccheri’s, the hypothesis of the right angle being
for Lambert the first, that of the obtuse angle the second, and that of the
acute angle the third hypothesis; [[. . .]]
Lambert goes much further than Saccheri in the deduction of new
propositions from the second and third hypotheses. The most remarkable
is the following.
The area of a plane triangle, under the second and third hypotheses, is
proportional to the difference between the sum of the three angles and two
right angles.
[[. . .]]
A remarkable observation is appended (§ 82): “In connexion with this
it seems to be remarkable that the second hypothesis holds if spherical
instead of plane triangles are taken, because in the former also the sum of
the angles is greater than two right angles, and the excess is proportional
to the area of the triangle.
“It appears still more remarkable that what I here assert of spherical
triangles can be proved independently of the difficulty of parallels.”
This discovery that the second hypothesis is realised on the surface of
a sphere is important in view of the development, later, of the Riemann
hypothesis (1854).
Still more remarkable is the following prophetic sentence: “I am al-
most inclined to draw the conclusion that the third hypothesis arises with
13 See also Becker [1964], p. 173.

an imaginary spherical surface” (cf. Lobachewsky’s Géométrie imaginaire,

1837).
Quotations 70.20. (1) Gauss [1863/1903], vol. 8, p. 177; pinched from Jammer
[1954/1993], p. 147.
I become more and more convinced that the necessity of our geometry
cannot be demonstrated, at least neither by, nor for, the human intellect.
In some future life, perhaps, we may have other ideas about the nature
of space which, at the present, are inaccessible to us. Geometry, therefore,
has to be ranked until such time not with arithmetic, which is of a purely
aprioristic nature, but with mechanics.
(2) Gauss-Bessel [1880], p. 490; Gauss to Bessel, January 27, 1829; my translation.14
Occasionally, in scattered free hours, I have also thought again about
another topic which with me is already close to 40 years old; I have in
mind the first grounds of geometry; [[. . .]]. Here too I have consolidated a
number of things, and my conviction that we cannot found the geometry
completely a priori has become, if possible, even stronger. Meanwhile, I will
probably not get around for yet a long time to work out my very extensive
investigations on it for public announcement, and perhaps this will never
happen in my lifetime, because I am afraid of the howling of the Boeotians,
if I wanted to fully express my view.
Comment. As regards the expression “Boeotian”, my 1976 copy of the Concise Ox-
ford Dictionary says on p. 108: “f. Boeotia in ancient Greece, proverbial for stupidity
of inhabitants”. (My 1993 edition of the New Shorter Oxford English Dictionary has
nothing more to offer.) The expression does not seem to be well known in the Eng-
lish language but I have every reason to wholeheartedly recommend it. The German
original is “Geschrei der Böoter”; translated as “hue and cry of the blockheads” in
the translation of E. Cassirer (by W.H. Woglom and C.W. Hendel) The Problem
of Knowledge. Philosophy, Science, and History since Hegel Yale University Press,
New Haven 1950; translated as “the clamor and cry of the blockheads” in Jammer
[1954/93], p. 148.
(3) Gauss-Bessel [1880], p. 493; Bessel to Gauss, February 10, 1829; my translation.
I should very much regret, had you let yourself refrain “by the howling
of the Boeotians” from explicating your geometrical views. Through that
which Lambert has said, and that which Schweikardt expressed orally, it has
become clear to me that our geometry is incomplete, and ought to receive a
correction, which is hypothetical and vanishes when the sum of the angles
=180˚. That would be the true geometry, the Euclidean one the practical,
at least for figures on the earth.
(4) Gauss-Bessel [1880], p. 497; Gauss to Bessel, April 9, 1830; my translation.
The ease with which you related to my views about geometry gave me true
pleasure, in particular since so few are open minded about it. According to
my deepest conviction the doctrine of space has quite a different status in
14 This set of quotation was instigated by Becker [1964], pp. 178 f.
our knowledge a priori than the pure doctrine of magnitude; our knowledge
of the former quite lacks that complete conviction of its necessity (thus also
of its absolute truth), which is peculiar to the latter one; we have to admit
with humility that even though the number is a mere product of our mind,
space also has a reality outside of our mind to which we cannot prescribe
its laws a priori.
Quotations 70.21. (1) Poincaré [1913a], p. 55.

Every conclusion supposes premises; these premises themselves either
are self-evident and need no demonstration, or can be established only by
relying upon other propositions, and since we can not go back thus to
infinity, every deductive science, and in particular geometry, must rest on
a certain number of undemonstrable axioms.
(2) Poincaré [1913a], pp. 55 f
It was long sought in vain to demonstrate [[. . .]] Euclid’s Postulate. What
vast effort has been wasted in this chimeric hope is truly unimaginable.
Finally, in the first quarter of the nineteenth century, and almost at the
same time, a Hungarian and a Russian, Bolyai and Lobachevski, established
irrefutably that the demonstration is impossible; they have almost rid us of
inventors of geometries ‘sans postulatum’; since the Académie des Sciences
receives only about one or two new demonstrations a year.
(3) Poincaré [1913a], p. 64.
[[W]]hat is the nature of the geometric axioms[[?]]
Are they synthetic a priori judgments, as Kant said?
They would then impose themselves upon us with such a force that we
could not conceive the contrary proposition, nor build upon it a theoretic
edifice. There would be no non-Euclidean geometry.
To be convinced of it take a veritable synthetic a priori judgment, the
following, for instance [[. . .]]:
If a theorem is true for the number 1, and if it has been proved that it
is true of n+1 provided it is true of n, it will be true of all positive whole
numbers.
Then try to escape from that and, denying this proposition, try to
found a false arithmetic analogous to non-Euclidean geometry—it can not
be done; one would even be tempted at first blush to regard these judgments
as analytic.
Comment. Notice that the mathematical induction which Poincaré gives as an ex-
ample of a “veritable synthetic a priori judgment” was regarded as purely analytic
or logical by Dedekind and Frege.15
70d. Formal axiomatics. It took two thousand years to make a considerable

improvement on Eukleides’ axiomatics: Pasch [1882], Hilbert [1899].
15 Cf. also Hilbert in quotations 76.13 in these materials.

Quotations 70.22. (1) Townsend [1902], p. 3; translation from Hilbert [1899], p. 2

(§ 1).
Let us consider three distinct systems of things. The things composing
the first system, we will call points and designate them by the letters
A, B, C, . . .; those of the second, we will call straight lines and designate
them by the letters a, b, c, . . .; those of the third, we will call planes and
designate them by the Greek letters α, β, γ, . . .. The points are called the
elements of linear geometry; and the straight lines, the elements of plane
geometry; . . .
We think of these points, straight lines, and planes as having certain
mutual relations, which we indicate by means of such words as “are situ-
ated,” “between,“ “parallel,” “congruent,” “continuous,” etc. The complete
and exact description of these relations follows as a consequence of the
axioms of geometry.
(2) Townsend [1902], p. 27; translation from Hilbert [1899], p. 34 (§ 9).
The axioms, which we have discussed in the previous chapter and have di-
vided into five groups, are not contradictory to one another; that is to say,
it is not possible to deduce from these axioms, by any logical process of rea-
soning, a proposition which is contradictory to any of them. To demonstrate
this, it is sufficient to construct a geometry where all of the five groups are
fulfilled.
To this end, let us consider a domain Ω consisting of all those algebraic
numbers which may be obtained by beginning with the number one and ap-
plying to it a finite number of times the four arithmetical operations √ (addi-
tion, subtraction, multiplication, and division) and the operation 1 + x2 ,
where ω represents a number arising from the five operations already given.
(3) Townsend [1902], p. 29; translation from Hilbert [1899], p. 36 (§ 9).
[[E]]very contradiction resulting from our system of axioms must also appear
in the arithmetic related to the domain Ω.
Comment. Readers who are interested in the actual considerations which lead from
the preceding quotation to this conclusion will have to consult Hilbert’s book.
With the discovery of a non-Euclidean geometry the question of mathematical

truth came up again and attention was led to axiomatics.
Quotations 70.23. (1) Poincaré [1913a], p. 61.

[[T]]he word existence has not the same sense when it refers to a mathemat-
ical entity and when it is a question of a material object. A mathematical
entity exists, provided its definition implies no contradiction, either in itself,
or with the propositions already admitted.
Comment. It is in this context that Poincaré talks about implicit axioms.
(2) Poincaré [1913a], p. 65.
The axioms of geometry [[. . .]] are neither synthetic a priori judgments
nor experimental facts.
They are conventions; our choice among all possible conventions is

guided by experimental fact; but it remains free and is limited only by
the necessity of avoiding all contradiction. Thus it is that the postulates
can remain rigorously true even though the experimental laws which have
determined their adoption are only approximative.
In other words, the axioms of geometry (I do not speak of those of
arithmetic) are merely disguised definitions.
(3) Poincaré [1913a], p. 81.
If Lobachevski’s geometry is true, the parallax of a very distant star will be
finite; if Riemann’s is true, it will be negative. These are results which seem
within the reach of experiment, and there have been hopes that astronom-
ical observations might enable us to decide between the three geometries.
But in astronomy ‘straight line’ means simply ‘path of a ray of light.’
If therefore negative parallaxes were found, or if it were demonstrated
that all parallaxes are superior to a certain limit, two courses would be
open to us; we might either renounce Euclidean geometry, or else modify
the laws of optics and suppose that light does not travel rigorously in a
straight line.
It is needless to add that all the world would regard the latter solution
as the more advantageous.
The Euclidean geometry has, therefore, nothing to fear from fresh ex-
periments.
Comment. As regards the claim that “it is needless to add“, compare my comment
to quotation 96.15 (1) regarding Tarski’s phrase “It would superfluous to stress”. It is
not the point here to criticize with hindsight, but to get to the nature of Poincaré’s
misjudgment: how, on the one hand, Poincaré could be so sure that “all the world
would” “suppose that light does not travel rigorously in a straight line” and, on the
other hand, all the world decided against him. My point is that self-evidence can
serve as a highly deceptive disguise, particularly deceptive, it seems to me, to people
who regard themselves as experts.
(4) Poincaré [1913a], p. 91.
[[B]]y natural selection our mind has adapted itself to the conditions of the
external world, that it has adopted the geometry most advantageous to the
species: or in other words the most convenient.
Comment. There is something curious about this position of Poincaré’s: it seems
to suggest that a process of evolution has taken place which is at the same time
completed, the most advantageous geometry has been found. Does Poincaré assume
that there is no other form of experience possible than the one at his time?
(5) Poincaré [1913b], p. 98.
Ich möchte [[. . .]] den Schluß ziehen, daß wir die Anschauung von einem
Kontinuum beliebiger Dimensionszahl haben, da wir die Fähigkeit haben,
ein physisches und mathematisches Kontinuum zu konstruieren; und ferner,
daß diese Fähigkeit aller Erfahrung vorausgeht, da ohne sie die Erfahrung
geradezu unmöglich wäre und sich auf rohe Sinnesempfindungen beschrän-

ken würde, die ungeeignet wären, in ein Ganzes eingeordnet zu werden.
Diese Anschauung nun ist nichts anderes, als unser Bewußtsein, daß wir
diese Fähigkeit besitzen. Doch diese Fähigkeit könnte sich in verschiede-
nem Sinn betätigen; sie könnte uns ebensogut einen Raum von vier wie
einen von drei Ausdehnungen aufzubauen gestatten. Es ist die Außenwelt,
die Erfahrung, die uns bestimmt, unsere Vorstellungen in der einen Rich-
tung mehr auszubilden als in der anderen.
Comment. Note the proximity to Kant.
Quotation 70.24. Einstein [1920], pp. 1 f.

Geometry sets out from certain conceptions such as “plane,” “point,”
and “straight line,” with which we are able to associate more or less definite
ideas, and from certain simple propositions (axioms) which, in virtue of
these ideas, we are inclined to accept as “true.” Then, on the basis of a
logical process, the justification of which we feel ourselves compelled to
admit, all remaining propositions are shown to follow from those axioms,
i.e., they are proven. A proposition is then correct (“true”) when it has
been derived in the recognised manner from the axioms. The question of
the “truth” of the individual geometrical propositions is thus reduced to
one of the “truth” of the axioms. Now it has long been known that the
last question is not only unanswerable by the methods of geometry, but
that it is in itself entirely without meaning. We cannot ask whether it is
true that only one straight line goes through two points. We can only say
that Euclidean geometry deals with things called “straight lines,” to each of
which is ascribed the property of being uniquely determined by two points
situated on it. The concept “true” does not tally with the assertions of
pure geometry, because by the word “true” we are eventually in the habit
of designating always the correspondence with a “real” object; geometry,
however, is not concerned with the relation of the ideas involved in it to
objects of experience, but only with the logical connection of these ideas
among themselves.
In a lecture before the Prussian Academy of Sciences, Einstein emphasized the

clarification that had been achieved by formal axiomatics.
Quotations 70.25. (1) Einstein [1921], p. 233.

[[A]]s far as the propositions of mathematics refer to reality, they are not
certain; and as far as they are certain, they do not refer to reality. It seems
to me that complete clarity as to this state of things became common prop-
erty only through that trend in mathematics which is known by the name
of “axiomatics.” The progress achieved by mathematics consists in having
neatly separated the logical-formal from its objective or intuitive content;
according to axiomatics the logical-formal alone forms the subject mat-
ter of mathematics, which is concerned with the intuitive or other content
associated with the logical-formal.
(2) Einstein [1921], p. 234.
[[S]]uch an expurgated exposition of mathematics makes it also evident that

mathematics as such cannot predicate anything about objects of our intui-
tion or real objects.
(3) Einstein [1921], pp. 234 ff
It is clear that the system of concepts of axiomatic geometry alone cannot

make any assertions as to the behavior of real objects of this kind, which
we will call practically-rigid bodies. To be able to make such assertions,
geometry must be stripped of its merely logical-formal character by the co-
ordination of real objects of experience with the empty conceptual schemata
of axiomatic geometry. To accomplish this, we need only add the proposi-
tion: solid bodies are related, with respect to their possible dispositions, as
are bodies in Euclidean geometry of three dimensions. Then the proposi-
tions of Euclid contain affirmations as to the behavior of practically-rigid
bodies.
Geometry thus completed is evidently a natural science; we may in
fact regard it as the most ancient branch of physics. [[. . .]] We will call this
completed geometry “practical geometry,” and shall distinguish it in what
follows from “purely axiomatic geometry.” The question whether the prac-
tical geometry of the universe is Euclidean or not has a clear meaning, and
its answer can only be furnished by experience. All length-measurements in
physics constitute practical geometry in this sense, so, too, do geodetic and
astronomical length-measurements, if one utilizes the empirical law that
light is propagated in a straight line, and indeed a straight line in the sense
of practical geometry.
[[. . .]] If we reject the relation between the body of axiomatic Euclidean
geometry and the practically-rigid body of reality, we readily arrive at the
following view, which was entertained by that acute and profound thinker,
H. Poincaré: Euclidean geometry is distinguished above all other conceiv-
able axiomatic geometries by its simplicity. Now since axiomatic geometry
by itself contains no assertions as to the reality which can be experienced,
but can do so only in combination with physical laws, it should be possible
and reasonable—whatever may be the nature of reality—to retain Euclidean
geometry. [[. . .]] If we reject the relation between the practically-rigid body
and geometry, we shall indeed not easily free ourselves from the convention
that Euclidean geometry is to be retained as the simplest.
Why is the equivalence of the practically-rigid body and the body of
geometry—which suggests itself so readily—rejected by Poincaré and other
investigators? Simply because under closer inspection the real solid bodies
in nature are not rigid, because their geometrical behavior, that is, their pos-
sibilities of relative disposition, depend upon temperature, external forces,
etc. Thus the original, immediate relation between geometry and physical
reality appears destroyed[[.]]
§ 71. LOGICISM — THE ULTIMATE FOUNDATION 1003
§ 71. Logicism — the ultimate foundation
Foundation, in the sense of reducing concepts to simpler ones, seems to come to an

end with logic: a bottom line seems to have reached.
71a. The rise of the reductionist conception of foundation. Arithmetics
and calculus proceeded quite independently of the development of axiomatics. The
axiomatic method which is characteristic of geometry was already well developed
in the 3rd century BC whereas the axiomatization of arithmetics is just a hundred
years old and is constituted by a new standard of precision.
The reductionist conception of foundation in mathematics is characterized by
the following stages:
(i) the elimination of the notion of indivisibles;
(ii) the arithmetization of analysis;
(iii) the set theoretical foundation of arithmetic;
(iv) the logical foundation of set theory.
There isn’t much to say about the elimination of the notion of indivisibles in the
context of this study although it has been linked as a reaction to “the second crisis in
the foundations of mathematics”, the one of ancient
√ Greek mathematics, constituted
by the discovery of the incommensurability of 2 and Zenon’s paradoxes.16 What
interests me here begins with the arithmetization of analysis and the set theoretical
foundation of arithmetic and Frege’s logical foundation of set theory (or rather
arithmetic).
Apart from the notion of set that underlies both, what interests me in the
arithmetization of analysis is Dedekind’s cut and in the set theoretical foundation
of arithmetic the principle of induction.
I begin with a couple of quotations regarding the background for the introduc-
tion of Dedekind’s cut.17
Quotations 71.1. (1) Beman [1909], pp. 9 f; translating Dedekind [1872], pp. 9 f.
[[T]]he way in which the irrational numbers are usually introduced is based
directly upon the conception of extensive magnitudes—which itself is no-
where carefully defined—and explains number as the result of measuring
such a magnitude by another of the same kind. Instead of this I demand
that arithmetic shall be developed out of itself.
That such comparisons with non-arithmetic notions have furnished the
immediate occasions for the extension of the number-concept may, in a
general way, be granted (though this was certainly not the case in the
introduction of complex numbers); but this surely is no sufficient ground
for introducing these foreign notions into arithmetic, the science of numbers.
Just as negative and fractional numbers are formed by a new creation, and
as the laws of operating with these numbers must and can be reduced to the
laws of operating with positive integers, so we must endeavor completely
16 Cf. §5 in Fraenkel et al. [1973], pp. 12–14 on “the three crises”.
17 A set theoretical formulation of Dedekind’s cut can be found in definition 86.3 in the
tools.
to define irrational numbers by means of the rational numbers alone. The

question only remains how to do this.
(2) Beman [1909], p. 11; translating Dedekind [1872], p. 10.
[[E]]very point p of the straight line produces a separation of the same into
two portions such that every point of one portion lies to the left of every
point of the other. I find the essence of continuity in the converse, i. e., in
the following principle:
“If all points of the straight line fall into two classes such that every
point of the first class lies to the left of every point of the second class,
then there exists one and only one point which produces this division of all
points into two classes, this severing of the straight line into two portions.”
Comment. This divisibility is the point for dialectic: in its extreme case, the divis-
ibility into true and false, it is incompatible with unrestricted abstraction. This is
the main topic.
The introduction of irrational numbers is already beyond the scope of the

present study. Readers might find Frege’s view on Dedekind’s approach in § 139
of the second volume of his Grundgesetze interesting.18 What is relevant for my
approach is the further step of a reduction of rational numbers to natural numbers
and then to logic.
Quotations 71.2. (1) Beman [1909], p. 4; translating Dedekind [1872], pp. 5 f.

I regard the whole of arithmetic as a necessary, or at least natural, conse-
quence of the simplest arithmetic act, that of counting, and counting itself
as nothing else than the successive creation of the infinite series of positive
integers in which each individual is defined by the one immediately preced-
ing; the simplest act is the passing from an already-formed individual to
the consecutive new one to be formed. The chain of these numbers forms in
itself an exceedingly useful instrument for the human mind, it presents an
inexhaustible wealth of remarkable laws obtained by the introduction of the
four fundamental operations of arithmetic. Addition is the combination of
any arbitrary repetitions of the above-mentioned simplest act into a single
act; from it in a similar way arises multiplication. While the performance
of these two operations is always possible, that of the inverse operations,
subtraction and division, prove to be limited. Whatever the immediate oc-
casion may have been, what ever comparisons or analogies with experience,
or intuition, may have led thereto; it is certainly true that just this limi-
tation in performing the indirect operations has in each case been the real
motive for a new creative act; thus negative and fractional numbers have
been created by the human mind; and in the system of all rational numbers
there has been gained an instrument of infinitely greater perfection.
(2) Beman [1909], pp. 31 f; translating Dedekind [1888], p. III.
In speaking of arithmetic (algebra, analysis) as a part of logic I mean to im-
18 Translated in Geach and Black [1952], pp. 173 f.
ply that I consider the number-concept entirely independent of the notions

or intuitions of space and time, that I consider it an immediate result from
the laws of thought. My answer to the problems propounded in the title of
this paper [[the nature and meaning of numbers; German: Was sind und was
sollen die Zahlen]]19 is, then, briefly this: numbers are free creations of the
human mind; they serve as a means of apprehending more easily and more
sharply the difference of things. It is only through the purely logical process
of building up the science of numbers and by thus acquiring the continuous
number-domain that we are prepared accurately to investigate our notions
of space and time by bringing them into relation with this number-domain
created in our mind. If we scrutinise closely what is done in counting an
aggregate or number of things, we are led to consider the ability of the
mind to relate things to things, to let a thing correspond to a thing, or
to represent a thing by a thing, an ability without which no thinking is
possible.
(3) Beman [1909], p. 33; translating Dedekind [1888], pp. IV f.
I feel conscious that many a reader will scarcely recognise in the shadowy
forms which I bring before him his numbers which all his life long have
accompanied him as faithful and familiar friends; he will be frightened by
the long series of simple inferences corresponding to our step-by-step un-
derstanding [[unseres Treppenverstandes]], by the matter-of-fact dissection
of the chains of reasoning on which the laws of numbers depend, and will
become impatient at being compelled to follow out proofs for truths which
to his supposed inner consciousness seem at once evident and certain. On
the contrary in just this possibility of reducing such truths to others more
simple, no matter how long and apparently artificial the series of inferences,
I recognise a convincing proof that their possession or belief in them is never
given by inner consciousness but is always gained only by a more or less
complete repetition of the individual inferences.
Comment. “Inner consciousness” is a translation for “innere Anschauung”. As re-
gards the aspect of the “shadowy forms”, cf. Russell in quotation 71.3 (2) below.
While Dedekind’s view is characterized by a preference for the aspect of succes-

sion or progression of numbers, i.e., their ordinal character, there is another view
which focuses on the aspect of one-to-one mappings of sets.
Quotations 71.3. (1) Cantor [1887], p. 380; my translation.

If we look at the definition of the finite cardinal number in Euclid we
have to acknowledge first of all that he, just like we do it, relates the number,
according to its true origin, to the set and does in no way turn the number
into a mere “symbol” which is associated to single objects in the process of
counting.
19 More literally translated: “What are and what is the point of numbers?”
(2) Russell [1903], p. 249.

Dedekind remarks in his preface [[as cited in quotation 71.2 (3) above]] that
many will not recognize their old friends the natural numbers in the shad-
owy shapes which he introduces to them. In this, it seems to me, the sup-
posed persons are in the right—in other words, I am one among them. What
Dedekind presents to us is not the numbers, but any progression: what he
says is true of all progressions alike, and his demonstrations nowhere—
not even where he comes to cardinals—involve any property distinguishing
numbers from other progressions. No evidence is brought forward to show
that numbers are prior to other progressions. We are told, indeed, that
they are what all progressions have in common; but no reason is given for
thinking that progressions have anything in common beyond the properties
assigned in the definition, which do not themselves constitute a new progres-
sion. The fact is that all depends upon one-one relations, which Dedekind
has been using throughout without perceiving that they alone suffice for the
definition of cardinals. The relation of similarity between classes, which he
employs consciously, combined with the principle of abstraction, which he
implicitly assumes, suffice for the definition of cardinals; for the definition
of ordinals these do not suffice[[.]]
Comment. The point is relevant in the context of my study. As far as I can see, it is
the use of classical logic which hides progression in contraction and which obscures
the role of progression as the first relevant concept in the definition of numbers; i.e.,
I regard ordinals as prior to cardinals (despite the fact that cardinals are treated
first in my presentation in chapter I of the tools).
(3) Dedekind (letter to H. Weber, 24 January 1888), translated from Becker [1964],
p. 244.
I must confess [[. . .]] that up to now I still keep regarding the ordinal number,
not the cardinal number (quantity [[Anzahl]]) as the original concept of
number.
I continue with Cantor’s and Dedekind’s characterizations of a set.
Quotations 71.4. (1) Wang [1974], p. 188; translating Cantor [1895], p. 481 ([1932],
p. 282).
By a “set” we shall understand any collection into a whole M of definite,
distinct objects m (which will be called the “elements” of M ) of our intuition
or our thought.
(2) Beman [1909], pp. 44 f; translating Dedekind [1888], pp. 1 f.
1. In what follows I understand by thing every object of our thought.
[[. . .]]
2. It very frequently happens that different things, a, b, c,... for some
reason can be considered from a common point of view, can be associated
in the mind, and we say that they form a system S ; we call the things
a, b, c,... elements of the system S, they are contained in S; conversely, S
consists of these elements. Such a system S (an aggregate, a manifold, a
totality) as an object of our thought is likewise a thing (1); it is completely

determined when with respect to every thing it is determined whether it is
an element of S or not.
Comment. Note that Dedekind is quite clear about sets (or systems, etc.) being
also objects. Also: the universal set is such a system, since it is determined for every
object that it is an element of the universal set.
(3) U. Petersen [2002] p. 1007; translating Frege [1893], p. 2.
Although here is a sentiment [[Ahnung]] of the correct contained in the
common point of view [[Gesichtspunkt]]; but that considering [[Auffassung]],
that putting together in the mind, is not an objective characteristic. I ask:
in whose mind? If they are put together in one mind, not in another, do
they then form a system?
Comment. This is Frege’s response to Dedekind [1888], pp. 1 f; simple but effective.
The point applies to Cantor’s phrase “Objekte unseres Denkens” as well.
(4) Dedekind [1888], p. 14.
Meine Gedankenwelt, d.h. die Gesamtheit S aller Dinge, welche Gegenstand
meines Denkens sein können, ist unendlich. Denn wenn s ein Element von
S bedeutet, so ist der Gedanke s′ , daß s Gegenstand meines Denkens sein
kann, selbst ein Element von S. Sieht man dasselbe als Bild ϕ(s) des Ele-
mentes s an, so hat daher die hierdurch bestimmte Abbildung ϕ von S die
Eigenschaft, daß das Bild S ′ Teil von S ist; und zwar ist S ′ echter Teil von
S, weil es in S Elemente gibt (z.B. mein eigenes Ich), welche von jedem
solchen Gedanken s′ verschieden und deshalb nicht in S ′ enthalten sind.
Endlich leuchtet ein, daß, wenn a, b verschiedene Elements von S sind, auch
ihre Bilder a′ , b′ verschieden sind, daß also die Abbildung ϕ eine deutliche
(ähnliche) ist [[. . .]]. Mithin ist S unendlich [[. . .]].
Comment. Dedekind points out in a footnote that a similar consideration can be
found in § 13 of the Paradoxien des Unendlichen by Bolzano, Leipzig 1851. An
English translation can also be found in Suppes [1960/72], p. 138 with a short remark
on Russell’s attitude.
At the turn of the century, the foundations of mathematics (analysis) seemed

to have reached perfection.
Quotation 71.5. Wang [1954], p. 242 ([1962], p. 560).

During the nineteenth century, the attempts to found analysis on a re-
liable basis went generally under the caption “arithmetization of analysis”.
It is well known that Cauchy, Weierstrass, Dedekind, Cantor all made im-
portant contributions to this program. Indeed, their results were so well re-
ceived among mathematicians that in 1900, Poincaré asserted: “Today there
remain in analysis only integers and finite or infinite systems of integers,
interrelated by a net of relations of equality or inequality. Mathematics, as
we say, has been arithmetized. . . . We may say today that absolute rigour
has been obtained”.
If by “arithmetization” is meant merely the elimination of geometrical

intuition, then the success is hardly disputable. If, on the other hand, by
“arithmetization” is meant a reduction of analysis to a theory of integers,
then the matter becomes more involved because not only integers but also
“finite or infinite systems of integers” are needed. Nowadays it would be
more customary to refer to these “systems” as sets or classes. What is ac-
complished is not the founding of analysis on the theory of integers alone,
but rather on the theory of integers plus the theory of sets.
Comment. The passage quoted is from Poincaré [1902]; it can also be found in
Fraenkel et al. [1973], p. 14.
The set theoretical foundation of mathematics was a great achievement, but

still, it did not explain why mathematics is applicable to the world without, appar-
ently, having its roots in the world. There was, however, an even further reaching
goal, which was quite capable of giving an answer to this question: logicism.
71b. The idea of a logical foundation of arithmetic. Although the basic

mathematical concepts had been traced back to the fundamental one of a set, the
question about the nature of mathematics, as put in terms of Kant: analytic or
synthetic, was still open. It was first of all Frege who tackled this question by
attempting a reduction of arithmetic to logic and then, independently of Frege,
Russell. In this section I shall highlight those aspects of their ideas which bear upon
my project.
Features 71.6. My emphasis is on the following aspects in Frege’s attempt at

reducing arithmetic to logic:
(FR1) a statement of number is an assertion about a concept;20
(FR2) difference between ‘Merkmal’ and ‘Eigenschaft’ of a concept;21
(FR3) the number as extension of a concept;22
(FR4) permission to pass from a concept to its extension;
(FR5) extensional standpoint in logic;
(FR6) the choice of symbol is arbitrary;23
(FR7) a definition cannot create objects;
(FR8) arithmetical equations can be applied only because they express
thoughts.24
I begin with a set of quotations which deal with the general issue of the logical
character of arithmetic.
20 This is part of the heading of § 46 of Frege’s [1884].

21 This is part of the heading of § 53 of Frege’s [1884].
22 This is the heading of § 68 of Frege’s [1884].
Quotations 71.7. (1) Austin [1950], p. 22e ; translating Frege [1884], p. 22.
A very emphatic declaration in favour of the analytic nature of the
laws of number is that of W. S. Jevons: “I hold that algebra is a highly
developed logic, and number but logical discrimination.”
§ 16. But this view, too, has its difficulties. Can the great tree of the
science of number as we know it, towering, spreading, and still continually
growing, have its roots in bare identities? And how do the empty forms of
logic come to disgorge so rich a content?
To quote Mill: “The doctrine that we can discover facts, detect the
hidden processes of nature, by an artful manipulation of language, is so
contrary to common sense, that a person must have made some advances
in philosophy to believe it.”
Comment. The question “how do the empty forms of logic come to disgorge so rich
a content?” is a central question in the present study. Recall in this context the
problem Hegel was facing: in which way can logic be more but purely formal?25 It
will come up again in chapters XXIX and XXX. Notice also Mill’s appeal to “common
sense”.
(2) Austin [1950], pp. 23e f; translating Frege [1884], pp. 23 f.
Instead of linking our chain of deduction direct to any matter of fact, we
can leave the fact where it is, while adopting its content in the form of a
condition. By substituting in this way conditions for facts throughout the
whole of a train of reasoning, we shall finally reduce it to a form in which
a certain result is made dependent on a certain series of conditions. This
truth would be established by thought alone or, to use Mill’s expression,
by an artful manipulation of language. It is not impossible that the laws
of numbers are of this type. This would make them analytic judgements,
despite the fact that they would not normally be discovered by thought
alone; for we are concerned here not with the way in which they are dis-
covered but with the kind of ground on which their proof rests [[. . .]] The
truths of arithmetic would then be related to those of logic in much the
same way as the theorems of geometry to the axioms. [[. . .]] [[T]]hen indeed
the prodigious development of arithmetical studies, with their multitudi-
nous applications, will suffice to put an end to the widespread contempt for
analytic judgements and to the legend of the sterility of pure logic.
Comment. Note a certain similarity to Kant’s consideration in [1781] (A), p. 27.26
(3) Austin [1950], p. 99e ; translating Frege [1884], p. 99.
I hope I may claim in the present work to have made it probable that the
laws of arithmetic are analytic judgements and consequently a priori. Arith-
metic thus becomes simply a development of logic, and every proposition
of arithmetic a law of logic, albeit a derivative one. To apply arithmetic in
the physical sciences is to bring logic to bear on observed facts[[.]]
25 Cf. sections 65c and d in these materials.
(4) Austin [1950], p. IVe ; translating Frege [1884], p. IV.

[[E]]ven an inference like that from n to n + 1, which on the face of it is
peculiar to mathematics, is based on the general laws of logic, and [[. . .]]
there is no need of special laws for aggregative thought.
Comment. Cf. quotation 70.21 (3) above for an opposite view held by Poincaré. The
point is a central one in the present study.
Grassmann attempts to obtain the law
a + (b + 1) = (s + b) + 1
by means of a definition, as follows:
“If a and b are any arbitrary members of the basic series, then by the
sum a + b is to be understood that member of the basic series for which the
formula
a + (b + e) = (s + b) + e
is valid.” e here is to be taken to mean positive unity. This definition can
be criticized in two different ways. First, sum is defined in terms of itself. If
we do not yet understand the meaning of a + b, we do not understand the
expression a+ (b + e) either. This criticism, however, can perhaps be evaded
if we say (admittedly going against the text) that what he is intending
to define is not sum but addition. In that case, the criticism could still
be brought that a + b would be an empty symbol if there were either no
member or several members of the basic series which satisfied the prescribed
condition. That this does not in fact ever happen, Grassmann simply
assumes without proof, so that the rigour of his procedure is only apparent.
Comment. The reference to Grassmann is given as Grassmann [1860], p. 4 (p. 301
in the reprint). Apparently, Frege did not see the point of what is called today a
definition by induction or recursive definition. But then, Grassmann’s definition is
not yet on the level of Dedekind’s definition in his [1888], p. 28 (126).
The basic line of the answer to Mill’s question is contained in the next quotation.
It is one of the most crucial parts of my own enterprise.
Quotation 71.8. Austin [1950], p. 99e ; translating Frege [1884], p. 99.

[[I]]n the external world, in the whole of space and all that therein is, there
are no concepts, no properties of concepts, no numbers. The laws of number,
therefore, are not really applicable to external things; they are not laws of
nature. They are, however, applicable to judgements holding good of things
in the external world: they are laws of the laws of nature. They assert not
connections between phenomena, but connections between judgements; and
among judgements are included the laws of nature.
Now the point is, what sort of objects is arithmetic talking about? How are
these logical objects given?
Number is not abstracted from things in the way that colour, weight
and hardness are, nor is it a property of things in the sense that they are.
But when we make a statement of number, what is that of which we assert
something? [[. . .]]
Number is not anything physical, but nor is it anything subjective (an
idea).
Number does not result form the annexing of thing to thing. It makes
no difference even if we assign a fresh name after each act of annexation.
[[. . .]]
[[. . .]] How are we to curb the arbitrariness of our ways of regarding
things, which threatens to obliterate every distinction between one and
many?
Comment. Does Frege mean to say that he takes “colour, weight, and hardness” as
primary qualities of things, or is this terminology just inappropriate in this context?
§ 46. It should throw some light on the matter to consider number in
the context of a judgement which brings out its basic use. While looking at
one and the same external phenomenon, I can say with equal truth both
“It is a copse” and “It is five trees”, or both “Here are four companies” and
“Here are 500 men”. Now what changes here from one judgement to the
other is neither any individual object, nor the whole, the agglomeration of
them, but rather my terminology. But that is itself only a sign that one
concept has been substituted for another. This suggests as the answer to
the first of the questions left open in our last paragraph, that the content
of a statement of number is an assertion about a concept. This is perhaps
clearest with the number 0. If I say “Venus has 0 moons”, there simply does
not exist any moon or agglomeration of moons for anything asserted of; but
what happens is that a property is assigned to the concept “moon of Venus”,
namely that of including nothing under it. If I say “the King’s carriage is
drawn by four horses”, then I assign the number four to the concept “horse
that draws the King’s carriage”.
(3) Austin [1950], pp. 79e –80e ; translating Frege [1884], pp. 79 f.
My definition is therefore as follows:
the Number which belongs to the concept F is the extension of the
concept “equal to the concept F ”.
(4) Austin [1950], p. 87e ; translating Frege [1884], p. 87. translation
0 is the number which belongs to the concept “unequal to itself”.
Comment. Together with the foregoing quotation we have: 0 is the extension of the
concept “equal to the concept “unequal to itself” ”, i.e.,
W V
0 = λx f (bij (f ) ∧ y (y ∈ x ↔ f (y) ∈ λz (z 6= z))) .

1 is the number which belongs to the concept “equal 0”.
Quotations 71.10. (1) Frege [1903], p. 155.

Wir haben uns an unsere Umwandlung der Allgemeinheit einer Gleichheit in
eine Werthverlaufsgleichheit erinnert, die uns das zu leisten verspricht, was
die schöpferischen Definition anderer Mathematiker nicht vermögen. Wir
haben die reelle Zahl als Grössenverhältnis aufgefasst und so die formale
Arithmetik im oben angegebenen Sinne ausgeschlossen. Damit haben wir
auf die Grössen hingewiesen als auf die Gegenstände, zwischen denen ein
solches Verhältnis stattfindet.
Da die Anzahlen nicht Verhältnisse sind, müssen wir sie von den po-
sitiven ganzen Zahlen unterscheiden. Darum ist es nicht möglich, das Ge-
biet der Anzahlen zu dem der reellen Zahlen zu erweitern; es sind eben
ganz getrennte Gebiete. Die Anzahlen antworten auf die Frage: „wieviele
Gegenstände einer gewissen Art giebt es?“ während die reellen Zahlen als
Maasszahlen betrachtet werden können, die angeben wie gross eine Grösse
verglichen mit einer Einheitsgrösse ist.
(2) Frege [1903], pp. 157 f.
§ 160. Das Wort „Grösse“ („Quantität“) bedeutet bei manchen Mathe-
matikern soviel wie „reelle Zahl“ oder „Zahl“ schlechtweg. Diese Gebrauchs-
weise hängt wohl damit zusammen, dass die Maasszahl und die durch sie
hinsichtlich einer Einheit (eines Maasses) bestimmte Grösse nicht immer
auseinander gehalten werden; sie kann für uns hier nicht in Betracht kom-
men.
Was Grösse sei, ist wohl noch niemals befriedigend gesagt worden.
Wenn wir die Erklärungsversuche durchgehen, stossen wir oft auf das Wort
„gleichartig“ oder ein ähnliches. Man verlangt von den Grössen, dass man
die gleichartigen unter ihnen vergleichen, addiren subtrahiren, auch eine
Grösse in ihr gleichartige Teile zerlegen könne. Mit dem Worte „gleichar-
tig“ ist offenbar gar nichts gesagt; denn in einer Hinsicht können Dinge
gleichartig sein, die in einer anderen ungleichartig sind. Die Frage also,
ob ein Gegenstand einem anderen gleichartig sei, lässt sich nicht mit „ ja“
oder „nein“ beantworten; es fehlt an dem ersten logischen Erfordernis, der
scharfen Begrenzung.
(3) Frege [1903], pp. 158 f.
Statt zu fragen: welche Eigenschaften muss ein Gegenstand haben, um
eine Grösse zu sein? muss man fragen: wie beschaffen muss ein Begriff sein,
damit sein Umfang ein Grössengebiet sei? Wir wollen nun statt „Begriffsum-
fang“ der Kürze wegen „Klasse“ sagen. Dann kann man die Frage auch so
stellen: welche Eigenschaften muss eine Klasse haben, um ein Grössengebiet
zu sein? Etwas ist eine Grösse nicht für sich allein, sondern nur, insofern es
mit andern Gegenständen einer Klasse angehört, die ein Grössengebiet ist.
(4) Frege [1903], p. 161.

Wir bedürfen [[. . .]] einer Klasse von Gegenständen, die in den Rela-
tionen unseres Grössengebietes zu einander stehen, und zwar muss diese
Klasse unendlich viele Gegenstände umfassen. Nun kommt ja dem Begriffe
endliche Anzahl eine unendliche Anzahl zu, die wir Endlos genannt haben;
aber diese Unendlichkeit genügt noch nicht. Nennen wir den Umfang eines
Begriffes, der dem Begriffe endliche Anzahl untergeordnet ist, eine K l a s s e
e n d l i c h e r A n z a h l e n , so kommt dem Begriffe Klasse endlicher Anzah-
len eine unendliche Anzahl zu, die grösser als Endlos ist; d. h. es lässt sich
der Begriff endliche Anzahl abbilden in dem Begriff Klasse endlicher An-
zahlen, aber nicht umgekehrt dieser in jenen.
Nun wären etwa zwischen den Klassen endlicher Anzahlen Relationen
nachzuweisen, welche als Angehörige eines Grössengebietes aufgefasst wer-
den könnten. Etwas anders wird sich die Sache freilich noch gestalten[[.]]
Note that a logical foundation of mathematics may be viewed as “ultimate”.

Apparently there is no way to go beyond logic.
71c. The making of a new logic. In order to reduce arithmetics to logic,
Frege had to provide a logic which would do the job. Aristoteles’ logic was in an
essential way incomplete: no relations. The whole thing is an interesting example
of the different way of thinking of a mathematician as compared to a philosopher
(Kant). At no stage (in print) did Kant actually make an effort to prove his claims.
In order to have an appropriate tool to achieve a logical foundation of arith-
metic, Frege had to go beyond the syllogistic logic of Aristoteles. In that he was
guided by certain mathematical analogies such as the analogy between function and
concept, as well as certain natural language considerations.
Remark 71.11. The making of modern logic also involved Boole, De Morgan,
Peano, Peirce, Schröder, and others, of whom only Peano is represented in the
following quotations (with a single one). This is partly due to the fact that my
prime interest lies in the emergence of paradoxes which, in turn, arose primarily
from Frege’s work. Readers interested in a broader view on the making of modern
logic might find Quine’s Selected Logic Papers enjoyable.
Quotations 71.12. (1) Geach and Black [1952], p. 138; translating Frege [1893],
p. VII.
It has often been said that arithmetic is only a more highly developed
logic; but that remains disputable as long as the proofs contain steps that
are not performed according to acknowledged logical laws, but seem to rest
on intuitive knowledge. Only when these are resolved into simple logical
steps can we be sure that arithmetic is founded solely upon logic.
(2) Geach and Black [1952]. p. 137; translating Frege [1893], p. VI.
The ideal of a strictly scientific method in mathematics, which I have tried
to realize here, and which perhaps might be named after Euclid, I should
like to describe in the following way.
It cannot be required that we should prove everything, because that

is impossible; but we can demand that all propositions used without proof
should be expressly mentioned as such, so that we can see distinctly what
the whole construction rests upon. We should, accordingly, strive to dimin-
ish the number of these fundamental laws as much as possible, by proving
everything that can be proved. Furthermore I demand—and in this I go
beyond Euclid—that all methods of inference used must be specified in
advance. Otherwise it is impossible to ensure satisfying the first demand.
(3) Frege [1884], p. 2.
Euclid gives proofs of many things which anyone would concede him without
question. And it was when men refused to be satisfied even with Euclid’s
standards of rigour that they were led to the enquiries set in train by the
Axiom of Parallels.
[[. . .]]
The aim of proof is, in fact, not merely to place the truth of a proposi-
tion beyond all doubt, but also to afford us insight into the dependence of
truths upon one another. After we have convinced ourselves that a boulder
is immovable, by trying unsuccessfully to move it, there remains the further
question, what is it that supports it so securely?
Comment. This is an important quotation in view of all those who have trouble
seeing formal deduction in any other role than that of securing informal reasoning,
as, for instance, Kreisel as documented in quotation 85.8 (2) in these materials.
(4) Geach and Black [1952], p. 138; translating Frege [1893], p. VII.
Since there are no gaps in the chain of inference, each axiom, assumption,
hypothesis, or whatever you like to call it, upon which a proof is founded,
is brought to light; and so we gain a basis for deciding the epistemological
nature of the law that is proved.
(4) Peano [1899], p. 11; pinched from van Heijenoort [1967b], p. 119, fn. 2; actually
translating Alessandro Padoa.
The proof of a proposition of logic has as its purpose, in general, not to
assure us of its truth, but rather to reduce this mode of reasoning to simpler
ones, which cannot be decomposed any further and which will be called
primitive propositions.
Quotations 71.13. (1) Van Heijenoort [1967b], pp. 12; translation of Frege [1879],
pp. 2 ff.
A distinction between subject and predicate does not occur in my way of
representing a judgment. [[. . .]] In ordinary language, the place of the subject
in the sequence of words has the significance of a distinguished place, where
we put that to which we wish especially to direct the attention of the
listener [[. . .]] Now, all those peculiarities of ordinary language that result
only from the interaction of speaker and listener [[. . .]] have nothing that
answers to them in my formula language, since in a judgment I consider
only that which influences its possible consequences. Everything necessary
for a correct inference is expressed in full, but what is not necessary is

generally not indicated; nothing is left to guesswork. In this I faithfully
follow the example of the formula language of mathematics, a language to
which one would do violence if he were to distinguish between subject and
predicate in it. [[. . .]]
In the first draft of my formula language I allowed myself to be misled by
the example of ordinary language into constructing judgments out of subject
and predicate. But I soon became convinced that this was an obstacle to
my specific goal and led only to useless prolixity.
Comment. Notice the point regarding the “interaction of speaker and listener” and
compare Searle in quotation 88.10 (3) in these materials.
(2) Long and White [1979], p. 120; translating Frege [1892–95], p. 130.
[[I]]t would be best to banish the words ‘subject’ and predicate’ from logic
entirely, since they lead us again and again to confound two quite different
relations: that of an object’s falling under a concept and that of one concept
being subordinated to another.
(3) Van Heijenoort [1967b], p. 13; translation of Frege [1879], pp. 4 f
The distinction between categoric, hypothetic, and disjunctive judg-
ments seems to me to have only grammatical significance.
The apodictic judgment differs from the assertory in that it suggests
the existence of universal judgments from which the proposition can be
inferred, while in the case of the assertory one such a suggestion is lacking.
By saying that a proposition is necessary I give a hint about the ground for
my judgment. But, since this does not affect the conceptual content of the
judgment, the form of the apodictic judgment has no significance for us.
Quotations 71.14. (1) Long and White [1979], p. 10; translating Frege [1880/81],
p. 10.
Among the various sorties Leibniz made upon his goal [[of a lingua charac-
terica]], the beginnings of a symbolic logic come closest to what seems to
be indicated by the phrase ‘calculus ratiocinator ’. They are to be found in
the essays:
Non inelegans specimen demonstrandi in abstractis and
Addenda ad specimen calculi universalis
In these Leibniz stuck very close to language. Just as the words we use
for the attributes of a thing follow one another, so he simply juxtaposes
the letters corresponding to properties in order to express the formation of
a concept. If, for instance, A means right-angled, B isosceles, C triangle,
then Leibniz represents right-angled isosceles triangle by ABC. He uses a
sign for identity, ∞, and the sign + with the definition:
‘A + B∞L significat A inesse ipsi L’.
This seems to coincide with the meaning recent logicians have given the
sign, according to which A + B represents the class of individuals which
belong to A or to B or to both. Since I am passing over less important
details, the only other fact I will mention is that Leibniz allows the words
‘non’ and ‘ens’ to occur in his formulae. In this project he surely has the
lingua characterica in mind, even though he made no express connection
with the attempts he made to represent a content.
Comment. Cf. quotations 70.11 above.
(2) Long and White [1979], pp. 12; translating Frege [1880/81], pp. 13 f.
Right from the start I had in mind the expression of a content. What I am
striving after is a lingua characterica in the first instance for mathematics,
not a calculus restricted to pure logic. But the content is to be rendered
more exactly than is done by verbal language. For that leaves a great deal
to guesswork, even if only of the most elementary kind. There is only an
imperfect correspondence between the way words are concatenated and the
structure of the concepts. The words ‘lifeboat’ and ‘deathbed’ are similarly
constructed though the logical relations of the constituents are different. So
the latter isn’t expressed at all, but is left to guesswork. Speech often only
indicates by inessential marks of by imagery what a concept-script should
spell out in full. At a more external level, the latter is distinguished from
verbal language in being laid out for the eye rather than for the ear. Verbal
script is of course also laid out for the eye, but since it simply reproduces
verbal speech, it scarcely comes closer to a concept-script than speech: in
fact it is at an even greater remove from it, since it consists in signs for
signs, not of signs for the things themselves. A lingua characterica ought,
as Leibniz says, peindre non pas les paroles, mais les pensées. The formula-
languages of mathematics come much closer to this goal, indeed in part
they arrive at it. But that of geometry is still completely undeveloped and
that of arithmetic itself is inadequate for its own domain; for at precisely
the most important points, when new concepts are to be introduced, new
foundations laid, it has to abandon the field to verbal language, since it
only forms number out of numbers and can only express those judgements
which treat of the equality of numbers which have been generated in dif-
ferent ways. But arithmetic in the broadest sense also forms concepts—and
concepts of such richness and fineness in their internal structure that in per-
haps no other science are they to be found combined with the same logical
perfection. And there are other judgements which arithmetic makes, beside
mere equations and inequalities. The reason for this inability to form con-
cepts in a scientific manner lies in the lack of one of the two components
of which every highly developed language must consist. That is, we may
distinguish the formal part which in verbal language comprises endings,
prefixes, suffixes and auxiliary words, from the material part proper. The
signs of arithmetic correspond to the latter. What we still lack is the logical
cement that will bind these building stones firmly together. Up till now ver-
bal language took over this role, and hence it was impossible to avoid using
it in the proof itself, and not merely in parts that can be omitted without
affecting the cogency of the patterns of inference, whose only purpose is to
make it easier to grasp connections.
Comment. According to the editors, “In 1881, this article was submitted by Frege
in turn to the Zeitschrift für Mathematik und Physik, the Mathematische Annalen
and the Zeitschrift für Philosophie und philosophische Kritik, but was in every case
rejected by the editors. It finally remained unfinished.”
[[A]]t first sight the proposition
“All whales are mammals”
seems to be not about concepts but about animals; but if we ask which
animal then we are speaking of, we are unable to point to any one in
particular. Even supposing a whale is before us, our proposition still does
not state anything about it. We cannot infer from it that the animal before
us is a mammal without the additional premiss that it is a whale.
By properties which are asserted of a concept I naturally do not mean the
characteristics which make up the concept. These latter are properties of the
things which fall under the concept, not of the concept. Thus “rectangular”
is not a property of the concept “rectangular triangle”; but the proposition
that there exists no rectangular equilateral rectilinear triangle does state
a property of the concept “rectangular equilateral rectilinear triangle”; it
assigns to it the number nought.
Quotations 71.16. (1) Geach and Black [1952], pp. 32 f; translating Frege [1891],
pp. 19 f.
It seems to be demanded by scientific rigour that we should have provisos
against an expression’s possibly coming to have no reference; we must see
to it that we never perform calculations with empty signs in the belief that
we are dealing with objects. People have in the past carried out invalid
procedures with divergent infinite series. It is thus necessary to lay down
rules from which if follows, e.g., what
‘⊙ + 1’
stands for, if ‘⊙’ is to stand for the Sun. What rules we lay down is a matter
of comparative indifference; but it is essential that we should do so—that
‘a + b’ should always have a reference whatever signs for definite objects
may be inserted in place of ‘a’ and ‘b.’ This involves the requirement as
regards concepts, that, for any argument, they shall have a truth-value as
their value; that it shall be determinate, for any object, whether it falls
under the concept or not. In other words: as regards concepts we have a
requirement of sharp delimitation; if this were not satisfied it would be
impossible to set forth logical laws about them. For any argument x for
which ‘x + 1’ were devoid of reference, the function x + 1 = 10 would
likewise have no value, and thus no truth-value either, so that the concept:
‘what gives the result 10 when increased by 1’
would have no sharp boundaries. The requirement of the sharp delimitation

of concepts thus carries along with it this requirement for functions in
general that they must have a value for every argument.
Comment. Here is one of these rash “it would be impossible”-claims. In view of what
I am going to do in Part G in the groundworks it is worthwhile pointing out that
dialectical logic, as conceived there, is just such a system of logical laws, viz., for
concepts with no sharp delimitation (not fuzzy, though). Compare also quotation
(6) in this lot.
If it is a question of the truth of something—and truth is the goal of logic—
we also have to inquire after meanings; we have to throw aside proper names
that do not designate or name an object, though they may have a sense;
we have to throw aside concept-words that do not have a meaning. These
are not such as, say, contain a contradiction—for there is nothing at all
wrong in a concept’s being empty—but such as have vague boundaries. It
must be determinate for every object whether it falls under a concept or
not; a concept word which does not meet this requirement on its meaning
is meaningless.
Comment. Since Frege’s times, the phrase “truth is the goal of logic” seems to have
been modified to “truth preservation is the goal of logic”.
[[E]]xception must be taken to the statement that the mathematician counts
as impossible only what is self-contradictory. A concept is still admissible
even though its defining characteristics do contain a contradiction: all that
we are forbidden to do, is to presuppose that something falls under it. But
even if a concept contains no contradiction, we still cannot infer that for
that reason something falls under it.
(4) Long and White [1979], p. 179; translating Frege [1906], pp. 193 f.
[[I]]f we ask, under what conditions a concept is admissible in science, the
first thing to stress is that consistency is not such a condition. The only
requirement to be made of a concept is that it should have sharp boundaries;
that is that for every object it holds that it either falls under the concept
or does not do so. Essentially this is nothing but the requirement that
the principle of non-contradiction should hold. But the admissibility of a
concept is entirely independent of the question whether objects fall under
it, and if so which, or in other words[[:]]27 whether there be objects, and if
so which, of which it can be truly asserted. For, before we can raise such
questions, we already need the concept. [[. . .]]
[[. . .]] It is of course self-evident that an object cannot have incon-
sistent properties. That the concept of a right-angled equilateral [[spatial
pentahedron]]28 contains a contradiction does not make it inadmissible. For
27 Semicolon in the translation, colon in the original.
28 This substitution is not just because I am German and pedantic. The word the translator
used evokes an association with a paradigm of obscenity.
we can see no reason why [[one]]29 should not be able to say of an object
that it is not a right-angled equilateral [[spatial pentahedron]], or why [[one]]
should not be permitted to say there are no right-angled equilateral [[spatial
pentahedrons]]. And before [[one]] arrives at such judgments, [[one]] must con-
sider the matter, and to do that [[one]] requires this concept. It is completely
wrongheaded to imagine that every contradiction is immediately recogniz-
able; frequently the contradiction lies deeply buried and is only discovered
by a lengthy chain of inference: throughout which you need the concept.
Comment. This is an important passage with regard to my approach; it is para-
phrased in 113.10.
(5) Austin [1950], pp. 105e –106e; translating Frege [1884], pp. 105 f.
[[By the way, how should one]] prove that a concept does not contain any
contradiction? It is by no means obvious; it does not follow that because we
see no contradiction there is none there, nor does a clear and full definition
afford any guarantee against it.
Comment. The first part of Austin’s translation “If such concepts were not admis-
sible, how could we ever” has been omitted. It has no counterpart in the original;
the corresponding German phrase is “Wie soll man übrigens”.
(6) Geach and Black [1952], p. 159; Geach translating Frege [1903], pp. 69 f.
§ 56
A definition of a concept (of a possible predicate) must be complete;
it must unambiguously determine, as regards any object, whether or not
it falls under the concept (whether or not the predicate is truly assertible
of it). Thus there must not be any object as regards which the definition
leaves in doubt whether it falls under the concept; though for us men, with
our defective knowledge, the question may not always be decidable. We
may express this metaphorically as follows: the concept must have a sharp
boundary. If we represent concepts in extension by areas on a plane, this
is admittedly a picture that may be used only with caution, but here it
can do us good service. To a concept without sharp boundary there would
correspond an area that had not a sharp boundary-line all round, but in
places just vaguely faded away into the background. This would not really
be an area at all; and likewise a concept that is not sharply defined is
wrongly termed a concept. Such quasi-conceptual constructions cannot be
recognized as concepts by logic; it is impossible to lay down precise laws
for them. The law of excluded middle is really just another form of the
requirement that the concept should have a sharp boundary. Any object ∆
that you choose to take either falls under concept Ψ or does not fall under
it; tertium non datur. E.g. would the sentence ‘any square root of 9 is odd’
have a comprehensible sense at all if square root of 9 were not a concept
with a sharp boundary? Has the question ‘Are we still Christians?’ really
got a sense, if it is indeterminate whom the predicate ‘Christian’ can truly
be asserted on, and who must be refused it?
29 The German “man” is translated into English “a man”.
§ 57
Now from this it follows that the mathematicians’ favourite procedure,
piecemeal definition, is inadmissible. The procedure is this: First they give
the definition for a particular case—e.g. for positive integers—and make
use of it; then, many theorems later, there follows a second definition for
another case— e.g. for negative integers and zero—; here they often commit
the further mistake of making specifications all over again for the case
they have already dealt with. Even if in fact they avoid contradictions, in
principle their method does not rule them out. [[. . .]] But the chief mistake
is that they are already using the symbol or word for theorems before it has
been completely defined—often, indeed, with a view to further development
of the definition itself. So long as it is not completely defined, or known in
some other way, what a word or symbol stands for, it may not be used in
an exact science—least of all with a view to further development of its own
definition.
Comment. This is an extremely crucial point; a point which I strongly contest. My

treatment can be found in chapter XXIX in the groundworks. The basic point:
Frege’s realism commits him to full definitions, whereas my dialectical position
commits me to a step by step procedure, where universes are created by the objects,
and new objects are being created by operations which lead out of a universe. Also of
interest in this context: the role of partial functions in the definition of computability
and Kripke’s approach to a definition of truth.
The next couple of quotations concerns the unsaturated nature of a concept

and its basic difference from that of an object.
Quotations 71.17. (1) McGuinness [1984], p. 282, footnote 14; translation of Frege
[1903], p. 373.
In §49 of his book, The Principles of Mathematics I (Cambridge, 1903),
Mr. B. Russell does not want to concede that a concept is essentially dif-
ferent from an object; concepts, too, are always supposed to be terms. He
supports his argument here with the contention that we find it necessary
to use a concept substantively as a term if we want to say anything about
it, e.g. that it is not a term. In my opinion, this necessity is grounded solely
in the nature of our language and therefore is not a properly logical one.
[[. . .]] It is clear that we cannot present a concept as independent, like an
object; rather it can only occur in connection. One may say that it can be
distinguished within, but that it cannot be separated from the context in
which it occurs. All apparent contradictions that one may encounter here
derive from the fact that we are tempted to treat a concept like an object,
contrary to its unsaturated nature. This is sometimes forced upon us by
the nature of our language. Nevertheless, it is merely a linguistic necessity.
Comment. Here is another “merely” so-and-so necessity.

(2) Long and White [1979], pp. 177 f; translating Frege [1906], p. 192.
[[L]]anguage brands a concept as an object, since the only way it can fit
the designation for a concept into its grammatical structure is as a proper
name. But in so doing, strictly speaking it falsifies matters. In the same
way, the word ‘concept’ itself is, taken strictly, already defective, since the
phrase ‘is a concept’ requires a proper name as grammatical subject; and so,
strictly speaking, it requires something contradictory, since no proper name
can designate a concept; or perhaps better still, something nonsensical.
Comment. It is in such context that I see Kant’s formulation in quotation 61.35 (2)
regarding noumena that “we have an understanding which problematically extends
further, but we have no intuition”.
(3) Geach and Black [1952], pp. 115 f; translating Frege [1904], p. 666.
The endeavour to be brief has introduced many inexact expressions into
mathematical language, and these have reacted by obscuring thought and
producing faulty definitions. Mathematics ought properly to be a model of
logical clarity. In actual fact there are perhaps no scientific works where you
will find more wrong expressions, and consequently wrong thoughts, than in
mathematical ones. Logical correctness should never be sacrificed to brevity
of expression. It is therefore highly important to devise a mathematical lan-
guage that combines the most rigorous accuracy with the greatest possible
brevity. To this end a symbolic language would be best adapted, by means
of which we could directly express thoughts in written or printed symbols
without the intervention of spoken language.
71d. The making of “Grundgesetz V”. In an article on function and con-

cept, Frege worked out his theory of a concept as a function and the extension of a
concept as the range of values of a function.
Quotation 71.18. Frege [1891], p. 17 (orig.: 2 f).

[[E]]in bloßer Ausdruck, die Form für einen Inhalt kann das Wesen der Sa-
che nicht sein, sondern nur der Inhalt selbst. Was ist nun der Inhalt, die
Bedeutung von »2 × 23 + 2«? Dieselbe wie von »18« oder von »3 × 6«. In
der Gleichung 2 × 23 + 2 wird ausgedrückt, daß die Bedeutung der rechtsste-
henden Zeichenverbindung dieselbe sei wie die der linksstehenden. Ich muß
hier der Ansicht entgegentreten, daß z. B. 2 + 5 und 3 + 4 zwar gleich, aber
nicht dasselbe seien. Es liegt dieser Meinung wieder jene Verwechslung von
Form und Inhalt, von Zeichen und Bezeichnetem zugrunde.
Comment. This is far away from Hegel’s idea of a dialectic of form and content, or
the “Selbständigkeit der Form”. Frege insists on the principle possibility of separating
form and content. I quote this, because it is this neglect for the independence of the
form which I consider the basic problem in Frege’s “Grundgesetz V”.
Quotations 71.19. (1) Geach and Black [1952], pp. 26 f; Geach translating Frege
[1891], pp. 21 f.
If we write
x2 − 4x = x (x − 4) ,
we have not put one function equal to the other, but only the values of
one equal to those of the other. And if we so understand this equation that
it is to hold whatever argument may be substituted for x, then we have
thus expressed that an equality holds generally. But we can also say: ‘the
value-range of the function x(x− 4) is equal to that of the function x2 − 4x,’
and here we have an equality between ranges of values. The possibility of
regarding the equality holding generally between values of functions as a
[particular] equality, viz. an equality between ranges of values, is, I think,
indemonstrable; it must be taken to be a fundamental law of logic.
We may further introduce a brief notation for the value-range of a
function. To this end I replace the sign of the argument in the expression
for the function by a Greek vowel, enclose the whole in brackets, and prefix
to it the same Greek letter with a smooth breathing. Accordingly, e.g.,
ἐ (ε2 − 4 ε)
is the value-range of the function x2 − 4x and
ἀ (α · [α − 4])
is the value range of the function x (x − 4), so that in
ἐ (ε2 − 4ε) = ἀ (α · [α − 4])
we have the expression for: the first range of values is the same as the
second.
(2) Geach and Black [1952], p. 31; Geach translating Frege [1891], p. 26 (orig.: 16).
[[W]]e can designate as an extension [[of a concept; German: Begriffsumfang]]
the value-range of a function whose value for every argument is a truth-
value.
(3) Geach and Black [1952], p. 31; Geach translating Frege [1891], p. 26 (orig.: 16).
A statement contains no empty place, and therefore we must regard
what it stands for as an object. But what a statement stands for is a truth-
value. Thus the truth-values are objects.
[[. . .]] Value-ranges of functions are objects, whereas functions them-
selves are not. [[. . .]] Extensions of concepts likewise are objects, although
concepts themselves are not.
Comment. The German word for the phrase “what . . . stands for” is “Bedeutung”
(4) Geach and Black [1952], p. 38; Geach translating Frege [1891], p. 34 (orig.: 26 f).
[[J]]ust as functions are fundamentally different from objects, so also func-
tions whose arguments are and must be functions are fundamentally dif-
ferent from functions whose arguments are objects and cannot by anything
else. I call the latter first-level, the former second-level, functions. In the
same way, I distinguish between first-level and second-level concepts.
(5) Furth [1964], p. 36; translating Frege [1893], p. 7.

I use the words
“the function Φ(ξ) has the same course-of-values as the function
Ψ(ξ)”
generally to denote the same as the words
“the functions Φ(ξ) and Ψ(ξ) have always the same value for the same
argument”.
[[. . .]] With [[. . .]] functions, whose value is always a truth-value, one may
[[. . .]] say, instead of “course-of-values of the function”, rather “extension of
the concept”, and it seems appropriate to call directly a concept a function
whose value is always a truth-value.
Although we have laid it down that the combination of signs
“ἐΦ(ε) = ἀΨ(α)”
has the same denotation as
V
“ α(Φ(α) = Ψ(α))”,30
this by no means fixes completely the denotation of a name like “ἐΦ(ε)”. We
have only a means of always recognizing a course-of-values if it is designated
by a name like “ἐΦ(ε)”, by which it is already recognizable as a course-of-
values. But we can neither decide, so far, whether an object is a course-of-
values that is not given us as such, and to what function it may correspond,
nor decide in general whether a given course-of-values has a given property
unless we know that this property is connected with a property of the
corresponding function. If we assume that
X(ξ)
is a function that never takes on the same value for different arguments,
then for objects whose names are of the form
“X(ἐΦ(ε))”
just the same distinguishing mark for recognition holds, as for objects signs
for which are of the form “ἐΦ(ε)”. To wit,
“X(ἐΦ(ε)) = X(ἀΨ(α))”
V
then also has the same denotation as “ α(Φ(α) = Ψ(α))”. From this it
follows
V that by identifying the denotation of “ἐΦ(ε) = ἀΨ(α)” with that of
“ α(Φ(α) = Ψ(α))”, we have by no means fully determined the denotation
of a name like “ἐΦ(ε)”—at least if there does exist such a function X(ξ)
whose value for a course-of-values as argument is not always the same as
the course-of-values itself. How may this indefiniteness be overcome? By its
being determined for every function when it is introduced what values it
takes on for courses-of-values as arguments, just as for all other arguments.
30 The notation of the quantifier has been adapted to the style in this book throughout the
present quotation.

[[W]]ith [[“ἐΦ(ε)”]] we are introducing not merely a new function-name, but
simultaneously answering to every name of a first-level function of one ar-
gument, a new proper name (course-of-values-name); in fact not just for
those [function-names] known already, but in advance for all such that may
be introduced in the future. To the inquiry whether a course-of-values-name
denotes something, we need only subject such course-of-values names as are
formed from denoting names of first-level functions of one argument.
Comment. This sounds somewhat like impredicativity to me; cf. also Thiel [1972],
p. 94. Anyway, this is taken from § 31 of the Grundgesetze in which Frege endeavours
to show that names in his system always have a reference. In view of Russell’s
paradox, this must fail. Bartlett [1961] must be mentioned here, although I can
only do so by hearsay.
(8) Geach and Black [1952], pp. 179 f; Geach translating Frege [1903], pp. 147 f.
What, then, did we do [[by introducing value-ranges]]? or rather, in the
first place, what did we not do? We did not enumerate properties and
then say: we construct a thing that is to have these properties. Rather
we said: If a (first-level) function (of one argument) and another function
are such as always to have the same value for the same argument, then
we may say instead that the range of values of the first is the same as
that of the second. We are then recognizing something common to the two
functions, and we call this the value-range of the first function and also the
value-range of the second function. We must regard it as a fundamental
law of logic that we are justified in thus recognizing something common to
both, and that accordingly we may transform an equality holding generally
into an equation (identity). This transformation must not be regarded as a
definition; neither the word ‘same’ or the equals sign, nor the word ‘value-
range’ or a complex symbol like ‘ἐΦ(ε),’ nor both together, are defined by
means of it.
(9) Geach and Black [1952], pp. 180 f; Geach translating Frege [1903], pp. 148 f.
People have indeed clearly already made use of the possibility of trans-
formation that I have mentioned; only they have asserted coincidence of
functions themselves rather than of value-ranges. [[. . .]]
Logicians have long since spoken of the extension of a concept, and
mathematicians have used the terms set, class, manifold; what lies behind
this is a similar transformation; for we may well suppose that what mathe-
maticians call a set (etc.) is nothing other than an extension of a concept,
even if they have not always been clearly aware of this.
What we are doing by means of our transformation is thus not really
anything novel; but we do it will full awareness, appealing to a fundamental
law of logic. [[. . .]]
If there are logical objects at all—and the objects of arithmetic are such
objects—then there must also be a means of apprehending, of recognizing,
them. This service is performed for us by the fundamental law of logic
that permits the transformation of an equality holding generally into an

equation. Without such a means a scientific foundation for arithmetic would
be impossible.
Comment. Notice: a set “is nothing other than an extension of a concept”.
Quotations 71.20. (1) Furth [1964], p. 6; translating Frege [1893], p. IX.

The primitive signs used in Begriffsschrift occur here [[in the Grundgesetze]]
also, with one exception. Instead of the three parallel lines I have adopted
the ordinary sign of equality, since I have persuaded myself that it has in
arithmetic precisely the meaning that I wish to symbolize. That is, I use
the word “equal” to mean the same as “coinciding with” or “identical with”;
and the sign of equality is actually used in arithmetic in this way. The
opposition that may arise against this will very likely rest on an inadequate
distinction between sign and thing signified.
Comment. This is a crucial point. In my dialectical logic I distinguish between
equality and identity, and I do so on the grounds of the failure of Frege’s “Grundge-
setz V”, the very law for the justification of which Frege employed the distinction
between sign and thing signified.
(2) Furth [1964], pp. 6 f. translating Frege [1893], p. X.
Formerly I distinguished two components in that whose external form is
a declarative sentence: (1) the acknowledgment of truth, (2) the content
that is acknowledged to be true. The content I called a ‘possible content
of judgment’. This last has now split for me into what I call ‘thought’
and ‘truth-value’, as a consequence of distinguishing between sense and
denotation of a sign. In this case the sense of a sentence is a thought, and
its denotation a truth-value.
(3) Furth [1964], p. 6; translating Frege [1893], p. X.
To the old primitive signs two more have now been added: the smooth
breathing, for the notation for the course-of-values of a function, and a
sign meant to do the work of the definite article of everyday language. The
introduction of the courses-of-values of functions is a vital advance, thanks
to which we gain far greater flexibility.
(4) Furth [1964], p. 45; translating Frege [1893], pp. 15 f.
The introduction of a notation for courses-of-values seems to be to
be one of the most important supplementations that I have made of my
Begriffsschrift since my first publication on this subject.
[[W]]e can transform the generality of an identity into an identity of courses-
of-values and vice versa. This possibility must be regarded as a law of logic, a
law that is invariably employed, even if tacitly, whenever discourse is carried
on about the extensions of concepts. The whole Leibniz-Boole calculus of
logic rests upon it.
(6) Furth [1964], pp. 49 f; translating Frege [1893], p. 19.

[[W]]e can serve our purpose by introducing the function
\\ξ
with the stipulation that two cases are to be distinguished:

1. If to the argument there corresponds an object ∆ such that the ar-
gument is ἐ(∆ = ε), then let the value of the function \\ξ be ∆ itself;
2. if to the argument there does not correspond an object ∆ such that
the argument is ἐ(∆ = ε), then let the value of the function be the
argument itself.
Accordingly \\ἐ(∆ = ε) = ∆ is the True, and “\ \ἐΦ(ε)” denotes the object
falling under the concept Φ(ξ) if Φ(ξ) is a concept under which falls one
and only one object; in all other cases “\
\ἐΦ(ε)” denotes the same as “ἐΦ(ε)”.
Comment. This is Frege’s “substitute for the definite article”.
Remark 71.21. Frege used a spiritus lenis to indicate the extension of a concept
Φ: ἐΦ(ε)
Quotations 71.22. (1) Geach and Black [1952], p. 147; translation of Frege [1893],
p. XXVI.
[[T]]he whole of the second part is really a test of my logical convictions.
[[. . .]] Those who have other convictions have only to try to erect a similar
construction upon them, and they will soon be convinced that it is not
possible, or at least is not easy. As a proof of the contrary, I can only admit
the production by some one of an actual demonstration that upon other
fundamental convictions a better and more durable edifice can be erected,
or the demonstration by some one that my premises lead to manifestly false
conclusions. But nobody will be able to do that.
Comment. This sound like Frege was quite convinced of his approach. He antici-
pated, however, the place where trouble might arise, as the next quotation shows.
(2) Geach and Black [1952], p. 138; translation of Frege [1893], p. VII.
A dispute can only arise, so far as I can see, because of my fundamental
law about ‘ranges of values,’ which perhaps has not yet been specifically
expressed by logicians, though it is in their minds when, e.g., they speak
of extensions of concepts. I hold that this is purely logical. In any case the
place is indicated where the decision has to be made.
Comment. This became the place, indeed, where the decision was made. When
Frege wrote this he was blissfully unaware of the complications that were to arise.
It does not only form the background to the following chapters XIX, XX, and XXI,
but also forms a cornerstone of my theory of dialectic and features prominently in
chapter XXVII in the groundworks.
(3) Geach and Black [1952], p. 137; translation of Frege [1893], p. VI.
In order to secure more flexibility and not fall into excessive prolixity, I
have taken the liberty of making tacit use of the interchangeability of the
sub-clauses (conditions) and of the possibility of amalgamating identical
sub-clauses[[.]]
Comment. In other words, a form of exchange and contraction. Now the latter, in
particular, is the point where the dispute arises with regard to Frege’s “Grundgesetz
V”, in the sense that both together, contraction and “Grundgesetz V”, cannot be
allowed on pain of inconsistency.
71e. On the philosophical significance of the logical turn in math-

ematics and Frege’s logicism. The positions of Dedekind and Frege are quite
different, despite the fact that they both emphasize the logical nature of arithmetic.
Although it is mainly Frege’s philosophical ideas which have come to be discussed
in philosophy, what is relevant for my enterprise is closer in character to Dedekind’s
analysis of number (ordinal).
I begin with a quotation which is not easily linked into, but which is nevertheless
of great importance for my approach to speculative philosophy.
Quotation 71.23. Beman [1909], p. 36; translating Dedekind [1888], p. VI.

[[T]]he greatest and most fruitful advances in mathematics and other sci-
ences have invariably been made by the creation and introduction of new
concepts, rendered necessary by the frequent recurrence of complex phe-
nomena which could be controlled by the old notions only with difficulty.
Comment. I am inclined to think that the step from a barter economy to a money
economy belongs into the same mould, although it may be quite contentious what
is to be understood by the phrase “rendered necessary by the frequent occurrence
of complex phenomena”. What I see as characteristic in such a step is a kind of
totalization over all possible acts of a certain kind. In the groundworks, I shall
introduce necessity as a kind of totalization over any finite number of repetitions of
assumptions of the same form.
In view of the label ‘logicism’ and its particular relevance for a brand of empiri-
cism, it is important to point out, firstly, that Frege never used this label himself,
and secondly, his logicism was restricted to arithmetic, i.e., did not extend to geo-
metry.
Quotations 71.24. (1) Austin [1950], pp. 111e –112e; translating Frege [1884],
pp. 111 f.
§ 102. It is common to proceed as if a mere postulation were equivalent
to its own fulfilment. We postulate that it shall be possible in all cases to
carry out the operation of subtraction, or of division, or of root extraction,
and suppose that with that we have done enough. But why do we not pos-
tulate that through any three points it shall be possible to draw a straight
line? Why do we not postulate that all the laws of addition and multiplica-
tion shall continue to hold for a three-dimensional complex number system
just as they do for real numbers? Because this postulate contains a con-
tradiction. Very well then, what we have to do first is to prove that these
other postulates of ours do not contain any contradiction. Until we have
done that, all rigour, strive for it as we will, is so much moonshine.
Comment. Very well then, what about Frege’s own system of postulates; or does he
feel exempted — because he is dealing with logic, perhaps? Or because he thought
he had established that names in his system always have a reference?31
(2) Geach and Black [1952], pp. 144 f; translating Frege [1893], pp. XIII f.
[[I]]t is of importance to make clear what definition is and what we can
reach by means of it. It is, it seems often credited with a creative power;
but really all there is to definition is that something is brought out, precisely
limited and given a name. The geographer does not create a sea when he
draws border lines and says: The part of the surface of the ocean, delimited
by these lines, I am going to call the Yellow Sea; and not more can the
mathematician really create anything by his act of definition. Nor can we
by mere definition magically give to a thing a property which it has not
got, apart from the property of now being called by whatever name one has
given it. But that an oval drawn on paper with pen and ink should acquire
by definition the property that, when it is added to one, one is the result,
I can only regard as scientific superstition. [[ ]]32 One might just as well
make a lazy pupil diligent by a mere definition. Confusion easily arises here
through our not making a sufficient distinction between concept and object.
If we say: ‘A square is a rectangle in which the adjacent sides are equal,’
we define the concept square by specifying what properties something must
have in order to fall under this concept. I call these properties ‘marks’ of
the concept. But it must be carefully noted that these marks of the concept
are not properties of the concept. The concept square is not a rectangle;
only the objects which fall under this concept are rectangles; similarly the
concept black cloth is neither black nor a cloth. Whether such objects exist
is not immediately known by means of their definitions. Now, for instance,
suppose we try to define the number zero by saying: ‘It is something which
when added to one gives the result one.’ With that we have defined a concept
by stating what property an object must have to fall under the concept.
But this property is not a property of the concept defined. It seems that
people often imagine that we have created by our definition something which
when added to one gives one. This is a delusion. Neither has the concept
defined got this property, nor is the definition a guarantee that the concept
is realized. That must first of all be a matter for investigation. Only when
we have proved that there exists one object and one only with the required
property are we in a position to give this object the proper name ‘zero.’ To
create zero is consequently impossible. I have already repeatedly explained
this but, as it seems, without result.
31 Cf. quotation 71.19 (7) above (including comment).
32 The translation is broken up into two paragraphs here, a break which can neither be found
in Frege’s original text nor does it make sense as regards the flow of thought.
Comments. (α) In the formalized theory of arithmetic zero is a primitive symbol.

Are we allowed to introduce logical connectives by definition?
(β) Frege apparently gives priority to the notion of concept against description (λ-
over ι-operator, or ε-operator; the ι-operator can be represented in some way in
the ideal calculus anyway). This is noteworthy in view of the various attempts to
locate the source of contradictions, in particular Kant’s and Russell’s.33 Also: Frege’s
analysis of concept in terms of function and Church’s λ-calculus (as opposed to an
iterative concept of set).
Quotations 71.25. (1) Geach and Black [1952], pp. 186 f; translating Frege [1903],
p. 100.
Whereas in meaningful arithmetic equations and inequations are sen-
tences expressing thoughts, in formal arithmetic they are comparable with
the positions of chess pieces, transformed in accordance with certain rules
without consideration for any sense. For if they were viewed as having a
sense, the rules could not be arbitrarily stipulated; they would have to be
so chosen that from formulas expressing true propositions could be derived
only formulas likewise expressing true propositions. Then the standpoint
of formal arithmetic would have been abandoned, which insists that the
rules for the manipulation of signs are quite arbitrarily stipulated. Only
subsequently may one ask whether the signs can be given a sense compati-
ble with the rules previously laid down. Such matters, however, lie entirely
outside formal arithmetic and only arise when applications are to be made.
Then, however, they must be considered; for an arithmetic with no thought
as its content will also be without possibility of application. Why can no
application be made of a configuration of chess pieces? Obviously, because
it expresses no thought. If it did so and every chess move conforming to the
rules correspond to a transition from one thought to another, applications
of chess would also be conceivable. Why can arithmetical equations be ap-
plied? Only because they express thought. How could we possibly apply an
equation which expressed nothing and was nothing more than a group of
figures, to be transformed into another group of figures in accordance with
certain rules. Now, it is applicability alone which elevates arithmetic from
a game to the rank of a science. So applicability necessarily belongs to it.
Is it good, then, to exclude from arithmetic what it needs in order to be a
science?
(2) McGuinness and Kaal [1980], p. 33; translating Frege [1976], pp. 58 f; from a
letter to Hilbert, 1.10.1895.
[[T]]he use of symbols must not be equated with a thoughtless, mechanical
procedure, although the danger of lapsing into a mere mechanism of formu-
las is more immediate here than with the use of words. [[. . .]] Yet I would
not want to regard such a mechanism as completely useless or harmful. On
the contrary, I believe that it is necessary. The natural course of events
33 Cf. my appropriation of Kant’s decision regarding the ‘conflict of reason with itself’ in the
third book.
seems to be as follows: what was originally saturated with thought hardens

in time into a mechanism which partly relieves the scientist from having to
think. Similarly, in playing music, a series of processes which were originally
conscious must have become unconscious and mechanical so that the artist,
unburdened of these things, can put his heart into the playing. I should
like to compare this to the process of lignification. Where a tree lives and
grows it must be soft and succulent. But if what was succulent did not in
time turn into wood, the tree could not reach a significant height. On the
other hand, when all that was green has turned into wood the tree ceases
to grow.
Quotations 71.26. (1) Austin [1950], pp. 101e f; translating Frege [1884], pp. 101 f.
I consider Kant did great service in drawing the distinction between syn-
thetic and analytic judgements. In calling the truths of geometry synthetic
and a priori, he revealed their true nature. And this is still worth repeat-
ing, since even to-day it is often not recognized. If Kant was wrong about
arithmetic, that does not seriously detract, in my opinion, form the value of
his work. His point was, that there are such things as synthetic judgements
a priori; whether they are to be found in geometry only, or in arithmetic as
well, is of less importance.
I must [[. . .]] protest against the generality of Kant’s dictum: without sen-
sibility no object would be given to us. Nought and one are objects which
cannot be given to us in sensation.
Comment. This strikes me as a clear rejection of the empiricist dictum nihil est in
intellectu quod non prius fuerit in sensu.
In [[some]] respect existence is analogous to number. Affirmation of exis-
tence is in fact nothing but denial of the number nought. Because existence
is a property of concepts the ontological argument for the existence of God
breaks down.
The next issue is that of Frege’s so-called realism. The following quotations are
meant to evoke a cautious attitude towards general formulations of what kind of
realist Frege was.
Quotations 71.27. (1) Furth [1964], pp. 23 f; translating Frege [1893], p. XXIV.
If we want to emerge from the subjective at all, we must conceive of knowl-
edge as an activity that does not create what is known but grasps what is
already there. The picture of grasping is very well suited to elucidate the
matter. If I grasp a pencil, many different events take place in my body:
nerves are stimulated, changes occur in the tension and pressure of muscles,
tendons, and bones, the circulation of the blood is altered. But the totality
of these events neither is the pencil nor creates the pencil; the pencil exists
independently of them. And it is essential for grasping that something be
there which is grasped; the internal changes alone are not the grasping. In
the same way, that which we grasp with the mind also exists independently
of this activity, independently of the ideas and their alterations that are
part of this grasping or accompany it; and it is neither identical with the
totality of these events nor created by it as a part of our own mental life.
Comment. This is a fairly strong claim and in view of what I shall do later it is
worth giving it some attention. What Frege seems to claim is that an object has to
exist independently and beforehand if non-subjective knowledge is to be possible.
Such a formulation remains incomplete, however, if it is not specified of what it is
independent.
I distinguish what I call objective from what is handleable or spatial or
actual. The axis of the earth is objective, so is the centre of mass of the solar
system, but I should not call them actual in the way the earth itself is so.
We often speak of the equator as an imaginary line; but it would be wrong
to call it an imaginary line in the dyslogistic sense; it is not a creature of
thought, the product of a psychological process, but is only recognized or
apprehended by thought.
It is in this way that I understand objective to mean what is independent of
our sensation, intuition and imagination, and of all construction of mental
pictures out of memories of earlier sensations, but not what is independent
of the reason, — for what are things independent of the reason? To answer
this would be as much as to judge without judging, or to wash the fur
without wetting it.
Comments. (α) This is an extremely interesting point in view of Kant’s concept of a
noumenon: Frege too seems to feel the need for a distinction between intuition and
reason (if I may neglect for the moment the distinction Kant makes between un-
derstanding and reason). Observe the close similarity between Frege’s formulation
(judge without judging; washing the fur without wetting it) and Hegel’s formulation
in quotation 65.2 (1) regarding the question “what the thing in itself ” is: determi-
nations without any determination; or more general, with Berkeley: conceiving the
inconceivable. See also variations 106.5 in the groundworks.
(β) Apparently, Frege isn’t quite as radical a realist as he is sometimes portrayed:
his notion of objective means independence of psychological factors, but not inde-
pendence of reason. But then, where and how does the dependence of reason make
itself felt?
(4) Furth [1964], pp. 15 f; translating Frege [1893], p. XVIII.
[[F]]or me there is a domain of what is objective, which is distinct from that
of what is actual, whereas the psychological logicians without ado take what
is not actual to be subjective. And yet it is quite impossible to understand
why something that has a status independent of the judging subject has
to be actual, i.e., has to be capable of acting directly or indirectly on the
senses. No such connection is to be found between the concepts [of being

objective and being actual][[.]]
(5) Furth [1964], pp. 24 f; translating Frege [1893], p. XXV.
While the mathematician defines objects, concepts, and relations, the psy-
chological logician is spying upon the origin and evolution of ideas, and to
him at bottom the mathematician’s defining can only appear foolish because
it does not reproduce the essence of ideation. He looks into his psychological
peep-show and tells the mathematician: “I see nothing at all of what you
are defining.” And the mathematician can only reply: “No wonder, for it is
not where your are looking for it.”
Comment. Husserl springs to mind, to mine at least.
The question “what is expressed in a statement of equality?” is of the central

relevance for Frege’s approach. The classic text is Sinn und Bedeutung (translated as
“On Sense and Reference” in Geach and Black [1952]) and the first two introductory
paragraphs are of such ‘beauty and clarity’ that I am going to quote them in full,
albeit in their translation — partly as a warning against a criterion of ‘beauty and
clarity’ in philosophy, because this is the point where Frege begins to pave his way
into paradox.34
Quotations 71.28. (1) Geach and Black [1952], pp. 56 f; translation of Frege
[1892b], pp. 25 ff.
Equality gives rise to challenging questions which are not altogether easy
to answer. Is it a relation? A relation between objects, or between names
or signs of objects? In my Begriffsschrift I assumed the latter. The reasons
which seem to favour this are the following: a = a and a = b are obviously
statements of differing cognitive value; a = a holds a priori and, according
to Kant, is to be labelled analytic, while statements of the form a = b often
contain valuable extensions of our knowledge and cannot be established a
priori. The discovery that the rising sun is not new every morning, but
always the same, was one of the most fertile astronomical discoveries. Even
to-day the identification of a small planet or a comet is not always a matter
of course. Now if we were to regard equality as a relation between that
which the names ‘a’ and ‘b’ designate, it would seem that a = b could not
differ from a = a (i.e. provided a = b is true). A relation would thereby be
expressed of a thing to itself, and indeed one in which each thing stands to
itself but to no other thing. What is intended to be said by a = b seems to
be that the signs or names ‘a’ and ‘b’ designate the same thing, so that those
signs themselves would be under discussion; a relation between them would
be asserted. But this relation would hold between the names or signs only
in so far as they named or designated something. It would be mediated by
the connexion of each of the two signs with the same designated thing. But
this is arbitrary. Nobody can be forbidden to use any arbitrarily producible
34 As regards the failure of extensionality in (higher order) logic, cf. theorem 126.78 on p. 1735
in the groundworks.
event or object as a sign for something. In that case the sentence a = b would
no longer refer to the subject matter, but only to its mode of designation;
we would express no proper knowledge by its means. But in many cases
this is just what we want to do. If the sign ‘a’ is distinguished from the
sign ‘b’ only as object (here, by means of its shape), not as sign (i.e. not
by the manner in which it designates something), the cognitive value of
a = a becomes essentially equal to that of a = b, provided a = b is true.
A difference can arise only if the difference between the signs corresponds
to a difference in the mode of presentation of that which is designated. Let
a, b, c be the lines connecting the vertices of a triangle with the midpoints
of the opposite sites. The point of intersection of a and b is then the same
as the point of intersection of b and c. So we have different designations for
the same point, and these names (‘point of intersection of a and b,’ ‘point
of intersection of b and c’) likewise indicate the mode of presentation; and
hence the statement contains actual knowledge.
It is natural, now, to think of there being connected with a sign (name,
combination of words, letter), besides that to which the sign refers, which
may be called the reference of the sign, also what I should like to call the
sense of the sign, wherein the mode of presentation is contained. In our
example, accordingly, the reference of the expressions the point of intersec-
tion of a and b’ and ‘the point of intersection of b and c’ would be the same,
but not their senses. The reference of ‘evening star’ would be the same as
that of ‘morning star,’ but not the sense.
Comments. (α) The italicizing of “this is arbitrary” is mine; it marks my point of
emphasis. Obviously, in this generality the formulation is untenable: we must not
use a symbol which is already in use otherwise.35
(β) Re “It is natural, now, . . .”: this phrase may be very English indeed, but I don’t
find it congenial to Frege’s thought; Frege’s concern just wasn’t anything “natural”;
the German original says: “Es liegt nun nahe . . .” which to my mind does not carry
the undertone of coercive conformity of the term natural.
(γ) Note the similarity of the phrase “mode of presentation”, German “die Art des
Gegebenseins”, to Kant’s “Vorstellungsarten”, also translated as modes of represen-
tation in Carus and Ellington [1977], p. 82.36
(2) Geach and Black [1952], p. 60; translating Frege [1892b], p. 42 (30, original).
The reference of a proper name is the object itself which we designate
by its means; the idea [[Vorstellung]] which we have in that case is wholly
subjective; in between lies the sense, which is indeed no longer subjective
like the idea, but is yet not the object itself.
(3) Geach and Black [1952], p. 63; translating Frege [1892b], pp. 45 f (33, original).
The thought loses value for us as soon as we recognize that the reference of
one of its parts is missing. We are therefore justified in not being satisfied
with the sense of a sentence, and in inquiring also as to its reference. But
35 Compare e.g., Curry et al. [1958], p. 20, quotation 78.5 (2) in these materials.
now why do we want every proper name to have not only a sense, but also
a reference? Why is the thought not enough for us? Because, and to the
extent that, we are concerned with its truth value. [[. . .]] It is the striving
for truth that drives us always to advance from the sense to the reference.
Comment. Aren’t we wonderful? The crown of the creation: striving for truth.
Mostly academics, I suppose, and mostly while they rush from one conference to
another. Always in hot pursuit of the truth. Anyway, it should be clear that I regard
this as the point where Frege is plastering over the gap where I hope to find the
basis for a notion of truth.
(4) Black [1955], p. 64; translating Frege [1892b], pp. 47 f (35, original).
If our supposition that the reference of a sentence is its truth value is correct,
the latter must remain unchanged when a part of the sentence is replaced
by an expression having the same reference. And this is in fact the case.
Leibniz gives the definition: ‘Eadem sunt, quae sibi mutuo substitui possunt,
salva veritate.’
Comment. This is indeed the question, and the whole point of my approach to
dialectical logic is to interpret Frege’s failure as a failure of this very doctrine enun-
ciated here.
(5) Black [1955], p. 65; translating Frege [1892b], p. 48 (35, original).
If now the truth value of a sentence is its reference, then on the one hand
all true sentences have the same reference and so, on the other hand, do all
false sentence. From this we see that in the reference of the sentence all that
is specific is obliterated. We can never be concerned only with the reference
of a sentence; but again the mere thought alone yields no knowledge, but
only the thought together with its reference, i.e. its truth value. Judgments
can be regarded as advances from a thought to a truth value.
Comment. On the one hand Frege says “We can never be concerned only with the
reference of a sentence”, but on the other hand he has an axiom of extensionality for
concepts37 and a form of unrestricted abstraction. This mixture, however, achieves
just the obliteration he seems to be keen to avoid.38
Quotations 71.29. (1) Geach and Black [1952], pp. 58 f; translation of Frege
[1892b], p. 28.
If words are used in the ordinary way, what one intends to speak of is
their reference. It can also happen, however, that one wishes to talk about
the words themselves or their sense. This happens, for instance, when the
words of another are quoted. One’s own words then first designate words
of the other speaker, and only the latter have their usual reference. We
then have signs of signs. In writing, the words are in this case enclosed in
quotation marks. Accordingly, a word standing between quotation marks
must not be taken as having its ordinary reference.
37Cf. quotation 71.29 (3) below.
38As regards the failure of extensionality in conjunction with unrestricted abstraction without
tertium non datur, cf. theorem 126.78 in the groundworks.
In order to speak of the sense of an expression ‘A’ one may simply use
the phrase ‘the sense of the expression“A” ’. In reported [[ungerade]] speech
one talks about the sense, e.g., of another person’s remarks. It is quite clear
that in this way of speaking words do not have their customary reference
but designate what is usually their sense. In order to have a short expres-
sion, we will say: In reported [[ungerade]] speech, words are used indirectly
[[ungerade]] or have their indirect [[ungerade]] reference. We distinguish ac-
cordingly the customary from the indirect [[ungerade]]; and its customary
sense from its indirect [[ungerade]] sense.
Comment. The German “ungerade” is commonly translated as “odd”, or, perhaps,
“uneven”. It is certainly not common usage to apply it to speech, sense or the use of
words, unlike “reported speech” in English. To my mind its meaning in the present
context would be better captured by “non-straight” since “gerade” means straight,
in any case something that isn’t quite as ordinary as “reported speech” in English
to preserve some of the peculiarity of Frege’s choice of words. I added the German
word to point out that in the translation there occur different words where there is
one and the same in the original.
(2) Geach and Black [1952], p. 68; translation of Frege [1892b], p. 51 f.
Dependent clauses expressing questions and beginning with ‘who,’ ‘what,’
‘where,’ ‘when,’ ‘how,’ ‘by what means,’ etc., seem at times to approxi-
mate very closely to adverbial clauses in which words have their customary
references. Theses cases are distinguished linguistically [in German] by the
mood of the verb. With the subjunctive, we have a dependent question and
indirect reference of the words, so that a proper name cannot in general be
replaced by another name for the same object.
Comment. Two things: firstly, the proximity of the words listed to the categories of
Aristoteles; secondly, the mention of the “indirect reference” (failure of substitutivity
of proper names). Preparation for my view of categories as intensional. Compare
also: Montague and Kalish, quotation 90.19 (1).
A concept-word means a concept, if the word is used as is appropriate
for logic. I may clarify this by drawing attention to a fact that seems to
weigh heavily on the side of extensionalist as against intensionalist logicians:
namely, that in any sentence we can substitute salva veritate one concept-
word for another if they have the same extension, so that it is also the case
that in relation to inference, and where the laws of logic are concerned,
that concepts differ only in so far as their extensions are different. The
fundamental logical relation is that of an object’s falling under a concept:
all relations between concepts can be reduced to this. If an object falls
under a concept, it falls under all concepts with the same extension, and
this implies what we said above. Therefore just as proper names can replace
one another salva veritate, so too can concept-words, if their extension is
the same.
Comment. Here is a crucial point for my theory of dialectic: substitutivity salva ver-
itate for concepts of the same extension does not hold when unrestricted abstraction
is available.
[[A]]lthough the relation of equality can only be thought of as holding for
objects, there is an analogous relation for concepts. Since this is a relation
between concepts I call it a second level relation, whereas the former relation
I call a first level relation. We say that an object a is equal to an object b
(in the sense of completely coinciding with it) if a falls under every concept
under which b falls, and conversely. We obtain something corresponding to
this for concepts if we switch the roles of concept and object. We could
then say that the relation we had in mind above holds between the concept
Φ concept X, if every object that falls under Φ also falls under X, and
conversely.
(5) Long and White [1979], pp. 122 f; translating Frege [1892–95], p. 133.
[[W]]e shall be well able to assert ‘what two concept-words mean is the
same if and only if the extensions of the corresponding concepts coincide’
without being led astray by the improper use of the word ‘the same’. And
with this statement we have, I believe, made an important concession to the
extensionalist logicians. They are right when they show by their preference
for the extension, as against the intension of a concept that they regard the
meaning and not the sense of the words as the essential thing for logic. The
intensionalist logicians are only too happy not to go beyond the sense; for
what they call the intension, if it is not an idea, is nothing other than the
sense. They forget that logic is not concerned with how thoughts, regardless
of truth-value, follow from thought, that the step from thought to truth-
value—more generally, the step from sense to meaning—has to be taken.
They forget that the laws of logic are first and foremost laws in the realm
of meanings and only relate indirectly to sense. [[. . .]]
[[. . .]] The meaning [[Bedeutung]] is [[. . .]] shown at every point to be the
essential thing for science. Therefore even if we concede to the intensionalist
logicians that it is the concept as opposed to the extension that is the
fundamental thing, this does not mean that it is to be taken as the sense
of a concept-word: it is its meaning, and the extensionalist logicians come
closer to the truth in so far as they are presenting—in the extension—a
meaning as the essential thing.
Comment. This is mainly regarding FR5, the extensional position: extensions are
the essential things in logic (and science).
(6) Geach and Black [1952], p. 70; translating Frege [1892b], pp. 53 f.
A logically perfect language (Begriffsschrift) should satisfy the conditions,
that every expression grammatically well constructed as a proper name out
of signs already introduced shall in fact designate an object, and that no
new sign shall be introduced as a proper name without being secured a
reference.
§ 72. THE DISCOVERY OF PARADOXES 1037
Comment. This is where I want to move in with my concept of logical experience;

somewhat in the sense of Hegel in quotation 66.23 (2) in these materials.
(7) Long and White [1979], p. 125; translating Frege [1892–95], pp. 135 f.
Logic must demand not only of proper names but of concept-words as well
that the step from the word to the sense and from the sense to the meaning
be determinate beyond any doubt. Otherwise we should not be entitled to
speak of a meaning at all. Of course this holds for all signs and combinations
of signs with the same function as proper names or concept-words.
Quotations 71.30. (1) Geach and Black [1952], p. 151; translation of Frege [1893],
p. 4.
The frequent use made of quotation marks may cause surprise. I use
them to distinguish the cases where I speak about the sign itself from
those where I speak about what it stands for. Pedantic as this may appear,
I think it necessary. It is remarkable how an inexact mode of speaking or
writing, which perhaps was originally employed only for greater convenience
or brevity and with full consciousness of its inaccuracy, may end in a confu-
sion of thought, when once that consciousness has disappeared. People have
managed to mistake numerals for numbers, names for the things named, the
mere devices of arithmetic for its proper subject-matter. Such experiences
teach us how necessary it is to demand the highest exactness in manner of
speech and writing.
(2) Geach and Black [1952], p. 151; translation Frege [1893], pp. 3 f.
It is not possible to give a regular definition of everything; [[for]] it must be
our endeavour to go back to what is logically simple and as such cannot
properly defined. I must then be satisfied with indicating by hints what I
mean.
Comment. I replaced “so” in the translation of Jourdain and Stachelroth by “for”
which I find more appropriate for the German “weil”.
Remark 71.31. Independently of Frege, Bertrand Russell came to essentially the

same conclusion that numbers are classes of classes. In this respect it is worthwhile
having a look at Russell [1903], sections 109–111; also appendix A, in which Russell
expounds Frege’s doctrine.
§ 72. The discovery of paradoxes
In the appendix A1 readers will find formal versions of some of the antinomies which
occurred in the logical foundation of set theory. The present paragraph is devoted
to a more historical presentation of this puzzling phenomenon, which is naturally
less rigorous but perhaps more suggestive. It is meant to provide a background for
a presentation of different reactions to the paradoxes in the next chapter. I wish to
point out, however, that it is in no way intended to be representative.
Remark 72.1. According to Copi [1971], p. 1, the “first of the modern paradoxes
to be published was that of Cesare Burali-Forti in 1897.” The reason that it is
not represented here is simply that it strikes me as considerably more complex in
its basic ideas and it doesn’t lend itself to a simplification of the kind that makes
Russell’s antinomy so attractive. Interested readers are referred to Russell [1908],
p. 61, for a sketch of the paradox. There is still room for the hope that the antinomy
of Burali-Forti (ordinals), and perhaps also that of the cardinality of the universal
class, can provide a link to the cosmological antinomies of Kant’s Critique of Pure
Reason, in so far as the absolute totality of a hierarchy is considered.39
72a. From transfinite cardinals to Russell’s antinomy. It is more than

one hundred years now that Georg Cantor discovered a strange result about infinite
sets. Whereas the odd and even numbers can be related in a one-to-one correspon-
dence to the natural numbers, and the natural numbers in turn can be related in a
one-to-one correspondence to the rational numbers and even to the algebraic num-
bers, there is no such one-to-one correspondence between the natural numbers and
the real numbers.
The above result can be extended to the power set of every set: there is no
one-to-one correspondence between the power set P(M) of a set M and the set M
itself.40 This means, in effect, that the infinite is well discriminated, similar to the
finite. There are larger and smaller infinite sets; there is a hierarchy (!) of cardinals.
Accordingly, Cantor developed an arithmetic of infinitely large number.
This, however, made a number of people very angry. Consequently, Cantor’s
treatment of infinite sets was heavily attacked for metaphysical reasons: the infinite
is conceived as actual. Probably many a mathematician felt satisfied, at least in-
wardly, when Cantor’s allegedly hazardous dealing with the concept of infinity was
punished with antinomies.
Quotations 72.2. (1) Van Heijenoort [1967b], p. 125; from a translation of Russell’s
letter to Frege, 16 June 1902.
Let w be the predicate: to be a predicate that cannot be predicated of
itself. Can w be predicated of itself? From each answer its opposite follows.
Therefore we must conclude that w is not a predicate. Likewise there is no
class (as a totality) of those classes which, each taken as a totality, do not
belong to themselves. From this I conclude that under certain circumstances
a definable collection does not form a totality.
Comment. Right at the end of the letter Russell included a symbolic formulation:
The above contradiction, when expressed in Peano’s ideography, reads
as follows:
w = cls ∩ x (x ∼ǫ x) .⊃: w ǫ w .=. w ∼ǫ w .
Comment. The interested reader may wish to go to Russell [1903], section 78, as
well as chapter X, for further reading regarding the early phase of the discovery of
39 Cf. Hessenberg in quotations 73.18 (1) and (2) below.
40 For a proof of this result see theorem 2.66 in the tools.
paradoxes. Frege’s reformulation of the paradox in a modern terminology can be

found in section 141b in appendix A1.
(2) Russell [1908], p. 60; and Whitehead and Russell [1910], p. 60.
Let w be the class of all those classes which are not members of themselves.
Then, whatever class x may be, “x is a w” is equivalent to “x is not an x”.
Hence, giving to x the value w, “w is a w” is equivalent to “w is not a w.”
Comment. This is generally referred to as “Russell’s antinomy”; it seems to have
been first discussed by Zermelo.
(3) Whitehead and Russell [1910], p. 60.
Let T be the relation which subsists between two relations R and S when-
ever R does not have the relation R to S. Then, whatever relations R and
S may be, “R has the relation T to S” is equivalent to “R does not have
the relation R to S.” Hence, giving the value T to both R and S, “T has
the relation T to T ” is equivalent to “T does not have the relation T to T .”
(4) Russell [1903], p. 101.
I was led to [[the paradox]] in the endeavour to reconcile Cantor’s proof that
there can be no greatest cardinal number with the very plausible supposition
that the class of all terms (which we have seen to be essential to all formal
propositions) has necessarily the greatest possible number of members.
Comment. A similar remark can be found in Russell’s letter to Frege, dated 24.
June 1902.
72b. Paradoxes proliferating. Once the news of Russell’s antinomy was

out, it seems, new ones popped up everywhere. While the set theoretical antinomies
appeared in a reasonably precise form right from the beginning, there were other
antinomies which only existed, at least for quite some time, in a natural language
form.
Quotations 72.3. (1) Russell [1908], p. 60; presenting Berry’s paradox.41 Also:
Whitehead and Russell [1910], p. 61.
The number of syllables in the English names of finite integers tends to in-
crease as the integers grow larger, and must gradually increase indefinitely,
since only a finite number of names can be made with a given finite num-
ber of syllables. Hence the names of some integers must consist of at least
nineteen syllables, and among these there must be a least. Hence ‘the least
integer not nameable in fewer than nineteen syllables’ must denote a defi-
nite integer; in fact, it denotes 111,777. But ‘the least integer not nameable
in fewer than nineteen syllables’ is itself a name consisting of eighteen syl-
lables; hence the least integer not nameable in fewer than nineteen syllables
can be named in eighteen syllables, which is a contradiction.
41 The last sentence of this quotation has the following footnote: “This contradiction was
suggested to us by Mr G. G. Berry of the Bodleian Library.”

(2) Van Heijenoort [1967b], p. 143; translating Richard [1905], p. 542

Let us write all permutations of the twenty-six letters of the French
alphabet taken two at a time, putting these permutations in alphabetical
order; then, after them all permutations taken three at a time, in alpha-
betical order; then, after them, all permutations taken four at a time, and
so forth. These permutations may contain the same letter repeated several
times; they are permutations with repetitions.
For any integer p, any permutation of the twenty-six letters taken p
at a time will be in the table; and, since everything that can be written
with finitely many word is a permutation of letters, everything that can be
written will be in the table formed as we have just indicated.
The definition of a number being made up of words, and these words
of letters, some of these permutations will be definitions of numbers. Let us
cross out from our permutations all those that not definitions of numbers.
Let u1 be the first number defined by a permutation, u2 the second, u3
the third, and so on.
We thus have, written in a definite order, all numbers that are defined
by finitely many words.
Therefore, the numbers that can be defined by finitely many words
form a denumerably infinite set.
Now, here comes the contradiction. We can form a number not belong-
ing to this set. “Let p be the digit in the nth decimal place of the nth
number of the set E; let us form a number having 0 for its integral part
and, in its nth decimal place, p + 1 if p is not 8 or 9, and 1 otherwise.” This
number N does not belong to the set E. If it were the nth number of the
set E, the digit in its nth place would be the same as the one in the nth
decimal place of that number, which is not the case.
I denote by G the collection of letters between quotations marks.
The number N is defined by the words of the collection G, that is, by
finitely many words; hence it should belong to the set E. But we have seen
that it does not.
(3) Russell [1908], pp. 60 f; presenting Richard’s paradox. Also: Whitehead and Rus-
sell [1910], p. 61.
Consider all decimals that can be defined by means of a finite number of
words; let E be the class of such decimals. Then E has ℵ0 terms; hence
its members can be ordered as the 1st, 2nd, 3rd,. . .. Let N be a number
defined as follows: If the nth figure in the nth decimal is p, let the nth figure
in N be p + 1 (or 0, if p = 9). Then N is different from all members of E,
since, whatever finite value n may have, the nth figure in the nth of the
decimals composing E, and therefore N is different from the nth decimal.
Nevertheless we have defined N in a finite number of words, and therefore
N ought to be a member of E. Thus N both is and is not a member of E.
Comment. Observe the similarity to the proof that the real numbers are not denu-
merable.
Remark 72.4. In Richard’s paradox diagonalization appears in its classic form. In

the formalized arithmetic of my tools it is the substitution function sin of §48
which captures the method of diagonalization. A precise definition of such a func-
tion, as a primitive recursive function, seems to have been first given in Gödel [1931]
for the explicit purpose of transferring Richard’s paradox into a meta-mathematical
result.
Quotations 72.5. (1) Tarski [1944], p. 80.

Let S be any sentence beginning with the words “Every sentence”. We
correlate with S a new sentence S* by subjecting S to the following two
modifications: we replace in S the first word, “Every”, by “The”; and we
insert after the second word, “sentence”, the whole sentence S enclosed in
quotation marks. Let us agree to call the sentence S “(self-)applicable” or
“non-(self-)applicable” dependent on whether the correlated sentence S* is
true or false. Now consider the following sentence:
Every sentence is non-applicable.
It can be easily shown that the sentence just stated must be both applicable
and non-applicable; hence a contradiction.
(2) Carnap [1931], p. 183.
By definition a property is “impredicable” if it does not belong to itself.
Now is the property “impredicable” itself impredicable? If we assume that
it is, then since it belongs to itself it would be, according to the definition of
“impredicable”, not impredicable. If we assume that it is not impredicable,
then it does not belong to itself and hence, according to the definition of
“impredicable”, is impredicable. According to the law of excluded middle, it
is either impredicable or not, but both alternatives lead to a contradiction.
(3) Ramsey [1925], p. 27; presenting Grelling’s paradox.
Some adjectives have meanings which are predicates of the adjective word
itself; thus the word ‘short’ is short, but the word ‘long’ is not long. Let
us call adjectives whose meanings are predicates of them, like ‘short’, au-
tological; others heterological. Now is ‘heterological’ heterological? If it is,
its meaning is not a predicate of it; that is, it is not heterological. But if
it is not heterological, its meaning is a predicate of it, and therefore it is
heterological. So we have a complete contradiction.
(4) Bachmann [1955/67], p. 113, fn. 1.
Angenommen, die Menge M ′ aller Mengen, die zu M äquivalent sind, exi-
stiere. Dann existiert [[. . .]] eine Menge N , die größere Mächtigkeit hat als
M ′ . Ferner existiert zu jedem X ∈ N die Menge PX aller Paare (X, Y ) mit
Y ∈ M , sodann die Menge P aller Mengen PX mit X ∈ N . Nun ist PX ∼ M
für jedes X ∈ N , also P ⊂ M ′ , ferner P ∼ N , Widerspruch.
CHAPTER XIX
EARLY REACTIONS TO THE PARADOXES
The emergence of an antinomy within Frege’s attempt at tracing arithmetic back to

logic, i.e., within his logical foundation of arithmetic, and the proliferation of para-
doxes in other areas more or less closely related to the foundations of mathematics,
gave a new impetus to discussions focusing on the foundations of mathematics.
This, in turn, led to the development of a number of new concepts and techniques
which will be of interest for my concern with the metaphysical antinomies. Apart,
of course, from the various inevitable speculations on the nature of mathematical
objects these are, in particular, the attempts at making the notion of self-reference
and vicious circle precise and the development of tools for investigating theories,
i.e., a concept of an exact metatheory.
The present chapter aims at sketching the historical-philosophical background
of the early reactions to essentially those paradoxes which I presented at the end
of the last chapter. It starts with the period immediately following the discovery
of the paradoxes right up to Gödel’s result of the incompleteness of all (sufficiently
rich) formalized theories, i.e., right up to the establishment of self-reference (fixed
point property) in formalized theories comprising arithmetic.
I shall not here try to make a new point or contribute in any way to the problems
involved. My aim is to provide the material for a later discussion from a dialectical
point of view.
§ 73. First reactions
The emergence of paradoxes in what seemed at the time a viable basis for math-
ematical reasoning is a challenging phenomenon. The reactions of mathematicians
were different according as to what their metaphysics of mathematics looked like.
First of all it seemed to lend support to those voices critical of the set theoretical
foundations of mathematics, in particular, of Cantor’s diagonal method and the use
of the actual infinite. The range of the reactions to the emergence of antinomies
reached from playing down the importance to building it up into a crisis √ of foun-
dation, comparable to that of the discovery of the incommensurability of 2. The
one I like most is the one which Russell reports of Poincaré; I have chosen it as the
motto for the groundworks.
73a. Frege’s response. The letter in which Bertrand Russell communicated

his antinomy of the class of classes which do not contain themselves as elements to
1042
§ 73. FIRST REACTIONS 1043
Gottlob Frege is dated 16 June 1902; Frege’s answer is dated 22 June 1902 and I
take the first quotation from it.
Quotations 73.1. (1) Van Heijenoort [1967b], p. 127; translation from Frege’s let-
ter to Russell, dated 22. June 1902.1
Your discovery of the contradiction caused me the greatest surprise and,
I would almost say, consternation, since it has shaken the basis on which I
intended to build arithmetic.
Comment. No buts; no excuses, no philosophical bullshit.2 I regard Frege’s reaction
as a paradigm of intellectual honesty.
(2) Van Heijenoort [1967b], p. 127; editing Russell’s response to van Heijenoort’s
request to publish the correspondence between him and Frege.
As I think about acts of integrity and grace, I realise that there is nothing
in my knowledge to compare with Frege’s dedication to truth. His entire
life’s work was on the verge of completion, much of his work had been
ignored to the benefit of men infinitely less capable, his second volume was
about to be published, and upon finding that his fundamental assumption
was in error, he responded with intellectual pleasure clearly submerging
any feelings of personal disappointment. It was almost superhuman and a
telling indication of that of which men are capable if their dedication is to
creative work and knowledge instead of cruder efforts to dominate and be
known.
“In the jaws of the press”, as Quine put it in his paper [1955], Frege wrote
an appendix to the second volume of his Grundgesetze der Arithmetik [1903], in
which he presented and discussed Russell’s antinomy, and also proposed a counter
strategy. The following quotations are taken from this appendix.
Quotation 73.2. Geach and Black [1952], p. 234; translating Frege [1903], p. 253.
Hardly anything more unfortunate can befall a scientific writer than to
have one of the foundations of his edifice shaken after the work is finished.
This was the position I was placed in by a letter of Mr. Bertrand Russell,
just when the printing of this volume was nearing its completion. It is a
matter of my Axiom (V). I have never disguised from myself its lack of the
self-evidence that belongs to the other axioms and that must properly be
demanded of a logical law. And so in fact I indicated this weak point in
the Preface to Vol. i (p. VII).3 I should gladly have dispensed with this
foundation if I had known of any substitute for it. And even now I do not
see how arithmetic can be scientifically established; how numbers can be
apprehended as logical objects, and brought under review; unless we are
permitted—at least conditionally—to pass from a concept to its extension.
1 The correspondence was in German and the original texts of the letters can be found in
Gabriel et al. [1976]. Sluga [1962] might also be enlightening.

2 Particularly in view of the categorical statement made in Frege [1893], p. XXVI; quota-
tion 71.22 (1) in these materials.

1044 XIX. EARLY REACTIONS TO THE PARADOXES
May I always speak of the extension of a concept—speak of a class? And if

not, how are the exceptional cases recognized? Can we always infer from one
concept’s coinciding in extension with another concept that any object that
falls under the one falls under the other likewise? These are the questions
raised by Mr. Russell’s communication.
I continue with more detailed aspect of Frege’s reaction to Russell’s paradox.
p. 253.
[[E]]veryone who has made use in his proofs of extensions of concepts, classes,
sets, is in the same position as I. What is in question is not just my particular
way of establishing arithmetic, but whether arithmetic can possibly be given
a logical foundation at all.
(2) Van Heijenoort [1967b], pp. 127 f; translation from Frege’s letter to Russell,
dated 22. June 1902.
[[W]]ith the loss of my Rule V, not only the foundations of my arithmetic, but
also the sole possible foundations of arithmetic, seem to vanish. Yet, I should
think, it must be possible to set up conditions for the transformations of the
generalization of an equality into an equality of courses-of-values such that
the essentials of my proofs remain intact. In any case your discovery is very
remarkable and will perhaps result in a great advance in logic, unwelcome
as it may seem at first glance.
Comment. Although Russell’s discovery has had a great impact on foundational
studies, the advance in logic that Frege was probably hoping for hasn’t happened
— so far, at least.
(3) Long and White [1979], p. 182; translating Frege [1906], p. 198.
‘[[T]]he extension of the concept square root of 1 ’ is here to be regarded as
a proper name, as is indeed indicated by the definite article. By permitting
the transformation, you concede that such proper names have meanings
[[Bedeutungen]]. But by what right does such a transformation take place, in
which concepts correspond to extensions of concepts, mutual subordination
to equality? An actual proof can scarcely be furnished. We will have to
assume an unprovable law here. Of course it isn’t as self-evident as one
would wish for a law of logic. And if it was possible for there to be doubts
previously, these doubts have been reinforced by the shock the law has
sustained from Russell’s paradox.
(4) Long and White [1979], p. 183; translating Frege [1906], p. 199.
[[A]]n extension of a concept is at bottom completely different from an ag-
gregate. The aggregate is composed of its parts. Whereas the extension of
a concept is not composed of the objects that belong to it. For the case is
conceivable that no objects belong to it. The extension of a concept simply
has its being in the concept, not in the objects which belong to it; these are
not its parts. There cannot be an aggregate what has no parts.
Comment. I include this quotation mainly because of Frege’s reason for rejecting
the paradigm of collection (aggregate) in favour of that of classification.4
p. 254.
What attitude must we adopt towards this? Must we suppose that the
law of excluded middle does not hold good for classes? Or must we suppose
there are cases where an unexceptionable concept has no class answering
to it as its extension?
(2) Geach and Black [1952], p. 236; Geach translating Frege [1903], p. 255.
Classes of proper objects would have to be distinguished from classes of
classes of proper objects; extensions of relations holding between proper
objects would have to be distinguished from classes of proper objects, and
from classes of extensions of relations holding between proper objects; and
so on. We should thus get an incalculable multiplicity of types; and in
general objects belonging to different types could not occur as arguments
of the same function. [[. . .]]
If these difficulties scare us off from the view that classes (including
numbers) are improper symbols; and if we are likewise unwilling to recog-
nize them as proper objects, i.e. as possible arguments for any first-level
function; then there is nothing for it but to regard class names as sham
proper names, which would thus not really have any reference. They would
have to be regarded as part of signs that had reference only as wholes.
Comment. This sounds like Russell’s “no-class theory” (contextual definition) to
me.5
(3) Geach and Black [1952], p. 239; translating Frege [1903], p. 257.
[[W]]e must take into account the possibility that there are concepts with
no extension (at any rate, none in the ordinary sense of the word).
Comment. Compare this to Russell’s no-class theory and his comments in Russell
[1937], p. x, regarding the abolition of classes.
(4) Geach and Black [1952], p. 241; Geach translating Frege [1903], pp. 260 f.
If in general, for any first-level concept, we may speak of its extension,
th[[e]]n the case arises of concepts having the same extension, although not
all objects that fall under one fall under the other as well.
This, however, really abolishes the extension of the concept, in the sense
we have given the word. We may not say that in general the expression ‘the
extension of one concept coincides with that of another’ stands for the
same thing as the expression ‘all objects that fall under the one concept
fall under the other as well, and conversely.’ We see from the result of our
deduction that it is quite impossible to give the words ‘the extension of the
4 Cf. section 83b in these materials for the role that the aspect of collection has taken later
in the foundation of set theory. Also, perhaps, quotations 83.8 (1), regarding the ‘logical’ versus
the ‘mathematical’ notion.
concept φ(ξ)’ such a sense that from concepts’ being equal in extension we
could always infer that every object falling under one falls under the other
likewise.
What Frege held responsible for the emergence of this antinomy within his log-
ical foundation of arithmetic was his “Grundgesetz (V)” which says (in a somewhat
adapted terminology):
^
λx C[x] = λx F[x] ↔ x (C[x] ↔ F[x]) .
It is with regard to this axiom schema that Frege, after having presented Russell’s
fatal result, makes the considerations of the following quotation.
Quotations 73.5. (1) Van Heijenoort [1967b], p. 128; translation of Frege’s letter
to Russell.
[[I]]t seems to me that the expression “a predicate is predicated of itself” is
not exact. A predicate is as a rule a first-level function, and this function
requires an object as argument and cannot have itself as argument (subject).
Therefore I would prefer to say “a notion is predicated of its own extension”.
If the function Φ(ξ) is a concept, I denote its extension (or the corresponding
class) by “ἐΦ(ε)” (to be sure, the justification for this has now become
questionable to me). In “Φ(ἐΦ(ε))” [[. . .]] we then have a case in which the
concept Φ(ξ) is predicated of its own extension.
Comment. The interesting point here for me is the remark “the justification for
this has now become questionable to me” which I take to relate above all to the
considerations in Sinn und Bedeutung as well as Funktion und Begriff.
(2) Geach and Black [1952], pp. 242 f; translating Frege [1903], p. 262.
[[W]]e see that the exceptional case is constituted by the extension itself, in
that it falls under only one of two concepts whose extension it is; and we see
that the occurrence of this exception can in no way be avoided. Accordingly
the following suggests itself as the criterion for equality in extension: The
extension of one concept coincides with that of another when every object
that falls under the first concept, except the extension of the first concept,
falls under the second concept likewise, and when every object that falls
under the second object, except the extension of the second concept, falls
under the first concept likewise.
Comment. In symbols:
^
λx C[x] = λx F[x] → y ((y 6= λx C[x] ∧ C[y] → F[y]) ∧ (y 6= λx F[x] ∧ F[y] → C[y])) .
Frege’s restriction on a characteristic property of extensions may be viewed as

an attempt to avoid the most obvious form of a vicious circle in the constitution of
the extension of a concept, though Frege gave no hint that he himself saw it this way.
To the contrary, there is a remark immediately following the passage just quoted
which seems to indicate that it is not the extension of a concept itself which Frege
considered affected by this restricting clause, but only its characteristic property,
viz. equality.
Quotation 73.6. Geach and Black [1952], pp. 242 f; translating Frege [1903], p. 262.
Obviously this cannot be taken as defining the extension of a concept,
but only as specifying the distinctive property of this second-level function.
Frege’s way out is unsuccessful.6 It is interesting, however, that Russell [1903],

p. 522, remarked that “it seems very likely that this is the true solution” and rec-
ommended to examine Frege’s argument.
Quotations 73.7. (1) Frege [1979], p. 263; translation of Frege [1969], p. 282 (diary
entries on the concept of numbers, 23. 3. 1924).
My efforts to become clear about what is meant by number have resulted
in failure. We are only too easily misled by language and in this particular
case the way we are misled is little short of disastrous.
(2) Frege [1979], p. 265; translation of Frege [1969], p. 283 (a draft on number, dated
1924).
My efforts to throw light on the questions surrounding the word ‘number’
and the words and signs for individual numbers seem to have ended in
complete failure. Still these efforts have not been wholly in vain. Precisely
because they have failed, we can learn something from them.
(3) Frege [1979], pp. 269 f; translating Frege [1969], pp. 288 f.7
One feature of language that threatens to undermine the reliability of
thinking is its tendency to form proper names to which no objects corre-
spond. [[. . .]] A particularly noteworthy example of this is the formation of
a proper name after the pattern of ‘the extension of the concept a’, e.g.
‘the extension of the concept star ’. Because of the definite article, this ex-
pression appears to designate an object; but there is no object for which
this phrase could be a linguistically appropriate designation. From this has
arisen the paradoxes of set theory which have dealt the death blow to set
theory itself. I myself was under this illusion when, in attempting to provide
a logical foundation for numbers, I tried to construe numbers as sets.
(4) Geach and Black [1952], p. 244; Geach translating Frege [1903], p. 265.
The prime problem of arithmetic may be taken to be the problem: How
do we apprehend logical objects, in particular numbers? What justifies us
in recognizing numbers as objects? Even if this problem is not yet solved
to the extent that I believed it was when I wrote this volume, nevertheless
I do not doubt that the way to a solution has been found.
73b. Russell’s Principles. In the appendix b to his Principles of Mathe-

matics, Russell sketches his proposal to escape the antinomies under the title “the
doctrine of types”. The following set of quotations is intended to give a rough idea.
6 Cf. Quine [1955]. My [1992a] (version 0.5 of the present study) included a version of this
paradox in section 97a. Cf. also remark 141.24 (1) in the appendix A1.
7 This quotation was instigated by Hoering [1975], p. 64.

It will [[. . .]] be necessary to distinguish (1) terms, (2) classes, (3) classes
of classes, and so on ad infinitum; we shall have to hold that no member of
one set is a member of any other set, and that xǫu requires that x should be
of a set of a degree lower by one than the set to which u belongs. Thus xǫx
will become a meaningless proposition; and in this way the contradiction is
avoided.
(2) Russell [1903], p. 523.
The doctrine of types is here put forward tentatively, as affording a possible
solution of the contradiction[[.]] [[. . .]]
Every propositional function φ(x)—so it is contended—has, in addition
to its range of truth, a range of significance, i.e. a range within which x
must lie if φ(x) is to be a proposition at all, whether true or false. This
is the first point of the theory of types; the second point is that ranges of
significance form types, i.e. if x belongs to the range of significance of φ,
then there is a class of objects, the type of x, all of which must also belong
to the range of significance of φ(x), however φ may be varied; and the range
of significance is always either a single type or a sum of several whole types.
(3) Russell [1903], p. 525.
Each of the types above enumerated is a minimum type; i.e., if φ(x) be a
propositional function which is significant for one value of x belonging to one
of the above types, then φ(x) is significant for every value of x belonging to
the said type. But it would seem—though of this I am doubtful— that the
sum of any number of minimum types is a type, i.e. a range of significance
for certain propositional functions. Whether or not this is universally true,
all ranges certainly form a type, since every range has a number; and so do
all the objects, since every object is identical with itself.
Outside the above series of types lies the type proposition; and from this
as a starting-point a new hierarchy, one might suppose, could be started;
but there are certain difficulties in the way of such a view, which render it
doubtful whether propositions can be treated like other objects.
498. Numbers, also, are a type lying outside the above series, and
presenting certain difficulties, owing to the fact that every number selects
certain objects out of every other type of ranges, namely those ranges which
have the given number of members.
(4) Russell [1903], p. 528.
To sum up: it appears that the special contradiction of [[the class of
classes which do not contain themselves as elements]] is solved by the doc-
trine of types, but that there is at least one closely analogous contradiction
which is probably not soluble by this doctrine. The totality of all logical
objects, or of all propositions, involves, it would seem, a fundamental logical
difficulty. What the complete solution of the difficulty may be, I have not
succeeded in discovering; but as it affects the very foundations of reasoning,
I earnestly commend the study of it to the attention of all students of logic.
73c. Cantor’s own view. Cantor’s attitude to the paradoxes was quite de-
cisive and there is no indication that he was caught unawares. The problem: these
ideas were written in letters to Dedekind and had only been published in 1932. At
the time when the paradoxes were discussed, they were not part of the discussion.
Features 73.9. In Cantor’s attitude to the antinomies I shall focus on the following
four aspects:
(CT1) consistent vs. inconsistent multiplicities;
(CT2) consistency is an axiom, even for finite multiplicities;
(CT3) independence of elementhood;
(CT4) it is a reversal of the correct, if one undertakes to base the concept of
number on that of the extension of a concept.
The first antinomy of set theory to be published was that of Burali-Forti. The
following quotation shows Cantor’s reaction.
Quotation 73.10. W. H. and G. C. Young [1929], p. 101, quoting from a letter by

Cantor, 9 March 1907;8 pinched from Wang [1974], p. 48.
‘What Burali-Forti has produced is thoroughly foolish. If you go back to
his articles in Circolo Matematico, you will remark that he has not even
understood properly the concept of a well-ordered set.’
Cantor was certainly aware of the presence of antinomies in his theory. This
is confirmed by the following passage from a letter to Dedekind, dated the 28th of
July 1899.
Quotation 73.11. van Heijenoort [1967b], p. 114; translating Cantor [1899], p. 443;
from a letter to Dedekind dated 28 July 1899.
[[A]] multiplicity can be such that the assumption that all of its elements
“are together” leads to a contradiction, so that it is impossible to conceive
of the multiplicity as a unity, as “one finished thing”. Such multiplicities I
call absolutely infinite or inconsistent multiplicities.
As we can readily see, the “totality of everything thinkable”, for exam-
ple, is such a multiplicity[[.]]
Comment. This concerns CT1, the distinction between consistent and inconsistent
multiplicities. Later in this letter, Cantor produces mathematical examples of in-
consistent multiplicities. The question remains, however, how are we to handle in-
consistent multiplicities. It is to be expected that they need more care, but of what
sort? Some hint is given in quotation 73.12 (2) below.
Given Cantor’s distinction between consistent and inconsistent multiplicities,

how can one know if the multiplicities Cantor uses are consistent ones? This is a
simple question and it has been asked by Cantor in another letter to Dedekind, one
month later (28th of August 1899), from which the following quotation is taken.
8 Cf. also Fraenkel [1932], p. 470, fn. 1.
Quotations 73.12. (1) U. Petersen [2002], p. 1050; translating Cantor [1899],

pp. 447 f; from a letter to Dedekind dated 28 August 1899.
. . . One has to ask the question, how I could possibly know that the well-
ordered multiplicities or sequences to which I assign the cardinal numbers
ℵ0 , ℵ1 , . . . , ℵ ω0 , . . . , ℵ ω1 , . . .
are really “sets” in the specified sense of the word, i.e. “consistent mul-
tiplicities”. Wouldn’t it be possible that already these multiplicities were
“inconsistent”, and that the contradiction of the hypothesis of a “being to-
gether of all of their elements” has not yet manifested itself? My answer to
this is that this question has to be extended to finite multiplicities as well
and that a thorough consideration leads to the result: it is not possible to
provide a “proof” of their “consistency”, even in the case of finite multiplic-
ities. In other words: the fact of the “consistency” of finite multiplicities is
a simple, unprovable truth, it is “The Axiom of Arithmetics” (in the old
sense of the word). And equally, the “consistency” of the multiplicities, to
which I assign the alefs as cardinal numbers, is “the axiom of the extended
transfinite arithmetics”.
Comment. This concerns CT2; it shows that Cantor was not just aware of the
possibility of inconsistent multiplicities, but even had some idea as to how to handle
them. As a matter of fact, Cantor used the multiplicity of all ordinal numbers (which
is an inconsistent multiplicity) to prove his well-ordering theorem (according to the
editor Zermelo, however, in an inconclusive way). Note also that Cantor did not
restrict the problem of consistency to infinite sets only, as most others did.
(2) Van Heijenoort [1967b], p. 114; translating Cantor [1899], p. 443; from a letter
to Dedekind dated 28 July 1899.
If [[. . .]] the totality of the elements of a multiplicity can be thought of
without contradiction as “being together”, so that they can be gathered
together into “one thing”, I call it a consistent multiplicity or a “set”. [[. . .]]
Two equivalent multiplicities either are both “sets” or are both incon-
sistent.
Every submultiplicity [[Teilvielheit]] of a set is a set.
Whenever we have a set of sets, the elements of these sets again form
a set.
Comment. The last two principles are curiously reversed to what was to become
the ‘iterative concept of set’.
(3) U. Petersen [2002], p. 1050; translating Cantor [1899], p. 448; from a letter to
Dedekind dated 31 August 1899.
. . . We want to assign equivalent “sets” to one and the same potency-
class, non-equivalent sets to different classes, and we consider the system
S of all classes that can be thought.
By a I mean simultaneously the cardinal number or potency of the sets
of the respective class, which just is one and the same for all these sets.
Let Ma be any determined set of the class a.
I maintain that the completely determined well-defined system S is no

“set”.
Proof. If S were a set, then
T = ΣMa ,
this sum carried out over all classes a, would be a set too; hence T would
also belong to a determined class, let us say to the class a0 .
But now there is the following theorem:
“If M is any set of the cardinal number a, then it is always possible to
derive from it another set M ′ , the cardinal number a′ of which is greater
than a.”
[[. . .]]
Let, therefore, a′0 be any cardinal number which is greater than a0 . Then
T with the potency a0 contains as part the set Ma′0 of greater potency a′0 ,
which is a contradiction.
The system T , hence also the system S are, therefore, n o sets. Thus
there are determined multiplicities which are not at the same time units, i.e.
multiplicities of such a kind that a real “being together of all its elements” is
impossible. These are what I call “inconsistent systems”, the others, however,
“sets”.
Quotation 73.13. [1932], p. 470, footnote 2, quoting Cantor.

Vielheiten unverbundener Dinge [[sind]] solche Vielheiten, bei denen die Ent-
fernung irgendeines oder mehrerer Elemente von keinem Einfluß auf das
Bestehenbleiben der übrigen Elemente ist.
Comment. This concerns CT3. Fraenkel comments that Cantor comes close here
to the ban of non-predicative definitions and thus unconsciously exposes his power
set, which is basic to the construction of his set theory, to criticism.
There is an interesting disagreement regarding the extension of concepts be-

tween Frege and Cantor to which I shall devote the last couple of quotations in this
section.
Quotations 73.14. (1) U. Petersen [2002], p. 1051; translating Cantor [1885],

p. 440.
[[T]]he “extension of a concept” is quantitatively in general something com-
pletely indeterminate; only in certain cases is the “extension of a concept”
quantitatively determined, then indeed, if it is finite, it is entitled to a cer-
tain number and, if it is infinite, a certain magnitude (Mächtigkeit). For
such a quantitative determination of the “extension of a concept”, however,
the concepts “number” and “magnitude” have to be given beforehand from
somewhere else, and it is a reversal of the correct, if one undertakes to base
the latter concepts on the concept “extension of a concept”.
Comment. In a note to the reprint of Cantor’s review, the editor, Ernst Zermelo,
founder of modern axiomatic set theory commented, inter alia, that Cantor does
only partially justice to Frege’s work and that introducing the extension of a concept
is basically inessential.9
(2) Frege [1885], p. 122, regarding Cantor’s review.
These remarks would be very fitting, and I should acknowledge them to
be quite justified, if it followed from my definition that, e.g., the number
of Jupiter’s moons was the extension of the concept ‘Jupiter’s moon’. But
they do not fit the definition I have given, according to which the number
of Jupiter’s moons is the extension of the concept ‘equinumerous to the
concept “Jupiter’s moon” ’[[.]]
73d. Blaming the actual infinite. Cantor’s dealing with the infinite was
regarded with considerable suspicion by many of his fellow mathematicians and it
is not surprising that the blame for the paradoxes was quickly laid on the so-called
actual-infinite .
Quotation 73.15. Poincaré [1906], p. 448.

Since long ago the notion of infinity had been introduced into mathe-
matics; but this infinite was what philosophers call a becoming. The math-
ematical infinite was only a quantity capable of increasing beyond all limit:
it was a variable quantity of which it could not be said that it had passed
all limits, but only that it could pass them.
Cantor has undertaken to introduce into mathematics an actual infinite,
that is to say a quantity which not only is capable of passing all limits, but
which is regarded as having already passed them.
Comment. More along this line can be found in quotations 74.10 (4) and (5) below.
Even without blaming the actual infinite for the antinomies it was still felt that
the role of the infinite in mathematics was in need of clarification.
Quotation 73.16. Benacerraf and Putnam (eds.) [1964], p. 134; translation of Hil-
bert [1926], pp. 161 f.
[[I]]n spite of the foundation Weierstrass has provided for the infinitesimal
calculus, disputes about the foundations of analysis still go on.
These disputes have not terminated because the meaning of the infinite,
as that concept is used in mathematics, has never been completely clari-
fied. Weierstrass’s analysis did indeed eliminate the infinitely large and the
infinitely small by reducing statements about them to [statements about]
relations between finite magnitudes. Nevertheless the infinite still appears
in the infinite numerical series which defines the real numbers and in the
concept of the real number system which is thought of as a completed to-
tality existing all at once.
Comment. Hilbert [1926] is not fully translated in Benacerraf and Putnam. Another
translation which includes Hilbert’s attempt to solve the continuum problem can
be found in van Heijenoort [1967b], pp. 369–392.
9 See Zermelo [1932], pp. 441 f.
Quotations 73.17. (1) Van Heijenoort [1967b], p. 143; translating Richard [1905].
Let us show that this contradiction is only apparent. We come back to
our permutations. The collection G of letters is one of these permutations;
it will appear in my table. But, at the place it occupies, it has no meaning.
It mentions the set E, which has not been defined. Hence I have to cross
it out. The collection G has meaning only if the set E is totally defined,
and this is not done except by infinitely many words. Therefore there is no
contradiction.
(2) Van Heijenoort [1967b], p. 142; translating Peano [1906].
Richard’s example does not belong to mathematics, but to linguistics; an
element that is fundamental in the definition of [[the diagonal number]] N
cannot be defined in an exact way (according to the rules of mathematics).
I close this section with a few quotations which are truly remarkable in the
context of the present study in so far as they clearly show an awareness of the link
between mathematics and metaphysics which constitutes the target of my enter-
prise.
Quotations 73.18. (1) U. Petersen [2002], p. 1053; translating Hessenberg [1906],

p. 633.
The paradox of the set [[of all ordinals]] reminds of those antinomies which,
according to Kant arise when we regard nature as a completed whole.
(2) U. Petersen [2002], p. 1053; translating Hessenberg [1906], p. 706.
If the ultrafinite paradoxes, in particular that of the set [[of all ordinals]]
cannot be removed, then there are grounds for the assumption that they
stand in connection with the antinomies which have been set up by Kant.
Comment. From the point of view of this study it is unfortunate that this line of
thought has not come to gain more influence on the subsequent development. In
other words, the representatives of transcendental idealism have failed to seize a
great opportunity.
(3) Kowalewsky [1918], pp. 757 f.
Der Widerstreit der rein vernünftigen und der anschaulichen Erkenntnis
bildet überhaupt eins der ältesten Probleme der Philosophie.
Schon die Eleaten und besonders Zenon haben die Schwierigkeiten ein-
gesehen, die mit der Annahme der transzendentalen Realität der Natur
unvermeidlich verbunden sind.
Die Paradoxien von Zenon kann man als den ersten Versuch bezeichnen,
die Idealität des Raumes und der Zeit zu beweisen[[.]]
Further reading 73.19. Bernays [1913], Kowalewsky [1918].

§ 74. Actual infinity, immutable classifications, and non-predicative def-

initions: Poincaré’s analysis of the paradoxes
Henri Poincaré followed Jules Richard in explaining the paradoxes in the sense of
discovering a flaw within seemingly correct arguments and coined the notion of non-
predicative definitions, which shortly afterwards was followed by Russell’s notion of
vicious circles. Poincaré says in his paper [1906] that he adopted the word “non-
predicative” from Russell [1906a]. Though neither of the notions is unequivocal,
one may still say that they are somewhat related to each other. Their realization,
however, within the frame of a formal system is open to a number of different inter-
pretations. The present and the following paragraphs are devoted to the historical
background, i.e., Poincaré’s and Russell’s original attempts towards arresting the
antinomies, in particular the amalgamation of the two original ideas of immutable
classifications (Poincaré) and vicious circles (Russell). In later sections, I shall dis-
cuss various other possibilities along this line.
As with probably most concepts, so with the concepts impredicative definition
and vicious circle: they did not come into being in their present shape, but emerged
from a cluster of more or less related ideas through a process of creative misun-
derstandings. In the case of the vicious circle principle, certain components of this
cluster are of particular interest for us. I shall therefore try to disentangle it to let
those ideas stand out better which I shall use later on. Since I do not intend to give
a complete analysis of the historical development of these two concepts, I confine
myself to quoting mainly those passages of the two authors which will be relevant
for my further investigations.
Features 74.1. In Poincaré’s analysis of the antinomies, I shall concentrate on the

following issues:
(PC1) the question of the compatibility of logic and the actual infinite;
(PC2) the concept of immutable classification;
(PC3) the concept of predicativity;
(PC4) the concept of vicious circle;
(PC5) the view of sets as based on classification;
(PC6) the connection between immutable classifications and infinite sets;
(PC7) the view of definitions as based on classifications;
(PC8) the view of logic as requiring fixed and unalterable objects.
In my presentation, I shall make use of Poincaré [1905], [1906], and [1909]. All
the features listed above shall play an important role in my later analysis of the
antinomies.
74a. Infinity, classifications, and logic. In his paper [1906], Poincaré chose
as a starting point for tackling the problem of antinomies the dealing with infinite
sets in the context of generalized statements.
Quotations 74.2. (1) Poincaré [1906], pp. 483 f.

In these definitions the word ‘all’ figures, as is seen in the examples cited
above. The word ‘all’ has a very precise meaning when it is a question of a
§ 74. POINCARÉ’S ANALYSIS OF THE PARADOXES 1055
finite number of objects; to have another one, when the objects are infinite
in number, would require there being an actual (given completely) infinity.
Otherwise all these objects could not be conceived as postulated anteriorly
to their definition, and then if the definition of a notion N depends upon all
the objects A, it may be infected with a vicious circle, if among the objects
A are some indefinable without the intervention of the notion N itself.
(2) Poincaré [1909], p. 45.
May the ordinary rules of logic be applied without change whenever we
consider collections comprising an infinite number of objects?
Comment. Cf. what Hegel says in quotation 66.16 (6) on “the calculation of the
infinite”. Cf. also the modification of induction in theorem 136.11 of the ground-
works which may give an idea of my way of dealing with infinity with the help of
a necessity operator.
Poincaré linked his attitude towards infinite sets to his view on the nature of
logic. This is a remarkable point in itself which I shall take up later. For the moment
I continue illustrating Poincaré’s view of logic by means of quotations.
Quotations 74.3. (1) Poincaré [1906], p. 484.

The rules of formal logic express simply the properties of all possible
classifications. But for them to be applicable it is necessary that these
classifications be immutable and that we have no need to modify them in
the course of the reasoning. If we have to classify only a finite number of
objects, it is easy to keep our classifications without change. If the objects
are indefinite in number, that is to say if one is constantly exposed to seeing
new and unforeseen objects arise, it may happen that the appearance of a
new object may require the classification to be modified, and thus it is we
are exposed to antinomies.
Comment. Here is the point: “new and unforeseen objects arise”. I postpone a dis-
cussion whether this is just a case of neophobia10 to §112 in the groundworks.
(2) Poincaré [1909], p. 45.
Formal logic is nothing but the study of the properties common to all
classifications; it teaches us that two soldiers who are members of the same
regiment belong by this very fact to the same brigade and consequently to
the same division; and the whole theory of the syllogism is reduced to this.
What is, then, the condition necessary for the rules of this logic to be valid?
It is that the classification which is adopted be immutable. We learn that
two soldiers are members of the same regiment, and we want to conclude
that they are members of the same brigade; we have the right to do this
provided that during the time spent carrying on our reasoning one of the
two men has not been transferred from one regiment to another.
10 The Shorter Oxford English Dictionary gives under the entry “neophilia”: “Love for, or
great interest in, what is new; a love of novelty.” It seems to me that the term “neophobia” is
perfectly complementing the term “neophilia”.
The antinomies which have been revealed all arise from forgetting this
very simple condition: a classification was relied on which was not im-
mutable and which could not be so[[.]]
Comment. Today, MostPeople will be inclined to regard Poincaré’s view on logic as
obsolete, but this doesn’t bother me since this MostPeople is an unreliable pundit
and may be gone by tomorrow.11 I find it an intriguing view and it actually led my
attention to classifications. Now it is Poincaré’s view of the nature of classifications
which needs more attention.
In order to illustrate his point Poincaré makes use of the following example
which he credits to Russell and which Russell, in turn, credits to Berry;12 I shall
present a formal version of this antinomy in the appendix A1. Since I am interested in
Poincaré’s view of the affair I cite here his own (translated) version of the antinomy
before I present his explanations.

What is the smallest integer which cannot be defined by a sentence with
fewer than one hundred French words? And furthermore does this number
exist?
Yes; for, with one hundred French words, we can construct only a finite
number of sentences, since the number of words in the French dictionary
is limited. Among these sentences, there will be some which will have no
meaning or which will not define any integer. But each of these will be
capable of defining at most a single integer. The number of integers capable
of being defined in this manner is therefore limited; consequently, there
are certainly some integers which cannot be so defined; and among these
integers, there is certainly one which is smaller than all the others.
No; for, if this integer did exist, its existence would imply a contradic-
tion, since it would be defined by a sentence with fewer than one hundred
French words; namely that very sentence which affirms that it cannot be
defined.
Next comes Poincaré’s analysis of this situation which I quote for the purpose
of illustrating his view of the role of classifications.

This reasoning rests on a classification of integers into two categories:
those which can be defined by a sentence with fewer than one hundred
French words and those which cannot be. In asking the question, we pro-
claim implicitly that this classification is immutable and that we begin our
reasoning only after having established it definitively. But that is not possi-
ble. The classification can be conclusive only when we have reviewed all the
sentences with fewer than one hundred words, when we have rejected those
which have no meaning, and when we have definitively fixed the meaning of
11 Although, of course, another MostPeople will then have taken his place.
12 Russell [1908], p. 60, footnote*.
those which possess a meaning. But among these sentences, there are some
which can have meaning only after the classification is fixed; they are those
in which the classification itself is concerned. In summary the classification
of the numbers can be fixed only after the selection of the sentences is com-
pleted, and this selection can be completed only after the classification is
determined, so that neither the classification nor the selection can ever be
terminated.
According to Poincaré’s view this kind of classification could be avoided in

principle in the case of finite sets but not in the case of infinite sets.
Observe that, so far, nothing is said why classifications may change, i.e., what
may affect a certain classification in such a way that it changes.
74b. Immutable classifications, predicative definitions, and sets. So
far, I have quoted Poincaré regarding classifications. According to Poincaré these
considerations extend to sets, collections and definitions. This is what the next
couple of quotations are meant to illustrate.

To define a set, a Menge, any collection whatever, is always to make a
classification, to separate the objects which belong to this set from those
which are not part of it.
Comment. Notice that Poincaré is very close to what Gödel called the logical con-
cept of set.13
(2) Poincaré [1909], p. 48.
Every definition is, in effect, a classification. It separates the objects which
satisfy the definition from those which do not, and it arranges them in two
distinct classes.
In the light of this I now continue quoting Poincaré concerning classifications.

[[E]]very time that new elements are added to the collection, this collection
is modified; it is therefore possible to modify the relation of this collection
with the elements already classified; and since it is in accordance with this
relation that these elements have been arranged in this or that drawer, it
can happen that, once this relation is modified, these elements will no longer
be in the correct drawer and that it will be necessary to shift them. As long
as there are new elements to be introduced, it is to be feared that the work
may have to be begun all over again; for it will never happen that there
will not be new elements to be introduced; the classification will therefore
never be fixed.
From this we draw a distinction between two types of classifications
applicable to the elements of infinite collections: the predicative classifica-
tions, which cannot be disordered by the introduction of new elements; the
13 See Wang [1974], p. 187; quotation 83.11 (2) in these materials.
non-predicative classifications in which the introduction of new elements

necessitates constant modification.
(2) Poincaré [1910], p. 47; quoted after Thiel [1972], p. 146.
[[W]]enn ich eine Menge von Gegenständen in eine Anzahl von Schachteln
einordnen soll, so kann zweierlei eintreten: entweder sind die bereits einge-
ordneten Gegenstände endgültig an ihrem Platze, oder ich muß jedesmal,
wenn ich einen Gegenstand einordne, die anderen oder wenigstens einen Teil
von ihnen wieder herausnehmen. Im ersten Falle nenne ich die Klassifikation
prädikativ, im zweiten nicht.
Comment. In the light of Cantor’s remark on ‘unverbundene Vielheiten’ in quo-
tation 73.13 above, impredicative classifications are clearly not sets in the sense
Cantor envisaged them.
Thus we ended up with a criterion for predicative classifications: they do not

change when new elements are added, i.e., predicative classifications are immutable
in a certain sense. But this still leaves us with the question, when a classification
actually is to be considered predicative, i.e., what is a criterion for a classification
to be immutable.

What we have just said about the classifications applies directly to the
definitions.
(2) Poincaré [1909], p. 48.
A definition like all classification, may or may not be predicative.
(3) Poincaré [1909], p. 49.
[[F]]rom a certain point of view, we should not say that a classification is
predicative in an absolute manner, but that it is predicative in relation to
a method of definition.
The next two quotations try to give an idea of how Poincaré takes the idea to
its conclusion via Richard’s paradox.

E is the aggregate of all the numbers definable by a finite number of words,
without introducing the notion of the aggregate E itself. Else the definition
of E would contain a vicious circle; we must not define E by the aggregate
E itself.
(2) Poincaré, [1906], p. 481.
Thus the definitions which should be regarded as not predicative are those
which contain a vicious circle.

A Menge is something about which we can reason; it is something fixed and
unalterable to a certain degree.
Comment. What does “to a certain degree” mean here? In chapter xxviii of my
[1992a] (version 0.5 of the present study) I discussed this question and proposed to
look at a set as being only, so to speak, “locally” unalterable, viz., in the context
of a certain reasoning, not just by itself (absolutely) unalterable. Since this is not
directly linked to the issue of dialectic, it is no longer part of this study; but this
does not mean that I distanced myself from the ideas put forward there.
(2) Poincaré [1909], p. 58.
We will then say that this set is not a Menge, if the corresponding classifi-
cation is not predicative; and that it is a Menge if this is the case or if it is
possible to reason as if it were.
(3) Poincaré [1909], pp. 50 f.
Antinomies have been arrived at because collections were considered which
contained objects in whose definition the notion of the collection itself is
inherent.
(4) Poincaré [1906], p. 483.
It is the belief in the existence of the actual infinite which has given
birth to those non-predicative definitions.
(5) Poincaré [1906], p. 484.
There is no actual (given completely) infinity. The Cantorians have forgot-
ten this, and they have fallen into contradiction. [[. . .]]
Logistic also forgot it, like the Cantorians, and encountered the same
difficulties. But the question is to know whether they went this way by
accident or whether it was a necessity for them. For me, the question is not
doubtful; belief in an actual infinity is essential in the Russell logic.
Comment. See quotation 75.3 for Russell’s return.
I close this paragraph with a couple quotations representing a somewhat differ-

ent line of thought.
Quotations 74.11. (1) Poincaré [1913b], p. 71.

Here is an object from which nothing could be derived until it had been
christened; all it needed was to be given a name and it worked wonders.
How can this happen? It is because by giving it a name we have asserted
implicitly that the object did exist (that is, was free from all contradiction)
and that it was entirely determined.
(2) Poincaré [1913a], p. 483.
[[N]]on-predicative definitions can not be substituted for the terms defined.
Under these conditions logistic is not sterile, it engenders antinomies.
Comment. The italicized bit (original) is an English translation of what I chose as
the motto for the groundworks. I do take it very seriously. The only problem is
to show that impredicativity cannot be avoided.
§ 75. The vicious circle principle: Russell’s analysis of the paradoxes
At this early stage, it seems that Russell and Poincaré developed their ideas in
close contact with — and opposition to — each other. Only Russell’s ideas, however,
reached the mature form of a full-fledged theory, the theory of types. Poincaré seems
to have been opposed to this sort of formalized logic altogether.
Features 75.1. In Russell’s analysis of the antinomies I shall concentrate on the

following features:
(RS1) self-reference and generalized propositions;
(RS2) the formulation of the vicious circle principle in terms of collections,
members, and totals.
(RS3) classes and functions;
(RS4) the reduction of classes to propositional functions;
(RS5) no class theory, meaninglessness (contextual definition).
As in the preceding section, these features shall play a role in my later analysis
of the antinomies.
75a. Generalized propositions. The vicious circle principle was introduced
in Russell [1908] and relevant passages of it can be found again, without any change,
in Whitehead and Russell [1910].
After stating the well-known paradoxes of the liar etc., Russell sets out to
analyze them.
Quotations 75.2. (1) Russell [1908], p. 61; Whitehead and Russell [1910], p. 61.
In all the above contradictions (which are merely selections from an
indefinite number) there is a common characteristic, which we may describe
as self-reference or reflexiveness.
Comment. This statement is remarkable; the idea of self-reference has not been
pursued any further. It seems as if Russell had a quite definite idea about the way
this self-reference or reflexiveness is established in the antinomies, viz. through the
use of generalization.
(2) Russell [1908], p. 61; Whitehead and Russell [1910], p. 62.
In each contradiction something is said about all cases of some kind, and
from what is said a new case seems to be generated, which both is and is
not of the same kind as the cases of which all were concerned in what was
said.
Comment. Note the similarity to Poincaré’s formulation in quotation 74.2 (1).
In contrast to Poincaré, Russell did not consider the actual infinite as being
responsible for the antinomies. The opposition is nicely illustrated by the following
quotation which responds to quotations 74.10 (4) and (5) above.
Quotation 75.3. Copi [1971], p. 14, translating Russell [1906a], p. 633.

Has this man [[the liar]] too forgotten that there is no actual infinite?
§ 75. RUSSELL’S ANALYSIS OF THE PARADOXES 1061
Although Russell’s point is not the actual infinite, it is still close: quantification.
Accordingly, he reformulates the paradox of the liar, for instance, with a quantifier.
Quotation 75.4. Russell [1908], pp. 61 f.

When a man says ‘I am lying’, we may interpret his statement as: ‘There
is a proposition which I am affirming and which is false’. All statements that
‘there is’ so-and-so may be regarded as denying that the opposite is always
true; thus ‘I am lying’ becomes: ‘It is not true of all propositions that either
I am not affirming them or they are true’; in other words, ‘It is not true
for all propositions p that if I affirm p, p is true’. The paradox results from
regarding this statement as affirming a proposition, which must therefore
come within the scope of the statement.
Comment. What is remarkable in this example is the way quantification comes
into the formulation of a paradox, which prima facie does not seem to involve
quantification. A version within a formalized theory that comes close to this line of
employing quantification can be found in section 142b in the appendix A1.
75b. Russell’s conception of vicious circles. Generalization leads on to

the notion of a totality.

[[A]]ll our contradictions have in common the assumption of a totality such
that, if it were legitimate, it would at once be enlarged by new members
defined in terms of itself.
Comment. In view of Cantor’s notion of ‘Vielheiten unverbundener Dinge’ in quo-
tation 73.13, one might feel inclined to add: “and thus fail to represent a ‘Vielheit
unverbundener Dinge’.” The task that remains, however, is to show how a peculiar
way of definition can establish a connection between the members of a ‘Vielheit’ in
such a way that removing one can have an effect on the membership of the others.
This, then, gives rise to the formulation of the vicious circle principle as it can
be found in Russell [1908] and Whitehead and Russell [1910].

These fallacies [[. . .]] are to be avoided by what may be called the ‘vicious-
circle principle’; i.e., ‘no totality can contain members defined in terms of
itself’. This principle, in our technical language, becomes: ‘Whatever con-
tains an apparent variable must not be a possible value of that variable’.
Thus whatever contains an apparent variable must be of a different type
from the possible values of that variable; we will say that it is of a higher
type. Thus the apparent variables contained in an expression are what de-
termines its type. This is the guiding principle[[.]]
Comment. What is so important about being defined in terms of itself? See next
quotation for Russell’s line of answer: not legitimate. I am not prepared to settle
for this line. Note that Russell changed “defined” to “can only be defined” in the
next quotation and “is only definable” in quotation 77.5. This is relevant in view
of Ramsey’s remark that “we may refer to a man as the tallest in a group” in
quotation 77.5 (2) below.
An analysis of the paradoxes to be avoided shows that they all result
from a certain kind of vicious circle. The vicious circles in question arise
from supposing that a collection of objects may contain members which can
only be defined by means of the collection as a whole. Thus, for example, the
collection of propositions will be supposed to contain a proposition stating
that “all propositions are either true or false”. It would seem, however, that
such a statement could not be legitimate unless “all propositions” referred
to some already definite collection, which it cannot do if new propositions
are created by statements about “all propositions”. We shall, therefore, have
to say that statements about “all propositions” are meaningless. More gen-
erally, given any set of objects such that, if we suppose the set to have a
total, it will contain members which presuppose this total, then such a set
cannot have a total.
After having thus stated the vicious circle principle, Russell sets out to a kind
of summary which I present in the next quotation.
Quotation 75.7. Whitehead and Russell [1910], p. 37.

The principle which enables us to avoid illegitimate totalities may be
stated as follows: “Whatever involves all of a collection must not be one
of the collection; or, conversely: “If, provided a certain collection had a
total, it would have members only definable in terms of that total, then the
said collection has no total”. We shall call this the “vicious circle principle”,
because it enables us to avoid the vicious circles involved in the assumption
of illegitimate totalities.
Comments. (α) This is virtually a repetition of Russell [1908], p. 63.
(β) The remarkable point here is that Russell explicitly formulates his vicious circle
principle in terms of ‘collections’, ‘sets’, ‘totals’ and their ‘objects’ and ‘members’,
while his examples are in terms of quantification.
75c. From totalities to functions. As a next step, Russell transfers the

considerations concerning collections to apply to functions.

[[A]] function is not a well-defined function unless all its values are already
well-defined. It follows from this that no function can have among its values
anything which presupposes the function [[. . .]]. [[T]]he function cannot be
definite until its values are definite. This is a particular case, but perhaps
the most fundamental case, of the vicious-circle principle.
The vicious circle principle as defined for totalities can now be easily transferred
to functions.
Quotations 75.9. (1) Whitehead and Russell [1910], p. 40.

[[T]]he values of a function cannot contain terms only definable in terms of
the function.
A function [[. . .]] presupposes as part of its meaning the totality of its values,
or, what comes to the same thing, the totality of its possible arguments.
75d. Theory of description and no-class theory. This is one of the most
important steps in Russell’s analysis, the contextual definition of the notion of
class which enables him to present a unified approach to all antinomies by limiting
variables.

The paradoxes of symbolic logic concern various sorts of objects: propo-
sitions, classes, cardinal and ordinal numbers, etc. All these sorts of objects,
as we shall show, represent illegitimate totalities, and are therefore capa-
ble of giving rise to vicious-circle fallacies. But by means of [[our]] theory
[[. . .]] which reduces statements that are verbally concerned with classes and
relations to statements that are concerned with propositional functions,
the paradoxes are reduced to such as are concerned with propositions and
propositional functions.
Comment. Functions here play a role comparable to that of classifications in Poin-
caré’s analysis: classes are based on them.
(2) Whitehead and Russell [1910], pp. 62 f.
[[A]] proposition about a class is always to be reduced to a statement about
a function which defines the class, i.e. about a function which is satisfied
by the members of the class and by no other arguments. Thus a class is
an object derived from a function and presupposing the function, just as,
for example,(x).φx presupposes the function φx̂ Hence a class cannot, by
the vicious-circle principle, significantly be the argument to its defining
function, that is to say, if we denote by “ ẑ(φz)” the class defined by φẑ, the
symbol “φ{ẑ(φz)}” must be meaningless.
(3) Russell [1908], p. 62.
That there is no such class results from the fact that, if we suppose that
there is, the supposition gives rise (as in the above contradiction) to new
classes lying outside the supposed total of all classes.
By saying that a set has “no total”, we mean primarily, that no significant
statement can be made about “all its members”. Propositions, as the above
illustration shows, must be a set having no total. The same is true, as we
shall shortly see, of propositional functions, even when these are restricted
to such as can significantly have as argument a given object a. In such
cases, it is necessary to break up our set into smaller sets, each of which is
capable of a total. This is what the theory of types aims at effecting.
Comment. The first sentence is a modified version of footnote * in Russell [1908],

p. 63; there he actually speaks of “nonsense”. Russell’s doctrine of orders and types
excludes all impredicative collections on the grounds that they are not meaningful
expressions, i.e., , if an impredicative collection has no total, it does not occur in
the formal system at all. This is a large claim. Why should no significant statement
be possible? Admittedly it may be possible that some statements among these are
not truth-definite, but why exclude them all? There are perfectly harmless ones like
A → A where A is about an impredicative collection. Later I shall discuss attempts
to escape this global verdict of meaninglessness. Compare Kant’s ‘sum-total’ of
appearances! What sort of statements can be made about this?
[[T]]he appearance of contradiction is produced by the presence of some word
which has systematic ambiguity of type, such as truth, falsehood, function,
property, class, relation, cardinal, ordinal, name, definition. Any such word,
if its typical ambiguity is overlooked, will apparently generate a totality
containing members defined in terms of itself, and will thus give rise to
vicious-circle fallacies. [[. . .]]
Thus the appearance of contradiction is always due to the presence of
words embodying a concealed typical ambiguity, and the solution of the
apparent contradiction lies in bringing the concealed ambiguity to light.
The symbols for classes like those for descriptions, are, in our system, incom-
plete symbols; their uses are defined, but they themselves are not assumed
to mean anything at all. [[. . .]] Thus classes, so far as we introduce them,
are merely symbolic or linguistic conveniences, not genuine objects.
Comment. In other words, a façon de parler.
(7) Russell [1919], pp. 181 f; quoted after Chihara [1972], p. 260, footnote 16.
We must seek a definition on the same lines as the definition of descriptions,
i.e., a definition which will assign a meaning to propositions in whose verbal
or symbolic expression words or symbols apparently representing classes oc-
cur, but which will assign a meaning that altogether eliminates all mention
of classes from a right analysis of such propositions.
(8) Russell [1944], p. 14; quoted after Chihara [1972], p. 260.
What was of importance in this theory was the discovery that, in analyzing
a significant sentence, one must not assume that each separate word or
phrase has significance on its own account. . . . It soon appeared that class-
symbols could be treated like descriptions, i.e., as non-significant parts of
significant sentences. This made it possible to see, in a general way, how a
solution to the contradictions might be possible.
Remark 75.11. Russell’s approach of introducing class terms is commonly de-

scribed as a ‘contextual definition’, e.g., in Quine [1970]. It renders an axiom of
extensionality redundant, simply because it does away with sets as independent
objects. For the latter reason, it is also a convenient way of avoiding the trouble
connected with intensional operators and substitutivity of identicals.14
75e. From the vicious circle principle to the hierarchy of propositions

and propositional functions. Once a kind of vicious circle was made responsible
for the emergence of the antinomies, a precise method to avoid such circles had to
be developed.
Quotations 75.12. (1) U. Petersen [2002], p. 1065; translating Russell [1906a],

p. 627.
Mr Poincaré believes that all the paradoxes stem from a kind of vicious
circle, and this is where I agree with him. But he doesn’t see the difficulty
in avoiding a vicious circle of this sort. I shall show that, if one wants to
avoid it, one has to adopt a theory similar to my “no-classes theory”; indeed,
it is to this end that I have invented it.
(2) U. Petersen [2002], p. 1065; translating Russell [1906a], p. 640.
It is important to observe that the vicious circle principle is not itself the
solution of the vicious circle paradoxes, but only the consequence which a
theory has to yield in order to provide a solution. In other words, one has
to construct a theory of expressions containing apparent variables which
would yield as a consequence the vicious circle principle.
Comment. I put some emphasis on this quotation in view of the distinction that
Kant makes between a critical decision and a resolution regarding the antinomy of
reason.15
In his paper [1908], Russell introduced a hierarchy of propositions and (propo-

sitional) functions which was to become known as the ramified theory of types as
opposed to the simple theory of types. The next set of quotations shows how Russell
introduces his hierarchy.

We are thus led to the conclusion, both from the vicious-circle principle
and from direct inspection, that the functions to which a given object a can
be an argument are incapable of being arguments to each other, and that
they have no term in common with the functions to which they can be argu-
ments. We are thus led to construct a hierarchy. Beginning with a and the
other terms which can be arguments to the same functions to which a can
be argument, we come next to functions to which a is a possible argument,
and then to functions to which such functions are possible arguments, and
so on.
Comment. Both times the crucial step is buried in the phrase “We are thus led
to . . . ” Observe, however, that the construction of a hierarchy is by no means a
necessary conclusion. I shall present a different option in the appendix A1.
14 See section 90b for some aspects of this issue.
15 See sections 62c and 62d in these materials
So far, this amounts to a simple theory of types. But the situation is more
complicated than that.
Quotation 75.14. Whitehead and Russell [1910], pp. 49 f.

[[T]]he hierarchy which has to be constructed is not so simple as might at first
appear. The functions which can take a as argument form an illegitimate
totality, and themselves require division into a hierarchy of functions. This
is easily seen as follows. Let f (φẑ, x) be a function of the two variables
φẑ and x. Then if, keeping x fixed for the moment, we assert this with all
possible values of φ , we obtain a proposition:
(φ) . f (φẑ, x) .
Here, if x is variable, we have a function of x; but as this function involves
a totality of values of φẑ, it cannot itself be one of the values included in
the totality, by the vicious-circle principle. It follows that the totality of
values of φẑ concerned in (φ) . f (φẑ, x) is not the totality of all functions in
which x can occur as argument, and that there is no such totality as that
of all functions in which x can occur as argument.
V
Comment. In the terminology of my tools this means that λy xVF[x, y] must not
be a possible value of the bound variable x, since, otherwise, λy x F[x, y] would
involve a reference to a totality, viz., any totality of all terms s such that F[s, t]
holds for fixed t, of which it is an element, i.e., it occurs among the s.

It follows from the above that a function in which φẑ appears as argu-
ment requires that “φẑ” should not stand for any function which is capable
of a given argument, but must be restricted in such a way that none of the
functions which are possible values of “φẑ” should involve any reference to
the totality of such functions. Let us take as an illustration the definition
of identity. We might attempt to define “x is identical with y” as meaning
“whatever is true of x is true of y”, i.e. “φx always implies φy.” But there,
since we are concerned to assert all values of “φx implies φy” regarded as
a function of φ , we shall be compelled to impose upon φ some limitation
which will prevent us from including among values of φ values in which
“all possible values of φ” are referred to. Thus for example “x is identical
with a” is a function of x; hence it is a legitimate value of φ in “φx always
implies φy,” we shall be able to infer, by means of the above definition, that
if x is identical with a, and x is identical with y, then y is identical with a.
Although the conclusion is sound, the reasoning embodies a vicious circle
fallacy, since we have taken “(φ) . φx implies φa” as a possible value of φx,
which it cannot be.
Remark 75.16. I distinguish between Russell’s vicious circle principle and the ram-
ified theory of types. The first is a philosophical analysis, the second a technical de-
vice to do justice to the philosophical analysis. This is in accordance with Russell’s
remark in quotation 75.12 (2) above.
§ 76. AXIOMATIC SET THEORY, FORMALISM, AND PROOF THEORY 1067
I close this paragraph on Russell’s analysis of the paradoxes with a quotation

regarding the non-logical character of the axiom of infinity which is needed, if the
simple theory of types is to serve as a foundation of arithmetic.
Quotation 75.17. Whitehead and Russell [1912], p. 183; quoted after Copi [1971],
p. 64.
It seems plain that there is nothing in logic to necessitate its truth or
falsehood, and that it can only be legitimately believed or disbelieved on
empirical grounds.
§ 76. Axiomatic set theory, formalism, and proof theory
In this paragraph I shall present relevant aspects of Zermelo’s axioms for set theory,
Hilbert ’s proof theory, and Brouwer ’s neo-intuitionism.
76a. Zermelo’s axioms for set theory. In the same year that Russell pub-
lished what was to become known as the ramified theory of types, Zermelo pub-
lished his axioms for set theory the formalized versions of which dominate today’s
axiomatic set theory.
Quotation 76.1. van Heijenoort [1967b], p. 200; translation of Zermelo [1908].

[[I]]n view of the “Russell antinomy” (1903, pp. 101–107 and 366–368) of the
set of all sets that do not contain themselves as elements, it no longer seems
admissible today to assign to an arbitrary logically definable notion a set,
or class, as its extension. Cantor’s original definition of a set (1895 ) as “a
collection, gathered into a whole, of certain well-distinguished objects of our
perception or our thought” therefore certainly requires some restriction; it
has not, however, been successfully replaced by one that is just as simple and
does not give rise to such reservations. Under these circumstances there is at
this point nothing left for us to do but to proceed in the opposite direction
and, starting from set theory as it is historically given, to seek out the
principles required for establishing the foundations of this mathematical
discipline. In solving the problem we must, on the one hand, restrict these
principles sufficiently to exclude all contradictions and, on the other, take
them sufficiently wide to retain all that is valuable in this theory.
Comment. The last sentence has become a somewhat standard idea for what is
considered a solution of the set theoretical paradoxes.
Zermelo presents axioms of extensionality, null, unit, pair, separation, power set,
union, choice, and infinity.16 In the following quotations I want to get an impression
of his reasoning as concerns the axiom of separation.
16 For their formulation within a formalized theory, see chapter XIII in the tools.
Quotations 76.2. (1) Van Heijenoort [1967b], p. 201; translation of Zermelo [1908].
4. A question or assertion E(x) is said to be definite if the fundamen-
tal relations of the domain, by means of the axioms and the universally
valid laws of logic, determine without arbitrariness whether it holds or not.
Likewise a “propositional function” [[. . .]] E(x), in which the variable term x
ranges over all individuals of a class K, is said to be definite if it is definite
for each single individual x of the class K. Thus the question whether aεb
or not is always definite, as is the question whether M =⊂ N 17 or not.
(2) Van Heijenoort [1967b], p. 202; translation of Zermelo [1908].
Axiom II. (Axiom of elementary sets [[. . .]].) There exists a (fictitious)
set, the null set, 0, that contains no element at all. If a is any object of
the domain, there exists a set {a} containing a and only a as element; if
a and b are any two objects of the domain, there always exists a set {a, b}
containing as elements a and b but no object distinct from both.
Axiom III. (Axiom of separation [[. . .]].) Whenever the propositional
function E(x) is definite for all elements of a set M , M possesses a subset
ME containing as elements precisely those elements x of M for which E(x)
is true.
By giving us a large measure of freedom in defining new sets, Axiom
III in a sense furnishes a substitute for the general definition of set that was
[[. . .]] rejected as untenable. It differs from that definition in that it contains
the following restrictions. In the first place, sets may never be independently
defined by means of this axiom but must always be separated as subsets from
sets already given; thus contradictory notions such as “the set of all sets” or
“the set of all ordinal numbers”, and with them the “ultrafinite paradoxes”,
to use Hessenberg’s expression (1906, chap. 24), are excluded. In the second
place, moreover, the defining criterion must always be definite in the sense
of [[above]] (that is, for each single element x of M the fundamental relations
of the domain must determine whether it holds or not), with the result that,
form our point of view, all criteria such as “definable by means of a finite
number of words”, hence the “Richard antinomy” and the “paradox of finite
denotation”, vanish.
Comment. As regards the reference to Hessenberg’s expression “ultrafinite paradox-
es”, cf. quotation 73.18 (2).
Axiom IV. (Axiom of the power set [[. . .]].) To every set T there cor-
responds another set UT , the power set of T , that contains as elements
precisely all subsets of T .
Axiom V. (Axiom of the union [[. . .]].) To every set T there corresponds
a set ST , the union of T , that contains as elements precisely all elements
of T .
17 In the symbolism of the tools: M ⊆ N .

Axiom VI. (Axiom of choice [[. . .]].) If T is a set whose elements all are
sets that are different from 0 and mutually disjoint, its union ST includes
at least one subset S1 having one and only one element in common with
each element of T .
[[. . .]]
Axiom VII. (Axiom of infinity [[. . .]].) There exists in the domain at least
one set Z that contains the null set as an element and is so constituted that
to each of its elements a there corresponds a further element of the form
{a}, in other words, that with each of its elements a it also contains the
corresponding set {a} as an element.
76b. Hilbert’s proof theory. The modern conception of formalism in math-

ematics with its permanent companion proof theory is intimately linked to the name
of David Hilbert. In a number of articles beginning with Über den Zahlbegriff [1900]
and Mathematische Probleme [1900] Hilbert developed his idea of proof theory.
Features 76.3. I shall focus on the following issues:

(HB1) axiomatic vs. genetic method;18
(HB2) in the beginning is the symbol;19
(HB3) we must supplement the finitary statements with ideal statements;20
(HB4) proof theory (metamathematics).
Literature 76.4. Bernays [1935a], Curry [1954], Hilbert [1917] (shift from axiom-
atic to proof theory), [1922], [1923], Kreisel [1958] (Hilbert’s programme), Schütte
[1965], [1980a], [1980b], Prawitz [1981].
Hilbert’s background is quite profound and it seems worthwhile illustrating

this with a few quotations. Hilbert’s earliest publication dealing with the question
of foundations of mathematics vis à vis the antinomies seems to be Hilbert [1904].
It was only thirteen years later that he then returned to this subject in print.
Quotations 76.5. (1) Van Heijenoort [1967b], p. 130; translating Hilbert [1904].
L. Kronecker, as is well known, saw in the notion of the integer the real
foundation of arithmetic; he came up with the idea that the integer—and,
in fact, the integer as a general notion (parameter value)—is directly and
immediately given; this prevented him from recognizing that the notion of
integer must and can have a foundation. I would call him a dogmatist, to the
extent that he accepts the integer with its essential properties as a dogma
and does not look further back.
18 See Hilbert [1900], also Hilbert and Bernays [1934], pp. 1 f and p. 20 (Existenzpostulate vs.
Konstruktionspostulate).
19 See Hilbert [1922], p. 163; cf. quotation 76.9 (2) below.
20 See Hilbert [1925], p. 174; cf. quotations 76.8 below.
(2) Van Heijenoort [1967b], p. 130; translating Hilbert [1904].

G. Frege sets himself the task of founding the laws of arithmetic by the
devices of logic, taken in the traditional sense. He has the merit of having
correctly recognized the essential properties of the notion of integer as well
as the significance of inference by mathematical induction. But, true to
his plan, he accepts among other things the fundamental principle that a
concept (a set) is defined and immediately usable if only it is determined for
every object whether the object is subsumed under the concept or not, and
here he imposes no restriction on the notion “every”; he thus exposes himself
to precisely the set-theoretic paradoxes that are contained, for example,
in the notion of the set of all sets and that show, it seems to me, that
the conceptions and means of investigation prevalent in logic, taken in the
traditional sense, do not measure up the rigorous demands that set theory
imposes. Rather, from the very beginning a major goal of the investigations
into the notion of number should be to avoid such contradictions and to
clarify these paradoxes.
G. Cantor sensed the contradictions just mentioned and expressed this
awareness by differentiating between “consistent” and “inconsistent sets”.
But, since, in my opinion he does not provide a precise criterion for this
distinction, I must characterize his conception on this point as one that still
leaves latitude for subjective judgment and therefore affords no objective
certainty.
Quotations 76.6. (1) Benacerraf and Putnam (eds.) [1964], p. 141; translating
Hilbert [1925], p. 169.
In the joy of discovering new and important results, mathematicians paid
too little attention to the validity of their deductive methods. For, simply as
a result of employing definitions and deductive methods which had become
customary, contradictions began gradually to appear. These contradictions,
the so-called paradoxes of set theory, though at first scattered, became
progressively more acute and more serious. In particular, a contradiction
discovered by Zermelo and Russell had a downright catastrophic effect when
it became known throughout the world of mathematics. Confronted by these
paradoxes, Dedekind and Frege completely abandoned their point of view
and retreated. Dedekind hesitated a long time before permitting a new
edition of his epoch-making treatise Was sind und was sollen die Zahlen
to be published. In an epilogue, Frege too had to acknowledge that the
direction of his book Grundgesetze der [[Arithmetik ]]21 was wrong. Cantor’s
doctrine, too, was attacked on all sides. So violent was this reaction that
even the most ordinary and fruitful concepts and the simplest and most
important deductive methods of mathematics were threatened and their
employment was on the verge of being declared illicit.
21 Hilbert’s original text is correct. For some reason the translation has “Mathematik” instead
of “Arithmetik”.
Comment. I see a certain connection to Kant A4, translated in N. K. Smith [1929],

pp. 46: “Mathematics gives us a shining example . . . ”.
(2) U. Petersen [2002], p. 1071; translating Hilbert [1922], p. 162.
Frege attempted the foundation of number theory on pure logic,
Dedekind on set theory as a chapter of pure logic: both did not achieve
their goal. Frege did not deal cautiously enough with the customary con-
cept formations of logic in their application to mathematics: thus he re-
garded the extension of a concept as something just given, in such a way
that he then felt entitled to take again these extensions without restriction
as objects themselves. As it were he fell for an extreme concept realism.
Dedekind fared similarly; his classic error consisted in taking the system
of all objects as starting point.
(3) Benacerraf and Putnam (eds.) [1964], p. 142; translation of Hilbert [1925],
pp. 170 f.
Does material logical deduction somehow deceive us or leave in the lurch
when we apply it to real things or events? No! Material logical deduction is
indispensable. It deceives us only when we form arbitrary abstract defini-
tions, especially those which involve infinitely many objects. In such cases
we have illegitimately used material logical deduction; i.e., we have not paid
sufficient attention to the preconditions necessary for its valid use. In rec-
ognizing that there are such preconditions that must be taken into account,
we find ourselves in agreement with the philosophers, notably with Kant.
Kant taught—and it is an integral part of his doctrine—that mathematics
treats a subject matter which is given independently of logic. Mathematics,
therefore, can never be grounded solely on logic. Consequently, Frege’s and
Dedekind’s attempts to so ground it were doomed to failure.
As a further precondition for using logical deduction and carrying our
logical operations, something must be given in conception, viz., certain
extralogical concrete objects which are intuited as directly experienced prior
to all thinking. For logical deduction to be certain, we must be able to see
every aspect of these objects, and their properties, differences, sequences,
and contiguities must be given, together with the objects themselves, as
something which cannot be reduced to something else and which requires
no reduction.
(4) Hilbert and Bernays [1934/68], pp. 15 f.
Es war zuerst Frege, der mit allem Nachdruck und mit scharfer, wit-
ziger Kritik die Forderung zur Geltung brachte, daß die Vorstellung der
Zahlenreihe als einer fertigen Gesamtheit durch einen Nachweis ihrer Wi-
derspruchsfreiheit gesichert werden müsse. Ein solcher Nachweis war nach
Freges Meinung nur im Sinne einer Aufweisung, als Existenzbeweis, zu
führen, und er glaubte, die Objekte für eine solche Aufweisung im Bereich
der Logik zu finden. Sein Verfahren der Aufweisung kommt darauf hinaus,
daß er die Gesamtheit der Zahlen definiert mit Hilfe der als existierend vor-
ausgesetzten Gesamtheit aller überhaupt denkbaren einstelligen Prädikate.
Aber die hierbei zugrunde gelegte Voraussetzung, welche ohnehin schon ei-
ner unbefangenen Betrachtung als sehr verdächtig erscheint, hat sich durch
die berühmten, von Russell und Zermelo entdeckten logischen und men-
gentheoretischen Paradoxien als unhaltbar erwiesen. Und das Mißlingen des
Fregeschen Unternehmens hat mehr noch als Freges Dialektik das Pro-
blematische an der Annahme der Totalität der Zahlenreihe zum Bewußtsein
gebracht.
Wir können nun angesichts dieser Problematik versuchen, an Stelle der
Zahlenreihe einen anderen unendlichen Individuenbereich für die Zwecke
der Widerspruchsfreiheitsbeweise zu verwenden, der nicht wie die Zahlen-
reihe ein reines Gedankengebilde, sondern aus dem Gebiet der sinnlichen
Wahrnehmung oder aber der realen Wirklichkeit entnommen ist. Sehen wir
aber näher zu, so werden wir gewahr, daß überall, wo wir im Gebiet der
Sinnesqualitäten oder in der physikalischen Wirklichkeit unendliche Man-
nigfaltigkeiten anzutreffen glauben, von einem eigentlichen Vorfinden einer
solchen Mannigfaltigkeit keine Rede ist, daß vielmehr die Überzeugung von
dem Vorhandensein einer solchen Mannigfaltigkeit auf einer gedanklichen
Extrapolation beruht, deren Berechtigung jedenfalls ebensosehr der Prü-
fung bedarf wie die Vorstellung von der Totalität der Zahlenreihe.
Hilbert [1904] proposes the axiomatic method as the way to cope with the
troubles.
Quotation 76.7. (1) Van Heijenoort [1967b], p. 131; translation of Hilbert [1904].
It is my opinion that all the difficulties touched upon can be overcome
and that we can provide a rigorous and completely satisfying foundation for
the notion of number, and in fact by a method that I would call axiomatic
and whose fundamental idea I wish to develop briefly in what follows.
(2) Van Heijenoort [1967b], pp. 135 f; translating Hilbert [1904].
The principles that must constitute the standard for the construction
and further elaboration of the laws of mathematical thought in the way
envisaged here are, briefly, the following.
I. Once arrived at a certain stage in the development of the theory, I
may say that a further proposition is true as soon as we recognize that no
contradiction results if it is added as an axiom to the propositions previously
found true [[. . .]]
II. In the axioms the arbitrary objects—taking the place of the notion
“every” and “all” in ordinary logic—represent only those thought-objects
and their mutual combinations that at this stage are taken as primitive or
are to be newly defined. [[. . .]]
III. A set is generally defined as a thought-object m, and the combi-
nations mx are called the elements of the set m, so that—contrary to the
usual conception—the notion of element of a set appears only as a subse-
quent product of the notion of set itself.

Point I expresses the creative principle that, in its freest use, justifies
us in forming ever new notions, with the sole restriction that we avoid a
contradiction. The paradoxes [[. . .]] become impossible by virtue of II and
III[[.]]
In Hilbert [1904], p. 134, one can already find the idea of proving the consistency
of an axiomatic theory by showing that all the axioms have a certain property which
is preserved by the rules, but not every proposition has it. As I mentioned before,
it took quite a while before Hilbert returned to this subject: Axiomatisches Denken
[1917], Über das Unendliche [1925] and Die Grundlagen der Mathematik [1927].
Quotations 76.8. (1) Benacerraf and Putnam (eds.) [1964], p. 138; translation of
Hilbert [1925], p. 174.
We encounter a [[. . .]] quite unique conception of the notion of infin-
ity in the important and fruitful method of ideal elements. The method
of ideal elements is used even in elementary plane geometry. The points
and straight lines of the plane originally are real, actually existent objects.
One of the axioms that hold for them is the axiom of connection: one and
only one straight line passes through two points. It follows from this axiom
that two straight lines intersect at most at one point. There is no theorem
that two straight lines always intersect at one point, however, for the two
straight lines might well be parallel. Still we know that by introducing ideal
elements, viz., infinitely long lines and points at infinity, we can make the
theorem that two straight lines always intersect at one and only one point
come out universally true. These ideal “infinite” elements have the advan-
tage of making the system of connection laws as simple and perspicuous as
possible. Moreover, because of the symmetry between a point and a straight
line, there results the very fruitful principle of duality for geometry.
Another example of the use of ideal elements are the familiar complex-
imaginary magnitudes of algebra which serve to simplify theorems about
existence and number of the roots of an equation.
Just as infinitely many straight lines, viz., those parallel to each other
are used to define an ideal point in geometry, so certain systems of infin-
itely many numbers are used to define an ideal number. This application
of the principle of ideal elements is the most ingenious of all. If we apply
this principle systematically throughout an algebra, we obtain exactly the
same simple and familiar laws of division which hold for the familiar whole
numbers 1, 2, 3, 4, . . . .
(2) Benacerraf and Putnam (eds.) [1964], p. 145; Hilbert [1925], p. 174.
Let us remember that we are mathematicians and that as mathemati-
cians we have often been in precarious situations from which we have √ been
rescued by the ingenious method of ideal elements. [[. . .]] Just as i = −1
was introduced to preserve in simplest form the laws of algebra (for exam-
ple, the laws about the existence and number of the roots of an equation);
just as ideal factors were introduced to preserve the simple laws of divisibil-
ity for algebraic whole numbers (for example, a common ideal divisor really
exists); similarly, to preserve the simple formal rules of ordinary Aristotelian
logic, we must supplement the finitary statements with ideal statements.
Quotations 76.9. (1) Prawitz [1981], p. 235; translating Hilbert [1917], p. 155.
[[W]]e must make the concept of specific mathematical proof itself object
of investigation, just as also the astronomer pays attention to his place of
observation, the physicist must care about the theory of his instrument,
and the philosopher criticizes reason itself.
Comment. This seems to be the first formulation of the idea that was to become
the basis of Hilbert’s proof theory. It can be found almost unchanged in quotation
76.9 (3) below which is taken from a paper of Hilbert’s published five years later.
(2) U. Petersen [2002], p. 1074; translating Hilbert [1922], pp. 162 f.
[[A]]bstract operating with general concept extensions and contents has
shown itself to be inadequate and unsafe. As precondition for the applica-
tion of logical inferences and the activating of logical operations something
has to be already given in intuition: certain extralogical discrete objects,
which exist intuitively as immediate experience prior to all thinking. If log-
ical inferring is to be safe, a complete overview of these objects in all their
parts has to be possible and their showing, distinction, succession exists im-
mediately intuitively for us at the same time with the objects as something
which cannot be further reduced to anything else. By taking this position,
the objects of number theory are to me — in exact contrast to Frege and
Dedekind — the symbols themselves, whose shape is generally and safely
recognizable by us independent of place and time and the special conditions
of the production of the symbol as well as minor differences in execution.
In this lies the solid philosophical attitude which I consider necessary for
the foundation of pure mathematics — as well as in general for all scientific
thinking, understanding and communicating: in the beginning — thus it
says — is the symbol.
In order to achieve our aims we have to make proofs as such the object
of our investigations; we are thus driven to some kind of a proof theory
which is about the operating with proofs themselves. For the concretely-
visualizable (“konkret-anschauliche”) number theory [[. . .]] the numbers were
the objective (“das Gegenständliche”) and presentable (“Aufweisbare”), and
the proofs of the theorems about numbers were already part of the con-
ceptual (“gedankliche”) realm. In our investigation now the proof itself is
something concrete and presentable: the contentual (“inhaltliche”) consider-
ations only take place with regard to the proof. Just like the physicist inves-
tigates his apparatus, the astronomer his position, just like the philosopher
practices critique of reason, the mathematician, according to my mind, has
to secure his theorems by means of a critique of proofs, and that is what
he needs this proof theory for.
(4) Benacerraf and Putnam (eds.) [1964], p. 151; translating Hilbert [1925], p. 190.
[[T]]he infinite is nowhere to be found in reality. It neither exists in nature
nor provides a legitimate basis for rational thought—a remarkable harmony
between being and thought. In contrast to the earlier efforts of Frege and
Dedekind, we are convinced that certain intuitive concepts and insights are
necessary conditions of scientific knowledge, that logic alone is not sufficient.
Operating with the infinite can be made certain only by the finitary.
The role that remains for the infinite to play is solely that of an idea—
if one means by an idea, in Kant’s terminology, a concept of reason which
transcends all experience and which completes the concrete as a totality—
that of an idea which we may unhesitatingly trust within the framework
erected by our theory.
Comment. Note the phrase “harmony of being and thought”; an allusion to Par-
menides?
(5) U. Petersen [2002], p. 1075; translating Hilbert [1923], p. 179.22
The basic idea of my proof theory is the following:
Everything that makes up mathematics in the present sense will be
strictly formalized, so that mathematics proper or mathematics in the more
narrow sense turns into a stock of formulas. These differ from the usual
formulas of mathematics only in this, that apart from the usual symbols
also the logical symbols, in particular those for “implies” (→) and for “not”
( ), occur. Certain formulas, which serve as building blocks of the formal
edifice of mathematics, are called axioms. A proof is a figure which as such
has to be visually (“anschaulich”) before us; it consists of inferences by virtue
of the schema of inference
S
S→T
,
T
where every time each of the premisses, i.e. of the respective formulas S
and S → T, is either an axiom or results directly from an axiom by means
of substitution, or coincides with the end formula S of an inference which
occurs previous in the proof or results from such an end formula by means
of substitution. A formula shall be called provable, if it is either an axiom
or results from an axiom by means of substitution or is the end formula of
a proof.
To the thus formalized mathematics proper comes a, so to speak, new
mathematics, a metamathematics, which is necessary for safeguarding the
former, in which — in contrast to the purely formal ways of inference of
mathematics proper — material23 inferring is applied, but only as proof
of the consistency of the axioms. In this metamathematics one operates
with the proofs of mathematics proper and these latter themselves make
22 See also Hilbert [1931], p. 489, or [1935], p. 192 (repr.), for an almost identical passage.
23 I use the word “material” for the German “inhaltlich” as is done in the translation of Hilbert
[1926]. An alternative would be “contentual”, which is used in the translations of Bauer-Mengelberg.
up the objects of material investigation. In this way the development of

the mathematical science as a whole takes place in constant change in two
different ways: by gaining new provable formulas from the axioms by way
of formal inference and on the other hand by adding new axioms together
with the proof of consistency by means of material inference.
Comment. Part of this quotation can already be found in Hilbert [1922], p. 169.
(5) Hilbert [1928], p. 15.
Wenn eine Aussage oder ein Beweis vorliegt, so muß er sich in allen
Teilen überblicken lassen. Die Aufweisung, das Wiedererkennen, die Un-
terscheidung und Aufeinanderfolge seiner einzelnen Teile ist unmittelbar
anschaulich für uns da.
Comment. Gödel’s remarks in quotation 94.12 lead on from here.
Remarks 76.10. (1) Proof theory (German: Beweistheorie), is in character a philo-

sophical theory employing mathematical methods. The use of the notion proof the-
ory in, e.g., Priest, Read and others for what I call syntax here, and Hodges [1983],
p. 107, calls a “formal proof system”, must not be confused with Hilbert’s introduc-
tion of the term “Beweistheorie” and the traditional use of the term in, e.g., Schütte,
Takeuti, Kreisel, Prawitz, Girard, Pohlers, etc. In the translation of Hilbert [1925]
in Benacerraf and Putnam [1964], “Beweistheorie” is translated as “theory of proof”.
(2) Paul Bernays, Hilbert’s close collaborator, remarks in his paper [1937], p. 83,24
that the separation of mathematics and meta-mathematics is formed after Kant’s
division of philosophy into critique and system.
Quotation 76.11. (1) Van Heijenoort [1967b], p. 473; translation of Hilbert [1927].
Russell’s and Whitehead’s theory of foundations is a general logical
investigation of wide scope. But the foundation that it provides for mathe-
matics rests, first, upon the axiom of infinity and, then, upon what is called
the axiom of reducibility, and both of these axioms are genuine contentual
assumptions that are not supported by a consistency proof; they are as-
sumptions whose validity in fact remains dubious and that, in any case, my
theory does not require.
(2) Van Heijenoort [1967b], p. 473; translation of Hilbert [1927].
Now with regard to the most recent investigations, the fact that re-
search on foundations has again come to attract such lively appreciation
and interest certainly gives me the greatest pleasure. When I reflect on
the content and the results of these investigations, however, I cannot for
the most part agree with their tendency; I feel, rather, that they are to a
large extent behind the times, as if they came from a period when Cantor’s
majestic world of ideas had not yet been discovered.
In this I see the reason, too, why these most recent investigations in
fact stop short of the great problems of the theory of foundations, for ex-
ample, the question of the construction of functions, the proof or refutation
24 Cf. quotation 78.3 in the next chapter.
of Cantor’s continuum hypothesis, the question whether all mathematical

problems are solvable, and the question whether consistency and existence
are equivalent for mathematical objects.
Further reading 76.12. Sieg [1999], Hilbert’s Programs: 1917–1922, looks to me

like a particular fine piece of historical scholarship; I think it is strongly recommend-
able for anyone who wants to get deeper into the details of Hilbert’s development,
but I must admit that I haven’t done it myself, so far, at least.
76c. Induction, recursion, and the problem of impredicativity.
Quotations 76.13. (1) Van Heijenoort [1967b], pp. 472 f; translation of Hilbert
[1927].
Poincaré already made various statements that conflict with my views;
above all, he denied from the outset the possibility of a consistency proof for
the arithmetic axioms, maintaining that the method of mathematical induc-
tion could never be proved except through the inductive method itself. But,
as my theory shows, two distinct methods that proceed recursively come
into play when the foundations of arithmetic are established, namely, on the
one hand, the intuitive construction of the integer as numeral [[. . .]], that is,
contentual induction, and on the other hand formal induction proper, which
is based on the induction axiom and through which alone the mathematical
variable can begin to play its role in the formal system.
Poincaré arrives at his mistaken conviction by not distinguishing be-
tween these two methods of induction, which are of entirely different kinds.
(2) Hilbert [1928], p. 10.
Eine unglückliche Auffassung Poincarés, betreffend den Schluß von n auf
n+ 1, die bereits zwei Jahrzehnte früher von Dedekind durch einen präzisen
Beweis widerlegt worden war, verrammelte den Weg zum Vorwärtsschrei-
ten. Ein neues Verbot, das Verbot der imprädikativen Aussagen, wurde
von Poincaré erlassen und aufrechterhalten, obwohl Zermelo sofort ein
schlagendes Beispiel gegen dieses Verbot angab und dieses Verbot außerdem
gegen die Resultate Dedekinds verstieß.
Comment. Cf. Nelson [1986], quotations 82.12 in these materials, for a revival of
concerns regarding induction.
The elementary aids for the formation of functions are apparently the
instantiation (i.e. substitution of an argument by a new variable or func-
tion) and the recursion (according to the schema of derivation of the value
of a function for n + 1 from that for n).
One might think that still other elementary methods of definition would
have to be added to these two processes of instantiation and recursion, e.g.
the definition of a function through stating its values up to a certain point,
from which on the function is to be constant; further the definition through
elementary processes which are obtained from calculations, as for instance
the remainder of a division or the greatest common factor of two numbers,

or also the definition of a number as the smallest amongst certain finitely
many numbers.
It turns out, however, that all such definitions can be represented as
special cases of the application of instantiation and recursion. The method
of finding the necessary recursions is essentially equivalent with that con-
sideration by means of which one recognizes the particular procedure of
definition as finite.
(4) Hilbert and Bernays [1934/68], p. 301.
Ein System von zwei Gleichungen der Gestalt
f(0) = a ,
f(n′ ) = h(n, f(n))
stellt eine Anforderung an die Funktion f(n) dar, deren Erfüllbarkeit nicht
aus der Struktur der Rekursionsgleichungen an und für sich hervorgeht, son-
dern auf den charakteristischen Eigenschaften der Strichfunktion beruht,
nämlich daß diese Funktion niemals zu dem Wert 0 führt, und daß zwei
verschiedenen Argumentwerten auch stets verschiedene Werte der Strich-
funktion entsprechen.
Die Forderung der Zulassung von rekursiven Definitionen ist also gleich-
bedeutend mit einer impliziten Charakterisierung der Strichfunktion. Und
zwar betrifft diese Charakterisierung gerade die beiden Eigenschaften der
Strichfunktion, auf Grund deren sie eine Abbildung im Sinne der Dede-
kindschen Unendlichkeitsdefinition liefert[[.]]
[[Die]] Methode, nach der man die Zahlentheorie formal, an Hand der Re-
kursionen und des Induktionsschemas, unter Vermeidung von gebundenen
Variablen entwickelt [[. . .]] soll als die rekursive Behandlung der Zahlentheo-
rie oder auch kurz als die „rekursive Zahlentheorie“ bezeichnet werden.
Diese rekursive Zahlentheorie steht insofern der anschaulichen Zahlen-
theorie [[. . .]] nahe, als ihre Formeln sämtlich einer finiten inhaltlichen Deu-
tung fähig sind. Diese inhaltliche Deutbarkeit ergibt sich aus der [[. . .]] Veri-
fizierbarkeit aller ableitbaren Formeln der rekursiven Zahlentheorie. In der
Tat hat in diesem Gebiet die Verifizierbarkeit den Charakter einer direkten
inhaltlichen Interpretation[[.]]
I close this section with two quotations which are meant to indicate a bit of the
grand style in the design of Hilbert’s proof theory.
Quotations 76.14. (1) Hilbert and Bernays [1934/68], p. 381.

Wie man weiß ist der große Fermatsche Satz [[. . .]] nicht bewiesen, und
man hat auch keine Methode, welche gestattet, für jede vorgelegte Ziffer k
zu entscheiden, ob für sie die Behauptung des großen Fermatschen Satzes
zutrifft. Das Reduktionsverfahren für das System (Z) [[der Zahlentheorie]]
müßte uns eine solche Methode liefern.
Auf entsprechende Weise müßte dieses Reduktionsverfahren die Ent-

scheidung über eine jede Frage der Lösbarkeit einer „diophantischen Glei-
chung“ enthalten, d.h. die Entscheidung über eine jede Frage, welche die
Lösbarkeit einer algebraischen Gleichung mit einer oder mehreren Unbe-
kannten und mit ganzzahligen Koeffizienten durch ganze positive Zahlen-
werte der Unbekannten betrifft. Auch die Frage, ob eine solche Gleichung
„unendlich viele“ Lösungen besitzt, oder auch, ob sie für beliebige Wer-
te einer der Unbekannten (bzw. beliebige Werte oberhalb einer gewissen
Schranke) stets eine Lösung (in den übrigen Unbekannten) besitzt, müßte
durch das Reduktionsverfahren entschieden werden.
(2) U. Petersen [2002], p. 1079; translating Hilbert [1917], p. 156.
I believe: anything which can possibly be subject to scientific thought
goes, as soon as it has matured to the stage of theory formation, to the
axiomatic method and thus indirectly to mathematics.
76d. Brouwer’s neo-intuitionism. I close this chapter with a short section

on Brouwer’s neo-intuitionism. Although hostile at first sight, Hilbert’s proof the-
ory and Brouwer’s neo-intuitionism do share some very basic features, even if, in
Hilbert’s approach, they are confined to the meta-mathematical level.
Quotations 76.15. (1) Brouwer [1912], p. 67.

On what grounds the conviction of the unassailable exactness of math-
ematical laws is based has for centuries been an object of philosophical
research; and two points of view may here be distinguished, intuitionism
(largely French) and formalism (largely German). In many respects these
two viewpoints have become more and more definitely opposed to each
other; but during recent years they have reached agreement as to this, that
the exact validity of mathematical laws as laws of nature is out of the ques-
tion. The question where mathematical exactness does exist, is answered
differently by the two sides; the intuitionist says: in the human intellect,
the formalist says: on paper.
In Kant we find an old form of intuitionism, now almost completely
abandoned, in which time and space are taken to be forms of conception
inherent in human reason. For Kant the axioms of arithmetic and geometry
were synthetic a priori judgments, i.e., judgments independent of experience
and not capable of analytical demonstration; and this explained their apod-
ictic exactness in the world of experience as well as in abstracto. For Kant,
therefore, the possibility of disproving arithmetical and geometrical laws
experimentally was not only excluded by a firm belief, but it was entirely
unthinkable.
Diametrically opposed to this is the view of formalism, which main-
tains that human reason does not have at its disposal exact images either
of straight lines or of numbers larger than ten, for example, and that there-
fore these mathematical entities do not have existence in our conception of
nature any more than in nature itself.
Comment. To the best of my knowledge, Hilbert never referred to his approach as

“formalism”.
(2) Brouwer [1912], p. 69.
[[T]]he most serious blow for the Kantian theory was the discovery of non-
euclidean geometry, a consistent theory developed from a set of axioms
differing from that of elementary geometry only in this respect that the
parallel axiom was replaced by its negative. For this showed that the phe-
nomena usually described in the language of elementary geometry may be
described with equal exactness, though frequently less compactly in the lan-
guage of non-euclidean geometry; hence, it is not only impossible to hold
that the space of our experience has the properties of elementary geometry
but it has no significance to ask for the geometry which would be true for
the space of our experience. [[. . .]]
However weak the position of intuitionism seemed to be after this pe-
riod of mathematical development, it has recovered by abandoning Kant’s
apriority of space but adhering the more resolutely to the apriority of time.
This neo-intuitionism considers the falling apart of moments of life into
qualitatively different parts, to be reunited only while remaining separated
by time, as the fundamental phenomenon of the human intellect, passing by
abstracting from its emotional content into the fundamental phenomenon
of mathematical thinking, the intuition of the bare two-oneness.
Comment. There might be a problem of translation here, but note the phrase “the
space of our experience”. To be strict, Kant speaks of “Anschauung” not experi-
ence (“Erfahrung”). Space is a form of intuition a priori. No Anschauung is strictly
speaking possible of a four-dimensional space; if geometry is taken in that sense,
Poincaré’s claim at the end of quotation 70.23 (3), that “Euclidean geometry has
. . . nothing to fear from fresh experiments”, is uncontested. But there is still the
problem of interpreting the result that empirical space is apparently not Euclidean.
(3) Brouwer [1912], p. 70.
[[T]]he intuitionist recognizes only the existence of denumerable sets, [[. . .]]
[[a]]nd in the construction of these sets neither the ordinary language nor
any symbolic language can have any other rôle than that of serving as a
non-mathematical auxiliary[[.]]
(4) Brouwer [1912], pp. 71 f.
In the domain of finite sets in which the formalistic axioms have an
interpretation perfectly clear to the intuitionists, unreservedly agreed to by
them, the two tendencies differ solely in their method, not in their results;
this becomes quite different however in the domain of infinite or transfinite
sets, where [[. . .]] the formalist introduces various concepts, entirely mean-
ingless to the intuitionist, such as for instance “the set whose elements are
the points of space,” “the set whose elements are the continuous functions
of a variable,” “the set whose elements are the discontinuous functions of
a variable,” and so forth. In the course of these formalistic developments it
turns out that the consistent application of the axiom [[of naive set theory]]
leads inevitably to contradictions.
Comment. Note the similarity to Kant as regards the doctrine of meaningless of
questions.
One of the amazing aspects: Weyl, in Göttingen, joined in.
Quotations 76.16. (1) [1921] quoted after D. van Dalen [1995], p. 146.
The antinomies of set theory are usually regarded as border skirmishes
that concern only the remotest provinces of the mathematical empire and
that can in no way imperil the inner solidity and security of the empire
itself or of its genuine central areas. Almost all the explanations given by
highly placed sources for these disturbances (with the intention of denying
them or smoothing them over), however, lack the character of a clear, self-
evident conviction, born of totally transparent evidence, but belong to that
sort of half to three-quarters hones attempts at self-deception that one so
frequently encounters in political and philosophical thought. Indeed, every
earnest and honest reflection must lead to the realization that the troubles
in the borderland of mathematics must be judged as symptoms, in which
what lies hidden at the center of the superficially glittering and smooth
activity come to light—namely the inner instability of the foundations upon
which the structure of the empire rests.
CHAPTER XX
FORMALIZATION, METAMATHEMATICS, AND THE

ESTABLISHMENT OF THE FIXED POINT PROPERTY
The ‘thirties saw the rise of a new phase of foundational research. Again, I shall
focus on those aspects which are relevant for my project rather than trying to give
a representative account.
Some of the early reactions to the antinomies were strongly characterized by
metaphysical attitudes and philosophically quite interesting. A threefold distinc-
tion of the positions in the foundations of mathematics emerged from these early
reactions and seems to have been fashionable ever since: logicism, intuitionism, and
formalism.1 Along with this distinction a view was established which blurred Frege’s
decisive step from collections (set theoretical, Cantor) to classifications (logical: ex-
tensions of concepts, Frege); a step which Cantor explicitly criticized in his review
of Frege’s [1884], albeit with a different background.2
§ 77. Easing the grip of orders and types
Whitehead and Russell’s three volumes of Principia Mathematica with its (two
variants of a) theory of types was still intended to save the logistic program of
reducing mathematics to logic. The theory of types, however, serves this purpose
rather poorly. In this first paragraph, I shall present number of attempts to escape
the strict rule of orders and types.
77a. Non vicious circles: Ramsey’s reconsideration of Russell. The
theory of types as presented in Russell [1908] is commonly referred to as ramified
theory of types, as opposed to the simple theory of types tentatively put forward in
Russell [1903]. Ramsey provided reasons for dropping the ramifications and adopting
the original simple type structure. Accordingly, the simple theory of types has also
been called the “Ramseyfied theory of types”.3
Features 77.1. In Ramsey’s analysis I shall focus on the following issues:

(RM1) there are two fundamentally distinct groups of antinomies;4
(RM2) truth functions of infinite number;5
1 See Wang [1987], p. 181, for instance.
3 I’ve got that from somewhere, but I can’t remember where.
4 See quotations 77.3 below.
1082
§ 77. EASING THE GRIP OF ORDERS AND TYPES 1083
(RM3) our inability to write propositions of infinite length is logically a mere

accident.6
(RM4) an ambiguity in the word ‘meaning’ has no relevance to mathematics
whatever.7
The sources used are Ramsey [1925] and [1926].
Remark 77.2. Ramsey’s position is of particular interest for me, since in all points
his view is a paradigm of incompatibility with the attitude I adopt in the present
study.
I begin with Ramsey’s distinction between set theoretical and semantical antin-
omies. According to Copi [1971], p. 14, Peano [1906], p. 157, already observed that
Richard’s paradox belongs to linguistics and not to mathematics. Ramsey continues
this line of thought.
Quotations 77.3. (1) Ramsey [1925], p. 20.

It is not sufficiently remarked, and the fact is entirely neglected in
Principia Mathematica, that these contradictions fall into two fundamen-
tally distinct groups, which we will call A and B. [[. . .]] Group A consists
of contradictions which, were no provision made against them, would oc-
cur in a logical or mathematical system itself. They involve only logical or
mathematical terms such as class and number, and show that there must
be something wrong with our logic or mathematics. But the contradictions
of Group B are not purely logical, and cannot be stated in logical terms
alone; for they all contain some reference to thought, language, or symbol-
ism, which are not formal but empirical terms.
(2) Ramsey [1926]. pp. 76 f.
We can easily divide the contradictions according to which part of the
theory is required for their solution, and when we have done this we find
that these two sets of contradictions are distinguished in another way also.
The ones solved by the first part of the theory are all purely logical; they
involve no ideas but those of class, relation and number, could be stated
in logical symbolism, and occur in the actual development of mathematics
when it is pursued in the right direction. Such are the contradiction of the
greatest ordinal, and that of the class of classes which are not members of
themselves. With regard to these Mr Russell’s solution seems inevitable.
On the other hand, the second set of contradictions are none of them
purely logical or mathematical, but all involve some psychological term,
such as meaning, defining, naming or asserting. They occur not in mathe-
matics, but in thinking about mathematics; so that it is possible that they
arise not from faulty logic or mathematics, but from ambiguity in the psy-
chological or epistemological notions of meaning and asserting. Indeed, it
seems that this must be the case, because examination soon convinces one
1084 XX. FORMALIZATION, METAMATHEMATICS, FIXED POINT PROPERTY
that the psychological term is in every case essential to the contradiction,

which could not be constructed without introducing the relation of words
to their meaning or some equivalent.
Comments. (α) What Ramsey claims is that the distinction which I also employ in
the appendix A1 for the sake of convenience when presenting a choice of antinomies
is indeed sufficient to avoid antinomies in the realm of logic and mathematics. With
regard to the semantic predicates,8 Ramsey’s point of view may be interpreted as
saying that the act of naming and the relation between a symbol and the object
which it denotes is alien to logic and mathematics and, therefore, that there is no
need to take account of them within formalized mathematical theories. This is a
crucial point in the present study because Gödel in fact provided such a relation
in the form of an arithmetization of a formalized theory and thereby was able to
obtain his incompleteness results.
(β) It seems to me that Ramsey has virtually abandoned the philosophical analysis
of the antinomies put forward by Poincaré and Russell.
The distinction of the two groups of antinomies is the key to Ramsey’s claim
that Russell’s so-called simple theory of types is sufficient to avoid mathematical
antinomies.
Quotation 77.4. Ramsey [1925], p. 24.

[[Russell’s theory of types]] consists really of two distinct parts directed
respectively against the two groups of contradictions. These two parts were
unified by being both deduced in a rather sloppy way from the ‘vicious
circle principle’, but it seems to me essential to consider them separately.
The contradictions of Group A are removed by pointing out that a
propositional function cannot significantly take itself as argument, and by
dividing functions and classes into a hierarchy of types according to their
possible arguments. Thus the assertion that a class is a member of itself is
neither true nor false, but meaningless. This part of the Theory of Types
seems to me unquestionably correct[[.]]
Comment. As regards Ramsey’s hierarchy of proposition and functions, compare
Ramsey [1925], pp. 46 f.
I proceed to Ramsey’s introduction of truth functions of infinite number. Ram-

sey defined a class of functions which he considered non-vicious, though in some way
still possibly circular. He called these functions predicative!function, apparently well
aware of the fact that this name had already been used by Russell. It is not clear
to me, however, whether or not he was familiar with Poincaré’s analysis of the
antinomies and his use of the notion of predicativity. What Russell called “predica-
tive function” Ramsey calls “elementary function”. Anyway, in what follows I quote
Ramsey, but I want to point out that I am not going to adopt his use of these words
for my purposes, because I give preference to Poincaré’s ideas with which they are
in apparent contradiction.
8 See §142 in the appendix A1 for examples.

A predicative function of individuals is one which is any truth-function of
arguments which, whether finite or infinite in number, are all either atomic
functions of individuals or propositions. This defines a definite range of
functions of individuals which is wider than any range occurring in Prin-
cipia. It is essentially dependent on the notion of a truth-function of an
infinite number of arguments; if there could only be a finite number of
arguments our predicative functions would be simply the elementary func-
tions of Principia. Admitting an infinite number involves that we do not
define the range of functions as those which could be constructed in a cer-
tain way, but determine them by a description of their meanings. They
are to be truth-functions—not explicitly in their appearance, but in their
significance—of atomic functions and propositions.
(2) Ramsey [1925], pp. 41 f.
But, it will be objected, surely in this there is a vicious circle; you cannot
include F x̂ = (φ) . f (φẑ, x̂) among the φ’s, for it presupposes the totality
of the φ’s. This is not, however, really a vicious circle. The proposition F a
is certainly the logical product of the propositions f (φẑ, â), but to express
it like this (which is the only way we can) is merely to describe it in a
certain way, by reference to a totality of which it may be itself a member,
just as we may refer to a man as the tallest in a group, thus identifying
him by means of a totality of which he is himself a member without there
being any vicious circle. The proposition F a in its significance, that is, the
fact it asserts to be the case, does not involve the totality of functions;
it is merely our symbol which involves it. To take a particularly simple
case, (φ) . φa is the logical product of the propositions φa, of which it is
itself one; but this is no more remarkable and no more vicious than is
the fact that p . q is the logical product of the set p, q, p . q, of which it is
itself a member. The only difference is that, owing to our inability to write
propositions of infinite length, which is logically a mere accident, (φ) . φa
cannot, like p . q, be elementarily expressed, but must be expressed as the
logical product of a set of which it is also a member. If we had infinite
resources and could express all atomic functions as ψ1 x, ψ2 x, then we could
form all the propositions φa, that is, all the truth-functions of ψ1 a, ψ2 a, etc.,
and among them would be one which was the logical product of them all,
including itself, just as p . q is the product of p, q, p∨q, p . q. This proposition,
which we cannot express directly, that is elementarily, we express indirectly
as the logical product of them all by writing ‘(φ) . φa’. This is certainly a
circuitous process, but there is clearly nothing vicious about it.
Comment. The emphasis here for me lies on the formulation “it is merely our sym-
bol” which is in some way involved in a circle. Recall Frege and his claim that we
must be free to choose any symbol (FR4).9 Suggestion for later: if something can
only be done in some particular way (such as being defined in a circular way), it is
9 See Frege [1892], p. 26; quotation 71.28 (1) in these materials.
worthwhile being cautious about calling this “a mere accident”. Observe also that
to “refer to a man as the tallest in a group” is very unlikely to be the only way to
identify him and thus that this is hardly a vicious circle in terms of Russell crite-
rion of only being definable in this way.10 Notice also the phrase: “If we had infinite
resources . . . ”, in particular in view of the proximity of my approach to dialectical
logic to ‘resource-conscious logic’.11
How then are the semantic antinomies avoided, according to Ramsey?

‘F ’R(F x̂) says that ‘F ’ means F x̂. Now this is certainly true for some
meaning of ‘means’, so to uphold our denial of it we must show some
ambiguity in the meaning of meaning, and say that the sense in which ‘F ’
means F x̂, i.e. in which ‘heterological’ means heterological, is not the sense
denoted by ‘R’, i.e. the sense which occurs in the definition of heterological.
We can easily show that this is really the case, so that the contradiction is
simply due to an ambiguity in the word ‘meaning’ and has no relevance to
mathematics whatever.
[[. . .]] [[W]]e have arbitrarily chosen the letter ‘F ’ for a certain purpose, so
that ‘F x’ shall have a certain meaning (depending on x). As a result of this
choice ‘F ’, previously non-significant, becomes significant; it has meaning.
But it is clearly an impossible simplification to suppose that there is a single
object F , which it means. Its meaning is more complicated than that, and
must be further investigated.
(2) Ramsey [1925], p. 45.
The essential point to understand is that the reason why
(∃φ) : ‘F ’R(φx̂)
can only be true if ‘F ’ is an elementary function, is not that the range
of φ is that of elementary functions, but that a symbol cannot have R to
a function unless it (the symbol) is elementary. The limitation comes not
from ‘R’.
Remarks 77.7. (1) To get an impression of how far Ramsey has moved away from
Poincaré’s initial analysis, try to imagine what Poincaré would say to an infinite
number of arguments!
(2) Interesting point: Russell’s sensitivity to the ambiguity of semantical concepts
such as meaning is dismissed by Ramsey without further discussion. This corre-
sponds to the point Frege arrived in sense and meaning: the symbol plays no role in
‘straight’ (“gerader”) speech, and arithmetic speech is straight.12 There is no need
for a distinction between equality (=) and identity (≡) in arithmetic.13
(3) It seems worthwhile mentioning in which way Ramsey is diametrically opposed
to my own approach. After reading the above quotation, the point is simple: for
10 See quotations 75.6 (2) and 77.5 in these materials.
11 Cf. Troelstra in quotation 99.9 (2) in these tools.
Ramsey it is merely a symbol which involves some crucial totality. For me there is a
conflict of form and content. There is no such thing as “merely a symbol” involving
something. This is the point where the positions clash: for the realist symbols are
merely symbols and don’t matter as regards the content. For the Hegelian there is
a dialectic of form and content which is a constitutive part of what things are. The
challenge is to formulate the point against Ramsey in a way that provides for an
alternative way of dealing with the antinomies.
77b. Quine’s New Foundations. In his paper [1937a], Quine presented an

elegant way to adhere in a certain way to a type distinction without having one’s
universe clogged up with an infinite repetition of entities of the same character but
of different types. The following quotations are taken from the reprint in Quine’s
from a logical point of view.
Quotation 77.8. (1) Quine [1953a], pp. 90 f.

Russell’s paradox[[. . .]] was overcome in Principia by Russell’s theory of
types. Simplified for application to the present system, the theory works
as follows. We are to think of all objects as stratified into so-called types,
such that the lowest type comprises individuals, the next comprises classes
of individuals, the next comprises classes of such classes, and so on. In ev-
ery context, each variable is to be thought of as admitting values only of a
single type. The rule is imposed, finally, that (αǫβ) is to be a formula only
if the values of β are of next higher type than those of α; otherwise (αǫβ) is
reckoned as neither true nor false, but meaningless. In all contexts the types
appropriate to the several variables are actually left unspecified; the context
remains systematically ambiguous, in the sense that the types of its vari-
ables may be construed in any fashion conformable to the requirement that
‘ǫ’ connect variables only of consecutively ascending types. An expression
which would be a formula under our original scheme will hence be rejected
as meaningless by the theory of types only if there is no way whatever of
so assigning types to the variables as to conform to this requirement on ‘ǫ’.
Thus a formula in our original sense of the term will survive the theory of
types if it is possible to numerals for the variables in such a way that ‘ǫ’
comes to occur only in contexts of the form ‘nǫn + 1’. Formulas passing this
test will be called stratified.
(2) Quine [1953a], pp. 91 f.
[[T]]he theory of types has unnatural and inconvenient consequences. Be-
cause the theory allows a class to have members only of uniform type, the
universal class V gives way to an infinite series of quasi-universal classes,
one for each type. The negation −x ceases to comprise all nonmembers of
x, and comes to comprise only those nonmembers of x which are next lower
in type than z. Even the null class Λ gives way to an infinite series of null
classes. The Boolean algebra no longer applies to classes in general, but
is reproduced rather within each type. The same is true of the calculus of
relations. Even arithmetic when introduced by definitions on the basis of
logic, proves to be subject to the same reduplication. Thus the numbers
cease to be unique; a new 0 appears for each type, likewise a new 1, and
so on, just as in the case of V and Λ. Not only are all these cleavages and
reduplications intuitively repugnant, but they call continually for more or
less elaborate technical maneuvers by way of restoring severed connections.
I will now suggest a method of avoiding the contradictions without
accepting the theory of types or the disagreeable consequences which it
entails. Whereas the theory of types avoid the contradictions by excluding
unstratified formulas from the language altogether, we might gain the same
end by continuing to countenance unstratified formulas but simply limiting
[[the axiom of abstraction]].
What is interesting in the context of my study is, of course, how Cantor’s

diagonal method fares under this treatment. The next quotation is to give Quine’s
explanation of the situation.
Quotation 77.9. Quine [1953a], p. 92, footnote 10.

Since everything belongs to V, all subclasses of V can be correlated
with members of V, namely, themselves. In view then of k, one might hope
to derive a contradiction. It is not clear, however, that this can be done.
Cantor’s reductio ad absurdum of such a correlation consists in forming the
class h of those members of the original class k which do not belong to
the subclasses to which they are correlated, and then observing that the
subclass h of k has no correlate. Since in the present instance k is V and the
correlate of a subclass ist that subclass itself, the class h becomes the class
of all those subclasses of V which do not belong to themselves. But [[the
modified axiom of abstraction]] provides no such class h. Indeed, h would
be ŷ ∼ (yǫy), whose existence is disproved by Russell’s paradox.
77c. Gödel on Russell’s theory of types. This final section is dedicated

to Gödel’s remarks on Russell’s theory of types and the vicious circle principle in
his paper [1944] on Russell’s Mathematical Logic.
Quotation 77.10. Gödel [1944], p. 131.

By analyzing the paradoxes to which Cantor’s set theory had led, [[Russell]]
freed them from all mathematical technicalities, thus bringing to light the
amazing fact that our logical intuitions (i.e., intuitions concerning such
notions as: truth, concept, being, class, etc.) are self-contradictory.
Comment. It should be kept in mind that the idea that underlies Gödel’s distinc-
tion of logical and set theoretical intuition is different from the one that underlies
Ramsey’s distinction of two kinds of antinomies in that Ramsey’s focused on epis-
temological notions where Gödel is concerned with logical ones.
Quotations 77.11. (1) Gödel [1944], p. 131, footnote 9.

If one wants to bring such paradoxes as “the liar” under this viewpoint,
universal (and existential) propositions must be considered to involve the
class of objects to which they refer.
(2) Gödel [1944], p. 133.

[[O]]ne defines (or tacitly assumes) totalities, whose existence would entail
the existence of certain new elements of the same totality, namely elements
definable only in terms of the whole totality.
(3) Gödel [1944], pp. 133 f.
In order to make [[the vicious circle]] principle applicable to the intensional
paradoxes, still another principle had to be assumed, namely that “every
propositional function presupposes the totality of its values” and therefore
evidently also the totality of its possible arguments. [Otherwise the concept
of “not applying to itself” would presuppose no totality (since it involves
no quantifications), and the vicious circle principle would not prevent its
application to itself.] A corresponding vicious circle principle for proposi-
tional functions which says that nothing defined in terms of a propositional
function can be a possible argument of this function is then a consequence.
(4) Gödel [1944], p. 134.
The logical system to which one is led on the basis of these principles
is the theory of orders in the form adopted, e.g., in the first edition of
Principia, according to which a propositional function which either contains
quantifications referring to propositional functions of order n or can be
meaningfully asserted of propositional functions of order n is at least or
order n + 1, and the range of significance of a propositional function as well
as the range of a quantifier must always be defined to a definite order.
In the second edition of Principia, however it is stated in the Introduc-
tion (pp. XI and XII) that “in a limited sense” also functions of a higher
order than the predicate itself (therefore also functions defined in terms of
the predicate [[. . .]]) can appear as arguments of a predicate of functions; and
in appendix B such things occur constantly. This means that the vicious
circle principle for propositional functions is virtually dropped.
(5) Gödel [1944], p. 149.
[[T]]he theory of types brings in a new idea for the solution of the para-
doxes, especially suited to their intensional form. It consists in blaming the
paradoxes not on the axiom that every propositional function defines a con-
cept or class, but on the assumption that every concept gives a meaningful
proposition, if asserted for any arbitrary object or objects as arguments.
The next set of quotations concerns Russell’s no class approach and his theory
of descriptions.
Quotations 77.12. (1) Gödel [1944], p. 141.

[[I]]n Russell the paradoxes had produced a pronounced tendency to build
up logic as far as possible without the assumption of the objective existence
of such entities as classes and concepts. This led to the formulation of the
[[. . .]] “no class theory,” according to which classes and concepts were to be
introduced as a façon de parler. But propositions, too, (in particular those
involving quantifications) were later on largely included in this scheme,
which is but a logical consequence of this standpoint, since e.g., universal

propositions as objectively existing entities evidently belong to the same
category of idealistic objects as classes and concepts and lead to the same
kind of paradoxes, if admitted without restrictions. As regards classes this
program was actually carried out; i.e., the rules for translating sentences
containing class names or the term “class” into such as do not contain
them were stated explicitly; and the basis of the theory, i.e., the domain
of sentences into which one has to translate is clear, so that classes can
be dispensed with (within the system Principia), but only if one assumes
the existence of a concept whenever one wants to construct a class. When
it comes to concepts and the interpretation of sentences containing this or
some synonymous term, the state of affairs is by no means as clear.
(2) Gödel [1944], p. 142.
This whole scheme of the no-class theory is of great interest as one
of the few examples, carried out in detail, of the tendency to eliminate
assumptions about the existence of objects outside the “data” and to replace
them by constructions on the basis of these data. The result has been in
this case essentially negative; i.e., the classes and concepts introduced in
this way do not have all the properties required for their use in mathematics
unless one either introduces special axioms about the data (e.g., the axiom
of reducibility), which in essence already mean the existence in the data
of the kind of objects to be constructed, or makes the fiction that one
can form propositions of infinite (and even non-denumerable) length, i.e.,
operates with truth-functions of infinitely many arguments, regardless of
whether or not one can construct them. But what else is such an infinite
truth-function but a special kind of an infinite extension (or structure)
and even a more complicated one that a class, endowed in addition with a
hypothetical meaning, which can be understood only by an infinite mind?
All this is only a verification of the view [[. . .]] that logic and mathematics
(just as physics) are built up on axioms with a real content which cannot
be “explained away.”
Comment. As regards infinite conjunctions: compare Ramsey in quotation 77.5 (1)
above.
(3) Gödel [1944], p. 130.
I cannot help feeling that the problem raised by Frege’s puzzling conclusion
has only been evaded by Russell’s theory of descriptions and that there is
something behind it which is not yet completely understood.
I continue with a number of considerations regarding a certain lack of conclu-

siveness of the vicious circle principle.

[[C]]orresponding to the phrases “definable only in terms of,” “involving,”
and “presupposing,” we have really three different principles, the second
and third being much more plausible than the first. It is the first form
which is of particular interest, because only this one makes impredicative

definitions impossible[[.]]
(2) Carnap [1931], p. 190.
If we reject the belief that it is necessary to run through individual cases
and rather make it clear to ourselves that the complete verification of a
statement about an arbitrary property means nothing more than its logical
(more exactly, tautological) validity for an arbitrary property, we will come
to the conclusion that impredicative definitions are logically admissible.
(3) Gödel [1944], pp. 135 f.
[[O]]ne may, on good grounds, deny that reference to a totality necessarily
implies reference to all single elements of it or, in other words, that “all”
means the same as an infinite logical conjunction. One may, e.g., follow
Langford’s and Carnap’s suggestion to interpret “all” as meaning analytic-
ity or necessity or demonstrability. There are difficulties in this view, but
there is no doubt that this way the circularity of impredicative definitions
disappears.
(4) Gödel [1944], p. 136.
[[O]]ne cannot say that an object described by reference to a totality “in-
volves” that totality, although the description itself does; nor would it con-
tradict the third form, if “presuppose” means “presuppose for the existence”
not “for the knowability”.
Comment. This strikes me as a fine example of the realistic approach, even though
I have little sympathy for realism.
I close this paragraph with a quotation from Gödel in which he expresses a

thought — or rather hope — that might still linger in the back of the mind of many
a logician.

It might even turn out that it is possible to assume every concept to be sig-
nificant everywhere except for certain “singular points” or “limiting points,”
so that the paradoxes would appear as something analogous to dividing
by zero. Such a system would be most satisfactory in the following re-
spect: our logical intuitions would then remain correct up to certain minor
corrections[[.]]
§ 78. Hilbert’s province
Kurt Grelling,14 Heinrich Behmann,15 Wilhelm Ackermann,16 Haskell B. Curry,17

and Kurt Schütte,18 all obtained their doctorates under Hilbert’s supervision with
contributions to the foundations of mathematics and logic.19
78a. Philosophical issues. The following first set of quotations is meant to
give an example of the misinterpretations to which Hilbert’s approach has been
subjected.

For [[Hilbert]] mathematics has to combine only pure symbols, and a true
mathematician should reason upon them without preconceptions as to their
meaning.
(2) Ramsey [1925], p. 5.
The formalists neglected the content altogether and made mathematics
meaningless[[.]]
(3) Ramsey [1925] p. 2.
[[T]]he formalist school, of whom the most eminent representative is now
Hilbert, have concentrated on the propositions of mathematics, such as
‘2 + 2 = 4’. They have pronounced these to be meaningless formulae to
be manipulated according to certain arbitrary rules, and they hold that
mathematical knowledge consists in knowing what formulae can be derived
from what others consistently with the rules.
Comment (to all three). Ouch. Cf. quotation 78.10, if you don’t know why that
hurts.
Paul Bernays, a philosopher-mathematician in the tradition of a certain neo-

Kantianism (“Neue Fries’sche Schule”, Leonard Nelson) and collaborator of David
Hilbert, introduced the label “platonism” to characterize a strong form of conceptual
realism.
Quotation 78.2. Benacerraf and Putnam [1964], p. 275; translation of Bernays

[1935b], p. 53.
[[T]]he tendency of which we are speaking consists in viewing the objects as
cut off from all links with the reflecting subject.
Since the tendency asserted itself especially in the philosophy of Plato,
allow me to call it “platonism”.
14 Die Axiome der Arithmetik mit besonderer Berücksichtigung der Beziehungen zur Men-
genlehre. 1910.
15 Die Antinomie der transfiniten Zahl und ihre Auflösung durch die Theorie von Russell
und Whitehead. 1922.

16 Begründung des „tertium non datur“ mittels der Hilbertschen Theorie der Widerspruchs-
freiheit. 1924.
17 Grundlagen der kombinatorischen Logik. 1930.
18 Untersuchungen zum Entscheidungsproblem der mathematischen Logik. 1934.
19 Cf. Hilbert [1935], p. 433.
§ 78. HILBERT’S PROVINCE 1093
Comment. Observe that in view of his remarks in the Grundlagen der Arithmetik as
quoted (in translation) in 71.27 (3) in these materials, Frege can hardly count as a
platonist: he does not claim an independence of his objects of research of reason; he
would have had to locate reason outside the subject to qualify as a platonist. Frege
speaks of independence of “our sensation, intuition, and imagination” and explicitly
excludes reason. In this respect Frege appears to be philosophically aware where
Cantor comes across as philosophically rather naive. I am more inclined to take
Cantor as a good example of a platonist in the sense of Bernays.
Quotation 78.3. Bernays [1937], p. 83.

Der Standpunkt, den Hilbert durch seine Beweistheorie einnimmt, ist da-
durch gekennzeichnet, daß er sowohl den Bedürfnissen der formalen Sy-
stematik wie denen der arithmetischen Evidenz gerecht werden will. Als
Mittel zur Vereinigung dieser Ziele dient ihm die Sonderung von Mathema-
tik und Meta-Mathematik, welche der Kantischen Teilung der Philosophie
in „Kritik“ und „System“ nachgebildet ist.
I close this section with a couple of quotations from (a translation of) the
introduction (§§1–2) to Gentzen’s first (published) consistency proof of classical
first order arithmetic: Gentzen [1936].
Quotations 78.4. (1) Szabo [1969], p. 137; translating Gentzen [1936], p. 6.

The assertion that a mathematical theory is consistent constitutes a propo-
sition about the proofs possible in that theory. It says that none of these
proofs leads to a contradiction. In order to carry out a consistency proof we
must therefore make the possible proofs in the theory themselves objects of
a new ‘metatheory’. The theory that has arbitrary mathematical proofs for
its objects is called ‘proof theory’ or ‘metamathematics’.
(2) Szabo [1969], pp. 137; translating Gentzen [1936], p. 7.
As the objects of our proof theory we take the proofs carried out in
mathematics proper. These proofs are customarily expressed in the words
of our language. They have the disadvantage that there are many different
ways of expressing the same proposition, and that an arbitrariness exists
in the order of the words, sometimes even ambiguity. In order to make an
exact study of proofs possible it is therefore desirable to begin by giving
them a uniform uniquely predetermined form. This is achieved by the ‘for-
malization’ of the proofs: the words of our language are replaced by definite
symbols, the logical forms of inference by formal rules for the formation of
new formalized propositions from already proved ones.
Comment. Contrast this with quotations 78.1 above.
78b. Formalization. In order for Hilbert’s proof theory to go ahead, the the-
ories under consideration had to be completely formalized. In this first section, I
shall consider different aspects of such a formalization.
Quotations 78.5. (1) Church [1934], p. 356

A system of symbolic logic must begin with a list of undefined symbols,
a list of formal axioms, and a list of rules of inference.
(2) Curry et al. [1958], p. 20.
It is clear that a formal system can be communicated only through a
presentation. It is also clear that the particular choice of symbolism does
not matter much. So long as we satisfy one indispensable condition —
namely that distinct names be assigned to distinct obs — we can choose the
symbolism in any way we like without affecting anything essential. We can,
therefore, regard a formal system as something independent of this choice,
and say that two presentations differing only in the choice of symbolism are
presentations of the same formal system. In this sense a formal system is
abstract with respect to its presentation.
Comment. Curry is explicit about the one point that is so easily skipped over by
many philosophers (including Frege): names cannot be chosen arbitrarily; they have
to satisfy the indispensable condition that different objects have different names,
otherwise a contradiction could be established immediately. Take John Smith, for
example; bilocation is the least that can established about him; he is probably also
red-headed, blond, black, white, fat, skinny, and so on.
Although it may be clear today, as Curry says, that a formal system can be
communicated only through a presentation, the need for a thorough listing of the
basic symbols and their possible combinations was not so clear in the beginning.
Concerning the Principia Mathematica, we find Gödel lamenting:

It is to be regretted that this first comprehensive and thorough going
presentation of a mathematical logic and the derivation of Mathematics is so
greatly lacking in formal precision in the foundations (contained in *1–*21
of Principia), that it presents in this respect a considerable step backwards
as compared with Frege. What is missing, above all, is a precise statement
of the syntax of the formalism. Syntactical considerations are omitted even
in cases where they are necessary for the cogency of the proofs [[. . .]] it
is necessary to have a survey of all possible expressions, and this can be
furnished only by syntactical considerations.
Quotations 78.7. (1) Curry et al. [1958], p. 20.

If we assign a unique determinate thing to each ob in such a way that
distinct things are always assigned to distinct obs, we have a representation
of the system.
(2) Curry et al. [1958], p. 21.
The idea of representation is to be contrasted with that of interpretation. By
an interpretation of a formal system we mean a correspondence between its
elementary statements and certain statements which are significant without
reference to the system.
The following couple of quotations concerns definition: contextual definition vs.

nominal (explicit) definition.
Quotations 78.8. (1) Church [1951], p. 9.

One method of abbreviation is by means of a nominal definition, which
introduces a particular new symbol to replace or stand for a particular
well-formed formula.
(2) Church [1956], p. 323.
In such a case, where a complex notation introduced by definition carries
the false appearance or suggestion that some part of the notation is to be
taken as denoting (or otherwise as having meaning in isolation), it is normal
to speak of contextual definition.
Quotations 78.9. (1) Hilbert and Ackermann [1938/50], p. 125.

[[W]]e are led to a natural extension of the predicate calculus of first order by
applying universal and existential quantifiers also to sentential and predicate
variables, and by distinguishing between free and bound variables of that
kind.
Comment. This concerns first order formalization vs. second order formalization.
(2) Bernays [1930], p. 52; my translation.
In particular the meaning of the principle of choice becomes fully com-
prehensible only with the logical formalism. We can express the principle
in the following form: if B (x, y) is a predicate (defined in some area) with
two objects and if there exists to every object x of the domain at least one
object y of this realm for which B (x, y) is satisfied, then there exists (at
least) one function y = f(x) such that for every object x of the domain of
B (x, y) the value f(x) is again an object of this realm and one that satisfies
B (x, f(x)).
If one reflects on what this claim says for the special case of a uni-
verse with only two elements, the objects of which we can represent by the
numbers 0,1 and in which only four different courses of values of functions
y = f(x) are to be considered, one finds that the claim results as a simple
application of the one of the distributive laws which hold for the relation
between conjunction and disjunction, viz. of the following elementary logi-
cal sentence: “If we have A and if we have in addition B or C, then we have
A and B, or we have A and C.”
Also in the case of any determinate finite number of objects of the
universe the claim of the principle of choice follows from this law of dis-
tributivity. The general claim of the principle of choice is hence nothing else
but the extension of an elementary logical law for conjunction and disjunc-
tion to infinite totalities and the principle of choice thus forms a supplement
to the logical rules which concern the general and the existential judgment,
i.e. rules of the existential reasoning the application of which to infinite to-
talities has also the meaning that certain elementary laws for conjunction
and disjunction are transferred to the infinite.
Quotation 78.10. Gödel [1931a], p. 201.

According to the formalist view one adjoins to the meaningful propo-
sitions of mathematics transfinite (pseudo-) assertions, which themselves
have no meaning, but serve only to round out the system, just as in geo-
metry one rounds out a system by the introduction of points at infinity.
78c. Normal forms, decidability, and completeness. The present section

is devoted to some of the ideas which stood behind the meta-logical research which
were carried out in the ‘twenties and early ‘thirties such as that by Post, Bernays,
Skolem, and Gödel.
Quotations 78.11. (1) Feferman et al. (eds.) [1986], p. 103; translating Gödel
[1930], p. 349.
Whitehead and Russell, as is well known, constructed logic and math-
ematics by initially taking certain evident propositions as axioms and de-
riving the theorems of logic and mathematics from these by means of some
precisely formulated principles of inference in a purely formal way (that is,
without making further use of the meaning of the symbols). Of course, when
such a procedure is followed the question at once arises whether the initially
postulated system of axiom and principles of inference is complete, that is,
whether it actually suffices for the derivation of every logico-mathematical
proposition, or whether, perhaps, it is conceivable that there are true propo-
sitions (which may even be provable by means of other principles) that
cannot be derived in the system under consideration.
Comment. In view of Feferman’s comments presented in quotations 85.19 (2) and
(3) in these materials, it seems questionable whether this question, at least his-
torically, “at once arises”. Anyway, the question that arises in the present study is
how to assess whether an unprovable proposition like the one that Gödel formulated
in his paper [1931] is indeed true. Or, differently put, the notion of classical truth
itself is at stake.
(2) Feferman et al. (eds.) [1986], p. 63; translating Gödel [1929].
If we replace the notion of logical consequence (that is, of being formally
provable in finitely many steps) by implication in Russell’s sense, more
precisely, by formal implication, where the [[functional]] variables are the
primitive notions of the axiom system in question, then the existence of a
model for a consistent axiom system (now taken to mean one that implies
no contradiction) follows from the fact that a false proposition implies any
other, hence also every contradiction (whence the assertion follows at once
by indirect argument).
Comment. There is a footnote to the last sentence, pointing out that this “seems
to have been noted for the first time by R. Carnap”.
Gödel’s completeness proof of classical first order logic virtually remained the
only one for almost twenty years. In 1949 Leon Henkin published a powerful new
method which he extended to the theory of types.
Quotation 78.12. Henkin [1950], p. 81.

The first order functional calculus was proved complete by Gödel in
1930. Roughly speaking, this proof demonstrates that each formula of the
calculus is a formal theorem which becomes a true sentence under every
one of a certain intended class of interpretations of the formal system.
For the functional calculus of second order, in which predicate variables
may be bound, a very different kind of result is known: no matter what
(recursive) set of axioms are chosen, the system will contain a formula
which is valid but not a formal theorem. This follows from results of Gödel
concerning systems containing a theory of natural numbers, because a finite
categorical set of axioms for the positive integers can be formulated within
a second order calculus to which a functional constant has been added.
By a valid formula of the second order calculus is meant one which ex-
presses a true proposition whenever the individual variables are interpreted
as ranging over an (arbitrary) domain of elements while the functional vari-
ables of degree n range over all sets of ordered n-tuples of individuals. Under
this definition of validity, we must conclude from Gödel’s result that the
calculus is essentially incomplete.
It happens, however, that there is a wider class of models which furnish
an interpretation for the symbolism of the calculus consistent with the usual
axioms and formal rules of inference. Roughly these models consist of an
arbitrary domain of individuals, as before, but now an arbitrary class of
sets of ordered n-tuples of individuals as the range for functional variables
of degree n. If we redefine the notion of valid formula to mean one which
can then prove that the usual axiom system for the second order calculus
is complete: a formula is valid if and only if it is a formal theorem.
A similar result holds for the calculi of higher order.
Comment. In my presentation of quantification this wider class of models is pro-
vided in §34 in chapter IX of the tools.
78d. Consolidation of Zermelo’s set theory.
Quotations 78.13. (1) Van Heijenoort [1967b], 292; translation of Skolem [1922],
p. 219.
A very deficient point in Zermelo is the notion “definite proposition”. [[. . .]]
So far as I know, no one has attempted to give a strict formulation of this
notion; this is very strange, since it can be done quite easily and, moreover,
in a very natural way that immediately suggests itself. In order to explain
this, [[. . .]] I mention the five basic operations of mathematical logic here
[[. . .]]:
(1× ) Conjunction [[. . .]]
(1+ ) Disjunction [[. . .]]
(2) Negation [[. . .]]
(3× ) Universal quantification [[. . .]]
(3+ ) Existential quantification [[. . .]]
As is well known, only three of these five operations are really needed,
since (1× ) and (1+ ), like (3× ) and (3+ ), are mutually definable by means
of (2).
By a definite proposition we now mean a finite expression constructed
from elementary propositions of the form aεb or a = b by means of the five
operations mentioned.
(2) Van Heijenoort [1967b], 295; translation of Skolem [1922], p. 223.
So far as I know, no one has called attention to this peculiar and ap-
parently paradoxical state of affairs. By virtue of the axioms we can prove
the existence of higher cardinalities, of higher number classes, and so forth.
How can it be, then, that the entire domain B can already be enumerated
by means of the finite positive integers? The explanation is not difficult
to find. In the axiomatization, “set” does not mean an arbitrarily defined
collection; the sets are nothing but objects that are connected with one
another through certain relations expressed by the axioms. Hence there
is no contradiction at all if a set M of the domain B is nondenumerable
in the sense of the axiomatization; for this means merely that within B
there occurs no one-to-one mapping Φ of M onto Z0 (Zermelo’s number
sequence). Nevertheless there exists the possibility of numbering all objects
in B, and therefore also the elements of M , by means of the positive inte-
gers; of course, such an enumeration too is a collection of certain pairs, but
this collection is not a “set” (that is, it does not occur in the domain B).
Thus, axiomatizing set theory leads to a relativity of set-theoretic no-
tions, and this relativity is inseparably bound up with every thoroughgoing
axiomatization.
The most important result above is that set-theoretic notions are rela-
tive. I had already communicated it orally to F. Bernstein in Göttingen in
the winter 1915–16.
The other important issue in the consolidation of Zermelo’s set theory may be
regarded as an elaboration of the concept of limitation of size. The birth of a clear
idea of the difference between iteration (mathematical) and classification (logical)
seems to have been with Gödel’s article What is Cantor’s continuum problem?
[1947], from which I take the next quotation.
Quotation 78.14. Gödel [1947], pp. 518 f.

As far as sets occur and are necessary in mathematics (at least in the
mathematics of today, including all of Cantor’s set theory), they are sets of
integers, or of rational numbers (i.e., of pairs of integers), or of real numbers
(i.e., of sets of rational numbers), or of functions of real numbers (i.e., of sets
of pairs of real numbers), etc.; when theorems about all sets (or the existence
of sets) in general are asserted, they can always be interpreted without any
difficulty to mean that they hold for sets of integers as well as for sets of real
numbers, etc. (respectively, that there exist either sets of integers, or sets of
real numbers, or . . . etc., which have the asserted property). The concept
of set, however, according to which a set is anything obtainable from the
integers (or some other well defined objects) by iterated application of the
operation “set of,” and not something obtained by dividing the totality
of all existing things into two categories, has never led to any antinomy
whatsoever; that is, the perfectly “naïve” and uncritical working with this
concept of set has so far proved completely self-consistent.
Comment. There is hardly any change to this crucial passage in the revised version
[1964]. Interesting: remark [1964], p. 272, footnote 40, concerning the concept of set
and ‘synthesis’ in Kant.
78e. Type free logics. Given that the iterative conception of set is a math-
ematical idea which sorts out a realm of well-determined collections, there is still
the logical aspect of unrestricted classification which requires in some way the ac-
knowledgment “that there are concepts with no extension (at any rate, none in the
ordinary sense of the word).” 20 The problem is, how can this be done?
There was quite an active period of research in type free logic in the early
fifties: Ackermann, Curry, and Schütte. The first edition of Schütte’s Proof Theory
(Beweistheorie) still contains a sub-system of analysis based on a type free logic.
Literature 78.15. The following is a list of papers covering essentially what is

known to me as far as type free logic is concerned in the years from 1930 until 1965.
Ackermann [1941], [1950], [1952], [1953], [1958], [1961], [1963];
Fitch [1936], [1948];
Łukasiewicz [1930];
Schütte [1953], [1954], [1958], [1960a];
Skolem [1957];
Prawitz [1965].
The logic I propose as a dialectical logic is a type free logic in the sense of
quotation 67.20 (2) below. This seems worth mentioning since not every logic which
comes under the label “type free” can claim to be that; cf. remark 78.19 below.
Quotation 78.16. Curry [1951], pp. 151 f.

12. Neben der deduktiven Vollständigkeit gibt es eine andere Art, die kom-
binatorische Vollständigkeit heiße. Ein System S besitzt diese Eigenschaft
dann und nur dann, wenn jede aus Termen des Systems und einem Hilf-
sterm (oder Variablen) x gebildete Formel M innerhalb S als Funktion von x
dargestellt werden kann. [[. . .]] Die kombinatorische Vollständigkeit erlaubt
uns also eine gewisse Freiheit bei der Konstruktion von Termen. Sie ist auch
für ein logisches System wünschenswert. Sie ist in der Tat viel wichtiger als
die deduktive Vollständigkeit. [[. . .]]
20 Frege [1903], p. 257; see quotation 73.4 (3) in these materials.

13. Diese zwei Arten von Vollständigkeit sind aber unverträglich. In der Tat
ist ein System mit einer Folgeverknüpfung derart, daß die Regel [[modus
ponens]] und die zwei Formeln
(4) A→A
(5) (A → (A → B)) → (A → B)
für beliebige Formeln A und B gelten, [[ist]] inkonsistent in dem Sinne, daß
jede Formel als beweisbar ausfällt. Denn aus einer beliebigen Formel B kann
man nach dem Muster des Russellschen Paradoxons eine Formel A derart
konstruieren, daß A und A → B gleich sind; daraus kann man dann mit
Hilfe von [[modus ponens]], (4), (5) leicht B ableiten. Dieses unangenehme
Paradoxon besteht ohne irgendeine Benutzung von Negation. Es sind also
nicht die Negation oder irgendwelche ihrer Eigenschaften, die den Ursprung
unserer Schwierigkeiten aus machen, sondern es ist die Folgeverknüpfung
selbst.
Quotations 78.17. (1) Ackermann [1958], p. 3; my translation.

It is one of the most essential results of the more recent foundational re-
search that there is no system of logic which is complete in every respect.
Equally relevant in this respect, besides the well known results of Gödel
[[[1931]]] which are based on an arithmetization of the paradox of the liar,
are the investigations of Kleene and Rosser [[[1935]]] which contain an arith-
metical version of the paradox of Richard. The latter investigation have
been supplemented by Curry [[[1941], [1942], and [1951]]]. Following him,
their result can be expressed in the following form: A logic cannot at the
same time be combinatorically and deductively complete, if consistency is
to be retained.
Comment. This is, essentially, what the present study is based on: the result that
logic is incomplete. Were it not for such an incompleteness, Hegel’s idea of a spec-
ulative philosophy would make no sense at all.
(2) Ackermann [1961], p. 3; my translation.

Under a type free logic is understood here, in accordance with the litera-
ture listed at the end [[of Ackermann’s paper, essentially Ackermann and
Schütte]] a logic which does not only possess formally the property of free-
dom of types, but beyond that an unrestricted axiom of comprehension.
Comment. The question is, how is the axiom of comprehension formulated, i.e.,
what notion of implication is employed to express the equivalence? Obviously, it
can’t be classical implication; and this leaves a lot of room for all sorts of weird
constructions of “non-classical” or “intensional” “implications”. This is why I pre-
fer a formulation in terms of the admissibility of rules, e.g., the N-style rules of
unrestricted abstraction (cf. definition 39.3 in the tools) must be admissible.
§ 79. UNDECIDABILITY, INCOMPLETENESS, COMPUTABILITY 1101
Quotation 78.18. Prawitz [1965], p. 95.

[[T]]he set-theoretical paradoxes are ruled out by the requirement that the
deductions shall be normal. [[. . .]]
[[. . .]] We have thus an example of a system for which the inversion prin-
ciple holds (also the λ-rules obviously satisfy the principle) and where we
hence can remove any given maximum formula, but where it is impossible
to remove all maximum formulas from certain deductions. If we consider
successive reductions of the quasi-proof ⊃-reductions and λ-reductions.
Remark. It appears from the works of Fitch that one can develop a
considerable portion of ordinary mathematics in F somewhat amplified. In
spite of this fact and the demonstrable consistency of F, the system has
serious disadvantages. Thus, although ⊃ E is a rule of the system, one
cannot in general infer that B is provable given that A and A ⊃ B are
provable, since there may be only a quasi-proof of B.
Comment. Cf. §§26 and 41 in the tools.
Remark 78.19. The “type-free formal system” presented in §11 of Feferman [1984]
is not a type free logic in the sense of my comment to quotation 78.17 (2) above. The
axiom of abstraction does not hold in the sense that there is no notion of implication
provided that would allow to express that F[t] generally ‘implies’ t ∈ λx F[x] and
vice versa. For example, it is not possible to establish both, R ∈ R → R ∈/ R and
R ∈/ R → R ∈ R and retain consistency. This is due to the circumstance that classical
logic holds for the usual connectives and the ‘axiom of abstraction’ is formulated
with a non-classical bi-implication ≡ which does not allow to conclude A → ¬A
and ¬A → A from A ≡ ¬A. Differently put, more in the vein of my comment to
quotation 78.17 (2), the N-style introduction rule of set formation is not admissible:
according to Feferman [1984], p. 98, ¬(r ∈ r) is provable his system S ′ (≡), where
r = {u|u ∈/ u}, but obviously r ∈ r cannot be provable.21 It would not have needed a
mention, I suppose, that Feferman regards his laws governing the connective ≡ as
natural.22
§ 79. Undecidability, incompleteness, and the notion of computability
In the last paragraph, I followed the idea of making theories precise by means of
formalization. I shall now focus on how this commitment to precision nurtured the
phenomenon of self-reference.
Metamathematical investigations, originally triggered by the emergence of an-
tinomies, have in turn produced strange phenomena which may be viewed as being
intimately linked to the phenomenon of antinomies. One may call it “the story of
how Epimenides became respectable”. Its central feature is the fact that the theory
of formalized systems became a branch of number theory, hence may be treated
21 Cf. Feferman [1984], p. 98, remarks (ii) and (iii).

22 See Feferman [1984], p. 96
within formalized arithmetics and thus self-reference has found its way into precise
theories.
Features 79.1. The salient philosophical features in Gödel’s incompleteness results

on which I shall concentrate are the following:
(GD1) contentually false propositions in consistent theories are possible, in
principle;
(GD2) it is conceivable that there exist finitary proofs that cannot be ex-
pressed in the formalism of Principia Mathematica.
The theoretical-historical side of the fairly short period from virtually the posing
of the problem of the completeness of what is commonly called today the (classical)
first order logic and (classical) first order Peano arithmetic by Hilbert on the 1928
Bologna-Congress until their solution by Gödel within the following two years is
beautifully worked out in Wang [1986]. Readers who are more closely interested in
that period are advised to consult Wang’s book on Gödel.
79a. The rise of self-reference in formalized theories. Self-reference, or

in more technical terms, the (indirect) fixed point property, took considerable time
to emerge from so simple a paradox as that of the liar. What it needed was the
idea of arithmetization which made possible a certain representation of metatheory
within the object-theory.
Arithmetization may have roots in Leibniz’ discovery of the possibility of ex-
pressing everything by means of numbers, in fact by means of only two digits.
Leibniz seemed to have envisaged the encoding of philosophical problems into equa-
tions over natural numbers and then applying the laws of arithmetic to solve them.
The result, translated back into language, must be true by virtue of the truth of
the laws of arithmetics.
Gödel’s procedure of arithmetization of metatheory depends on the specific
features of a formalized theory. The actual procedure of this arithmetization is
tedious. It is, however, important; this is why its fundamentals are provided in §12
in the tools.
Quotations 79.2. (1) Van Heijenoort [1967b], p. 597 (also: Feferman et al. (eds.)
[1986], p. 147); translating Gödel [1931], p. 174.
The formulas of a formal system (we restrict ourselves here to the system
PM ) in outward appearance are finite sequences of primitive signs (vari-
ables, logical constants, and parentheses or punctuation dots), and it is
easy to state with complete precision which sequences of primitive signs
are meaningful formulas and which are not. Similarly, proofs, from a formal
point of view, are nothing but finite sequences of formulas (with certain
specifiable properties). Of course, for metamathematical considerations it
does not matter what objects are chosen as primitive signs, and we shall
assign natural numbers to this use. Consequently, a formula will be a finite
sequence of natural numbers, and a proof array a finite sequence of nat-
ural numbers. The metamathematical notions (propositions) thus become
notions (propositions) about natural numbers or sequences of them;† there-

fore they can (at least in part) be expressed by the symbols of the system
PM itself. In particular, it can be shown that the notions “formula”, “proof
array”, and “provable formula” can be defined in the system PM ; that is,
we can, for example, find a formula F (v) of PM with one free variable v
(of the type of a number sequence) such that F (v), interpreted according
to the meaning of the terms of PM, says: v is a provable formula.
(2) Footnote to the above quotation, relating to † .
In other words, the procedure described above yields an isomorphic image
of the system PM in the domain of arithmetic, and all metamathematical
arguments can just as well be carried our in this isomorphic image. [[. . .]]
Comment. This is worth a moment’s reflection in terms of the gnoseological diagram
101.19 on p. 1407 in the groundworks: the question in that framework will be,
whether it is possible to produce an isomorphic image of the world in the mind.
(3) Hilbert and Bernays [1939/70], p. 216, footnote 1; my translation.
The idea of representing the symbolic formulas in this a way as a succession
of numerals and thus translate the rules for the formation of symbolic for-
mulas and for the performance of formal proofs into arithmetic rules, had
already been envisaged by Hilbert in the context of his considerations re-
garding Cantor’s continuum problem. At that time, however, the realization
of this idea seemed to be encumbered with too many complications.
(4) Hilbert and Bernays [1939/70], p. 217 f; my translation.
We can, however, accomplish arithmetization also by imitating arithmeti-
cally the grammatical structure of formulas instead of the way they are
written. The difference of the procedure consists in that, that we do not
assign numbers to the predicates, symbols, and function symbols, as well
as the logical symbols, but arithmetical functions.
Quotation 79.3. Hilbert and Bernays [1939/70], p. 294; my translation.

Observe that the removal of the contradiction in the case of the formaliza-
tion of the modified antinomy of the liar is an essentially different one from
the one in the case of the original antinomy. While in the case of the formal
reproduction of that antinomy the contradiction ceases to apply because a
certain combination of words of the language is not translatable into the
formalism, in the case of the modified antinomy all linguistic termini under
consideration can also be represented in the formalism, and the avoidance
of the antinomy is here based on the fact that some inference, admissible in
linguistic argumentation, turns, under the translation into the formalism,
into a non admissible inference in the formalism.
Quotations 79.4. (1) Feferman et al. (eds.) [1986], pp. 203 & 205; translating
Gödel [1931a], pp. 150 f.
[[I]]n all the well-known formal systems of mathematics—for example, Prin-
cipia mathematica (together with the axioms of reducibility, choice and
infinity), the Zermelo-Fraenkel and von Neumann axiom systems for set
theory, and the formal systems of Hilbert’s school—there are undecidable
arithmetical propositions. [[. . .]] For all formal systems for which the ex-
istence of undecidable arithmetical propositions was asserted above, the
assertion of the consistency of the system in question belongs to the propo-
sitions undecidable in that system. That is, a consistency proof for one of
these systems [[Σ]] can be carried out only by means of methods of infer-
ence that are not formalized in [[Σ]] itself. For a system in which all finitary
(that is, intuitionistically unobjectionable) forms of proof are formalized,
a finitary consistency proof, such as the formalists seek, would thus be
altogether impossible. However, it seems questionable whether one of the
systems hitherto set up, say Principia mathematica, is so all-embracing (or
whether there is a system so all-embracing at all).
(2) Feferman et al. (eds.) [1986], p. 201; translating Gödel [1931a], p. 147.
[[The formalist]] view presupposes that if one adjoins to the system S of
meaningful propositions the system T of transfinite propositions and ax-
ioms and then proves a theorem of S by making a detour through theorems
of T , this theorem is also contentually correct, hence that through the ad-
junction of the transfinite axioms no contentually false theorems become
provable. This requirement is customarily replaced by that of consistency.
Now I would like to point out that one cannot, without further ado, regard
these two demands as equivalent. For, if in a consistent formal system A
(say that of classical mathematics) a meaningful proposition p is provable
with the help of the transfinite axioms, there follows from the consistency of
A only that not-p is not formally provable within the system A. Nonethe-
less it remains conceivable that one could ascertain not-p through some
sort of contentual (intuitionistic) considerations that are not formally rep-
resentable in A. In that case, despite the consistency of A, there would
be provable in A a proposition whose falsity one could ascertain through
finitary considerations.
(3) Feferman et al. (eds.) [1986], p. 203; translating Gödel [1931a], p. 148.
(Assuming the consistency of classical mathematics) one can even give
examples of propositions (and in fact of those of the type of Goldbach or
Fermat) that, while contentually true, are unprovable in the formal system
of classical mathematics. Therefore, if one adjoins the negation of such a
proposition to the axioms of classical mathematics, one obtains a consistent
system in which a contentually false proposition is provable.
Comment. This is a very important point in the present study. From a dialectical
point of view it cannot be established that, e.g., the negation of Gödel’s sentence is
a false proposition.
(4) Wang [1986], p. 91, on Gödel’s letter to Zermelo of 12 October 1931.
He does not see, G[[ödel]] remarks, the essential point of his result in
that we cannot include the whole mathematics in a formal system (or we
can go beyond any given formal system); that already follows from Cantor’s
diagonal procedure and does not exclude the possibility of completeness of

certain subsystems of mathematics. Rather the essential point of his result
is that every formal system of mathematics (which includes addition and
multiplication) contains rather simple propositions that are expressible in
it but undecidable by it.
Comment. Note that there is no mention of truth.
As regards the interpretation of the second result, the unprovability of consis-

tency in the formalized theory itself, Gödel himself pointed out that he did not
regard it as a refutation of Hilbert’s programme.
Quotations 79.5. (1) Van Heijenoort [1967b], p. 615 (also: Feferman et al. (eds.)
[1986], p. 195); translating Gödel [1931], p. 197.
I wish to note expressly that Theorem XI [[Gödel’s second incompleteness
theorem]] (and the corresponding results for M [[set theory]] and A [[classical
mathematics]]) do not contradict Hilbert’s formalistic viewpoint. For this
viewpoint presupposes only the existence of a consistency proof in which
nothing but finitary means of proof is used, and it is conceivable that
there exist finitary proofs that cannot be expressed in the formalisms of
P [[Principia mathematica]] (or of M or A).
Comment. This is important in view of the general attitude (expressed, e.g., in
quotations 84.20 (1) and (2) in these materials) claiming somewhat a collapse of
Hilbert’s program.
(2) Hilbert and Bernays [1934], p. VII; my translation.
I wish to point out that the opinion which temporarily appeared on the
scene that certain more recent results of Gödel implied the impossibility
of my proof theory is shown erroneous. What that result shows in fact is
only that for further reaching consistency proofs one has to exploit the
finitary position more effectively than this is necessary in the consideration
of elementary formalisms.
Comment. There are interesting remarks to this passage by the editors in the Col-
lected Works by Gödel (Feferman et al. (eds.) [1986]).
(3) Hilbert [1930b], p. 194; my translation.
If it has been established that the formula
A (z)
always, if Z is a given numeral, turns into a correct numerical formula, then
the formula
(x) A(x)
may be put as a starting formula.
Let me remind you that the statement (x) A(x) reaches much further
than the formula A (z), where z is an arbitrarily given numeral. The reason
is that in the first case one may not only substitute for x in A(x) a numeral,
but also an expression of number character formed in our formalism[[.]]
Comment. I wonder if this is what led to Schütte’s way of proving cut elimination
in semi formal arithmetic.
(4) Wang [1986], p. 87.
In his review of [[Hilbert’s paper [1931] quoted above]], G[[ödel]] merely says
that the rule, ‘structurally, is of an entirely new kind.’ According to Car-
nap’s ‘diary,’ G[[ödel]] said on 21 May 1931 that Hilbert’s program would
be compromised by acceptance of this rule.
(5) Tarski [1956], p. 294
[[S]]uch a rule, on account of its ‘infinitistic’ character, departs significantly
from all rules of inference hitherto used, [[. . .]] it cannot easily be brought
into harmony with the current view of the deductive method, and finally
[[. . .]] the possibility of its practical application in the construction of de-
ductive systems seems to be problematic in the highest degree[[.]]
79b. Church, Rosser, Kleene. According to Feferman et al. (eds.) [1986],

p. 338, “Gödel lectured on his 1931 results at the Institute for Advanced Study
[[Princeton]] from February to May 1934. Notes were taken by S. C. Kleene and J. B.
Rosser.” This seems to have been the starting point for a number of generalizations,
extensions, and other new results based on Gödel’s new techniques the main results
of which are the undecidability of arithmetic and of first order logic. As far as the
technical side of these results is concerned, the reader is referred to §48 in the tools.
In this section I shall only pick out a few remarks which are suitable to shed some
light on the philosophical implications of these result.
Quotation 79.6. Kleene [1952], p. 246.

[[B]]y selecting a particular enumeration of the formal objects, or a particular
correlation of distinct natural numbers to the distinct formal objects (not
using every number), and then talking about the correlated numbers instead
of the formal objects, metamathematics becomes a branch of the arithmetic
of the natural numbers.
Quotation 79.7. Church [1941], p. 1.

Underlying the formal calculi which we shall develop is the concept of a
function, as it appears in various branches of mathematics, either under
that name or under one of the synonymous names, “operation” or “transfor-
mation.” The study of the general properties of functions, independently of
their appearance in any particular mathematical (or other domain), belongs
to formal logic or lies on the boundary line between logic and mathematics.
[[. . .]]
A function is a rule of correspondence by which when anything is given
(as argument ) another thing (the value of the function for that argument)
may be obtained. That is, a function is an operation which may be ap-
plied on one thing (the argument) to yield another thing (the value of the
function).
Comment. How does this relate to Frege’s Function and Concept ?23
Quotation 79.8. Hilbert and Ackermann [1938/50], p. 124.

[[I]]t should be noted that the impossibility of a general decision procedure
does not mean that we can find definite formulas whose universal validity
has been proved not to be decidable. To assume the existence of such a
proof would in fact lead to an immediate contradiction. From such a proof
it would follow that the formula would not be deducible from the axiom
system of [[classical quantificational logic]]. But by the completeness theorem
[[37.9 in the tools]], the satisfiability of the contradictory of the formula
could then be proved. Thus the universal validity of the formula would be
decided after all, viz. in the negative.
Comment. This situation does not extend to logic without contraction.
Quotations 79.9. (1) Church [1934], p. 357.

The Richard paradox can be said to consist in the following problem.
How is it possible that a system of symbolic logic, in which the set of all
formulas is enumerable, should be adequate for any branch of mathematics
which deal with the members of a non-enumerable set (in particular for
elementary number theory)?
(2) Church [1934], p. 360.
[[I]]f there is no formalization of logic as a whole, then there is no exact
description of what logic is, for it is in the very nature of an exact description
of logic that it implies a formalization. And if there is no exact description
of logic, then there is no sound basis for supposing that there is such a thing
as logic.
[[. . .]]
Fortunately, however, there is a way out of this condition of nihilism.
The theorem which led us to such pessimistic conclusions does not really
apply to all systems of symbolic logic but only to systems which satisfy
certain condition. And one of these conditions is, either that there shall be
a unique symbol for implication between propositional functions, or that
there shall be a set of symbols for implication and an effective way by which
we can always determine whether a given formula is one of the symbols for
implication. For, in the contrary case, there would be no effective way of
picking out from a list of theorems those which had the form N (x) ⊃x
N (f (x)), and hence we could escape from our second dilemma in the same
way that we did from our first one.
Therefore we seek a system of symbolic logic in which the notion of
implication between propositional functions is obtained by definition, and
in which there are a variety of notions of implication, obtainable by different
definitions. In the case of each definition we desire that it shall be possible
by an intuitive argument to prove the character of the defined symbol as
23 Cf. Church’s statement regarding his acquaintance with Frege’s work in his letter to the
editors of Scott [1980] as quoted there on p. 260.

an implication symbol. But there shall be no uniform means of determining

whether a given formula is an implication symbol.
A system of this sort not only escapes our unpleasant theorem that is
must be either insufficient of oversufficient, but I believe that it escapes the
equally unpleasant theorem of Kurt Gödel.
I close this section with a quotation regarding the third “truth value” in Kleene’s
“three valued logic”.
Quotation 79.10. Kleene [1952], pp. 332 f.

[[W]]e shall introduce new senses of the propositional connectives, in which,
e.g. Q(x) ∨ R(x) will be defined in some cases when Q(x) or R(x) is unde-
fined.
It will be convenient to use truth tables, with three “truth values” t
(‘true’), f (‘false’) and u (‘undefined’), in describing the senses which the
connectives shall now have.
Some remarks are appropriate to justify our use of truth tables here
form the finitary standpoint, and to explain how we are led to choose the
particular tables given below.
We were justified intuitionistically in using the classical 2-valued logic,
when we were using the connectives in building primitive and general recur-
sive predicates, since there is a decision procedure for each general recursive
predicate; i.e. the law of the excluded middle is proved intuitionistically to
apply to general recursive predicates.
Now if Q(x) is a partial recursive predicate, there is a decision procedure
for Q(x) on its range of definition, so the law of excluded middle or excluded
“third” (saying, that for each x, Q(x) is either t or f) applies intuitionistically
on the range of definition. But there may be no algorithm for deciding,
given x, whether Q(x) is defined or not (e.g. there is none when Q(x) is
µyT1 (x, x, y) = 0). Hence it is only classically and not intuitionistically that
we have a law of the excluded fourth (saying that for each x, Q(x) is either
t, f or u).
The third “truth value” u is thus not on a par with the other two t and
f in our theory.
79c. Semantics and the concept of truth in formalized theories. In

the case of Gödel’s first incompleteness theorem it is usually said that a true wff is
not provable. What does true mean here?
The main obstacle in defining a notion of truth are the so-called semantic
antinomies.
Remark 79.11. In view of the aim of the present study it is helpful to keep in mind
that “correct” should be substituted for “true” if one wants to avoid confusion with
the notion of truth in speculative philosophy.24
24 As regards the latter, see quotations 65.16 (2) and (3). As regards the distinction between
“correct” and “true“ see also quotations 65.10 (4) and (5) in these materials.
Quotation 79.12. Wang [1986], p. 51 (partly quoting from Carnap’s diary).

On 2 July 1931, G[[ödel]] proved ‘in the [[Schlick/Vienna]] circle’ that
arithmetic truth is not definable (in arithmetic). ‘There are in arithmetized
metalogic ordinary concepts which are not definable. This is proved by
deriving a contradiction (after Richard) from supposing the concept avail-
able; in the argument we have to grant the consistency of arithmetic. True
Zahlformel (without variable) is not definable.’
Quotations 79.13. (1) Hilbert and Bernays [1939/70], p. 263; my translation.

The method of arithmetization of metamathematics was developed by
Gödel for the purpose of establishing two general theorems which state the
deductive incompleteness of every well defined, however not too narrow,
logical-mathematical formalism.
The idea of the proof by means of which Gödel obtained these theorems
provides at the same time a method of mathematical tightening of those
logical and set theoretical paradoxes in which the relation of designation
and designated object plays a significant role[[.]]
(2) Hilbert and Bernays [1939/70], p. 271 f; my translation.
A remarkable case of this kind is the impossibility of representing the
notion “value of a number-determining expression” which, again, can be
established for a deductive formalism under some very general conditions.
[[. . .]]
[[. . .]]
[[. . .]] assumption: d1 ) There exists a term e(a), in which the number
variable a is the only variable which occurs free, and which is such that, if n
is the number of a term t which contains no free variable, then the equation
e(n) = t
is derivable[[. . .]].
Comment. Interested readers might find it worthwhile looking at Wang [1955] for
further elaborations along this line.
Quotation 79.14. Tarski [1944], p. 81, note 16.

To define recursively the notion of satisfaction, we have to apply a certain
form of recursive definition which is not admitted in the object-language.
Hence the “essential richness” of the meta-language may simply consist in
admitting this type of definition. On the other hand, a general method is
known which makes it possible to eliminate all recursive definitions and to
replace them by normal, explicit ones. If we try to apply this method to the
definition of satisfaction, we see that we have either to introduce into the
meta-language variables of a higher logical type than those which occur in
the object-language; or else to assume axiomatically in the meta-language
the existence of classes that are more comprehensive than all those whose
existence can be established in the object-language.

Die rekursiven Definitionen für Summe und Produkt lassen sich [[. . .]]
im Formalismus des Systems [[CPA1 ]] nicht so wie explizite Definitionen
durch Ersetzungen eliminieren. Wohl aber lassen sich die weiteren Funk-
tionsdefinitionen durch primitive Rekursionen im Formalismus [[CPA1 ]] bei
Hinzunahme des ι-Symbols auf explizite Definitionen zurückführen; d.h. es
lassen sich für die rekursiv eingeführten Funktionszeichen anstatt ihrer re-
kursiven Definitionen solche expliziten Definitionen angeben, mittels deren
die Rekursionsgleichungen jener rekursiven Definitionen ableitbar sind.
(2) Hilbert and Bernays [1939/70], p. 289; my translation.
A deductive formalism for which a proof of consistency succeeds cannot,
therefore, comprise the whole range of possible proofs for number theoretic
propositions.
Disregarding the question of the proof of consistency, the thought of
a characterization of mathematics in general as a deductive formalism, as
it was temporarily suggested by logistic systems, appears inappropriate al-
ready on the basis of the stated deductive incompleteness [Unabgeschlossen-
heit] of every sufficiently expressive and sufficiently delimited deductive for-
malism.
In our presentation of the starting problematic and the aim of the proof
theory we have avoided right from the start to introduce the idea of a total-
system of mathematics in a philosophically basic meaning, but confined
ourselves to characterize the actually existent systematics of analysis and
set theory as such, which form a suitable framework for the classification
of the geometric and physical disciplines. This purpose can be satisfied by
a formalism without having the property of full deductive completeness
[Abgeschlossenheit].
Our view that the material [inhaltliche] standpoint which underlies the
development [Ausbildung] of the systems of analysis and set theory is one of
extrapolating formation of ideas, is also well compatible with the deductive
incompleteness [Unabgeschlossenheit] of the system: the ways of inference in
the system are oriented towards the idea of a closed, completely determined
actuality and represent this idea formally; but from this it does not follow
that the deductive (metamathematically to be acknowledged) structure has
to have the property of deductive completeness [Abgeschlossenheit].
§ 80. Proof theory, ordinal analysis, and truth definitions
In the beginning, logic was axiomatic, somewhat modeled on Eukleides’ geometry.

It might not come as a surprise, therefore, that logic has to struggle with similar
problems such as the consistency and completeness of its axioms. These problems
may not be too difficult to handle in the case of so-called first-order logic, but once
formalized theories are based on an axiomatic formulation of logic, the difficulties
increase considerably. But logic does not have to be formulated as an axiomatic
system. To have realized this and developed an alternative is the achievement of
§ 80. PROOF THEORY, ORDINAL ANALYSIS, AND TRUTH DEFINITIONS 1111
Gerhard Gentzen. His so-called sequential formulation of logic is one of the central
tools in my approach to dialectical logic.
Features 80.1. I shall focus on the following aspects in Gentzen’s analysis of logical
reasoning:
(GN1) isolation of structural rules;
(GN2) symmetrical formulation of rules for constants;
(GN3) cut elimination.
80a. From axiomatics to the calculus of sequents. The sequential formu-

lation of formalized theories is probably the most important tool in my enterprise
of a foundation of dialectic. Its importance stems from the fact that it allows con-
sistency proofs, via cut elimination, in which the length of formulas plays no role.
The present section is meant to provide a little background information regarding
the emergence of the sequential approach from the axiomatic approach.
Quotation 80.2. Hilbert and Ackermann [1928/59], p. 25.

Die Auswahl der Grundformeln des Axiomensystems und der Regeln
zur Ableitung neuer Formeln ist [[. . .]] weitgehend der Willkür überlassen. Es
gibt da keine eindeutige Lösung. Gewisse Forderungen werden wir allerdings
an die Grundformeln und die Ableitungsregeln stellen, z. B. daß sie nicht
zu kompliziert sind und daß wir mit möglichst wenig Axiomen und Regeln
auskommen.
Comment. Hilbert and Ackermann’s Grundzüge der theoretischen Logik is particu-
larly interesting because it had its first edition in 1928 and was completely reworked
in 1958. By that time Gentzen’s work had revolutionized the approach to logic and
Ackermann has taken this into account by presenting a modified system of Schütte’s
which takes advantage of Gentzen’s improvements.
Quotations 80.3. (1) Szabo [1969], p. 68; translating Gentzen [1934], p. 176.
1. My starting point was this: The formalization of logical deduction,
especially as it has been developed by Frege, Russell, and Hilbert, is rather
far removed from the forms of deduction used in practice in mathematical
proofs. Considerable formal advantages are achieved in return.
In contrast, I intended first to set up a formal system which comes as
close as possible to actual reasoning. The result was ‘calculus of natural
deduction’[[.]]
(2) Szabo [1969], pp. 68; translating Gentzen [1934], p. 177.
2. A closer investigation of the specific properties of the natural calculus
finally led me to a very general theorem which will be referred to below as
the ‘Hauptsatz ’.
The Hauptsatz says that every purely logical proof can be reduced to
a definite, though not unique, normal form. Perhaps we may express the
essential properties of such a normal proof by saying: it is not roundabout.
No concepts enter into the proof other than those contained in its final
result, and their use was therefore essential to the achievement of that
result.
The Hauptsatz holds both for classical and for intuitionist predicate
logic.
In order to be able to enunciate and prove the Hauptsatz in a convenient
form, I had to provide a logical calculus especially suited to the purpose.
For this the natural calculus proved unsuitable. For, although it already
contains the properties essential to the validity of the Hauptsatz, it does
so only with respect to its intuitionist form, in view of the fact that the
law of excluded middle [[. . .]] occupies a special position in relation to these
properties.
[[. . .]]
The Hauptsatz permits of a variety of applications. To illustrate this I
shall develop a decision procedure [[. . .]] for intuitionist propositional logic
[[. . .]], and shall in addition give a new proof of the consistency of classical
arithmetic without complete induction[[.]]
(2) Szabo [1969], pp. 88; translating Gentzen [1934], pp. 195 f.
Intuitively speaking, these properties of derivations without cuts may be
expressed as follows: The S-formulae become longer as we descend lower
down in the derivation, never shorter. The final result is, as it were, grad-
ually built up from its constituent elements. The proof represented by the
derivation is not roundabout in that it contains only concepts which recur
in the final result[[.]]
Remark 80.4. As Gentzen pointed out,25 his Hauptsatz was partly anticipated by
Herbrand.
80b. Cut elimination and consistency proofs. What is so interesting
about the cut rule? One aspect is that the eliminability of the cut rule is closely
linked to the consistency of a formalized theory. In the case of LK observe that
except for the cut rule all the rules are such that the lower sequent is always longer
than each of the upper sequents and every subformula of the upper sequent occurs
again in the lower sequent. The cut rule is the only rule in Gentzen’s (symmetrical)
calculus of sequents which enables one to get to shorter wffs. In particular: without
cut no empty sequent. This is why cut-elimination became so important for proof
theory. Its first really new application was Gentzen’s proof of the consistency of first
order Peano arithmetic (PA).
Quotations 80.5. (1) Szabo [1969], pp. 260 f; translating Gentzen [1938], p. 26.
It is to be shown that every derivation is consistent ; this may be para-
phrased by saying that no derivation has an empty endsequent. For from
a contradiction, → A and → ¬A, we can first of all derive the sequents
→ ¬A and ¬A →, and from them, by means of a cut, the empty sequent.
(Conversely, from the empty sequent every arbitrary sequent can be derived
by ‘thinnings’.)
25 Gentzen [1934], p. 177, footnote 3 ), and p. 409, footnote 6 ).
It makes sense that we should begin by proving the consistency of simple

derivations, then of more complex ones, using the consistency of the simpler
derivations, and so forth. We thus proceed ‘inductively’. It is furthermore
not implausible that this procedure repeatedly requires the examination of
an already infinite sequence of derivations before a more complex class can
be tackled; for example, first all derivations consisting of only one sequent,
then all derivations consisting of two sequents, etc. This actually means
that we are applying a ‘transfinite induction’. In practice the pattern of
this analysis is of course considerable more involved than in the case of the
given example.
Comment. Gentzen’s explanation of the involvement of transfinite ordinals in the
consistency proof: one has to finish infinite sets of deductions. Question: what con-
stitutes these sets? My favorite is the role of contraction in the (part-) elimination
of cuts, more precisely, in the bifurcation step.26
(2) Szabo [1969], pp. 280 f; translating Gentzen [1938], p. 40.
[[T]]he purpose of ω α1 +1 in the evaluation of a CJ -inference figure [[. . .]]: in
the reduction the figure is broken up into a number of cuts; and in some
sense the n-fold multiple of one and the same derivational section occurs.
In order to achieve a decrease in the ordinal number, we must therefore
choose as the ordinal number of the original derivational section up to the
CJ -inference figure the ‘limit number ’ of all ‘n-fold multiples’ of the ordinal
number of the upper sequent, i.e., ω α1 +1 = ‘ω α1 · ω’. (The expressions in
‘ ’ serve of course only as illustrations; they are not even defined in this
context.)
(3) Szabo [1969], p. 283; translating Gentzen [1938], p. 42.
The main idea is: in the reduction the same derivational section occurs twice,
although both times somewhat simplified. In the general case, however,
α < α1 + α2 , where α1 and α2 are assumed to be smaller than α. For the
exponential expression, however, it holds that ω a > ω a1 + ω a2 .
Comment. Here is the key to understanding why contraction is so relevant in my
approach: “in the reduction the same derivational section occurs twice”. Contraction
is one source of this phenomenon, induction another.
(4) Schütte [1951], pp. 369 f; my translation.
The proof theoretical seizure (Erfassung) of a mathematical theory is
carried out by means of a “codificate”, i.e. by means of a system which does
not only formally represent the mathematical relations, but also all logical
forms of inference occurring in the theory under discussion. The proof of a
mathematical theorem appears in a codificate as an arrangement of symbols
set up according to certain rules, a so-called “deduction”. Because of the
finitude of our thinking, every mathematical proof contains only a finite
number of inferences. This fact finds expression in the codificates which
only take into account finite deductions.
26 See the intuitive consideration 46.9 in the tools.
There are, however, reasons for calling in codificates with infinite de-
ductions, and that, in particular, for the use of “infinite induction”. By
this the form of inference is meant which concludes from the validity of a
sentence for every single number the corresponding generalized statement.
[[. . .]] But even such codificates can be brought into accord with the finitude
of our thought. For it is only to be required that the infinite deductions are
always constructed in a particular law-governed way, so that they can be
described on the basis of this in a unique way. The finitude of our thought
is then not expressed in the deduction itself, but in the metamathematical
seizure of the deduction.
Codificates which contain the infinite induction are distinguished from
the more confined codificates in which only finite deductions are admitted
in particular by “ω-completeness”. [[. . .]]
A further advantage is offered by the infinite induction, because by
adding this form of inference, deductions in codificates which contain num-
ber theory can also be carried out without detour in the sense of Gentzen,
i.e. the schema of inference of “cuts” can be eliminated.
Comment. A footnote refers to Hilbert’s schema of inference in his paper [1930b],
the relevant part of which can be found in quotation 79.5 (3) above.
80c. ‘Finitary means’, Gödel’s second theorem and ordinal analysis.

Against the odds, or so it seems, Gentzen managed to outwit the limitations placed
on Hilbert’s programme by Gödel’s second incompleteness theorem.
Quotation 80.6. Hilbert and Bernays [1934], p. 34 f.

Die Überschreitung des finiten Standpunktes findet bereits in den
Schlußweisen der Zahlentheorie statt, indem hier Existenzaussagen über
ganze Zahlen [[. . .]] zugelassen werden ohne Rücksicht auf die Möglichkeit
einer tatsächlichen Bestimmung der betreffenden Zahl, und indem man Ge-
brauch macht von der Alternative, daß eine Aussage über ganze Zahlen
entweder für alle Zahlen zutrifft oder daß es eine Zahl gibt, für die sie
unzutreffend ist.
Diese Alternative, das „tertium non datur“ für ganze Zahlen, kommt
implizite auch zur Anwendung bei dem „Prinzip der kleinsten Zahl“, welches
besagt: „Wenn eine Aussage über ganze Zahlen für mindestens eine Zahl
zutrifft, so gibt es eine kleinste Zahl, für die sie zutrifft.“
Das Prinzip der kleinsten Zahl hat in seinen elementaren Anwendungen
finiten Charakter.
Comment. As regards a formal version of the ‘least number principle’, see 46.20
in the tools. In view of my own approach: it is contraction which fails the least
number principle, and that even in its elementary applications.
Quotations 80.7. (1) Szabo [1969], pp. 238 f; translating Gentzen [1938], p. 9.
[[Gödel’s]] theorem has frequently been taken as conclusive proof that Hil-
bert’s programme is unrealizable. This view is based on the conviction —
and there seemed to be some evidence in its favour — that the ‘finitist’ or
‘constructive’ forms of inference by which the consistency proofs were to

be carried out, merely represent part of the precisely formalizable forms of
inference occurring in elementary number theory. If this were the case then,
according to Gödel’s theorem, these forms of inference would no longer be
sufficient to prove the consistency of number theory. I am of the opinion,
however, that there are forms of inference which are still in harmony with
the constructivist interpretation of infinity and which, nevertheless, tran-
scend the framework of formalized number theory, indeed, by their very
nature, these techniques can presumable be extended beyond the frame-
work of any formally delimited theory. [[. . .]] They are closely connected
with the ‘transfinite induction’ occurring in set theory, but this does not
mean that they are subject to the same contingencies as that rule; on the
contrary, they are proved constructively in a way entirely independent of
set theory.
(2) Szabo [1969], p. 287; translating Gentzen [1943], p. 140.
The impossibility of proving transfinite induction up to the ordinal
number ε0 with elementary number-theoretical techniques may be inferred
indirectly from the following two facts:
1. Gödel’s theorem: The consistency of elementary number theory can-
not be proved with the techniques of that theory.
2. The consistency of elementary number theory has been proved by
applying transfinite induction up to ε0 , together with exclusively elementary
number-theoretical techniques.
[[. . .]]
On the other hand, it is known that transfinite induction up to any
ordinal number below ε0 is provable in elementary number theory.
Comment. The birth of ordinal analysis. Cf. Pohlers quotation 85.17 (1) in the next
chapter for a formulation some forty odd years later.
Quotations 80.8. (1) Wang [1952], pp. 243 f.

Let S be a system containing the usual second-order predicate calculus
with the usual number theory as its theory of individuals, and S’ be a
system related to S as an (n + 1)th order predicate calculus is to an nth
except that we do not use variables of the (n + 1)th type in defining classes
of lower types. Tarski’s assertions seem to lead us to believe that we can
prove the consistency of S in S’. On the other hand, it is known that if S is
consistent then S has a model in the domain of natural numbers. But if S
has such a model, then, we seem also to be able to argue, S ′ has a model in
S because S contains both natural numbers and their classes. Therefore, we
can formalize (so it appears) these arguments in S ′ and prove within S ′ that
if S is consistent then S ′ is. If that be the case we shall have a proof of the
consistency of S ′ within S ′ and therefore, by Gödel’s theorem on consistency
proofs, S ′ (and probably also S) will be inconsistent. Indeed, since we need
at least a system like S ′ to develop analysis and since these reasonings
do not depend on peculiar features of the systems under consideration, we
shall be driven to the conclusion that practically every system adequate to

analysis is inconsistent.
In trying to examine exactly where the above arguments break down,
we have found it helpful to formalize more explicitly certain truth defi-
nitions and consistency proofs with such definitions. The results of such
formalizations [[. . .]], it is thought, bring out more clearly than usual cer-
tain features in the procedures of constructing truth definitions and proving
consistency. For example, the use of impredicative classes is dispensable for
defining truth but does not seem so for proving consistency[[.]]
Comment. A formal treatment of the kind Wang refers to is provided in §50 of the
tools for first order arithmetic within a system of second order arithmetic.
(2) Wang [1952], pp. 263.
In each of many different forms of set theory, we say that we can develop
the ordinary number theory. Sometimes within a same set theory we can
also develop number theory in more than one way. Naturally the number
theory which we obtain is in each case relative to the axioms of the set
theory as well as to the definitions we adopt for the arithmetic notions.
If we consider each set theory as a theory for the set concept, then the
number theory and the arithmetic concepts we obtain in each case are also
relative to the underlying set concept. In particular, in each system of set
theory which contains number theory, the principle of induction becomes
a set-theoretical principle derivable from the axioms of the system; and
whether induction is applicable to a certain sentence of the system depends
both on the strength of the axioms of the system and the definitions for
the arithmetic notions such as those of the number zero, for the successor
function, and for the predicate of being a natural number (or for the class
of all natural numbers).
CHAPTER XXI
PHILOSOPHY OF MATHEMATICS AND LOGIC TODAY
When Kant formulated his question “How are a priori synthetic judgments possi-
ble?” 1 he was apparently quite confident that he had provided the essential ingre-
dient for an understanding of the nature of mathematical knowledge.
When Frege set out to derive fundamental laws of arithmetic from what he saw
as a system of pure logic, he seems to have been convinced that he was about to
settle the question — against Kant.
When Whitehead and Russell wrote their Principia Mathematica they wanted
to defend logicism in the face of the antinomies.
When Hilbert designed his proof theory, he expressed his hope to achieve final
justification for the application of tertium non datur in mathematics which found
its expression in his famous “there is no ignorabimus in mathematics”.2
Since then studies in the foundations of mathematics have produced an ex-
tremely complex picture. In this chapter I shall try to put together those features
of the more recent topics in the philosophy of mathematics which bear upon my
attempt at bringing together dialectic and the paradoxes in logic, semantics, and
the foundations of mathematics.
Common to all these approaches is a new precaution towards abstraction and
unrestrained formation of concepts, in particular the actual infinite.
Discussions in the philosophy of mathematics often suffer from lack of depth in
either philosophy or mathematics. I find Gödel’s paper [1944] a great example of a
combination of both disciplines.
A considerable amount of work that is being done is revisionist. In every para-
graph of the present chapter, there is a section entitled ‘Looking back’ which tries
to take them into consideration.
§ 81. The general issue of foundation
I begin with the classical question in the philosophy of mathematics: how is it that
mathematics is applicable to the external world without being involved in empirical
observation? To make it short, there isn’t much progress. No powerful new ideas
1 Kant [1787] (B), p. 19, see quotation 60.14 in these materials.
2 Benacerraf and Putnam [1964], p. 150; translation of Hilbert [1926], p. 180. The allusion is
to Emil du Bois-Reymond (*1818−†1896, German physiologist, teaching in Berlin) who in a talk
on the limits of natural science (Über die Grenzen des Naturerkennens) asserted that there were
quite a number of “naturwissenschaftlicher Probleme” the solution of which would never succeed.
He closed his talk with the words “Ignoramus et ignorabimus”.
1117
1118 XXI. PHILOSOPHY OF MATHEMATICS AND LOGIC TODAY
have been put forward, but a lot has been written about the various metaphysical
attitudes circulating in the foundations of mathematics.
Features 81.1. As regards the general issue of the foundations of mathematics, I

am interested in the following topics:
(FM1) the question of the independent existence of mathematical objects;
(FM2) the “explanation” of concepts in terms of other concepts.
81a. Looking back. Hindsight is a curious thing. Sometimes it seems to be

more like a collective madness or blindness. The following quotations are meant to
throw some light on the genesis of our own hindsight.
Quotations 81.2. (1) E. T. Bell [1940/45], p. 5.

It is not known where or when the distinction between inductive infer-
ence—the summation of raw experience—and deductive proof from a set
of postulates was first made, but it was sharply recognized by the Greek
mathematicians as early as 550 b.c. [[. . .]] [[T]]here may be some grounds for
believing that the Egyptians and the Babylonians of about 2000 b.c. had
recognized the necessity for deductive proof.
(2) Struik [1967], p. 51.
Euclid’s treatment is based on a strictly logical deduction of theorems
from a set of definitions, postulates, and axioms.
Comment. Admittedly, Eukleides’ treatment of geometry was an amazing achieve-
ment. But in view of the precision aimed at and achieved by Frege,3 and Hilbert’s
axiomatic geometry, I am concerned not to blur the difference: Eukleides’ treatment
just wasn’t strict logical deduction, at least not according to modern standards.
Nevertheless, it set the standard for more than twenty centuries.
(3) Wang [1987], p. 177.
If we look at Euclid’s axiomatic theory for geometry from our present per-
spective, we notice two things: it is not a formal system, and it is open
to different interpretations. In the first respect, it is different from the (el-
ementary or first-order) formal systems of arithmetic and set theory, fa-
miliar to logicians today, which also admit different models, although not
by intention. In the second respect, Euclid’s theory is satisfied by a whole
spectrum of models from the full uncountable plane to its subset of points
constructible from two points by ruler and compass. Even its intention ap-
pears to be indefinite; in this regard it is different from Dedekind’s theory
A (the ‘second-order Peano arithmetic’ [[. . .]]), which is also not a formal
system. It may be said that the development of the axiomatic method
within mathematics has, since Euclid, but particularly since the beginning
of the nineteenth century, striven for the dual goal of finding mathematical
systems that are both formal and categorical. G[[ödel]]’s incompleteness the-
orem tells us that we cannot find axiomatic theories (of much mathematical
significance) that simultaneously satisfy both requirements.
§ 81. THE GENERAL ISSUE OF FOUNDATION 1119
(4) Gardner [1981], p. 19.

Poincaré held the opinion that if optical experiments seemed to show
physical space was non-Euclidean, it would be best to preserve the sim-
pler Euclidean geometry of space and assume that light rays do not follow
geodesics. Many mathematicians and physicists, including Russell, agreed
with Poincaré until relativity theory changed their mind.
Comment. The point that interests me is what actually made them change their
mind.
Gödel’s complaints as presented in quotation in 78.6 regarding the lack of for-

mal precision notwithstanding, Whitehead and Russell’s three volumes of Principia
Mathematica can be said to have provided a frame for foundational studies for
twenty years after the first volume appeared.
Quotations 81.3. (1) E. T. Bell [1940/45], pp. 558 f.

The Boolean tradition ended, in a sense, in 1905 with the completion
of the four-volume, 2,033-page treatise of F. W. K. E. Schröder (1841-1902,
German), Vorlesungen über die Algebra der Logik, 1890, 1891, 1895, 1905.
Symbolic logic as it existed at the turn of the century was here organized
and expounded with that thoroughness which the professorial world had
learned to anticipate from German scholarship in the nineteenth century.
Had not the originality of Frege and Peano inaugurated a renaissance in
mathematical logic, Schr[[ö]]der’s massive master-piece might have weath-
ered the centuries as the tombstone of Leibniz’ dream. But mathematical
logic was too vigorous to submit to premature burial under anybody’s two
thousand pages of meticulous erudition [[. . .]] When, if ever, it becomes pos-
sible to print 2,000 or 10,000 pages on mathematical logic that will retain
their vitality unimpaired for a generation, mathematics itself will be dead.
[[. . .]]
Peano’s project in the Formulaire of translating technical mathematics
into the symbols of mathematical logic which he had invented was partly
responsible for the most comprehensive logical symbolization of mathemat-
ics yet (1945) attempted, the Principia mathematica (1910, 1912, 1913)
of Whitehead and Russell. Here was executed in minute detail the pro-
gram outline by Russell in his Principles (1903), “to prove that all pure
mathematics deals exclusively with concepts definable in a small number
of fundamental logical concepts, and that all its propositions are deducible
from a very small number of fundamental logical principles[[.”]]
Comment. And, if I may add, Principia Mathematica stopped short of 2,000 pages
— by less than a hundred pages, not to threaten, of course, the liveliness of math-
ematical logic.
(2) Hodges [1983], p. 4.
Whitehead and Russell’s Principia Mathematica [1910] [[. . .]] contains no-
tation, axioms and theorems which we now regard as part of first-order
logic, and for this reasons it was quoted as a reference by Post, Langford,
Herbrand and Gödel up to 1931[[.]]
(3) Fraenkel et al. [1973], p. 299
In Peano’s original system the successor operation is regarded as primitive,
whereas the operations of addition and multiplication are introduced by
definitions, through the recursion equations [[as formulated in definitions
8.7 (1) and (2) in the tools]]. We know now that this way of introduc-
ing addition and multiplication is legitimate only within an appropriate
predicate calculus of at least second order; in the framework of first-order
predicate calculus these equations are taken to be the additional axioms
[[. . .]]
Peano himself acknowledges his debt to Dedekind whose characteriza-
tion of the natural numbers, however, is not quite axiomatic in the modern
sense. Dedekind, in his turn, recognizes a kinship to Frege’s work, though
Frege himself tended rather to stress the differences. A definition of addi-
tion and multiplication through informally stated recursion equations was
already given by Peirce in 1881 but was not known to either Frege, Dedekind
or Peano.
Comment. Recently Grassmann has joined the club and it looks like he can make a
reasonable claim of a priority of twenty years as against Peirce: see Odifreddi [1989],
p. 22; also Frege [1884], p. 8, where Frege criticizes the approach in Grassmann.
(4) Goodstein [1965], p. 117.
Frege’s definition of number in terms of classes is a purely technical device,
of value and importance in constructing arithmetic in a particular type of
axiom system, but no more than that.
Comment. For someone who agrees with this line, the present study is probably
rather misguided. I do, indeed, take the basic line of Frege’s definition of number as a
great philosophical achievement, viz., that numbers are determinations of concepts;
and the contradictions which follow from the basic law of logic (“Grundgesetz V”)
which enabled him to define numbers as determinations of concepts, I take as a
philosophically even more important result.
(5) Hodges [1983], p. 116.
The efforts of various nineteenth-century mathematicians reduced all the
concepts of real and complex number theory to one basic notion: classes.
So when Frege, in his Grundgesetze der Arithmetik I [1893], attempted a
formal system which was to be adequate for all arithmetic and analysis,
the backbone of his system was a theory of classes. One of his assumptions
was that for every condition there is a corresponding class, namely the class
of all the objects that satisfy the condition. Unfortunately this assumption
leads to contradictions, as Russell and Zermelo showed. Frege’s approach
has now been abandoned.
Comment. This is pretty much the standard story, school knowledge. There is noth-
ing really wrong with it, except that all subtleties have fallen prey to hindsight
interpretation. As a result it is virtually useless for providing an understanding of

what I want to do in this study; except, perhaps, to point out that I stick with
Frege’s approach, albeit not as a theory of classes, but of concepts.
Regarding Russell’s approach, in particular quotation 75.12 (2) in these mate-
rials.
Quotations 81.4. (1) Chihara [1972], pp. 254 f.
Having agreed with Poincaré as to the source of the paradoxes—the
supposed source being, in each instance, a violation of the vicious-circle
principle—one might think that this would have satisfied Russell’s desire for
a solution to the paradoxes. This was not so: for Russell, the vicious-circle
principle was “purely negative in its scope” ([[[1908]]], p. 63); he felt that an
adequate solution to the paradoxes must provide a positive theory which
would “exclude” totalities in accordance with the vicious-circle principle.
Comment. The point is relevant for me in view of Kant’s distinction between a
decision and a resolution. I employ a somewhat similar distinction in the ground-
works. It is a point which seems to have attracted little attention and this is why
I quoted Chihara here; I do not, thereby, want to claim that my distinction reflects
the one made by Russell.
(2) Chihara [1972], p. 263, footnote 23.
[[T]]he theory of descriptions led Russell to his “no-class” theory, which in
turn was thought to be the discovery in the quest for a solution of the
paradoxes since it seemed to provide him with the crucial step in finding a
justification for his type restrictions, and thus for his vicious-circle principle,
and it allowed him to treat all the paradoxes, including the set-theoretical
ones, in terms of propositions and propositional functions.
Quotations 81.5. (1) Curry [1980], p. 96.
In current mathematics the tendency is to think of a function in the sense of
Dirichlet, i.e. essentially as a set of ordered pairs, whereby the first elements
range over a set, the domain of the function, and when the first element of
a pair is given, the second is uniquely determined. Thus the notion of set
is more fundamental than that of function, and the domain of a function is
given before the function itself. In combinatory logic, on the other hand, the
notion of function is fundamental; a set is a function whose application to an
argument may sometimes be an assertion, or have some other property; its
members are those arguments for which the application has that property.
The function is primary; its domain, which is a set, is another function.
Thus it is simpler to define a set in terms of a function than vice versa; but
the idea is repugnant to many mathematicians, and probably Scott.
Comment. Compare also Church forty years earlier in quotation 79.7. What I am
doing in the groundworks is to take the notion of set, or rather concept, as
primary and I find it quite easy to define the notion of a function (given unrestricted
abstraction, of course) in terms of it. As regards Scott’s response to the suggestion
of repugnancy, cf. the next quotation.
(2) Scott [1980], pp. 254 f.

As regards the question of which comes first: the function or the set, it is
not a question of repugnance or prejudice on my part that causes me to
formulate constructions within set theory but a problem of helplessness.
And I can pinpoint rather narrowly where I think the trouble lies.
For the sake of argument think of a set as a truth-valued function. (I
know this over simplifies Curry’s approach, but a more subtle view is not
needed for the point I will make.) Instead of X ∈ A we will write A(X);
to assert X ∈ A means to assert A(X)=true. On the other hand, to as-
sert X ∈/ A means to assert A(X)=false. The domain of the function A is
“everything”—but now the rub starts. The Russell Paradox shows that the
domain cannot really be everything if we were to allow full comprehension.
There is no way around Cantor’s Theorem that there are more functions
than there are arguments and values.
Comment. Repugnance is a tricky thing to diagnose, simply because it is part of
the constitution of repugnance to deny its own presence. The best criterion of re-
pugnance I know of is blindness towards some possibility. In the case of Scott it
would be the blindness towards the possibility of treating unrestricted comprehen-
sion consistently on the basis of a non-classical logic. Cf. in this respect Myhill in
quotation 96.7 in these materials. As regards my own approach: the way around
Cantor’s theorem is employing a logic which doesn’t subscribe to full tertium non
datur. Details can be found in the groundworks.
(3) Scott [1980], p. 227.
Who first suggested that a function could be regarded as a set of ordered
pairs? By 1914 both Wiener and Hausdorff were doing just that (for re-
lations as well as functions and with pairs reduced to sets as well), but I
do not find any earlier references in their works or in some other books I
consulted.
Comment. The point is an interesting one because the view of relations as sets of
ordered pairs with the reduction of the notion of the ordered pair to that of sets is
one which I adopt completely.
(4) Barendregt [1977], pp. 1092 f.
Church originally designed the λ-calculus as part of a general system
of functions intended to be a foundation of mathematics. The paradox of
Kleene and Rosser [[[1935]]] showed that this system was inconsistent. After
[[Church [1941]]], Church seemed to have lost interest in using the λ-calculus
to provide a foundation for the whole of mathematics. Curry et al. [1958,
1972], on the other hand, have developed various systems of illative combi-
natory logic intended as an ultimate foundation. These systems, however,
have not been developed enough to be a satisfactory basis for mathematics.
(5) Curry [1980], p. 96.
Combinatory Logic is a branch of mathematical logic which is con-
cerned with the ultimate foundations. It is not an independent system of
logic, competing with the theory of types, abstract set theory, mereology,
or what not; nor does it attempt to form a consistent system adequate for
this or that portion of classical mathematics. Rather it forms a common
substratum for a variety of such theories.
I began this paragraph with what I called the classical question in the philos-
ophy of mathematics. But the subject of the philosophy of mathematics seems to
have changed.
Quotations 81.6. (1) Bachmann [1955], p. 5.

Es zeigt sich, daß sich nicht die gesamte mathematische Wirklichkeit
in ein formales System einordnen läßt, da das mathematische Denken im-
mer wieder über sich hinausführt. Man kann aber jedes „nicht-kategorische“
Axiomensystem (d. h. ohne Vollständigkeitsaxiom) durch neue Axiome er-
weitern, so daß man eine Ineinanderschachtelung von Systemen erhält [[. . .]],
d. h. man kann, um allen Bedürfnissen der Mengenlehre gerecht zu werden,
immer wieder neue Bereiche von Dingen schaffen (im Gegensatz zur Can-
torschen und Finslerschen Theorie, wo der Bereich der mathematischen
Objekte als fertig vorliegend betrachtet wird), wobei aber der Cantorsche
Bereich nie ausgeschöpft werden kann (wenn dieses Problem überhaupt
sinnvoll ist).
(2) Goodstein [1965], p. 118.
We must concede that Frege was right when he observed that we cannot
capture a concept in a system of axioms, for axiom systems are not cate-
gorical. Is there then some other way in which mathematical concepts may
be defined? The axiomatic method consists in specifying certain properties
of the concept from which other properties may be derived; the method
is purely formal because no property of the concept which is not specified
by the axioms, may be used in the derivation of new properties. When
we contrast an axiomatic with an intuitive approach to a problem we are
contrasting a derivation from premisses named in advance with a deriva-
tion from hidden, i.e. unexpressed premisses. To say that no axiom-system
suffices to capture the concept is not to contrast the axiomatic with the in-
tuitive but says, in effect, that the concept cannot be made precise. We have
the feeling that the concept is somehow in us, only we cannot quite bring
it out; but because we are ready to accept, or reject, certain consequences
of our assumptions does not mean that we know all possible consequences
in advance of drawing them out. To know what one wants is often no more
than a readiness to accept or reject what is offered, not to have at hand a
model by which the acceptance or rejection is to be made.
In general, it seems, the reductionist conception dominates; but the question

what to reduce mathematics to is open. Answers are cheap and so we enjoy a broad
variety of possible positions.4
4 The situation seems characteristic of societies with capitalist production mode: before you
know what you want, your money (efforts) has gone into things which were cheap.
Anyway, the wonderful thing about all these positions is that they can be
changed. The next quotation is just for the records.
Quotation 81.7. Fraenkel et al. [1973], p. 336.

Quine, starting as a logicist, has for many years tried to uphold a nom-
inalistic position but he now feels that conceptualism is a position into
which he can lapse when tired of his quixotic attempts at nominalistic
reconstruction[[.]]
I close this section with a couple of evergreen lines, just as a reminder of the
resilience of platitudes.
Quotations 81.8. (1) Passmore [1957/66], p. 147.

Frege attempts — in his The Foundations of Arithmetic (1884) and Fun-
damental Laws of Arithmetic (1893–1903) — to make arithmetic secure by
deriving it from the laws of logic: his philosophy grows out of the problems
which that attempt engenders.
Comment. Here we go again: the stubborn prejudice that Frege aimed at making
arithmetics secure. Contrast what Frege says in quotation 71.12 (3) in these mate-
rials. There seems to be some problem in seeing a philosophical issue, beyond that
simple minded one of ‘securing’ a science, such as to establish the analytic nature of
arithmetic, or to get an “insight into the dependence of truths upon one another.” 5
(2) E. T. Bell [1940/45], p. 73.
Of all changes that mathematical thought has suffered in the past 2,300
years, the profoundest is the twentieth-century conviction, apparently final,
that Plato’s conception of mathematics was and is fantastic nonsense of
no possible value to anyone, philosopher, mathematician, or mere human
being.
Comment. Who would have doubted, if only for a moment, that a conviction gained
in the twentieth-century is final anyway?
81b. Mathematics, realism, and the external world. I have presented

the basic idea of Frege’s logical foundation of arithmetic of how the laws of numbers
apply to the outside world. As an answer to an eminently philosophical, or even
metaphysical question, this seems to have been lost out of sight.
Quotation 81.9. Yourgrau [1969], p. 78.

Mathematics ‘flirted’ with so-called real objects of experience and thereby
evolved a formalism that was evidently adequate to them. In other words,
the physical object (or entity or formula) appeared to possess a mathemat-
ical structure. Some authors went so far as to propound a preestablished
harmony between mathematics and physical reality! Einstein was bothered
by this strange relationship between the analytic, viz. the propositions of
5 “Einsicht in die Abhängigkeiten der Wahrheiten von einander”; Frege [1884], p. 2. Cf. quo-

mathematics, and the synthetic, viz. the facts of the empirical world: How
can it be that mathematics, a product of human thought independent of
experience, is so admirably adapted to the objects of reality? For Whittaker
the answer was simple: The world is rationally formed! It is a regrettable
circumstance that the connection between mathematics and physical reality
is so heavily cumbered with plausible as well as highly fictitious arguments
and explanations. For instance, geometry and analysis have been compared
with a mirror (or image) of the visible world, and theory of numbers has
been regarded as an expression of the audible word. Further, it can be
shown that the theory of partial differential equations stems from physical
situations of a more refined nature.
Quotations 81.10. (1) Wang [1985], p. 460.

In a famous passage [[. . .]] Gödel compares physical objects with sets
(mathematical objects) and says that “even our ideas referring to physical
objects contain constituents qualitatively different from sensations or mere
combinations of sensations.” He also says that sets and physical objects
both have the function of “the generating of unities out of manifolds”: this
is clearly the essence of sets and in the case of physical objects we have
the Kantian idea of generating “one object out of its various aspects.” In
conversations, Gödel speaks of sets as “quasi-spatial” objects.
(2) Wang [1987], p. 52.
G[[ödel]] saw a strong analogy between theoretical physics and set theory.
Physics is confirmed by sense perceptions; set theory by its consequences
in elementary arithmetic. The fundamental insights in arithmetic, which
cannot be reduced to anything simpler, are analogous to sense perceptions.
(3) Wang [1987], p. 20, quoting from an unsent letter of Gödel’s to Grandjean.
[[T]]he view that mathematics is syntax of language [[. . .]], understood in any
reasonable sense, can be disproved by my results.
Putnam [1969], p. 216, argued that “the ‘necessary’ truths (or rather ‘truths’)
of Euclidean geometry turned out to be falsehoods”, and that for empirical reasons.
Quotation 81.11. Putnam [1969], p. 219.

The old scientific use of the term ‘straight’ line rested on a large number
of laws. Some of these laws were laws of pure geometry, namely, Euclid’s.
Others were principles from physics, e.g. ‘light travels in straight lines’,
‘a stretched threat will lie in a straight line’, etc. What happened, in a
nutshell, was that this cluster of laws came apart. If there are any paths
that obey the pure geometrical laws (call them ‘E-paths’), they do not obey
the principles from physics, and the paths that do obey the principles from
physics — the geodesics — do not obey the old principles of pure geometry.

The most fundamental division is between objectivistic mathematics and
constructive mathematics. The former includes all number theory, classical
analysis, and Cantor’s higher infinities. The latter can be delineated in
three somewhat different ways. The first position deals only with natural
numbers, using just computable functions and quantifier-free methods of
proof. The second is intuitionism which admits quantifiers but denies the
law of excluded middle. The third is predicative set theory which permits
quantifiers and the general law of excluded middle, but rejects impredicative
definitions on the ground that they violate the vicious-circle principle.
(2) Wang [1974], p. 53.
Elsewhere I have further elaborated the four domains mentioned above,
adding a fifth which is the most restrictive position and might be called
ultrafinitism. It distinguishes even finite numbers into manageable and un-
manageable ones and urges that only the manageable ones are intuitively
evident. Roughly speaking, finitism and intuitism [[sic]] accept only the po-
tentially infinite, predicativism accepts the set of natural numbers as actual
but not the higher infinities, while objectivism accepts the actual infinities.
(3) Wang [1986], p. 51.
On 10 June [[1931]], G[[ödel]] made some observations [[recorded in Carnap’s
diary]] similar to what he told me [[Wang]] in the 1970s. It is arbitrary where
one wishes to put the limit: (1) only concrete propositions, (2) number vari-
ables and mathematical induction, (3) Hilbert (in metamathematics), (4)
Brouwer, (5) classical mathematics. Which formulas and rules (in mathe-
matics) to admit as ‘meaningful’ in the first place, because certain ideas
are associated with them: this is entirely a matter of free resolution. So one
can equally acknowledge all of classical mathematics. There is no plausible
boundary (Unterschied), even though a well-defined limit can be indicated
at different places.
Quotation 81.13. Wang [1974], p. 37.

In the evolution of axiom systems there has emerged a sharp criterion of
formalization in terms, not of meaning and concepts, but of notational
features of terms and formulas. The criterion goes like this: There is a
mechanical procedure to determine whether a given notational pattern is a
symbol occurring in the system, whether a combination of there symbols is
a meaningful formula (a sentence), or an axiom or a proof of the system.
Thus, the formation rules, i.e. rules for specifying sentences, are entirely
explicit in the sense that theoretically a machine can be constructed to pick
out all sentences of the system if we use suitable physical representation
of the basic symbols. The axioms and rules of inference are also entirely
explicit. Every proof in each of these systems, when written out completely,
consists of a finite sequence of lines such that each line is either an axiom
or follows from some previous lines in the sequence by a definite rule of
inference. Therefore, given any proposed proof, presented in conformity
with the formal requirements for proofs in these systems, we can check its
correctness mechanically. Theoretically, for each such formal system, we can
also construct a machine which continues to print all the different proofs of
the system from the simpler ones to the more complex, until the machine
finally breaks down through wear and tear. If we suppose that the machine
will never break down, then every proof of the system can be printed by the
machine. Moreover, since a sentence is a theorem if and only if it is the last
line of a proof, the machine will also, sooner or later, print every theorem
of the system.
Since Frege formulated his view regarding the nature of objectivity of mathe-
matical objects,6 this view went through a variety of formulations which, though
sometimes quite elucidating, are not always satisfactory with regard to a theory of
knowledge.
Since the original suggestion of the label “platonism” in Bernays [1935b], as
quoted in 78.2, the conception of mathematical realism has gone through various
reformulations, with a strong tendency on the human being, as the obvious candidate
for the role of the reflecting subject.
Quotations 81.14. (1) Feferman [1964], p. 1.

From one point of view, often identified as the Platonistic or Cantorian con-
ception, sets have an existence which is independent of human definitions
and constructions.
(2) Feferman [1985], p. 41
[[M]]athematical objects have an existence which is independent of us and
of any means of definition and construction.
(3) von Kutschera [1964], p. 15; my translation.
[[S]]trict platonism, as taken above all by Frege [[. . .]] consists, briefly, in
the assumption that the abstract objects of logic, concepts, functions and
classes, exist independently of human thought and are given to the latter
as objects of scientific research. As opposed to this the “Principia” start
from the conceptualistic assumption that such abstract logical entities are
created by thought. There are then certain principles according to which
entities are created from each other and only entities constructible according
to these principles may be viewed as existing.
Comment. Frege speaks of independence of “Empfinden” (sensation), “Anschauen”
(intuition), “Vorstellen” (imagination), and “Entwerfen innerer Bilder aus den Erin-
nerungen früherer Empfindungen” (construction of mental pictures out of memories
of earlier sensation). Kutschera’s version of independence of human thought seems
to suggest that he does not think human thought capable of reason, because Frege
explicitly expresses that he does not consider things independent of the reason.7
6 See quotations 71.27 in these materials.

7 See Frege [1884], p. 36, quotation 71.27 (3) in these materials.
Allusions to human thought, or human definitions dominate. This, however, is by

no means necessary in order to accommodate for a certain relativity of mathematical
truth as the following quotation indicates.
Both Gödel and Hardy speak of a mathematical reality. For example, Hardy
says, ‘I believe that mathematical reality lies outside us, that our function
is to discover or observe it.’ This sounds obscure. It may be suggested that
the Fermat arithmetic and the non-Fermat arithmetic define two concepts
of number, and that it is meaningless to ask whether Fermat’s conjecture
is true or false, because a mathematical formula has no well-determined
meaning independently of any calculus to which it belongs. If Hardy’s re-
mark is meant to refute this suggestion, then it is right, at least insofar as
natural numbers are concerned.
On the other hand, the question whether the parallel postulate is true
may seem to depend on the particular kind of geometry we wish to use. Here
we seem to have a case where our meaning is not entirely determinate (to
imagine lines meeting at infinity, etc.) so that it may be possible to have
different determinations. In other words, when pressed, we are not able
to say whether the parallel postulate is true or false. Indeed, this example
probably led people to refuse to speak of truth outside a calculus. But there
is no reason to suppose that this is the general rule.
(2) Wang [1974], p. 162.
The question, ‘Do numbers and sets exist,’ seems unclear. We are not
interested in certain specific objects but rather any objects satisfying certain
conditions, so that isomorphic structures are equally acceptable. Instead of
asking whether numbers exist, one may ask whether it is possible that all
condition (true sentences) about numbers can be simultaneously satisfied.
Quotations 81.16. (1) Goodstein [1965], p. 77.
Cantor’s proof of the non-denumerability of the class of real numbers is now
known to establish only relative non-denumerability. For every formalised
theory of real numbers can be shown to have a denumerable model; that
is to say there is an interpretation of the predicates of the system under
which every true statement about classes of integers becomes a true state-
ment about integers. Since the integers are denumerable it follows that
the class of real numbers in the system is denumerable. Cantor’s proof of
non-denumerability is therefore a proof of the incompleteness of the formal
system, a proof that the function which is known to enumerate the class
of real numbers is undefinable in the system. Even without appeal to the
known existence of a denumerable model the relativity of Cantor’s proof is
apparent. For the proof starts by saying ‘let a1 , a2 , . . . be an enumeration of
the real numbers’ and this statement of course has a concealed existential
premiss ‘If the class of real numbers is denumerable in our formalism’. The
conclusion to be drawn from the familiar reductio ad absurdum proof is
therefore that the class is not denumerable in our formalism. To pass from
this to absolute non-denumerability we must know that our system of class

logic is complete, i.e. that all enumerations are definable in it, and precisely
this proves not to be the case.
(2) Van Heijenoort [1967], pp. 290 f.8
For Skolem the discrepancy between an intuitive set-theoretic notion
and its formal counterpart leads to the “relativity” of set-theoretic notions.
Thus, two sets are equivalent if there exists a one-to-one mapping of the
first onto the second; but this mapping is itself a collection of ordered pairs
of elements. If, in a formalized set theory, this collection exists as a set, the
two given sets are equivalent in the theory; if it does not, the sets are not
equivalent in the theory and, when one set is that of the natural numbers as
defined in the theory, the other becomes “nondenumerable”. The existence
of such a “relativity” is sometimes referred to as the Löwenheim-Skolem
paradox. But, of course, it is not a paradox in the sense of an antinomy; it
is a novel and unexpected feature of formal systems.
(2) Hodges [1983], p. 88.
Skolem’s own explanation of why his argument debunks axiomatic set-
theoretic foundations is very obscure. He says in several places that the
conclusion is that the meaning of ‘uncountable’ is relative to the axioms of
set theory. I have no idea what this means.
Comment. I wonder whether Hodges would find it more comprehensible had Skolem
talked about a relativity to human thought instead.
Quotations 81.17. (1) Barker [1969], p. 2.

By realism as a philosophy of mathematics we may understand the doc-
trine that laws of mathematics are fundamentally to be regarded as literal
descriptions of objects of some kind. [[. . .]] [[N]]ominalism [[. . .]] may be de-
fined as the view that there are no abstract, non-spatio-temporal entities
whatsoever. [[. . .]] [[C]]onceptualism [[. . .]] may be defined as the view that
there are such abstract entities but that they are brought into being by our
mental activity. [[. . .]]
[[P]]latonism is a form of realism according to which mathematics has for
its subject-matter a realm of real, non-spatial, non-mental, timeless objects.
It claims that these objects can be grasped by human reason; that we can be
acquainted with them through the ‘eye of reason’, as it were, even though
our five senses provide us no avenue of contact with them.
(2) Barker [1969], pp. 4 f.
From the viewpoint of a realist philosophy of mathematics, the incom-
pletability theorem can be regarded not as calling into question the inde-
pendent reality of mathematical entities such as sets or numbers, but rather
as indicating an essential limitation in the expressive power of symbolism;
the limitation being that no symbolism can fully succeed in characterizing a
system of objects as rich as the natural numbers. The realist may continue
8 This quotation was instigated by Putnam [1980], p. 45.
to believe that sets and numbers are real entities even though he acknowl-
edges that no consistent list of axioms can be formulated that will describe
them completely.
(3) Goodstein [1963], p. 215.
The case against classical philosophical realism in mathematics is over-
whelmingly strong. In geometry the well-known consistency proofs of non-
Euclidean geometry relative to the Euclidean makes it impossible for only
one of the two geometries to be valid and yet both cannot mirror the real
world. The neo-realist argues from this, not that the elements of geometry
are concepts, but that no formal system can adequately express the whole
of geometry, which is something revealed only to the intuition.
Comment. I suggest that this be read in connection with Adler’s claim that some
questions “require sacred theology, founded on revelation, for their adequate solu-
tion” 9 and Putnam [1980], p. 474.10
81c. Logicism, intuitionism, and formalism. The distinction that charac-

terized the early factions in the foundational discussion seems to have lost much of
its appeal since Carnap, Heyting, and von Neumann stated their basic tenets at the
Königsberg conference in 1930. In this section I shall have a cursory look at what
remains of the classical threesome.
Quotations 81.18. (1) Chihara [1972], pp. 246 f.

In Frege’s system, concepts are stratified into different levels: there
are first-level concepts, second-level concepts, third-level concepts, and it
would seem that one could go on to higher and higher level concepts. The
variables of the system are restricted so that no variable that ranges over
objects can also range over concepts; furthermore, no variable that ranges
over concept of one level can also range over concepts of another level.
(One might say that Frege’s system is a simple type theory for concepts.)
However, Frege could see no reason for sorting the objects of the system
into different ranges: object variables ranged over all the objects. Since the
extensions of concepts were all lumped together as objects, Frege’s sporting
of entities into different ranges for variables did not save his system from
Russell’s paradox.
Like Frege before him, Russell too evidently felt that logic could not pre-
vent variables from ranging over all the objects in the universe: restricting
variables to only part of the totality of objects must have seemed artificial
and unmotivated to Russell. This is where the “no-class” theory came in: it
allowed Russell to drop the assumption that classed are objects or things[[.]]
The project of reducing arithmetic to logic is quite independent of conceptual

realism.
9 See quotation 68.28 in these materials.

10 See quotation95.28 (4) in these materials.

Even if Frege was largely mistaken in his views of the nature of mathematics,
by his failure to recognise the relativity of mathematical concepts in his
search for an absolute logic of mathematics, nevertheless his programme to
make explicit the rules of proof marks the opening of the modern era. In
the penultimate stage of this evolution mathematics and logic are entirely
divorced from physics and psychology. Mathematics is conceived of as a
purely formal calculus, a game with initial positions and rules for moving
the pieces, an uninterpreted calculus; and the centre of interest passes from
mathematics to metamathematics conceived of as the comparative study
of formal systems by means once again of an intuitive logic. Like a snake
swallowing its tail, intuitive logic operates on its own codification. But
mathematics itself of course is formalisable and the final stage is perhaps
an infinite hierarchy of formal systems.
(2) Fraenkel et al. [1973], p. 183.
Even if Frege had been entirely successful in reducing arithmetic to
logic — which he was not in view of the emergence of antinomies in his
system —, it is not always sufficiently realized today and has not been suf-
ficiently realized by Frege himself that this would still not have meant that
all of analysis were reducible to logic, in spite of the already accomplished
arithmetization of analysis. This is so because the arithmetization means
reduction to “integers and finite or infinite systems of integers”, in Poincaré’s
terms [[. . .]], hence not only to integers but also to what we would now call
sets of integers. The reduction could therefore be regarded as completed
only if, in addition to arithmetic, also the general theory of sets — or at
least the theory of sets of integers and, probably, of sets of sets of integers
etc. — were “reduced” to logic. This, however, was never done nor even
attempted by Frege. This really final step was attempted only by Russell
(and Whitehead).
Comment. There is a point that puzzles me. On the one hand Frege regarded arith-
metic which he took to include the real numbers (witness his treatment of the real
numbers in the second volume of his Grundgesetze) as analytic. On the other hand
he agreed with Kant as regards the synthetic character of geometry, at least at the
time of the Grundlagen (witness quotation 71.26 (1) in these materials). How is
this compatible with the existence of Descartes’ analytic geometry, i.e., the possi-
bility of interpreting geometry within analysis? Would Frege not have regarded a
translation of geometry into analysis as incompatible with the possibility of analysis
being analytic and geometry synthetic, or is it possible that he did not regard all
of analysis as analytic? Cf. also quotation 86.3 (2) in these materials.
Quotation 81.20. Wang [1974], p. 24.

With regard to the reducibility thesis, we have four alternatives: to deny
that mathematics is reducible to logic, to conclude that mathematics is
analytic, to conclude that logic is synthetic, to examine more closely the
concepts involved such as logic and analytic.
Quotations 81.21. (1) Shoenfield [1962], p. 132.

The Frege-Russell definition of natural number in terms of sets, although
of great interest for the study of axiomatics, is hardly such an explanation
[[of one concept in terms of an other]]; for the notion of an arbitrary set is
surely much more complicated than that of natural number.
Comment. To be in accordance with the distinction of higher order logic (predi-
cation) and set theory (collection) which will be presented in the next paragraph,
this would have to run: “The Frege-Russell definition of natural number in terms
of extensions of concepts”. The notion of set, employed by Frege and Russell was
actually a logical notion, based on the idea of predication, not the mathematical
notion, based on the idea of collecting previously given objects.
(2) Wang [1953], p. 385 ([1962], p. 478).
While Dedekind and Frege tried to provide foundations for number
theory by what are essentially set-theoretic considerations, Kronecker and
Poincaré insisted that natural numbers are the most basic items in our
mathematical thinking. It now seems clear that we understand better the
nature of natural numbers than that of arbitrary sets. We know in a pretty
definite sense the consistency of number theory but not that of set theory.
We can in a definite sense obtain or derive number theory in set theory but
not vice versa.
Comment. I take this as a challenge for any theory of (higher order) logic, and to
meet this challenge I provide a logical foundation of primitive recursive arithmetic
in chapter XXXIV.
(3) Feferman [1986], p. 449.
That there is a fundamental difference between our understanding of
the concept of natural numbers and our understanding of the set concept,
even for sets of natural numbers, seems to me undeniable.
There is the following footnote (2).
The idea that sets are prior to natural numbers and that the latter are to
be ‘defined’ in terms of the former ridiculously turns this on its head.
Comment. Do I detect a similarity to Cantor’s review of Frege?11
Quotations 81.22. (1) Fraenkel et al. [1973], pp. 210 f.

[[A]]nalysis and geometry as developed since the 17th century and especially
since the beginning of the 19th century have — so [[intuitionists]] argue —
utterly disregarded the peculiar traits of infinity and their consequences for
mathematics. The supposedly strict methods introduced into real number
theory and calculus during the 19th century from Cauchy to Weierstrass
and Cantor, far from reaching the desired goal, have rather raised to an
elaborate system the erroneous tendency of treating infinity with methods
created for finite domains.
According to this view the antinomies appearing at the turn of the cen-
tury are but a secondary symptom, evolving at a rather accidental spot[[.]]
(2) Fraenkel et al. [1973], p. 275.
The Brouwerian believed that this conception was wholly wrong from the
beginning. They accused it of misunderstanding the nature of mathematics
and of unjustifiedly transferring to the realm of infinity methods of reason-
ing that are valid only in the realm of the finite. By regaining the right
perspective, mathematics could be constructed on a basis whose intuitive
soundness could not be doubted. The antinomies were only the symptoms of
a disease by which mathematics was infected. Once this disease was cured,
one need worry no longer about the symptoms. All Russellians thought that
our naiveness consisted in taking for granted that every grammatically cor-
rect indicative sentence expresses something which either is or is not the
case, and some — among them Russell himself — believed, in addition,
that through some carelessness a certain type of viciously circular concept
formation had been allowed to enter logico-mathematical thinking. By re-
stricting the language — and proscribing the dangerous types of concept
formation — the known antinomies could be made to disappear. Their faith
in the consistency of the resulting, somewhat mutilated, systems was less
strong than that of the Brouwerians, since certain intuitively not too well
founded devices had to be used in order to restore at least part of the lost
strength and maneuverability. Zermelians, finally, thought that our blunder
consisted in naively assuming that to every condition there must correspond
a certain entity, namely the set of all those objects that satisfy this condi-
tion. By suitable restriction of the axiom of comprehension, in which this
assumption is formulated, they tried to construct systems which were free
of the known antinomies yet strong enough to allow for the reconstruction
of a sufficient part of classical mathematics.
(3) Heyting [1956], p. 102; quoted after Fraenkel et al. [1973], p. 323.
[[N]]o formal system can be proved to represent adequately an intuitionistic
theory. There always remains a residue of ambiguity of the signs, and it can
never be proved with mathematical rigour that the system of axioms really
embraces every valid method of proof.
(4) Feferman [1964], p. 2.
The other extreme is what we shall refer to as the predicative conception.
According to this, only the natural numbers can be regarded as ‘given’ to us
[[. . .]]. In contrast, sets are created by man to act as convenient abstractions
(façons de parler ) from particular conditions or definitions.
The next set of quotations concerns formalism.
Quotations 81.23. (1) Prawitz [1981], p. 252.

Hilbert took a formalist position only towards sentences that contained
reference to infinite totalities, which sentences he called ideal, while he
considered the other sentence, completely meaningful.
(2) Pollard [1990], p. 65.

Crudely put, formalism as a philosophy of mathematics is the view
that the proper and, indeed, only possible objects of mathematical research
are mathematical theories (rather than the items to which those theories
purportedly refer).
Comment. This is very crudely put, indeed, and I’m afraid, it is also what has
become somewhat standard. Contrast this to Gödel in quotation 78.10 in these
materials and Prawitz above.

(1a) In want of a complete trust in the usual understanding of the
classical conceptions, one may seek a constructive interpretation
in terms of which all principles of classical mathematics become
valid, in other words, a constructive reduction of the entire classi-
cal mathematics. This is e.g. Gentzen’s and Takeuti’s position.
(1b) Without any distrust in classical mathematics, one may seek a
constructive foundation of it as the fulfilment of the axiomatic
tradition — this may be Hilbert’s position [[. . .]].
(2) Doubting the classical principles, one may seek a constructive
foundation of mathematics, not by an interpretation of all clas-
sical concepts, but by developing mathematics constructively as
far as possible at the possible expense of sacrificing certain classi-
cal principles. This is the line first taken by Brouwer.
(3) A third line is to maintain that classical mathematics can be made
or is already intelligible in its own non-constructive terms and
therefore completely reliable. This line is represented by Kreisel
[[. . .]], who refers to the cumulative hierarchy when it comes to
interpreting set-theory[[.]]
Quotation 81.25. Girard [1987], p. 38.

Since the time of Hilbert, no new foundational scheme has been proposed.
Certainly people know too much to present a naive ontology of mathematics
(and perhaps not yet enough to present a really challenging explanation of
mathematical activity).
Comment. Note the phrase “explanation of mathematical activity”. This is different

from asking why mathematics is so secure.
§ 82. Self-reference, the vicious circle principle, and the theory of types
82a. Preliminary remarks. There are certain reasons for not being satisfied
with the Ramified Theory of Types as an appropriate formal representation of the
vicious circle principle.
§ 82. SELF-REFERENCE, VICIOUS CIRCLE PRINCIPLE, THEORY OF TYPES 1135
Quotations 82.1. (1) Wang [1959], p. 231 ([1962], p. 641).

Mixing of types is forbidden on the ground that otherwise we would get
meaningless propositions. This does not follow from the vicious-circle prin-
ciple since given a total of objects and a total of sets of these objects, the
vicious-circle [[principle]]12 is not violated if we merge the two totals into
one and use variables ranging over the larger new total.
(2) Copi [1971], p. 103.
The Ramified Theory of Types has also been criticized on the basis of its
connection — or lack of connection — with the vicious circle principle.
As the Kneales remark, ‘. . . the principle is dubious and difficult to apply’
(Kneale, 1962, 666). It has been objected that some of the specifications
ruled out as impredicative on the basis of that principle really do not violate
it. The implication here is that the Ramified Theory of Types goes beyond
the restrictions actually imposed by the vicious circle principle.
82b. Transfinite types. In the present section I can do no more than to give
a few hints; I have never done any work on transfinite types; I still think, however,
a short note regarding this possibility should be included in these materials. The
interested reader may find Fraenkel [1973], pp. 175 ff, helpful for a first orientation;
also: Wang [1954] and [1959].
Quotation 82.2. Carnap [1958], p. 113.13

Can this multiplicity of arithmetics be avoided without giving up the dis-
tinction of types? [[. . .]] One possible way of avoiding this multiplicity con-
sists in adding transfinite levels. Using the transfinite ordinal numbers of
set theory, we designate the lowest level that is higher than all the finite
level ω + 1, etc. There is then put forward a rule of formation specifying
that a predicate of any transfinite level can take argument-expressions of
any lower level whatever[[.]]
Quotation 82.3. Feferman [1964], p. 14.

A return to ‘constructive’ mathematics via a transfinite ramified theory,
among other devices, has been particularly urged during the last few years
by Lorenzen [[[1958]]] and Wang [[[1962]]]. Despite the philosophical attrac-
tiveness of their position, they did not succeed in establishing it definitively.
First of all, the exact nature of their proposals has never been made com-
pletely clear, even taken a broad view of the nature of formalization. In
particular, these authors did not show how the transfinitely conceived se-
quence Mα (α < ω1 ) is to be dealt with in a system of proofs, each of
which consists of finitely many symbols. Secondly, and more importantly,
they failed to meet the old objection that ramified analysis is a parody of
classical analysis, not a significant part of it.
12 My addition, but I can’t see that the sentence makes sense without the word “principle”.
13 This quotation was instigated by Copi [1971], p. 63.
82c. The vicious circle principle.

All the antinomies, whether logical or semantic, share a common feature
that might be roughly and loosely described as self-reference. In all of them
the crucial entity is defined, or characterized, with the help of a totality to
which it belongs itself. There seems to be involved a kind of circularity in
all the argumentations leading up to the antinomies, and it is obvious that
attempts should have been made to see therein the culprit.
I know of only one attractive, albeit unsuccessful, later attempt to handle vicious
circles: Hintikka’s paper [1956]. Hintikka set out from the following considerations:
inclusive and exclusive interpretation of variables.
Quotations 82.5. (1) Hintikka [1956], p. 242.

A definition is said to be impredicative if it presupposes (or involves, or
makes use of) a totality to which the definiendum belongs. What exactly
is meant by this is not quite clear. The most natural interpretation is that
a definition of x is impredicative if it makes use of bound variables one of
the possible values of which is x itself.
Comment. In terms of the definition of a set
W
M :≡ ιy x (x ∈ y ↔ F[x]) ,
what Hintikka’s version of the vicious circle principle
V says is that M must not be
among the possible values of the bound variables of x (x ∈ a ↔ F[x]).
(2) Hintikka [1956], p. 242.
This interpretation is virtually implied by Russell’s dictum that ‘whatever
involves all of a collection must not be one of the collection.’
(3) Fraenkel et al. [1973], p. 199.
Hintikka’s system keeps our logical intuitions intact up to minor corrections;
these minor corrections themselves are not even real deviations from our
logical intuitions but embodiments of the exclusive interpretation which is
the appropriate one in the case of the axiom of comprehension, though our
intuitions are certainly not sharp enough to make this clear without the
“help” of the antinomies.
The presence of a misunderstanding in Hintikka’s approach has been revealed

as early as 1959 by Hao Wang.
Quotation 82.6. Wang [1962], pp. 640 f ([1959], p. 230).

Recently Hintikka 1956 suggests that there are two nonequivalent vi-
cious-circle principles, a weaker one by Russell and a stronger one by
Poincaré. The weaker one is interpreted to mean that bound variables are
permissible in a condition introducing a new object x as long as x itself
does not belong to their range. This amusing interpretation seems to be
based on a strenuous misunderstanding. According to [[Hintikka’s view]] it
is clear that if the constitution of x depends on the range, then not only the
range must not contain x itself, but it must not contain anything definable
in terms of x either. Thus, in Hintikka’s proposed theory, any condition
F y can define a set x of all objects distinct from x and satisfying F y, pro-
vided only all bound variables in F y are restricted by the clause “distinct
from x”. Hintikka himself and others have since shown the inconsistency
of the principle. The following simple argument would both establish the
inconsistency and bring out the point that his principle does violate the
vicious-circle principle. Let ιy be the unit class of y, ῐy be z if ιz is y and
the empty class otherwise. Then the set K defined by
y 6= K → (y ∈ K ≡ (z)(z 6= K → (y = ιz → y ∈/ ῐz)))
leads to the contradiction ιιK ∈ K ≡ ιιK ∈/ K. The source of the trouble is
that, although z cannot be K itself, it can be, for example, ιK which is
definable only in terms of K.
Comment. In view of the numerous formulations of the vicious circle principle in
terms of a variable not taking a certain set as value this is worth emphasizing.14
Quotations 82.7. (1) Hazen [1983], p. 334.

Perhaps the key idea arising out of the debate [[of Russell and Poincaré]] was
the notion that the contradictions were attributable to a kind of circularity
of definition. This, however, was not the ordinary kind of circularity, in
which a word is defined, perhaps through a chain of intermediate definitions,
in terms of itself.
(2) Quine [1969], p. 242.
Definition, in the clearest sense, is what occurs when a new notation is
introduced as short for an old one. No question of legitimization can arise
in connection with definition, so long as a mechanical procedure is provided
for expanding the new notation in all cases uniquely into old notation. Now
what Poincaré criticized is not the definition of some special symbol as short
for ‘{x : x ∈/ x}’, but rather the very assumption of the existence of such a
class; the assumption of the existence of a class y fulfilling ‘(x)(x ∈ y . ≡
. x ∈/ x)’. We shall do better to speak not of impredicative definition but of
impredicative specification of classes, and, what is the crux of the matter,
impredicative assumptions of class existence.
And what of the vicious circle? A circular argument seduces its victim
into granting a thesis, unawares, as a premiss to its demonstration. A cir-
cular definition smuggles the definiendum into the definiens, in such a wise
as to prevent expansion into primitive notation. But impredicative specifi-
cation of classes is neither of these things. It is hardly a procedure to look
askance at, except as one is pressed by the paradoxes to look askance at
something or other.
14 Section 97b of my [1992a] (version 0.5 of the present study) contained a formal version of
an antinomy in Hintikka’s system.

Comment. Contrast quotation 74.11 (1) where Poincaré explicitly says that the
object had to be “christened” to work wonders, and that “because by giving it a
name we have asserted implicitly that the object did exist”.
(3) Wang [1952], p. 56.
The most important examples of impredicative definitions are probably
those involved in the following situations. (1) In defining the least upper
bound of a bounded class of real numbers (say, each as a class of rational
numbers) as the real number which is the union of all the real numbers
of the class, we refer to the totality of all real numbers. (2) In proving
Cantor’s theorem, we assume there is a one-one correspondence between
the members of a class and all its subclasses and consider the subclass
consisting of all the members of the given class which do not belong to
their corresponding subclasses; in this definition of the special subclass we
refer to the totality of all the subclasses of the given class.
Quotations 82.8. (1) Wang [1959], p. 229 ([1962], pp. 639 f).
Instead of the phrase “definable only in terms of”, Russell also speaks
of “involving” and “presupposing”. It is pointed out in Gödel 1944 that
these different versions are not equivalent for one who thinks of classes
as pluralities existing independently of our knowledge and definitions. He
would say that there are acceptable definitions which, though violating [[the
version formulated in terms of definability]], do not violate the alternative
formulations. It seems, however, clear that the original intention is to deal
with definition (intensions) rather than entities (like Cantor) or proofs
(like Brouwer). That is why Russell regarded the different formulations
as equivalent. [[. . .]]
On the other hand, there is also a sense in which one is concerned with
objects rather than definitions. This is seen from the qualification of “terms
of” by “only”. In other words, the principle is directed to the introduction
of new objects. Once a range has been satisfactorily introduced, there is
no objection against using an impredicative characterization to identify or
select an object in the domain. For example, if we assume the totality of all
positive integers given, there is no objection against using the least number
operator to specify a principle of selection from the given totality; or, if we
assume a class of real numbers given, there is no objection against speaking
of a maximum of the class.
Impredicative characterizations are objected to not just as such but
only as a means for initially introducing an object. One does not object to
using in a proposed characterization bound variables whose predetermined
range contains any object which may satisfy the characterization, but only
to conditions which propose to introduce new objects and yet make use of
bound variables which contain the objects to be introduced. In the latter
case, the range of the variables are not yet determined at the time and may
be affected by the definition itself or others, yet to be introduced, which
depend on it. There is then a circle involved in the process of determining
the range of the variable: the range of the variable depends on the consti-
tution of the object (usually a set) to be introduced by the condition, but
the structure of the object in turn depends on the range of the variables.
Principle 5.1 rejects certain uses of bound variables, viz., those whose range
contains members which are definable only by using these bound variables.
This is equivalent to saying that bound variables can only be used in the
introduction of some new object when their range is already determined at
the time. That means, the range must not change with the introduction of
new objects. In other words, new objects are only to be introduced stage
by stage without disturbing the arrangement of things already introduced
or depending for determinedness on objects yet to be introduced at a later
stage.
(2) Chihara [1972], pp. 285 f.
From the point of view of those who think that there really are sets which ex-
ist independently of human thoughts and practices, the vicious-circle prin-
ciple is false. (Of course, if one thinks that there are such things as sets, as
distinct from propositional functions, one would reject the “no-class” theory
outright.)
The following quotations reflect attempts to (re)formulate the vicious circle

principle.
Quotations 82.9. (1) Feferman [1964], p. 2.

In order, for example, to predicatively introduce a set S of natural numbers
x we must have before us a condition F(x), in terms of which we define S
by
V
(1.1) x [x ∈ S ↔ F(x)] .
However, before we can assert the existence of such S, it should already have
been realized that the defining condition F(x) has a well-determined meaning
which is independent of whether or not there exists a set S satisfying (1.1)
(but which can depend on what sets have been previously realized to exist).
In particular, to determ[[in]]ing what members S has, we should not be led
via F(x) into a vicious-circle which would return us to the very question we
started with. Conditions F(x) which do so are said to be impredicative; it
should be expected that most conditions F(x) involving quantifiers ranging
over ‘arbitrary’ sets are of this nature.
(2) Quine [1969], p. 241.
Poincaré tried to account for Russell’s paradox as the effect rather of
a subtle fallacy than of a collapse of irreducible principles. He attributed it
to what he called a vicious circle. The defining characteristic of the para-
doxical class y is ‘(x)(x ∈ y . ≡ . x ∈/ x)’, and the paradox comes, as we know,
of letting the quantified variable ‘x’ here take y itself as a value. It is, he
suggested, illegitimate to include a class y, or any classes whose specifica-
tion might presuppose y, in the range of a quantification that is used in
specifying y itself. He called the suspect procedure impredicative. We must

not presuppose y in defining y.
(3) Church [1976], p. 747.
To avoid impredicativity the essential restriction is that quantification
over any domain (type) must not be allowed to add new members to the
domain, as it is held that adding new members changes the meaning of
quantification over the domain in such a way that a vicious circle results.
(4) Hazen [1983], p. 335.
In order for the definiens to have a definite meaning, the quantified variables
occurring in it must have definite ranges. In the sort of definition under
consideration, therefore, the propriety of the definition will depend on the
existence, and presence in the range of the bound variables of the definiens,
of the very set to be defined, and the set, supposing sets to be ontologically
dependent on their definitions, will only exist if the definiens has a definite
meaning.
Comment. This seems to come very close to Feferman [1964].
I continue with quotations regarding the nature of this presupposition.
Quotations 82.10. (1) Kreisel [1960], p. 371.

[[L]]’objet défini, c’est-à-dire l’ensemble défini, est utilisé dans la propriété
qui le définit. Plus exactement, l’objet défini constitue une valeur parti-
culière d’une variable qui intervient dans l’expression de la propriété. Si
l’ensemble A est défini par
(X) [X ∈ A ≡ P (X)] ,
une valeur de la variable X est l’ensemble A lui-même.
(2) Chihara [1972], p. 249.
[[A]]lthough one does not find complete unanimity as to how “impredicative
specification” is to be defined, the general idea can be given as follows: a
specification of a set A by means of the schema
(x) (xǫA ↔ φx)
is impredicative if the set A, were it to exist, or any set presupposing the
existence of A, falls within the range of a bound variable in the specification.
(3) Hazen [1983], p. 334.
[[A]] typical example would be a definition of a set of, say, natural numbers
in which the definiens contained quantified variables ranging over all sets of
natural numbers — a definition of a set or property, to generalize, in which
the set defined is itself supposed to be among the values of the quantified
variables appearing in the definiens.
Quotation 82.11. Quine [1969], p. 242.

[[W]]e are not to view classes literally as created through being specified —
hence as dated one by one, and as increasing in number with the passage of
time. Poincaré proposed no temporal implementation of class theory. The
doctrine of classes is rather that they are there from the start. This being
so, there is no evident fallacy in impredicative specification. It is reasonable
to single out a desired class by citing any trait of it, even though we chance
thereby to quantify over it along with everything else in the universe. Im-
predicative specification is not visibly more vicious than singling out an
individual as the most typical Yale man on the basis of averages of Yale
scores including his own.
So the ban urged by Russell and by Poincaré is not to be hailed as the
exposure of some hidden but (once exposed) palpable fallacy that underlay
the paradoxes. Rather it is one of various proposals for so restricting the
law of comprehension:
(Ey)(x)(x ∈ y . ≡ . Fx)
as to thin the universe of classes down to the point of consistency.
Still the proposal is less arbitrary than some alternatives, in that it real-
izes a constructional metaphor: it limits classes to what could be generated
over an infinite period from unspecified beginnings by using, for each class, a
membership condition mentioning only preexisting classes. Metaphor aside,
the distinctive feature of such a set theory is that its universe admits of a
(transfinite) ordering such that every class that is specified by a member-
ship condition at all is specified by one in which the values of all variables
are limited to things earlier in the ordering.
Quotations 82.12. (1) Nelson [1986], p. 1.

The reason for mistrusting the induction principle is that it involves an
impredicative concept of number. It is not correct to argue that induction
only involves the numbers from 0 to n; the property of n being established
may be a formula with bound variables that are thought of a ranging over all
numbers. That is, the induction principle assumes that the natural number
system is given.
Comment. This impredicativity comes out beautifully in the second order theory
CQ2 A{:1} of section 47c.
(2) Nelson [1986], p. 2.
It appears to be universally taken for granted by mathematicians, what-
ever their views on foundational questions may be, that the impredicativity
inherent in the induction principle is harmless—that there is a concept of
number given in advance of all mathematical constructions, that discourse
within the domain of numbers is meaningful. But numbers are symbolic con-
structions; a construction does not exist until it is made; when something
new is made, it is something new and not a selection from a preexisting
collection. There is no map of the world because the world is coming into
being.
(3) Nelson [1986], pp. 79 f.

There is a barrier separating the predicative from the impredicative, a bar-
rier as absolute as that between the genetic and the formal, and more
sharply delineated than the debatable demarcation between the finite and
the infinite.
As syntactical methods for establishing certain inductions have failed,
we may be tempted to turn to semantics and appeal to the set ω of all
natural numbers. Of course it is a theorem of axiomatic set theory that
formalized Peano Arithmetic is consistent, but this is not what people have
in mind who argue that Peano Arithmetic is consistent because its axioms
are true statements about ω. What is at issue here is not the familiar
construct of formal mathematics, but a belief in the existence of ω prior
to all mathematical constructions. What is the origin of this belief? The
famous saying by Kronecker that God created the numbers, all else is the
work of Man, presumably was not meant to be taken seriously. Nowhere
in the book of Genesis do we find the passage: And God said, let there
be numbers, and there were numbers; odd and even created he them, and
he said unto them, be fruitful and multiply; and he commanded them to
keep the laws of induction. No, the belief in ω stems from the speculations
of Greek philosophy on the existence of ideal entities or the speculations
of German philosophy on a priori categories of thought. An appeal to ω to
justify the induction principle is no more secure than are these philosophical
systems, and yet it is hard to relinquish.
Comment. I wouldn’t see much hope for the laws of induction, if it were just God’s
command to the numbers to keep them. He wasn’t particularly lucky with the laws
he commanded us humans to keep, was he? (Apart from the recommendation to be
fruitful and to multiply; but he employed some rather mean trick to achieve this
particular goal.)
(4) Nelson [1986], pp. 81.
In addition to philosophical objections to impredicative methods, one
may have mathematical misgivings about their use. Perhaps the place to
look for possible trouble is not in the upper branches of set theory, but
rather at the very roots in arithmetic where impredicativity first appears.
Comment. Contrast this with Quine’s attitude as presented in quotation 96.20 (2)
in these materials.
(5) Nelson [1986], p. 174.
Mathematicians have always operated on the unchallenged assumption
that it is possible in principle to express 2b as a numeral by performing the
recursively indicated computations. To say that it is possible in principle
is to say that the recursion will terminate in a certain number of steps—
and this use of the word “number” can only refer to the primitive notion;
the steps are things that are counted by a sequence of tally marks. In what
number of steps will the recursion terminate? Why, in somewhat more than
2b steps. The circularity in the argument is glaringly obvious.
§ 83. SET THEORY 1143
§ 83. Set theory
It will be clear that my interest in set theory covers only a fraction of what is dealt
with in the philosophy of mathematics regarding set theory.
Features 83.1. I shall focus on the following aspects:

(IC1) separation of logic and mathematics: classification vs. collection;
(IC2) flight from logic;
(IC3) the doctrine of limitation of size;
(IC4) the iterative concept of set.
Comment. Re IC1. This goes against Poincaré’s view of sets as based on classifi-
cation (PC5). It seems important to understand this attempt to keep two things
apart: mathematical intuition as regard sets and logical intuition as regards con-
cepts. As regards concepts, intuition seems bankrupt indeed; this, of course, affects
the reducibility of the notion of set to that of concepts, i.e., the logical foundation of
arithmetic (set theory, mathematics). It is at this point that my dialectical approach
comes in.
83a. Aspects of the reception of the antinomies today. One of the roots
of the separation of logic and mathematics can be found in the classification of the
antinomies.

To be sure, Russell’s antinomy was not the first one to appear in a basic
philosophical discipline. From Zenon of Elea up to Kant and the dialectic
philosophy of the 19th century, epistemological contradictions awakened
quite a few thinkers from their dogmatic slumber and induced them to re-
fine their theories in order to meet these threats. But never before had an
antinomy arisen at such an elementary level, involving so strongly the most
fundamental notions of the two most “exact” sciences, logic and mathemat-
ics.
Euphoric assessments of Ramsey’s distinction of paradoxes, like that of Carnap

quoted in 86.12, seem to have been largely abandoned today. The next couple of
quotations is meant to give an impression of this tendency.
Quotations 83.3. (1) Rogers [1971], p. 144.

In 1926 F. P. Ramsey (1903—1930) proposed a distinction of the paradoxes
known at that time into two types: logical, or mathematical paradoxes, and
epistemological, or semantical paradoxes. Ramsey argued that paradoxes of
the latter sort, by virtue of making reference to language (meaning, truth,
definability) cannot be stated within mathematics, in which there is no
reference to such matters, and thus that there is no need to consider them
at all in attempting to devise ways of avoiding paradox within mathematics.
As we shall see, this reasoning is not as conclusive as Ramsey apparently
thought, but his classification is helpful and has been widely used.
(2) Goddard and Johnston [1983], pp. 499 f.

The basis of Ramsey’s classification is between those paradoxes which can
be set up entirely within a formal system (in fact set theory) and those which
require additional (semantic) concepts. Or we may say, perhaps, that the
logical paradoxes for Ramsey are those such that the domain over which the
quantifiers range contains only formal objects (numbers, classes, etc.), while
the domain in the case of the semantic paradoxes contains nonformal objects
such as people, catalogues, statements, etc. But this seems to be the least
important feature of the paradoxes, even though Ramsey’s classification had
the advantage of simplifying the type-theoretical solution of the “logical”
paradoxes. What is important about the paradoxes is their common feature
not their differences. This common feature is that the presuppositions in
each of them have the same formal structure. And it is the formal structure
of the presupposition, not its interpretation over a domain, which lies at
the heart of the paradoxes. [[. . .]]
Ramsey’s distinction is therefore misleading. And it would be equally
misleading to argue in the converse direction from the fact that some para-
doxes can be removed by one kind of known solution while others cannot
(e.g., Russell’s can be removed by simple type theory while Grelling’s re-
quires order theory), to the conclusion that the paradoxes should be classi-
fied differently when they yield to different solutions, for this, too, ignores
the similarity of formal structure in the different paradoxes.
(3) Goddard and Johnston [1983], p. 501.
Why is [[the wff α]] (∃x)(y)((y, x) ≡ ∼ (y, y)) contradictory?
Since f (y, x) ≡ ∼ f (y, y) is equivalent to ∼ (f (y, x) ≡ f (y, y)), the
formula which is the scope of the innermost quantifier is the negation of
a condition for the equality of x and y. Thus, from the identity schema
(y = x) ⊃ (A(y, x) ≡ A(y, y)), we have (y = x) ⊃ (f (y, x) ≡ f (y, y)); hence
∼ (f (y, x) ≡ f (y, y)) ⊃ (y 6= x), for all x and y: i.e., (x)(y) ((f (y, x) ≡
∼ f (y, y)) ⊃ (y 6= x)) is a thesis. At the same time, if we take an instantia-
tion case of α, say (y) (f (y, I) ≡ ∼ f (y, y)), there is nothing which prevents
us from taking I to be a value for y, indeed we are required to include I
in the values over which the unrestricted universal quantifier ranges. But
that is to take y = x for that particular instantiation. Thus, we have a
condition which entails y 6= x, for all x and y, yet the quantification over
that condition requires us to include the case for which y = x for some x
and y.
Comment. Compare remark 38.48 (regarding the LK-deduction of the formula α
above in 38.47 (3)).
(4) Priest [1995], p. 156.
Ramsey’s distinction depends on the relatively superficial fact of what vo-
cabulary is used in the paradoxes, and, in particular, whether this belongs
to mathematics properly so called. But worse, this is a notoriously shift-
ing boundary. Ramsey was writing before the heyday of metamathematics.
Had he been writing ten years later, it would have been clear that a num-
ber of items of vocabulary occurring in paradoxes of Group B do belong to
mathematics.
The next set of quotations concerns the way antinomies are avoided in set
theory.
Quotations 83.4. (1) Rogers [1971], p. 153.

In order to block [[Russell’s]] contradiction, what Zermelo did was to assert
the existence not of the set of all sets which satisfy the condition A, but
only of the set of all sets which are elements of some given set and satisfy
the condition A. If we assert the existence of the set of all sets which satisfy
A, for every choice of A, then we are postulating the existence of sets which
are ‘too big’, so to speak. The reasoning lying behind the axiom schema
of separation, in which the phrase ‘z ∈ x’ occurs, is that we must in some
way restrict the size of the set we are to postulate in connection with the
condition A. The restriction which this axiom schema embodies is that the
set postulated be a subset of some set x already known to exist.
(2) Fraenkel et al. [1973], pp. 135 f.
The way we introduced and motivated [[von Neumann-Bernays set theory]]
is not the way this was done historically for set theory with classes. The first
axiom system for set theory with classes was put forth by von Neumann in
1925. The main technical difference between his system and [[von Neumann-
Bernays set theory]] is that he used the notion of function as the basic notion
rather than those of set and class. [[. . .]]
What von Neumann regards as the first main feature of his theory is
the following. In ZF the guiding principle in writing down the axioms is the
doctrine of limitation of size, i.e. we do not admit very comprehensive sets
in order to avoid the antinomies. Von Neumann regards as the main idea of
his set theory the discovery that the antinomies do not arise from the mere
existence of very comprehensive sets, but from their elementhood, i.e. from
their being able to be members of other sets.
Comment. Observe here: doctrine of limitation of size. How is this related to the
iterative conception of set?
(3) Fraenkel [1973], p. 142.
[[von Neumann’s]] addition of proper classes to the universe of set theory
results from his discovery that it is not the existence of certain classes that
leads to the antinomies, but rather the assumption of their elementhood,
i.e., their being members of other classes; therefore he introduces these
classes as proper classes which are not members of classes. This is not a
completely satisfactory solution of the problem of the existence of collec-
tions as objects, since now, even though proper classes are real objects, col-
lections of proper classes do not exist. The existence of real mathematical
objects which cannot be members of even finite classes is a rather pecu-
liar matter, even though in actual mathematical treatment such classes are
rarely needed, and can, in case of need, be represented by some classes of

sets. One cannot just blame the antinomies for this peculiar situation; we
shall see that if one proceeds carefully enough then the assumption that
proper classes can be members of classes or of other objects can be seen
not to cause any contradiction.
(4) Crossley et al. [1972], p. 60.
Why is the Russell collection not a set? It turns out that it is not a set
because there are too many things that want to be in it. R is just too big
a collection of things to be a set.
Comment. Something strikes me as terribly unsound in this quotation. It may be the
use of the term collection which seems incompatible with the spirit of the iterative
conception of set or it may just be the way of talking about things that “want to
be in” a set.
Once mathematicians recovered from the first shock, it seems, they started
reflecting on the situation and came to the conclusion that what Russell had done
was neither intended by Cantor nor was it mathematical practice.
It is often claimed that it was the concept of set which led into antinomies. This
is not really correct. What actually led into antinomies was the logical foundation
of the concept of number which employed extensions of concepts. This is quite a
different matter. Logic deals with concepts, mathematics with sets. The interpreta-
tion of sets as extensions of concepts which lay at the heart of logicism was taken in
by the antinomies. This, however, does not raise a genuine mathematical problem.
As Gödel stated with respect to the antinomies:

For someone who considers mathematical objects to exist independently
of our constructions and of our having an intuition of them individually,
and who requires only that the general mathematical concepts must be
sufficiently clear for us to be able to recognize their soundness and the
truth of the axioms concerning them, there exists, I believe, a satisfactory
foundation of Cantor’s set theory in its whole original extent and meaning,
namely, axiomatics of set theory interpreted in the way sketched below.
It might seem at first that the set-theoretical paradoxes would doom to
failure such an undertaking, but closer examination shows that they cause
no trouble at all. They are a very serious problem, not for mathematics,
however, but rather for logic and epistemology.15
(2) Myhill [1984], p. 129.
Gödel said to me more than once, “There never were any set-theoretic para-
doxes, but the property-theoretic paradoxes are still unresolved”[[.]]
So we are being told to be careful and try to keep apart mathematical intui-
tion and logical intuition. The first is concerned with collections, the second with
15 In the original [1947], p. 518: “They are a very serious problem, but not for Cantor’s set
theory”.
descriptions and/or concepts. It is the logical intuition which found itself trapped.
Therefore, we should try to regain and make precise the genuine mathematical in-
tuition when undertaking a foundation of set theory. This point of view is common
to modern set theoretical investigations; the following couple of quotations is meant
to give an impression.
Quotations 83.6. (1) Scott [1974], p. 207.

It must be understood from the start that Russell’s paradox is not to
be regarded as a disaster. It and the related paradoxes show that the naive
notion of all-inclusive collections is untenable. That is an interesting result,
no doubt about it. But note that our original intuition of set is based on
the idea of having collections of already fixed objects. The suggestion of
considering all-inclusive collections only came in later by way of formal
simplification of language. The suggestion proved to be unfortunate, and
so we must return to the primary intuitions.
Comment. Note above all that here is someone who speaks of “our original intu-
ition”. “Needless to add” that this “our” refers to some “rest of us”.
(2) Myhill [1984], p. 129.
The current heresy bespeaks at once a (hopefully temporary) victory of
pragmatism and generally sloppy thinking over philosophical analysis, and
at the same time a failure of literary and exegetic scholarship in the tradi-
tional sense. It holds that Cantor had a “naïve” conception of set according
to which every well-formed formula with one free variable determines a set.
Comment. See the following quotation.
Quotations 83.7. (1) Stegmüller [1979], p. 12; my translation.

[[T]]he classical mathematicians were mistaken when they assumed the un-
restricted axiom of comprehension with the help of which set theoretic
antinomies were provable. And what happened then, and still happens, is
in principle always the same: One tries this and that, until one obtains a
sufficiently rich system of which one hopes that it will be shown to be free
of contradictions.
Comment. This could well be what Myhill has in mind. I shall refer to it as dogmatic
skepticism.
(2) Fraenkel et al. [1973], p. 15.
The occurrence of the antinomies showed that the naive concept of set as
appearing in Cantor’s “definition” of set, and in the most general conclusions
derivable from it, cannot form a satisfactory basis for set theory, much less
for mathematics as a whole. [[. . .]] The discovery of the antinomies called
therefore for a re-examination of the concept of set or, rather, of the way
this concept was handled. This re-examination resulted in a great divergence
in the diagnosis of the ills of Cantor’s set theory, and, naturally, different
diagnoses led to the recommendation of different cures.
(3) Tait [1981], p. 540.

The paradoxes of set theory arose not from the introduction of such trans-
finite objects as functions of some given type A → B, but from Frege’s as-
sumption that the universe—or at least the mathematical universe—could
itself be treated as a type V of mathematical object (so that one could then
construct objects for example of type V →V which are consequently also
of type V ). This was a very bold conjecture, and it is not at all surprising
that it was almost immediately refuted. But, contrary to an often expressed
view, it is not an assumption that was implicit in mathematical practise,
even in the work of Cantor[[.]]
(4) Scott [1974], p. 208.
The truth is that there is only one satisfactory way of avoiding the para-
doxes: namely, the use of some form of the theory of types. That was at
the basis of both Russell’s and Zermelo’s intuitions. Indeed the best way
to regard Zermelo’s theory is as a simplification and extension of Russell’s.
(We mean Russell’s simple theory of types, of course.)
(5) Scott [1980], p. 257.
[[I ]]t is impossible to eliminate from logic and mathematics ALL
type distinctions. [[. . .]] [[N]]o magic so far has ever made a set A and its
powerset PA equal. [[. . .]][[T]]he kind of type difference that Russell recog-
nized will always be present somewhere in a theory of logical objects.
Comment. No magic was ever needed to prove that there is a bijection between the
universal set and its power set — not even tertium non datur.16 Where a certain
magic is needed — and that in combination with tertium non datur — is when it
comes to proving that a set and its powerset are not equal.17 This magic is commonly
called “Cantor’s diagonal method” and it employs type confusion, self-application, or
whatever you want to call the kind of thing that is excluded by Russell’s vicious circle
principle (which yields a ramified type distinction) in conjunction with classical
logic. Not much imagination is needed to declare that the universal set is not a set,
just a pragmatic attitude. Where more than magic is needed is to get an academic
mind to reconsider a preconceived idea. In view of the last two quotations it might
be worthwhile pointing out with regard to dialectical logic, firstly that it has no type
distinctions à la Russell, secondly that it is not in general provable that there is no
one-to-one correspondence between a set and its powerset, and thirdly that it is in
fact provable that the powerset of the universal set equals the universal set itself.
What dialectical logic has instead is some kind of a modal operator. Higher order
magic is probably required to make someone like Scott see that this is an alternative
way of drawing conclusions from Cantor’s diagonal method. It is an approach that
will give rise to something like a distinction of types, viz., in form of categories, but
it would be far fetched to call it a “kind of type difference that Russell recognized”.
Main point: it is a distinction that will be derived within the theory, not imposed
from the start.
16 See 128.11iii in section 128a of the groundworks.
17 See section 53c in the tools.
(6) Scott [1980], p. 232.

[[T]]he project is doomed as soon as x 7→ 1 + x is allowed; this function has
no fixed point, but in a full theory of combinators all functions have fixed
points.
Comment. In my theory of dialectical logic in the groundworks, x 7→ 1 + x does
have a fixed point. The question is always what you take as the realm: 2 − 3, for
instance, has no solution in the natural numbers; x 7→ 1 + x has no fixed point in the
natural numbers either, but it has a fixed point in unrestricted abstraction. But, I
suppose, this isn’t what Scott would consider an acceptable realm — which I would
just regard as another a form of begging the question, or call it repugnance.18
(7) Cohen [1974], p. 10.
For me the essential point is the existence of infinite totalities. The at-
titude towards infinite sets has traditionally been the great dividing line
between mathematicians. The famous antinomies in logic never played a
role in mathematics simply because they were totally alien to the type of
reasoning one normally uses. One never considers all possible objects in the
universe nor the length of descriptions, etc. These difficulties belong more
properly to the development of the notions of a formal system. Similarly,
Zeno’s paradoxes do not really strike us as presenting serious difficulties in
the way they were intended. In general, I would say that many of these prob-
lems were historically associated with the transition period from classical
philosophy to mathematics as we know it today.

The reactions of Frege and Cantor to the paradoxes were sharply different
and can be described as the bankruptcy theory versus the misunderstand-
ing theory. The difference can undoubtedly be attributed to their different
conceptions of set (the logical versus the mathematical notion).
(2) Wang [1974], p. 48.
‘The superstitious fear and awe of mathematicians in face of the contra-
diction.’ But Frege was a logician and Cantor was a mathematician. Cantor
was not a bit worried about the contradictions. [[. . .]] Admittedly Cantor’s
well-known definition of the term ‘set’ is difficult, yet it cannot be denied
that the definition does exclude, through the mildly ‘genetic’ element, the
familiar derivation of contradictions.
Comment. The opening quotation is from Wittgenstein [1956], I.20, p. 53. As re-
gards the classification of Frege as a logician: according to Flew [1979/83], p. 126,
Frege was professor of mathematics in Jena. Dummett in the Concise Encyclopedia
of Western Philosophy and Philosophers lists him as ‘German philosopher’. (To the
best of my — second hand — knowledge he never got beyond ‘honorary professor’,
albeit ‘ordinary’, of mathematics in Jena which means that he wasn’t all that suc-
cessful in the academic pecking order; cf. also Kreiser in quotation 92.29 in these
materials.)
(3) Wang [1974], p. 191.

In the extreme cases, the proponents of the misunderstanding theory pro-
pose to uncover flaws in seemingly correct arguments, while the bankruptcy
theorists find our basic intuition proven to be contradictory and seek to re-
construct or salvage what they can, by ad hoc devices if necessary. The
basic intuitive concept is often called naive set theory and identified with
the belief in an absolute comprehension principle according to which any
property defines a set. That some notion like this was actually seriously
developed by Frege was a historical accident often advanced as evidence
that we do have such a contradictory intuition. The principle, if correct, in
fact appears to be the sort of thing which belongs to the domain of logic.
[[. . .]] Viewed in the light of Cantor’s development of set theory, however, it
is not at all clear that we do have such a contradictory intuition.
Comment. I would find it interesting to know what counts as an historical accident.
I suggest this should be seen in connection with Frege’s mere linguistic necessity 19
and Ramsey’s “merely our symbol” 20 .
(3) Wang [1974], pp. 45 f.
It is remarkable that the Russell-Zermelo contradiction led Frege to doubt
whether arithmetic can be given a reliable foundation at all. Actually the
contradictions in no way make it necessary to abandon all definitions of
number in terms of sets. What is affected is only the project of formalizing
a general theory of sets, and that only according to the Fregean conception
which seems to make sets a part of logic. We were at first struck by the
fact that natural numbers, real numbers, and many other things can all be
gotten out of sets. Then we found that contradictions can be gotten out
of sets too. Many people conclude, therefore, that we are now to design a
calculus which includes as much of the other things as possible but not the
contradictions.
If we do not think in terms of a system, why can we not treat a proof of
contradiction as just another piece of mathematics which could be judged
interesting or uninteresting more or less in the same manner as other math-
ematical proofs? True, the conclusion, being a contradiction, cannot be sig-
nificant in the same way as an ordinary theorem is. The proof establishes
either more or less than usual. It either shows “the unreliability of our ba-
sic logical intuition”, or reveals some confusion on the part of the owner
of the proof. Dividing both sides of the correct equation 3 · 0 = 2 · 0 or
3(2 − 2) = 2(2 − 2) by 0, we can easily get the contradiction: 3 = 2. Such
a discovery does not excite us because it is well established that the re-
striction c 6= 0 is essential in inferring a = b from ac = bc. Why can one
not discard the contradictions in set theory as easily? The reason, on the
surface, is the lack of any comparably simple and natural restriction which
19 See Frege [1903], p. 373; quotation 71.17 in these materials.

20 See Ramsey [1925], p. 41; quotation 77.5 (2) in these materials.
would do the job. This is, it is sometimes said, an indication of the more
basic fact that our concept of set is not sufficiently clear.
I close this section with a quotation of what the Encyclopedia of Philosophy
says on resolving the paradoxes.
Quotation 83.9. van Heijenoort [1967a], p. 49
Since the last years of the nineteenth century the paradoxes have exerted
a profound influence on the development of logic. For a while their effect
on logic and the foundations of mathematics seemed devastating. After the
advent of the theory of types and of axiomatic set theory they were, so
to speak, domesticated, but they remained a constant source of concern
to logicians. At first they were often considered to be due to the breach
of some specific rule of logic, which explains Russell’s invocation of the
vicious-circle principle. Such injunctions to avoid breaches of logical rules
undoubtedly guided Russell, Zermelo, and others in their construction of
systems in which the known paradoxes could not be reproduced. However,
no rule could be formulated that would by itself eliminate the paradoxes,
and only the paradoxes. Even the notion of circularity could not be given
a precise form that would be a necessary and sufficient condition for the
existence of a paradox. It is impossible to characterize a circular argument
in such a way that every circular argument leads to a paradox and every
paradox is the result of a circular argument.
Comment. Note the formulation “necessary and sufficient condition for the existence
of a paradox”! Though it does not seem to be common usage, it seems to express
what mathematicians feel: the antinomies are a nuisance which should be eliminable
without loss.
83b. The iterative conception of set. Once the idea of collection is clearly
separated from that of classification (Gödel in quotation 78.14 in the last chapter)
the stages of collection have to be specified. This seems to be largely what the
iterative conception of set concentrates on.
Remark 83.10. A word of warning: I do not want to discuss whether there really
was such a clear intuition about sets as collections of previously given things among
the founders of the concept of set. My point is simply to pick up a strong point
in current foundational discussions which clarifies a fundamental difference between
mathematical and logical intuition, no matter if this was clear to the mathematicians
at the end of the last century or even in the early times of axiomatic set theoretical
foundations.
A set is a collection of previously given objects; the set is determined when
it is determined for every given object x whether or not x belongs to it.
The objects which belong to the set are its members, and the set is a single
object formed by collecting the members together. The members may be
objects of any sort: plants, animals, photons, numbers, function, sets, etc.
According to the iterative concept, a set is something obtainable from

some basic objects (such as the empty set, or the integers, or individuals,
or some other well-defined urelements) by iterated applications of the op-
eration ‘set of’ which permits the collecting together of any multitude of
‘given’ objects (in particular, sets) or any part thereof into a set.
(2) Wang [1974], p. 187.
This iterative concept of set is of course quite different from the dichotomy
concept which regards each set as obtained by dividing the totality of all
things into two categories (viz. those which have the property and those
which do not). Following Gödel, one may speak of the two concepts as the
mathematical versus the logical.
(3) Shoenfield [1977], pp. 322 f.
As a first approximation, a set is a collection of objects. Thus a set is formed
by selecting certain objects, called the members of the set; and the set is
completely determined by its members.
The objects which are members of sets may be of any kind. In partic-
ular, we want to consider a set as an object and thus to allow it to be a
member of another set. [[. . .]]
[[. . .]]
We have now reached the following point: a set x is formed by choosing
the sets which are to be members of x. Are there any restrictions on the
sets which we may pick? There are, as the paradoxes of set theory show.
Let us recall the Russell paradox. Let r be the set whose members are
all sets x such that x is not a member of x. Then for every set x,
x∈r↔x∈
/x (1)
Substituting r for x, we get a contradiction.
The explanation is not really difficult. When we are forming a set z by
choosing its members, we do not yet have the object z, and hence cannot
use it as a member of z. The same reasoning shows that certain other sets
cannot be members of z. Hence y is not available as an object when z is
formed, and therefore cannot be a member of z.
Putting the matter in a positive way, a set z can have as members only
those sets which are formed before z.
Comment. There is a footnote regarding the notion before: “We should interpret
before here in a logical rather than temporal sense. It is similar to what we mean
when we say that one theorem must be proved before another.”
(4) Wang [1974], p. 188.
The full concept of class (truth, concept, being, etc.) is not used in mathe-
matics, and the iterative concept, which is sufficient for mathematics, may
or may not be the full concept of class. Therefore, the difficulties in these
logical concepts do not contradict the fact that we have a satisfactory math-
ematical foundation of mathematics in terms of the iterative concept of set.
In relation to logic as opposed to mathematics, Gödel believes that the
unsolved difficulties are mainly in connection with the intensional para-

doxes (such as the concept of not applying to itself) rather than with either
the extensional or the semantic paradoxes. In terms of the contrast be-
tween bankruptcy and misunderstanding as considered below, Gödel’s view
is that the paradoxes in mathematics, which he identifies with set theory,
are due to a misunderstanding, while logic, as far as its true principles are
concerned, is bankrupt on account of the intensional paradoxes.
Comment. This is pretty much the position I want to challenge. I do not think
that mathematics, not even its arithmetical subsystem, can escape the subversive
influence of impredicative classification. No special favour (“Extrawurst”) for math-
ematics!
(5) Drake [1974], p. 14.
The notion that every abstraction should give a set seems to arise if
we think of a complete universe given, and a set as being any partition of
the universe (so that a set is only something about which we can say, of
anything in the universe, either that it belongs to the set or that it does
not). If we call such things properties, then it is clear that the paradoxes are
a real problem to be dealt with before a thoroughgoing theory of properties
can be developed. (The consensus of opinion seems to be that in fact it
just cannot be developed, but some system of levels must be resorted to—
that we cannot consistently regard properties as always being members of
the universe to which they apply.) But it should not be forgotten that the
paradoxes never applied to any type structure, and in this sense they are
not paradoxes of set theory.
(6) C. Parsons [1977], p. 335.
One can state in approximately neutral fashion what is essential to the ‘iter-
ative’ conception: sets form a well-founded hierarchy in which the elements
of a set precede the set itself.
(7) C. Parsons [1977], pp. 336 f.
The idea that the elements of a set are prior to the set is highly persua-
sive as an approach to the paradoxes. If we suppose that the elements of a
set must be ‘given’ before the set, then no set can be an element of itself.
And there can be no universal set. The reasoning leading to the Russell and
Cantor paradoxes is cut off.
However, one does not deal so directly with the Burali-Forti paradox.
Why should it not be that all ordinals are individuals and therefore ‘prior’
to all sets, so that there is no obstacle of this kind to the existence of a set
of all ordinals? To be sure, once we look at things in this way it becomes
persuasive to view the Burali-Forti argument just as a proof that there is
no set of all ordinals.
What is important in this context: If objects are well-defined they form truth-
definite propositions no matter how they are referred to.
An intuitive justification of the axioms of ZF can be found in Shoenfield [1977],

pp. 325 ff. I am not interested in further details regarding the steps of the iterative
concept of set and close this section with the following collection of quotations.
Quotations 83.12. (1) Shoenfield [1962], p. 133, footnote 3.

[[F]]inite sets present no problems, and the set of natural numbers is concep-
tually, the simplest infinite set. However the principal difficulties concerning
predicativity already arise in this simplest case.
Comment. Cantor in his letter to Dedekind dated 28 August 189921 expresses more
caution regarding the consistency of finite sets; at least he doesn’t say that it
presents no problem. To him it is a simple, but unprovable truth.
(2) Boolos [1971], p. 223.

We may capture part of the content of the idea that at any stage every
possible collection (or set) of sets formed at earlier stages is formed (if it has
not yet been formed) by taking as axioms all formulas p(s)(∃y)(x)(xǫy ↔
(χ&(∃t)(tEs & xFt)))q where χ is a formula [[. . .]] in which no occurrence of
‘y’ is free. Any such axiom will say that for any stage there is a set of just
those sets to which χ applies that are formed before that stage.
Comment. With Boolos I shall call these axioms specification axioms. They form
the crucial point of the iterative conception of set.
(3) Boolos [1971], p. 227.

We do not believe that the axioms of replacement or choice can be inferred
from the iterative conception.
(4) Boolos [1971], p. 230.
It seems that, unfortunately, the iterative conception is neutral with
respect to the axiom of choice. It is easy to show that, since, as is now
known, neither the axiom of choice nor its negation is a theorem of ZF,
neither the axiom nor its negation can be derived from the stage theory.
Of course the stage theory, which is supposed to formalize the rough de-
scription, could be extended so as to decide the axiom. But it seems that
no additional axiom, which would decide choice, can be inferred from the
rough description without the assumption of the axiom of choice itself, or
some equally uncertain principle, in the inference. The difficulty with the
axiom of choice is that the decision whether to regard the rough description
as implying a principle about sets and stages from which the axiom could
be derived is as difficult a decision, because essentially the same decision,
as the decision whether to accept the axiom.
21 See Cantor [1932], p. 448; quotation 73.12 in these materials.

83c. Set theory, higher order logic, and the foundations of mathemat-
ics. The philosophical background of contemporary research in the foundations of
mathematics, the philosophy of mathematics and logic, hardly goes beyond some
rudimentary analytical philosophy. The tradition of transcendental idealism in the
foundations of mathematics22 does not seem to have been continued. One of the
few philosophically minded researchers in the foundations of mathematics was Hao
Wang. The first of the following group of quotations is a beautiful example of linking
Kant into set theory.
Quotations 83.13. (1) Wang [1954], pp. 262 f; partly paraphrasing Kant [1873],
p. 342.23
[[T]]o ask whether the totality of all sets of positive integers is denumerable
(in the absolute sense) is very much like asking, as a common man though
perhaps not as a physicist, whether or not the world is bounded in time
and space. The totality of all sets or of all sets of positive integers is like
Kant’s thing-in-itself, while the constructible sets correspond to all possible
experience. To parrot Kant: Now if I inquire after the quantity of the total-
ity, as to its number, it is equally impossible, as regards all my notions, to
declare it indenumerable or to declare it denumerable. For neither assertion
can be contained in mental construction, because construction of an inde-
numerable totality or a closed denumerable totality incapable of further
expansion, is impossible; these are mere ideas. The number of the totality,
which is determined in either way, should therefore be predicated of the
transcendent totality itself apart from all constructive thinking. We cannot
indeed, beyond all possible construction, form a definite notion of what the
transcendent totality of all sets may be. Yet we are not at liberty to abstain
entirely from inquiring into it; for construction never satisfies reason fully,
but in answering questions, refers us further and further back, and leaves
us dissatisfied with regard to their complete solution. . . . The enlarging of
our views in mathematics, and the possibility of new discoveries, are in-
finite. But limits cannot be mistaken here, for mathematics refers to the
constructible only, and what cannot be an object of intuitive contemplation
such as the totality of all laws, lies entirely without its sphere, and it can
never lead to them; neither does it require them.
(2) Wang [1974], p. 43.
It might seem puzzling that, e.g. the Peano axioms, in particular, an al-
ternative explicit formulation with only a finite number of axioms should
contain so many surprises. The essential thing is, of course, the possibility
of iterated applications of the same old rules in an unbounded number of
combinations. This is also why proving the consistency of such a system is
no easy matter.
22 Cf. quotations 73.18 in these materials; also Bernays.

23 See quotation 62.15 (1); Wang, apparently, used a different translation.
Comment. As regards an answer from a dialectical point of view to the question

why proving the consistency of arithmetic is so much more complicated than, e.g.,
first order logic, see my considerations in chapter XXIX.
(3) Wang [1974], pp. 49 f.
It seems reasonable to suppose that if a theory is consistent, it must have
some interpretation. It may be very difficult to fabricate a model, but how
can a theory be consistent and yet be satisfied by no model whatsoever? The
fundamental theorem of logic gives a sharper answer for theories formulated
as formal systems within the framework of pure logic, i.e. the theory of
quantifiers: any such theory, if consistent, has a relatively simple model in
the theory of positive integers, simple in the sense that rather low level
predicates in the arithmetic hierarchy would suffice.
Comment. This, I should have thought, is enough to raise suspicion against pure
logic.
(4) Wang [1974], pp. 50 f.
The arithmetic translations of theorems in the usual systems of set theory
are often no longer theorems of the usual system of arithmetic. As a result, a
consistency proof of the axioms of arithmetic does not settle the consistency
question for classical analysis or set theory. Even in the consistency proof
of arithmetic, there appears to be an indeterminacy in the notion of finitist
proofs.
Moreover, there is a choice between different axiom systems of arith-
metic not only in the simple sense that alternative equivalent formulations
of, say, the Euclidean geometry are familiar, but in the deeper sense that
extensions of the usual set of arithmetic axioms seem to be just as natu-
ral, e.g. the addition of transfinite induction to the first epsilon number.
This tends to indicate that there is something absolute in the concept of
number and we only gradually approximate it through mental experimen-
tations. Or, at least, we have no full control over our intentions and mental
constructions which, once in existence, tend to live a life of their own.
In a different direction, the existence of consistent systems which have
no standard models (e.g. are omega inconsistent) points to a certain dis-
crepancy between existence and consistency. The usual axioms require that
certain sets or numbers exist but remain mum on what things to exclude.
On account of this, we can add unnatural numbers to the natural numbers
without violating the axioms, and, indeed, consistently add new axioms to
require that there must be unnatural numbers too. One might argue with
reason that although these unnatural numbers are required by the axioms
of a consistent system, they should not be said to exist. Such a position
would foil the unqualified identification of consistency with existence.
Quotations 83.14. (1) Wang [1954], p. 244 (italics mine).
In inventing set theory, the two most remarkable jumps which Cantor made
were: the invention of transfinite ordinal numbers of his second number
class, and the use of indenumerable and impredicative sets. The first is now
known to be harmless and useful (especially in certain metamathematical

considerations), while the second remains a mystery which has shed little
light on any problems of ordinary mathematics. There is no clear reason
why mathematics could not dispense with impredicative or absolutely inde-
numerable sets.
(2) Goodstein [1963], p. 210.
[[I]]f we do not assume that the totality of subsets forms a set (and this
is nothing but an assumption) then all that the diagonal process proves is
that from any sequence of subsets we can construct another subset, just as
from any natural number we can construct another by adding one.
(3) Takeuti [1975], p. 131.
[[I]]n the case of natural numbers we can actually complete the creation of
natural numbers and construct a complete universe of natural numbers,
namely ω. This is so, because in the case of natural numbers +1 is the sole
operation for creating new objects and its behavior is quite clear. In the
case of set theory the principle of creation is powerful and inexact.
(4) Wang [1955], pp. 232 f.
[[T]]here is no axiom system in which we can get all the real numbers or
the classes of positive integers. This follows easily from Cantor’s famous
argument for non-denumerability. Thus, given any axiom system, we can
enumerate all the classes of positive integers which can be proved to exist
in the system, either by applying Löwenheim’s theorem or by reflecting on
the fact that the theorems of existence in the system can be enumerated.
Hence, if we define with Cantor a class K of positive integers such that for
each n, n belongs to K if and only if n does not belong to the nth class in
the enumeration, then the existence of K cannot be proved in the system.
In other words, although in the system we can also speak of all the classes
of positive integers, we cannot really formalize without residue the intuitive
notion of “all” with regard to classes of positive integers; in each formalized
axiom system, there is always some class of positive integers that is left out.
Comment. I take this to indicate some dialectic of form and content. In this context,
compare also Goodstein [1963], Putnam [1980].
Quotation 83.15. Drake [1974], pp. 19 f.

Quine [1937] has introduced a form of “set theory” which is not based on
any intuitive idea of what a set is, but simply a formal system, paralleling
Frege’s but borrowing features from Russell’s formulation of the simple
theory of types in order to block derivation of the paradoxes. Since it has
been shown that this system cannot have an intuitive basis in the cumulative
type structure (the integers of a model cannot be the standard integers; as
is sometimes said, it cannot have a “standard model”; see Rosser and Wang
[1950]), we cannot regard it as being a set theory, in our sense, at all[[.]]
Quotation 83.16. Cocchiarella [1993], p. 32.

Quine: Higher-order logic “has the fault of diverting attention from the
major cleavages between logic and set theory. It encourages us to see the
general theory of classes and relations as a mere prolongation of quantifi-
cation theory in which the hitherto schematic predicate letters are newly
admitted into quantifiers and into other positions that we hitherto reserved
to ‘x’, ‘y’, etc.” Indeed, “the orthodox terms ‘first-order predicate logic’ and
‘higher-order predicate logic’ have the same deplorable source,” namely, “a
blurred view of . . . the distinction between schematic letters and variables,
and . . . between use and mention.”
Note: It is false to claim, and misleading to suggest that the use of
predicate variables as dummy schema letters antedates their use in the
predicate quantifiers of second-order logic. Frege, the founding father of
first-order predicate logic, developed that logic only as an integral part of
second-order predicate logic, which he also founded.
It is also false and misleading to suggest, that the development of
higher-order logic is a mistake based on a confusion between schematic let-
ters and variables, and between mention and use. Worse, it [[is]] ad hominem:
for it is only by seeing “dimly” and having a “blurred view” that one can
make this mistake, according to Quine.
Note: It is false to claim that by allowing quantifiers to be affixed to
predicate variables we are thereby committed to nominalizing those vari-
ables (and predicate expressions in general) and allowing them to occur as
abstract singular terms in the “positions that we hitherto reserved for ‘x’,
‘y’, etc.”
And even when we extend second-order predicate logic and allow pred-
icate expressions to be nominalized and occur as abstract singular terms
on a par with individual variables, it is false to claim that the objects de-
noted by those singular terms must also by [[sic]] the values assigned to the
predicate variables in their role as predicates.
Frege did not admit abstract singular terms at all into his original for-
mulation of second-order logic in the Begriffsschrift. It was only later in
the Grundgesetze that such terms were introduced through the smooth-
breathing abstraction operator as a nominalizing device. That was a neces-
sary step to take in order “to obtain objects out of concepts, namely extents
of concepts or classes,” but Frege was quite explicit in seeing it as an added
step — for he insists that “we can treat the principal parts of logic without
speaking of classes, as I do in my Begriffsschrift.”
The next quotation concerns Gödel’s later view on concepts, classes and ex-
tensionality. It is a particularly interesting one in view of the approach I take to
dialectical logic.
Quotation 83.17. Wang [1986], p. 286.

G[[ödel]]’s concept of class seems to have changed somewhat over the years.
In his later conversations [[. . .]], he took classes only as a subsidiary deriv-
ative of concepts. He proposed also to delete from the Russell paper the
sentence, ‘It might even be that the axiom of extensionality or at least
something near to it holds for concepts’ ([[Wang [1974]]], p. 220). In 1975 he
believed that the axiom of extensionality holds only for concepts in excep-
tional cases and that, therefore, classes are of no central importance for the
theory of concepts.
Comment. This is an extremely relevant point in view of my enterprise, and if it
is only to give me the reassurance that I am not completely off the planet with
what I am doing in the groundworks, in particular regarding the global failure
of extensionality, as formulated in theorem 130.19.
The last set of quotations in this section concerns the continuum hypothesis
and the axiom of choice.
Quotations 83.18. (1) Suppes [1960/72], p. 252.

The independence of the continuum hypothesis and many related as-
sertions show the intuitive incompleteness of the standard axioms of set
theory—incompleteness to a degree hardly expected. To a certain extent
the situation is comparable to that in geometry after the independence of
the parallel postulate was established. We now have many geometries and
presumably we may expect may set theories. However, the intuitive conse-
quences for the foundations of mathematics are not certain. Will there be
a turning away from set theory as the standard foundation of mathemat-
ics? The answer may be affirmative, but it is likely that the main lines of
development of Zermelo-Fraenkel set theory will retain a permanent place
nearly comparable to that given Euclidean geometry.
Comment. Regarding my approach to dialectic it may be worthwhile pointing out
that the continuum hypothesis doesn’t play a role in my enterprise, simply because
the relevant theorem 3.45 cannot be proved on the basis of dialectical logic.
(2) Jech [1977], p. 346.
[[W]]hy has this simple (if not self-evident) axiom [[of choice]] generated so
much controversy? For no postulate since Euclid’s Parallel Axiom aroused
so much excitement in mathematical circles and led to so many philosoph-
ical arguments about the foundations of mathematics.
The answer lies in the nonconstructive nature of the axiom of choice.
The axiom postulates existence of a set C which has certain properties
(namely chooses one element xi in each Ai ), but does not give the slightest
hint how to construct such a set. On the other hand, all other axioms of
set theory assert that certain constructions on sets result in new sets, and
that various totalities of elements, defined in a certain way, are indeed sets.
For instance, the power set axiom states that for every set X, the collection
P(X) of all subsets of X is a set.
Comment. In view of Bernays’ remark in quotation 78.9 (2) in these materials, it
may be worthwhile pointing out that the axiom of choice is not acceptable for my
theory of dialectic, simply because it hides contraction.
(3) Van Benthem and Doets [1983], p. 298.

The key quantifier combination in Frege’s predicate logic expresses depen-
dencies beyond the resources of traditional logic: ∀∃. This dependency may
be made explicit using a Σ11 formula:
∀x∃yϕ(x, y) ↔ ∃f ∀xϕ(x, f (x)).
This introduction of so-called Skolem functions is one prime source of Σ11
statements. The quantification over functions here may be reduced to our
predicate format as follows:
∃X(∀xyz(X(x, y) ∧ X(x, z) → y = z) ∧ ∀x∃y(X(x, y) ∧ ϕ(x, y))).
Even at this innocent level, the connection with set theory shows up [[. . .]]:
the above equivalence itself amounts to the assumption of the Axiom of
Choice (Bernays).
§ 84. Metamathematics
84a. Quantification, theoretical constants, and some philosophy of

logic. Quantification is commonly seen as a paradigm of infinity and this makes
it a topic worth some attention in the context of my enterprise. I should like to
emphasize, however, that I do not share this view of quantification. Readers who
want to know how I treat infinite formulas are referred to chapter XXXIII in the
groundworks.
I begin with some revision of history.
Quotations 84.1. (1) Barwise [1980], p. 98.

In a sense, the study of infinite formulas goes back to the 1880’s. Gottlob
Frege and Charles Saunder [[sic]] Peirce are responsible for the introduction
of the quantifiers ∀ and ∃ into symbolic logic.
(2) Barwise [1980], p. 99.
In his 1925 paper, Ramsey argues that it is really an accident that an
infinite formula cannot be written down explicitly, and that this accident
should be ignored by logic. What he saw as accidental, however, Hilbert
saw as necessary[[.]]
Comment. As regards the reference to Ramsey, see quotation 77.5 (2) in these ma-
terials. Contrast this with the fact that Hilbert introduced the ω-rule in his pa-
per [1931], somewhat as a reaction to Gödel’s result. Cf. also Gödel’s reaction to
Hilbert’s move as reported by Wang (from Carnap’s diary) in quotation 79.5 (4).
(3) Barwise [1980], pp. 100.
The Gödel Incompleteness Theorem gave an example of a sentence ∀xA(x)
about natural numbers such that each of A(0), A(1), A(2), . . . is provable
with a finite proof from the first-order version of the axioms for natural
numbers, but such that ∀A(x) is not provable with a finite proof. This
suggests adding the following infinitary rule of proof, called the ω-rule (in
§ 84. METAMATHEMATICS 1161
the past often called the rule of complete induction, or Carnap’s rule): For
any formula B, if each of B(0), B(1), B(2), . . . are provable, then conclude
∀xB(x).
If the point of such a rule is lost on the reader, if he feels that there is
no possibility of ever actually applying such a rule, he is in good company.
The rule was first mentioned by Hilbert and Tarski, who both felt the same
way. [[. . .]]
The first person to have taken the rule seriously seems to have been
Rudolf Carnap in 1935 in his work on the syntax and semantics of language.
[[. . .]] Carnap had in mind a fixed countable universe of discourse where
everything had a name, say a1 , a2 , . . . .
Comment. The part in double square brackets is essentially the Tarski quotation
79.5 (5) in these materials. As regards Hilbert’s position: Hilbert [1930b], p. 194,
actually calls this rule “finit” (finitary), given every numerical instantiation is cor-
rect, and he seems quite happy to include it in his proof theory. Compare also
Schütte in quotation 80.5 (3). According to Wang [1986], p. 87, drawing on various
private reports, it was Gödel who disapproved of Hilbert introducing this rule in
his proof theory, actually for pretty much the same reason as Barwise thinks it is
suggested — call it pragmatism. Anyway, it should suit Ramsey. In my view it rep-
resents a fine example of what Myhill, in quotation 83.6 (2) above, calls a “victory
of pragmatism and generally sloppy thinking over philosophical analysis”.
Quotation 84.2. Hodges [1983], pp. 59 f.

Quantifiers did provoke one quite new proof-theoretic contrivance. In the
1920s a number of logicians (notably Skolem, Hilbert, Herbrand) regarded
quantifiers as an intrusion of infinity into the finite-minded world of propo-
sitional logic, and they tried various ways of – so to say – deactivating
quantifiers. Hilbert proposed the following: replace ∃xφ everywhere by the
sentence φ[εxφ/x], where ‘εxφ’ is interpreted as ‘the element I choose among
those that satisfy φ’. The interpretation is of course outrageous, but Hil-
bert showed that his ε-calculus proved exactly the same sequents as more
conventional calculi.
Comment. What actually is it that is so outrageous about this interpretation?
Quotations 84.3. (1) Quine [1970], pp. 65 f.

Pioneers in modern logic viewed set theory as logic; thus Frege, Peano,
and various of their followers, notably Whitehead and Russell. Frege, White-
head and Russell made a point of reducing mathematics to logic; Frege
claimed in 1884 to have proved in this way, contrary to Kant, that the
truths of arithmetic are analytic. But the logic capable of encompassing
this reduction was inclusive of set theory.
(2) Quine [1970], p. 68.
Followers of Hilbert have continued to quantify predicate letters, obtain-
ing what they call a higher-order predicate calculus. The values of these
variables are in effect sets; and this way of presenting set theory gives it a
deceptive resemblance to logic. One is apt to feel that no abrupt addition to

the ordinary logic of quantification has been made; just some more quanti-
fiers, governing predicate letters already present. In order to appreciate how
deceptive this line can be, consider the hypothesis ‘(∃y)(x)(xǫy. ≡ F x)’. It
assumes a set {x : F x} determined by an open sentence in the role of ‘F x’.
This is the central hypothesis of set theory, and the one that has to be re-
strained in one way or another to avoid the paradoxes. This hypothesis itself
falls dangerously out of sight in the so-called higher-order predicate calcu-
lus. It becomes ‘(∃G)(x)(Gx ≡ F x)’, and thus evidently follows from the
genuinely logical triviality ‘(x)(F x ≡ F x)’ by an elementary logical infer-
ence. Set theory’s staggering existential assumptions are cunningly hidden
now in the tacit shift from schematic predicate letter to quantifiable set
variable.
The next set of quotations concerns questions of a delimitation of logic.

The problem of logical constants is the central concern of what has
been called ‘philosophical logic.’ It may be formulated in several ways: to
find really distinctive features of the forms and constants of logic; to give
a general intelligible account of the concept of a logical particle; to find a
natural place for logic in an overall view of human knowledge; to extract
logic from reflecting on the idea of a proposition.
(2) Wang [1974], pp. 20 f.
In one sense (pure or formal) logic is concerned with sentences which are
valid, i.e. hold independently of any special subject matter or are true in
all possible worlds. There is an ambiguity in this conception which turns
on a confusing question which boils down to whether pure sets are to be re-
garded as forming a special subject matter. It seems clear that a sufficiently
important concept of logic does result by excluding sets which inevitably
involve concepts such as infinity, uncountability. Since we intend to disre-
gard modal logic altogether, we thus arrive at the first and narrowest sense
of logic: (elementary or pure) logic is nothing but quantification theory or
(first order) predicate calculus, with or without equality.
A second sense corresponds roughly to what is commonly known as
mathematical logic which includes, besides pure logic, also model theory,
recursion theory, the axiomatic treatment of integers, reals, and sets. In
either case, logic is closely mixed up with metalogic or metamathematics.
The third and broadest sense is far less definite. It is the study of pure
reason or the treatment of what is rational. In this broad sense, it could
include a logic of discovery, a logic of development, some form of inductive
logic, or some form of dialectic logic.
(3) Wang [1974], p. 143.
The many attractive properties of the first order or restricted predicate
calculus (quantification theory) have suggested the convenient identification
of it with first order logic, pure logic, or just logic. Given the identification,
we can determine the realm of logical truths as the theorems of the calculus
(or the substitution instances of these theorems). It is natural to ask how
such an identification could be justified. And this is a question on which one
can hardly expect any definitive answers. The question might be separated
into two parts: how do we choose the logical constants (more exactly, the
logical grammar), and how, after we have chosen the logical constants, do
we determine the realm of logical truths.
(4) Wang [1994], p. 269.
Gödel, like Frege, believes that logic is primarily a theory of concepts.
If we assume or can prove that every set is the extension of some concept,
then set theory is derivable from the theory of concepts. In any case, Gödel
shares Frege’s belief that set theory is a part of logic. That is why I see
Gödel’s conception of logic as a natural development of Frege’s conception
of logic. The main difference is that Frege did not make much use of Cantor’s
work but Gödel took the tradition of Cantor fully into consideration.
(5) Wang [1987], p. 53.
I tend to agree with Dedekind and Hilbert (and G[[ödel]]) that logic includes
set theory. If we use this concept of logic, then we are justified in accepting
the familiar argument for proving Peano’s axioms categorical. Hence, each
proposition is ‘decided’ by the axioms (i.e., either true or false). But this
decidability or determination is different from the more strict requirement
of an algorithm to carry out the decision in each case.
In continue with a set of quotation regarding the nature of logic and analyticity.
Quotations 84.5. (1) Smullyan [1968], p. VII.

We use the term “analytic” to apply to any proof procedure which obeys
the subformula principle (we think of such a procedure as “analysing” the
formula into its successive components).
(2) Hacking [1979], pp. 299 f.
Let there be assumed a notion of classical truth and falsity for elemen-
tary sentences lacking logical constants, and let the deducibility relation be
complete for classical logical consequence among the class of such sentences.
Then one is in a certain sense able to read off the semantics of the logical
constants from the operational rules. Given the underlying notions of truth
and logical consequence, the syntactic rules determine a semantics.
Quotations 84.6. (1) Barwise [1985], p. 5.

As logicians we do our subject a disservice by convincing others that
logic is first-order logic and then convincing them that almost none of the
concepts of modern mathematics can really be captured in first-order logic.
Paging through any modern mathematics book, one comes across concept
after concept that cannot be expressed in first-order logic.
(2) Barwise [1985], p. 6.

The first order thesis [[. . .]] confuses the subject matter of logic with one of
its tools. First-order logic is just an artificial language constructed to help
investigate logic, much as the telescope is a tool constructed to help study
heavenly bodies. From the perspective of the mathematician in the street,
the first-order thesis is like the claim that astronomy is the study of the
telescope.
Comment. Is this supposed to be an allusion to Hilbert’s remark about the as-
tronomer paying attention to his place of observation, the physicist taking care
about the theory of his instrument, and the philosopher criticizing reason itself?24
Barwise seems to be caught in heavy shadow-boxing. I’d like to know who he is
fighting with.
For some of the “arguments that first order logic is more fundamental than the
alternatives”, see R. M. Martin [1965], pp. 279 ff.
The next set of quotations concerns the distinction of first and second order
theories.

If we regard Zermelo’s theory as a functional calculus of the first order, von
Neumann’s system may be considered to be a functional calculus of the
second order.
(2) Shapiro [1985], p. 725.
The difference between, say, second-order real analysis and first-order
real analysis is that in the former it is asserted that the completeness prop-
erty applies to every subset of the domain, whether it can be defined in the
language of real analysis or not; whereas in the latter, it can only be shown
that the completeness property applies to subsets of the domain that are
definable in the given first-order language.
(3) Hodges [1983], p. 77.
Second order logic can be translated wholesale into a kind of two-sorted
first order logic [[. . .]]
How can one distinguish between a proof calculus for second-order logic
on the one hand, and on the other hand a first-order proof calculus which
also proves the sentences in ∆? The answer is easy: one can’t.
(4) Hodges [1983], p. 88.
[[I]]t is simply meaningless to classify mathematical statements absolutely
as ‘first-order’ or ‘not first-order’. One and the same statement can per-
fectly well express a second-order condition on structure A but a first-order
condition on structure B.
24 Hilbert [1918], p. 155.

(5) Van Benthem and Doets [1983], p. 315.

One weak spot in popular justifications for employing higher-order logic
lies precisely in the phrase ‘all predicates’. When we say that Napoleon has
all properties of the great general, we surely mean to refer to some sort of
relevant human properties, probably even definable ones. In other words,
the lexical item ‘property’ refers to some sort of ‘things’, just like other
common nouns. [[. . .]]
Thus, there arises the logical idea of re-interpreting second-order logic,
or even higher-order logic as some kind of many-sorted first-order, with
various distinct kinds of objects: a useful though inessential variation upon
first-order logic itself. To be true, properties and predicates are rather ab-
stract kinds of ‘things’; but then, so are many other kinds of ‘individual’
that no one would object to. The semantic net effect of this change in
perspective is to allow a greater variety of models for Lx , with essentially
smaller ranges of predicates than the original ‘full’ ones. Thus more poten-
tial counter-examples become available to universal truths, and the earlier
set of Lx validities decreases; so much so, that we end up with a recursively
axiomatizable set. This is the basic content of the celebrated introduction
of ‘general models’ in Henkin [1950]: the remainder is frills and laces.
Comment. This concerns Henkin [1950], completeness in the theory of types. Cf.
quotation 78.12 in the last chapter.
(6) Rogers [1971], pp. 84 f.

With the help of the logic of second order we are able to state a number
of very interesting second-order theories; in particular, the second-order
arithmetic of natural numbers, and the second-order algebra of real num-
bers. These second order formulations are on a number of counts superior
to the corresponding first-order theories of arithmetic and algebra. For one
thing, these second-order theories contain only finitely many axioms of a
specifically mathematical nature, while the corresponding first-order theo-
ries require an infinite number of such axioms. Further, as we shall see, these
second-order theories possess the very important property of being categor-
ical, in the sense that all of their principal models are alike in structure; the
corresponding first-order theories, however, are not categorical. And these
important differences between first and second-order theories hold not just
for arithmetic and algebra, but various other mathematical theories as well,
including various theories of geometry. [[. . .]]
As against these various advantages of second-order theories over first-
order theories, there is one big disadvantage, however: second-order logic
is in an important sense incomplete. That is, there is no consistent and
effectively defined set of axioms in which all of the valid formulas of the
second-order logic are derivable. Due in part to this reason, second-order
theories have not in the past received as much attention from logicians as
first-order theories.
Quotation 84.8. Hazen [1983], p. 360.

There is no translation of the language of first-order arithmetic into that of
the ramified theory of types which will make the translations of the axioms
of first-order arithmetic, including the instances of the axiom scheme of
induction, come out as theorems of the ramified system (even with the
axiom of infinity).
Quotations 84.9. (1) Kreisel [1958], p. 161.

Usually Gödel’s incompleteness theorems are taken as showing a lim-
itation on the syntactic approach to an understanding of the concept of
infinity. We note the superficially paradoxical fact [[. . .]] that the complete-
ness of the predicate calculus, so to speak the “adequacy” of the syntactic
approach for predicate logic, leads to a limitation too. This is a point of
view emphasized by Skolem.
Comment. Watch out for what the learned calls a “superficially paradoxical fact”.
(2) Tharp [1975], p. 7.
Strangely, compactness seems to be frequently ignored in discussions
of the philosophy of logic. It is strange since the most important theories
have infinitely many axioms. With only completeness it seems possible a
priori, that a logic might not prove all logical consequences of these theories.
Compactness amounts to the condition that if X l.i. A then Γ l.i. A for
some finite subset Γ of X. Since completeness ensures that if Γ l.i. A then
Γ ⊢ A one may conclude that if the system is both compact and complete,
all logical consequences of a set of hypotheses are provable. We claim that
that is the philosophical point at issue: if something follows, it can be known
to follow.
Comment. This strikes me as very important for me, but I don’t know how.
Quotation 84.10. Garson [1984], p. 249.

The novice may wonder why quantified modal logic (QML) is considered
difficult. QML would seem to be easy: simply add the principles of first-
order logic to propositional modal logic. Unfortunately, this choice does
not correspond to a intuitively satisfying semantics. From the semantical
point of view, we are confronted with a number of decisions concerning the
quantifiers, and these in turn prompt new questions about the semantics
of identity, terms, and predicates. Since most of the choices can be made
independently, the number of interesting quantified modal logics seems be-
wilderingly large.
84b. Model theory and completeness. The characteristic feature of model

theory is the establishment of a relation between a formal theory and certain prop-
erties of the members of a set of independently existing objects.
Quotation 84.11. Hodges [1983], p. 60.

[[T]]he main problem is to show that every consistent theory has a model.
This involves constructing a model — but out of what? Spontaneous cre-
ation is not allowed in mathematics; the pieces must come from somewhere.
Skolem [1922] and Gödel [1930] made their models out of natural numbers,
using an informal induction to define the relations. A much more direct
source of materials was noticed by Henkin [1949] and independently by
Rasiowa and Sikorski [1950]: they constructed the model of ∆ out of the
theory ∆ itself [[. . .]] they factored out a maximal ideal in a ring[[.]].
Quotations 84.12. (1) Van Dalen [1980], p. V.

Logic appears in a ‘sacred’ and in a ‘profane’ form. The sacred form is
dominant in proof theory, the profane form in model theory.
Comment. Tell me what is sacred and what is profane to you, and I tell you what
sect you belong to. This quotation may well indicate the present tendency, but I
am still inclined to consider model theory ideology-riddled.
(2) Crossley et al. [1971], p. 20.
Model Theory is the study of relations between languages and the world, or
more precisely between formal languages and the interpretations of formal
languages.

[[I]]n contrast to proof theory, model theory tries to study a notion of truth
and logical consequence independently of the question of how we come to
know that a sentence is true or a logical consequence of given premisses —
it is true that most presentations of model theory contain proof-theoretical
elements, but pure model theory may be presented without mentioning a
deductive apparatus for the language studies (as most consistently done by
Kreisel and Krivine [[1966]]).
(2) Hodges [1983], pp. 66 f.
[[I]]f a first-order sequent is correct by the Traditional Logician’s definition,
then it is correct by the Proof Theorist’s too. Since the converse is straight-
forward to prove, we have a demonstration that the Traditional Logician’s
notion of validity exactly coincides with the Proof Theorist’s. The proof of
this result uses nothing stronger than the assumption that the axioms of
first-order Peano arithmetic have a model.
The Traditional Logician’s notion of logical implication is quite infor-
mal — on any version it involves the imprecise notion of a ‘valid English
argument’. Nevertheless we have now proved that it agrees exactly with
the mathematically precise notion of logical implication given by the Proof
Theorist. (Cf. Kreisel [1967].) People are apt to say that it is impossible to
prove that an informal notion and a formal one agree exactly. Since we have
just done the impossible, maybe I should add a comment. Although the no-
tion of a valid argument is vague, there is no doubt that (i) if there is a
formal proof of a sequent, then any argument with the form of that sequent
must be valid, and (ii) if there is an explicitly definable counterexample to

the sequent then there is an invalid argument of that form. We have shown,
by strict mathematics, that every finite sequent has either a formal proof
or an explicitly definable counterexample. So we have trapped the informal
notion between two formal ones. Contrast Church’s thesis, that the effec-
tively computable functions (informal notion) are exactly the recursive ones
(formal). There is no doubt that the existence of a recursive definition for
a function makes the function effectively computable. But nobody has yet
thought of any kind of mathematical object whose existence undeniably
implies that a function is not effectively computable. So Church’s thesis
remains unproved.
Comment. I’ve got a sneaking suspicion that I am on the wrong side again; at
least on the wrong side of catholicism. Anyway, my problem is the role of the least
number operator in the definition of recursive functions.
Quotations 84.14. (1) Montague [1965], p. 136.

[[W]]henever we speak of the standard models of a first-order theory T we
have in mind a related second-order theory U ; the standard models of T
are then identified with the models of U .
(2) Boolos [1979], pp. 35 f.
Sentences of PA will be called true or false when they are true or false in
the standard model of the language of PA.
84c. Self-reference, the fixed point property, and incompleteness. In

the center of the incompleteness and undecidability results stands what I called
the indirect fixed point property. It was virtually provided in Gödel [1931], but it
seems that it had to wait for Montague [1962] to be formulated as the “principle of
self-reference” in full generality.25
Quotations 84.15. (1) Smoryński [1981], p. 357.

One of the great curiosities of [[self-reference in arithmetic]] is how long
it took (perhaps better: is taking) for the subject to develop. Even the most
obvious and central fact—the Diagonalisation Theorem—seems to have had
difficulty in surfacing. It is not to be found in many of the basic textbooks
[[. . .]] and it is only stated in its most rudimentary form in most others
[[. . .]] Yet, diagonalisation in arithmetic is fifty years old and was stated in
proper generality in print in 1962 [[Montague [1962]]]—long before most of
the available textbooks were written.
Comment. What Smoryński calls the “Diagonalisation Theorem” corresponds to
theorem 48.22 (indirect fixed point property) in the tools.
25 See also the formulation of the (indirect) fixed-point property in theorem 48.22 in the
tools.
(2) Smoryński [1982], p. 444.

Let ψυ be any formula with only υ free such that, for all sentences φ,
PA ⊢ ψ(p¬φq) ↔ ¬ψ(puq).
Then ψ and ¬ψ commute, but share not fixed point.
The proof is immediate. [[. . .]]
Just as it is not the case that every function has a fixed point or that
every pair of commuting functions possessing fixed points will share a fixed
point, it is not, as we have just seen, the case that commuting formulae will
share a fixed point.
Comment. The proof is only immediate given classical logic. In the groundworks,
I shall establish that this does not hold for a system of dialectical logic with the
tru-predicate, for instance. By the way, part of my approach is that every function
does indeed have a fixed point,26 without succumbing to Scott’s prophecy of doom;27
not necessarily, however, in the realm of the natural numbers.

It is often said that Gödel’s formula (∀m)G(m) is true but unprovable.
The reason for saying that it is true is presumably that since each of G(0),
G(1), G(2), . . . is provable, and so true, therefore G(m) is true for all m,
which is just another way of saying that (∀m)G(m) is true. Of course if we
do mean nothing more by saying that (∀m)G(m) is true than that G(m)
is true for all m then it is certainly true to say that (∀m)G(m) is true but
unprovable. But the expression is a rather misleading one. The relationship
between the formal system and the metalanguage which is established by
recursion assures us that if R(m) is a primitive recursive predicate such
that R(m) holds for some m then certainly R(m) is provable in A, or
rather the formula in A which represents R(m) is provable. But it is the
essence of Gödel’s theorem itself that although G(m) is primitive recursive
the formula (∀m)G(m) does not express the notion ‘for all m, G(m)’ in
the formal system. As we have seen there is an interpretation of the system
in which G(0), G(1), G(2), . . . are not all the instances of (∀m)G(m) and
therefore the truth of these instances is not to be identified with the truth
of the formula (∀m)G(m).
Comment. The ‘truth’ of something like the Gödel sentence is a central issue in my
approach to dialectic, but I have not yet made up my mind on it. I shall discuss
aspects of it in §115 in the materials.
(2) Goodstein [1963], p. 219.
Another common mistake is to suppose that (∀m)G(m) is true because
it truly affirms of itself that it is non-demonstrable. We recall that G(m)
is an abbreviation for the formula which is obtained by substituting the
number a of the formula
(∀m)¬Pr(m, Stn (ν/n))
26 Cf. theorem 130.8 in the groundworks.
for the variable n in this formula.

The number of the resulting formula is of course
Sta (ν/a)
as we already have had the occasion to remark; hence the fact that
(∀m)G(m) is unprovable, i.e. that formula number Sta (ν/a) is unprovable,
tells us that
¬Pr(m, Sta (ν/a))
is provable for each value of m, but this of course tells us nothing about
the formula
(∀m)¬Pr(m, Sta (ν/a))
that is, nothing about (∀m)G(m). The Gödel numbering established a code
in which each instance of the numerical formula G(m) says that m is not
the number of the proof of (∀m)G(m), but (∀m)G(m) itself says nothing
at all in the code. Thus the Gödel sentence is neither an example of self-
reference nor of self-description. Even the sense in which we can say that
the formula of A which we are denoting by
Pr(x, y) (P)
says that x is the number of the proof of formula number y is in need of
clarification. As a formula of A, P says nothing at all. As the representative
in A of a certain arithmetical relation it says that y is the exponent of
the greatest power of the greatest prime number which divides x; and only
as a sentence of the code which the Gödel numbering establishes, does
this arithmetical relation say that x is the number of the proof of formula
numbery.
(3) Goodstein [1965], p. 20.
The code has been used and mentioned, and there is no self-reference.
Comment. This is what it comes down to, for me. This use and mention, however,
is not a confusion, it is pretty inescapable once you accept a certain amount of
arithmetic and a formalized theory incorporating this. The question for me then
is, does it make sense to stick to classical logic if use and mention ‘confusion’ is
inescapable? In this respect: the proof that Gödel’s sentence is true requires tertium
non datur in the sense discussed in section 120b in the groundworks.
(4) Odifreddi [1989], p. 165.
[[S ]]elf-reference is never direct: it comes from a controlled confusion of two
levels of meaning for integers, which are seen both as numbers and as names
for formulas.
Quotations 84.17. (1) Crossley et al. [1972], p. 37.

The diagonal method is often regarded with suspicion; however in reality it
is perfectly concrete and non-paradoxical.
Comment. What kind of “reality” would that be? A Platonic universe, or just the
‘reality’ of the human mind to get used to anything, if it is repeated often enough.
Contrast Boolos in quotation 84.19 below.
(2) Kripke [1975], p. 692.

Gödel put the issue of the legitimacy of self-referential sentences beyond
doubt; he showed that they are as incontestably legitimate as arithmetic
itself.
Quotations 84.18. (1) von Kutschera [1964], p. 93; my translation.

[[S]]imilar circular structures as those which underlie the antinomy of the
“Liar” occur, via the arithmetization of [[the theory]] and the possibility of
formal expressibility of primitive recursive predicates in [[the theory]], in the
proof of Gödel.
(2) von Kutschera [1964], p. 99; my translation.
If, however, “P (t, t)” [[von Kutschera’s formulation of an undecidable
number theoretic predicate]] is meaningless under the assumption of the
indirect proof, then this proof for the unsolvability of the halting problem
for TM is untenable[[.]]
Comment. This is one of the few attempts of which I know to question the proof
methods employed in undecidability and incompleteness results. The observation is
challenging, although it is placed within the rather dogmatic tradition of meaning-
lessness of expressions.
Quotation 84.19. Boolos [1993], pp. 54 f.

Löb’s theorem is utterly astonishing for at least five reasons. In the
first place, it is often hard to understand how vast the mathematical gap is
between truth and provability. [[. . .]]
Secondly, Bew seems here to be working like negation. After all, if ¬S →
S is provable, then so is S; proving S by proving ¬S → S is called reductio
ad absurdum (or, sometimes, the law of Clavius). Moreover, inferring S
solely on the ground that (S → S) is demonstrable is known as begging the
question, or reasoning in a circle. To one who conflates truth and provability,
it may then seem that Löb’s theorem asserts that begging the question is
an admissible form of reasoning in PA.
Thirdly, one might have thought that at least on occasion, PA would
claim to be sound with regard to an unprovable sentence S, i.e., claim that
if it proves S, then S holds. But Löb’s theorem tells us that it never does
so[[. . . .]]
Fourthly, one might very naturally suppose that provability is a kind of
necessity, and therefore, just as (p → p) always expresses a truth if the
box is interpreted as “it is necessary that” [[. . .]] Bew(p(Bew(pSq) → S)q)
would also always be true or at least true in some cases in which S is false
and not true only in the rather exceptional cases in which S is actually
provable.
Finally, it seems wholly bizarre that the statement that if S is provable,
then S is true is not itself provable, in general. For isn’t it perfectly obvious,
for any S, that S is true if provable? Why are we bothering with PA if its
theorems are false? And how could any such (apparently) obvious truth not
be provable?
Comment. In particular the last point doesn’t seem to be made as strongly as pos-
sible; it is not just an “(apparently) obvious truth”, but one that can be proved in
a second order extension of PA. Anyway, I wished philosophers of Hegelian persua-
sion had gained the slightest inkling of that “utterly astonishing” situation depicted
above.
84d. Evaluating incompleteness and undecidability results. Gödel’s

and Hilbert’s remarks concerning the relevance of Gödel’s second incompleteness
theorem notwithstanding, logicians generally read Gödel’s second incompleteness
result simply as a defeat of Hilbert’s programme.
Quotations 84.20. (1) G. H. Moore [1982], p. 259.

Although Gödel’s Second Incompleteness Theorem dashed Hilbert’s
hopes of establishing the consistency of arithmetic and set theory finitisti-
cally, Bernays continued to develop Hilbertian proof theory.
Comment. Why Bernays? It was Gentzen, Ackermann, Schütte, Takeuti, Prawitz,
to name a few, who carried on with consistency proofs. Even Gödel himself par-
ticipated with his paper [1958] for Bernays. Compare also Gödel’s own remarks in
(2) Barwise [1980], p. 99.
In spite of the fact that Gödel’s Theorem proved that Hilbert’s Program
could not succeed, that it could never provide a foundation for mathematics,
there remained in many quarters a preoccupation with seeing just how far
the program could be pushed.
Comment. It seems to me that Gödel [1958] belongs into these quarters too.
(3) Wang [1993], p. 924; reporting from conversations with Gödel between October
1971 to March 1972.
Hilbert had an attitude of rationalistic optimism which has not been refuted
by Gödel’s Theorem. And Gödel shares this rationalistic optimism. This
rationalism is directed more to the conceptual aspect than to the real world.
(4) Resnik [1974], p. 136.
Gödel’s two famous theorems about formal systems for number theory are
usually interpreted as undermining Hilbert’s program. The first, the incom-
pleteness theorem, has less bearing upon the program than is often credited
to it. If Hilbert had claimed that every sentence of mathematics is true or
false and that truth is to be identified with provability in some particu-
lar formal system, then the existence of undecidable sentences for systems
(otherwise) adequate for mathematics would have undermined his program.
But Hilbert only ascribes literal truth-values to real sentences, and every
formal system to which Gödel’s theorems apply is complete with respect
to its real sentences. Thus the undecidable sentences are ideal sentences.
Moreover, since they are far removed from ordinary mathematics there is
no compelling reason for a Finitist to include them among his formal theo-
rems. (The argument that there are undecidable but true sentences is not
finitistically meaningful, because the sentences involved are ideal.)

In each of certain systems a proposition p can be found and interpreted as
saying that p itself is not a theorem. It can also be proved that the propo-
sition p cannot be proved in the system (if the system is consistent). Some
people think that this already establishes the unprovability of consistency in
the same system. This involves a misunderstanding because, although if p is
not provable, then the system is consistent, it remains to be demonstrated
that if there is any proposition not derivable, p is also not derivable. It is not
sufficient to say that p is intuitively equivalent to the proposition expressing
consistency: 1 the question whether the natural proposition expressing con-
sistency is provable remains to be decided; 2 there are propositions which
intuitively also express consistency of the system but are provable.
(2) Wang [1974], p. 260.
It may be thought that the famous Gödel proposition, which essentially
says of itself that it is not provable in a certain system, violates the vicious-
circle principle. If so, it would be pretty bad for the vicious-circle principle,
since Gödel’s construction is a perfectly sound procedure. Actually, how-
ever, the self-reference is achieved by using considerations from outside the
system. Gödel is defining by a non-objectionable method something which
ordinarily can only be defined by a vicious circle. This is like proving some
set defined by an impredicative definition actually equivalent to a set de-
fined by a predicative definition, thereby making the set non-objectionable.
More exactly, if we say, “This proposition is not provable,” we are using
self-reference, and defining a proposition by referring to itself (or a totality
including itself), but when we find a way of doing the matter as Gödel
does, it is justified. We no longer define a proposition but just interpret a
proposition, and prove results by means of this interpretation.
Comment. This is a crucial point; a possible interpretation. But what then is ref-
erence more than a possible interpretation? Cf. my Stages 110.10 in the ground-
works.
(3) Copi [1971], p. 105.
Efforts to defend the vicious circle principle lead in a very interesting direc-
tion. Fraenkel and Bar-Hillel consider and criticize four arguments intended
to justify it: first it avoids the semantic paradoxes (but there are other ways
of doing so); second, it rules out self-reference (but self-reference by itself
is unobjectionable, as witness the highly useful self-reference in Gödel’s
arithmetization of syntax)[[.]]
Comment. What interests me here is the view of Gödel’s arithmetization as estab-
lishing self-reference.
(4) Wang [1987], p. 63.

For each of a wide class of formal systems S (such as first-order Peano
arithmetic), G[[ödel]] constructs a sentence P that is true but not provable
in the system S. For numerical formulas such as 2 + 3 = 5, 2 × 3 > 5,
we have a clear idea of which are true and which are false. Wittgenstein
does not question this. The sentence P goes beyond such simple statements
F (n) by one generalization, being essentially of the form ‘For all positive
integers n, F (n).’ G[[ödel]] shows that each of F (1), F (2), etc., is provable
in S and therefore true, but P itself is not provable in S. It now seems clear
that, since F (1), F (2), all are true, then P is true by the meaning of ‘all
n.’ This, however, is the main step with which Wittgenstein is not willing
to go along. The objection is presumably that we are taking the infinite
collection of positive integers or F (1), F (2), etc., as a completed (rather
than a potential) whole. The proof assumes the consistency of the system
S, an assumption that is generally taken for granted and that in particular
nobody questions when S is the Peano arithmetic. But surprisingly for this
context Wittgenstein brings in his favorite disdain: ‘the superstitious fear
and awe of mathematicians in face of contradictions’ (Remarks, p. 53).
Tarski [1930] presented a decision method for elementary algebra. In view of

the results of Gödel and Church regarding the incompleteness and undecidability
of arithmetic this may look confusing since elementary algebra includes the real
numbers. The following quotation points out how these results are compatible.
Quotation 84.22. Rogers [1971], p. 135.

[[W]]ithin elementary algebra we are not able to refer in any way to the set
of all natural numbers. There is no formula within R which has one free
variable and is satisfied by a real number x if an only if x is a real number
which is a natural number. For this reason, we are not able to make all
of the assertions here that we are able to make within elementary arith-
metic. For example, within elementary algebra we are not able to say that
a certain condition is not satisfiable within the natural numbers, because
of our inability to refer to just those numbers. It is this fact which blocks
the immediate transfer of Gödel’s result and the Church-Rosser result to
elementary algebra.
Quotation 84.23. Wang [1974], p. 10; quoting from a letter by Gödel.

[[T]]here was another reason which hampered logicians in the application
to metamathematics, not only of transfinite reasoning, but of mathemati-
cal reasoning in general [and, most of all, in expressing metamathematics
in mathematics itself — added April, 1972]. It consists in the fact that,
largely, metamathematics was not considered as a science describing ob-
jective mathematical states of affairs, but rather as a theory of the human
activity of handling symbols.
Quotation 84.24. Wang [1987], p. 23.

The philosophical aspect of G[[ödel]]’s completeness and incompleteness
results can best be understood in the contexts of Hilbert’s formalism and of
logical positivism. If it had turned out that mathematical formal systems
were complete and decidable and their consistency were provable by finitist
means, one would have made an astonishing advance, conclusive in our
understanding of mathematics and tremendous in our understanding of
human knowledge. [[. . .]]
G[[ödel]]’s negative solutions of the problems show that we cannot have
such exhaustive clean answers to these fundamental questions about the
nature of mathematical knowledge. He demonstrated only that mathemat-
ics is not ‘syntax,’ but we are still a long way from a fairly clear picture of
what mathematics is.
Comment. I don’t think it could have ever “turned out that mathematical formal
systems were complete and decidable” because there is no room for our own existence
as humans, and the ones who do mathematics, in that sort of beastly paradise. Cf.
section 108e in the groundworks. But, of course, this is not a conclusive argument.
Quotations 84.25. (1) Enderton [1977], p. 535 f.

The initial application of recursive functions was to prove incomplete-
ness theorems of logic. For that purpose, no deep results on the internal
structure of the class of recursive functions are required. And in fact a
more restricted class, such as the primitive recursive functions [[. . .]] would
suffice.
But [[. . .]] there are other applications for recursive functions. And if
for no other reason, the recursive functions would be studied for their own
interest as the effectively computable functions. And the basic fact that
gets such a study off the ground is the possibility of encoding machines into
integers that can then be supplied as input to other (or the same) machines.
There is a direct analogy here with actual digital computers. The ear-
liest such computers were programmed by setting switches and inserting
wires into plugboards. It was then realized (by von Neumann) that for a
suitable constructed computer, the program could be coded into machine
words (i.e. integers) and stored in the machine in the same manner as
data — the stored program computer. The first practical benefit of this
approach to programs is the speed at which new programs can be loaded
into the computer to replace old programs. But a more significant benefit
(for our purposes) is the possibility of executive programs, e.g. operating
systems. An executive program accepts another program as incoming data.
The executive program might then study the incoming object program and
see that its instructions are carried out.
(2) Enderton [1977], p. 546.
Why is recursive function theory part of mathematical logic? If logi-
cians had not invented recursive functions, computer scientists would have
developed the subject later. But it was not a mere historical accident that
recursive fun[[c]]tions were invented by logicians. There are certain aspects

of logic that inevitably involve the notions of constructiveness and effec-
tiveness.
A basic concept of logic is that of a proof. Now a proof, viewed ab-
stractly, is a series of statements that “establishes” without doubt the truth
of its conclusion, given the truth of its assumptions. But to establish con-
vincingly the truth of the conclusion, the proof must be verifiable by others.
There must be some procedure by which an outsider can verify the correct-
ness of the proof, without having to supply brilliant insight. That is, it must
be possible to verify the correctness of proofs by an effective procedure. The
set of proofs must be recursive.
One last quotation in this section, regarding Church’s thesis.
Quotation 84.26. Barendregt [1997], p. 187.

Church’s Thesis is plausible but cannot be proved, nor even stated in
(classical) mathematical terms, since it refers to the undefined notion of in-
tuitive computability. On the other hand, Church’s Thesis can be refuted. If
ever a function will be found that is intuitively computable but (demonstra-
bly) not lambda definable, then Church’s Thesis is false. For more than 60
years this has not happened. This failure to find a counterexample is given
as an argument in favor of Church’s Thesis. I think it is fair to say that
most logicians do believe Church’s Thesis. One may wonder why doubt-
ing Church’s thesis is not a completely academic question. This becomes
clear by realizing that [[Skolem [1923]]] had introduced the class of primitive
recursive functions that for some time was thought to coincide with that
of the intuitively computable ones. But then [[Ackermann [1928]]] showed
that there is a function that is intuitively computable but not primitive
recursive.
§ 85. Proof theory
Gentzen’s work on natural deduction set an immensely influential agenda for proof
theory: cut elimination and/or normalization. Today, Gentzen-style proof theory is
no longer predominantly focused on the issue of consistency. The main point for my
enterprise is “umwegloses Schließen”.
85a. Looking back. Proof theory did not start with Gentzen’s ideas and the
first couple of quotations is dedicated to a prior historical level which almost seems
to have fallen into oblivion.
Quotations 85.1. (1) Kreisel [1958], pp. 163 f.

[[Hilbert]] wanted proof itself to be made the object of mathematical study
([[1935]], p. 165). Though this is needed for a syntactic (combinatorial) for-
mulation of independence results, by itself it is not the crucial point for
proof theory. For, clearly, the traditional independence proofs by means of
§ 85. PROOF THEORY 1177
models, as in the case of the parallel axiom, or even the impossibility proofs
for certain constructions by means of ruler and compass, are applicable to
formalized systems. Thus the consistency of the rules of set theory is proved
as follows: when read in the intended manner, the axioms of, e.g. Zermelo’s
set theory are true of the concept of set and the rules of proof are such that
true statements are transformed into true one. Hence the formal system is
consistent.
No, from the point of view of technique the crucial point is that from an
early stage Hilbert had in mind a new type of analysis in which the detailed
structure of the proof is considered. In particular, in the consideration of
the so-called transfinite symbols εx A(x), one does not define “models” for
them once and for all, but different numbers are substituted for a given
symbol depending on the particular proof of the system which is analyzed.
Briefly: instead of constructing a model for a system as a whole he gives a
method for constructing a model for each particular proof of the system.
(2) Curry and Feys [1958], pp. 275 f.
In the work of Hilbert it was an all-important problem to demonstrate
the consistency of systems of classical arithmetic and analysis. It seems
likely that back of this insistence on consistency there was a bit of Ger-
man idealistic philosophy, a wish to justify classical mathematics on an
absolutely certain a priori basis. He felt that he would vindicate classical
mathematics absolutely if he could prove its consistency by finite argument,
and from his standpoint such a proof of consistency was vital.
(3) Dreben and Denton [1970], p. 419.
[[T]]he oldest and most naive idea in proof theory: a set of axioms is con-
sistent if it has a model. Hence it is to be contrasted with that approach
initiated by Gentzen in 1938 [[. . .]] and continued by Schütte [[[1951], . . . ,
[1960]]] in which proofs, that is, formal derivations, are subject to various
purely syntactic manipulations, and questions of interpretation play little
role.

The attack on reductio ad absurdum was started in 1913 by the Dutch
mathematician L. E. J. Brouwer. At that time, Brouwer, who had already
won a world-wide reputation by this work on topology, was one of the
editors of the Mathematische Annalen and he opened the attack by rejecting
all papers offered to the Annalen which applied the tertium non datur to
propositions the truth or falsehood of which could not be decided in a finite
number of steps. The Editorial Board met this emergency by resigning—
and then re-electing themselves, minus Brouwer. Incidentally, the Dutch
Government so resented this slight on their leading mathematician that
they founded a rival mathematical journal, with Brouwer in charge.
The defence of reductio ad absurdum was undertaken by Hilbert, who
regarded Brouwer’s criticism as a personal affront. Hilbert planned to dispel
all doubts of the validity of the indirect method, by proving that classical
mathematics is free from contradiction, and contains no insoluble problems;

this proof of course using only arguments accepted by everyone.
The formalist programme—as this ambitious plan was called—was
abandoned in 1931 following the publication of Gödel’s remarkable proof
that every sufficiently rich mathematical system contains problems which
are demonstrably insoluble. Gödel also established the impossibility of a
proof of freedom from contradictions within the framework of classical
mathematics. The question was re-opened, however, some five years later
when Gentzen succeeded in proving that classical number theory is free
from contradiction, but his proof transcended the limitations laid down in
the formalist programme.
(2) E. T. Bell [1940/45], p. 566.
The “frog-and-mouse battle” of the 1920’s–30’s between the formalists
and the intuitionists [[. . .]] was a battle to the death between Hilbert’s for-
malism and Brouwer’s intuitionism for the possession of mathematics. It
doesn’t seem to have occurred to either combatant that while he was en-
gaged in trying to exterminate his enemy, some ragged camp follower might
take off with the prize; or that it might not make the slightest difference
to mathematics whether the battle for him was won, lost, or drawn. The
entire fracas bore a singular resemblance to the wars of the Middle Ages
over subtle questions of religious dogma that later and saner generations
perceived to have pseudo questions devoid of meaning.
Comment. Not to be misunderstood: I include this piece as a pretty good example
of a pretty poor judgment in conjunction with pretty poor writing.28 It doesn’t seem
to have occurred to Bell that he might have been too close historically to judge the
relevance of the dispute in the fundations of mathematics, although most of the
methods typically employed in metamathematical investigations, in particular that
of coding, recursion, and even the notion of (Turing-) computability, had already
made their appearance. Recursion theory would have just been in the making and
the idea of a universal computer had not yet been born; not surprisingly today’s
fruitful co-operation of proof theory and computer science could hardly have been
anticipated by Bell.
(3) Smoryński [1991], p. 141.
Of course, Bell is generally quoted more for the passion of his prose than
the soundness of his opinions.

The separation of the deductive role of the different logical constants
was partly achieved already by Hilbert in some of his axiomatic formulations
of sentential logic. Gentzen is able to complete this separation by separating
also the role of implication from that of the other constants.
28 Lovers of fantastic writing might be thrilled to read further that “the noise of battle rolled”
and Hilbert “shouted” and physicists “stood back to avoid the missiles”.
Comment. The axioms in Hilbert and Bernays [1934] show a nice systematization
the origin of which may be Hilbert [1904].29
Quotation 85.4. Fraenkel et al. [1973], p. 338 f.
Skolem was able to develop a great part of classical arithmetic in this theory
[[of primitive recursive arithmetic]] and Gödel succeeded in showing that
it suffices for the arithmetization of the elementary syntax of any formal
system.
Comment. There is a footnote worth quoting:
In Goodstein [[1957b]], this theory has found its authoritative textbook.
Nothing is presupposed, not even propositional calculus. For the philosophy
behind it, see Goodstein [[1952]].
The point here is, for me, at least, whether full classical logic is taken as the basic
system, as it is often done with PRA,30 or not. (The classical propositional calculus
with bounded quantification can be expressed anyway.)
Quotation 85.5. Rabin [1977], p. 598.
The study of decidability should be viewed as a component, or natu-
ral outgrowth, of Hilbert’s Program for the foundations of mathematics.
Hilbert envisaged a codification of the various branches of mathematics by
systems of axioms, with an axiomatized logic serving as a common basis for
deduction of consequences (theorems) from the axioms. Hilbert hoped that
such a formalization would turn the derivation of mathematical results into
a mechanical game with strings of symbols. According to Hilbert’s plan, this
would give us such a comprehensive survey of all formal theorems within
any mathematical discipline, that we would be able to demonstrate that no
formal statement and its negation are jointly provable, thereby demonstrat-
ing the consistency of mathematics. Also implied by Hilbert’s Programme is
the belief that the process of theorem-proving is mechanizable or, in modern
parlance, that mathematical theories are decidable.
85b. Formalism and the conception of proof theory. Proof theory began
as part of a foundational doctrine sometimes labeled formalism, alongside with logi-
cism and intuitionism. The lack of understanding for the goals of Hilbert’s brand of
‘formalism’ and his idea of proof theory and the distortion of many of the early ac-
counts seem to have been overcome to a considerable extent, but, it seems to me, not
completely. The demand for a thorough investigation of mathematical proofs stands
beyond all metaphysical disagreement which recommends the use of formalization.
Quotations 85.6. (1) Kreisel [1958], pp. 157 f.
[[T]]here is no evidence in Hilbert’s writings of the kind of formalist view
suggested by Brouwer when he called Hilbert’s approach “formalism.” In
particular, we could say that Hilbert wanted to eliminate the use of transfi-
nite concepts from proofs of finitist assertions instead of referring to symbols
29 See van Heijenoort [1967b] p. 132.
30 Cf. remark 45.58 in the tools.
and formulæ[[not containing transfinite symbols]]. The symbols were a means

of representation. The real opposition between Brouwer’s and Hilbert’s ap-
proach was not at all between formalism and intuitive mathematics, but
between (i) the conception of what constitutes a foundation and (ii) be-
tween two informal ways of reasoning, namely finitist and intuitionist.
(2) Resnik [1974], p. 134 f.
Although it is not clear how Hilbert’s finitism is to be formulated, for our
present purposes, we can take it to be captured by means of quantifier free
primitive recursive arithmetic. [[. . .]]
Although a significant amount of number theory can be developed on
this basis, it cannot capture all of classical mathematics. Hilbert’s solu-
tion to this problem was to assert that all other sentences of mathematics
should be construed as meaningless instruments designed for the fruitful
manipulation of real sentences. In analogy with the use of ideal elements in
geometry — e.g., points at infinity in which parallel lines meet — he called
such sentences ideal sentences.
(3) Feferman [1977], p. 351.
Proof theory was conceived by Hilbert as the means to carry out his grand
program to secure the foundations of mathematics by purely mathematical
means which were to be of the most elementary and evident kind. The
logical structure of mathematical practice had been successfully mirrored
within a variety of formal systems S for algebra, geometry, number theory,
analysis, and set theory, all logically based in the predicate calculus (the
logic of propositional connectives and quantifiers). The consistency of each
such S is a combinatorial proposition which may be shown to be equivalent
to a number- theoretical statement of the form (for all natural numbers
n) f (n) = g(n), where f, g are effectively computable. Hilbert expected to
be able to derive such statements using only quantifier-free logic, recursive
definitions of functions and proof by induction. Each derivation of a specific
numerical statement by these means has a finite model. This differs from
the situation where quantifiers are essentially involved in the reasoning:
thus even where the variables of S range over natural numbers one may say
that application of the reasoning of the predicate calculus in S implicitly
involves infinitistic concepts.
Comment. This is an interesting piece of substitution of modern concepts in an
old doctrine. Hilbert just didn’t put it this way; but that’s what it amounts to, in
hindsight.
(4) Schütte [1977], p. 2.
We start from the premise that in mathematics Tertium non datur is not
meaningful as a general basic principle. But if one gives up this principle as
a method of proof then one loses the right to indirect proof procedures such
as one cannot do without in classical mathematics. In order to guarantee
the correctness of classical mathematics in a logically unobjectionable way
there only remains the possibility of basing mathematical statements not on
provability as such but on formal deducibility. This deducibility which may

be regarded as a notional provability must be given axiomatically so that
it is sufficient for the basic theorems of classical mathematics. In this way
one does not need to interpret the concept of deducibility in a meaningful
way, but only to establish its formal admissibility by means of a consistency
proof. This is in fact Hilbert’s programme: to establish the consistency of
mathematics in its classical form.
Hilbert called the theory which solves this problem proof theory (Be-
weistheorie). The object of a proof-theoretic investigation is a mathematical
theory which is precisely determined not only by its mathematical notions
and axioms but also by the basic logical concepts and modes of inference
which it allows. Now one must distinguish clearly between the laws of the
theory which are to be investigated and those of the theory under which the
investigation is to be carried out. The notions and theorems of the theory
under investigation are represented in a formal system and treated purely
formally without reference to their meaning, while the proof-theoretic in-
vestigation is concerned with the logical meaning of the notions and in-
ferences modes. Thus the formal theory is complemented by a meaningful
metatheory (proof theory). The metatheory is also called “metalogic” or
“metamathematics”.
(5) Takeuti [1975], p. 1.
Mathematics is a collection of proofs. This is true no matter what stand-
point one assumes about mathematics — platonism, anti- platonism, in-
tuitionism, formalism, nominalism, etc. Therefore, in investigating “math-
ematics”, a fruitful method is to formalize the proofs of mathematics and
investigate the structure of these proofs. This is what proof theory is con-
cerned with.
(6) Prawitz [1981], p. 236.
[[I]]nstead of proofs in a more absolute sense, [[Hilbert]] wanted to study
formal proofs, i.e. derivations in given formal systems, and, secondly, he
thought of the subject as a strictly mathematical one.
Quotations 85.7. (1) Smoryński [1977], p. 823.

Even if Hilbert had faith in Zermelo’s set theory, he could not use it: For,
he had not to secure mathematics but to stop a Putsch. So Hilbert proposed
his Conservation Program: To justify the use of abstract techniques, he
would show — by as simple and concrete a means as possible — that
the use of abstract techniques was conservative — i.e. that any concrete
assertion one could derive by means of such abstract techniques would be
derivable without them.
(2) Smoryński [1977], p. 824.
Hilbert’s Consistency Program is a natural outgrowth of and successor to
Hilbert’s Conservation Program. There are two reasons for this:
(i) Consistency is merely the assertion that some string of symbols is
not derivable. Since derivations are simple combinatorial manipulations,
this is a finitistically meaningful statement and ought to have a finitistic

proof.
(ii) Proving consistency of the formal system encoding the abstract
concepts already establishes the conservation result!
Quotations 85.8. (1) Kreisel and Takeuti [1977], p. 35; to my mind featuring
Kreisel as “one of us”.
One of us would go so far as to say that [[the]] principal epistemological
value of [[Hilbert’s programme]] was this: it provided the concepts in terms
of which more or less natural assumptions about mathematical reasoning
could be formulated precisely enough to be put in their place. The assump-
tions are verified for the bulk of mathematical practice in as much as this
practice can be formalized in remarkably weak (sub)systems (of set theory).
The assumptions are refuted in principle since they do not apply to all pos-
sible valid mathematical reasoning. Also — and, for one of us, this is philo-
sophically by far the most significant result of work on Hilbert’s programme
— the original epistemological claims for the programme, concerning its rel-
evance to certainty, have not been established. As a matter of empirical fact,
our confidence in a part of practice was not increased, when abstract (set
theoretic) concepts were successfully eliminated ; certainly much less than by
a more precise analysis of which sets we are talking about (which segments
of the cumulative hierarchy). This fact of mathematical experience seems
— again, to one of us — just as convincing evidence against those assump-
tions about mathematical reasoning which are behind Hilbert’s programme
as its theoretical refutation mentioned above. Both of us agree that the pro-
gramme and, in particular, its distinction between finitist and non-finitist
reasoning are natural, especially when on begins to reflect on mathematics;
but we cannot agree on the objective epistemological significance of this
distinction.
On the other hand neither of us doubts the permanent value of proof
theory if this theory is separated from those philosophical doubts (about
the validity of currently used principles) which provided the original raison
d’être for Hilbert’s programme. What we envis[[a]]ge is an analysis of the
structure of proofs, providing concepts in terms of which we can state facts
(about proofs) which we really want to know.
(2) Kreisel [1987], p. 401.
The particular philosophical pollution associated with proof theory began
with such pretentious—and therefore simple minded—claims as Frege’s and
Hilbert’s that so-called complete formalization, especially with finitist meta-
mathematics, is needed for reliability of proofs. Later there was Turing’s
suggestion, elaborated ad nauseam by epigones of Gentzen, that ordinals are
the measure of depth for theorems, if not for all things. The flashy precision
of such classifications pollutes the intellectual atmosphere by blinding the
observer to their obvious inadequacy for understanding the (two) delicate
aspects of mathematical reasoning just mentioned.
Comment. To do justice to Frege, compare this formulation with what he said in

his [1884], p. 2,31 and to be able to see the point of the present study, keep in mind:
complete formalization is not employed for the reliability of proofs, but for the “in-
sight into the dependence of truths upon one another” ( Frege ibid.). This is a good
opportunity to warn against the “philosophical pollution” caused by simplification
and/or lack of understanding of a philosophical standpoint; a danger which Hilbert’s
proof theory is particularly exposed to, since it is directed at mathematics, and the
philosophical horizon of some mathematicians matches that of business executives
talking of the “philosophy” of their firm.
(3) Feferman [1977], p. 354.
It goes against the grain to prove results by specially restricted methods
when they are recognized already to be true, unless one sees sharpenings of
the conclusions or interesting side-products.
Comment. Gödel’s sentence is recognized already to be true, and what I am doing
in this study goes against the grain of mathematical school knowledge in so far as
I go through the proof of its truth and point out where it depends on a certain
application of tertium non datur. I have little doubt about the appeal of such an
enterprise to schoolmen and I therefore wish to express a general warning: beware
of what is “recognized already to be true” and even more of those who wear it on
their sleeves. According to my experience, it is the most boring people who love to
talk about what is ‘interesting’. The capacity to see beyond school knowledge is not
one of schoolmen.
(4) Wang [1960], pp. 257 f.
The suspiciously aggressive term “mechanical mathematics” is not un-
attractive to a mathematical logician. It is a common complaint among
mathematicians that logicians, when engaged in formalization, are largely
concerned with pointless hairsplitting. It is sufficient to know that proofs
can be formalized. Why should one take all the trouble to make exact
how such formalizations are to be done, or even to carry out actual for-
malizations? Logicians are often hard put to it to give a very convincing
justification of their occupation and preoccupation. One lame excuse which
can be offered is that they are of such a temperament as to wish to tab-
ulate all scores of all base ball players just to have a complete record in
the archives. The machines, however, seem to supply, more or less after
the event, one good reason for formalization. While many mathematicians
have never learned the predicate calculus, it seems hardly possible for the
machine to do much mathematics without first dealing with the underly-
ing logic in some explicit manner. While the human being gets bored and
confused with too much rigour and rigidity, the machine requires entirely
explicit instructions.
It seems as though that logicians had worked with the fiction of man as
a persistent and unimaginative beast who can only follow rules blindly, and
then the fiction found its incarnation in the machine. Hence, the striving
31 Cf quotation 71.12 (2) in these materials.
for inhuman exactness is not pointless, senseless, but gets direction and
justification.
Comment. What Wang does not seem to consider is that the fiction of man as a
mathematically unimaginative beast may also find its incarnation in mathematical
idiots, like myself, who greatly profit from not having to understand thoughts — or
“content of mathematical signs” as Engeler says in the next quotation — but can
happily juggle symbols to make them fit their own purpose.
(4) Engeler [1983], p. 6
Notation relieves mathematicians from the need always to think of the
content of mathematical signs and allows them instead literally to compute
with the abstractions themselves.
Comment. Yes, it seems that whatever mathematicians have mastered has been
made accessible (in principle) to mathematical idiots; or, actually, ‘mastering’ in
mathematics just means to make it accessible to mathematical idiots. In general, I
find this a mixed blessing, but I cannot deny that I greatly profited from it — see
preceding comment.
85c. Cut elimination, consistency, and the finitist standpoint.

So far as the present state of mathematics is concerned, speculations on
inconsistent systems are rather idle. No formal system which is widely used
today is under very serious suspicion of inconsistency. The importance of
set-theoretical contradictions has been greatly exaggerated in some quar-
ters. When the non-Euclidean geometries were discovered and found to be
unintuitive, it was natural to look for consistency proofs by modeling con-
siderations. And then it was a short step before one asked for the basis
on which the model itself is founded. When Kronecker thought of classi-
cal analysis as a game with words, it was again natural that he raised the
question whether such a game was even consistent. But the more modern
search for consistency proofs is differently motivated and has a more serious
purpose than avoiding contradictions: it seeks for a better understanding
of the concepts and methods.
(2) Feferman [1987], p. 479.
The efficacy for proof theory of the Hilbert-Gentzen program of ‘securing’
mathematics by consistency proofs is undeniable. This has led, and con-
tinues to lead, to an enormous body of results and methods, especially
concerning informative ordinal analysis of non-constructive theories and
their reduction to constructive theories. But critical examination of this
program leaves one feeling uncomfortable about just what is accomplished.
In particular, is one’s conviction about the consistency of PA or Σ11 -AC,
or Π11 -CA, or Σ12 -AC increased by the proof-theoretical results for these
systems? It would be stretching things to say that the consistency of such
systems was an open question which was finally settled to one’s satisfaction
by the ‘consistency proofs’. [[. . .]]
Hilbert’s program is one of the large global foundational schemes, each

of which has been riddled with criticism and is more suspect than the
mathematics which it is supposed to assure. We have grown tired of them,
but logicians have come up with no new schemes to take their place.
Quotations 85.10. (1) Takeuti [1975], p. 85.

[[I]]t seems quite reasonable to characterize Hilbert’s finitist standpoint as
that which can be formalized in primitive recursive arithmetic.
Comment. Cf. Hilbert and Bernays in quotation 76.13 (5) in these materials.
(2) Tait [1981], p. 524.
[[F]]initist reasoning is essentially primitive recursive reasoning in the sense
of Skolem [[[1923]]].
(3) Wang [1987], p. 56.
I am under the impression that according to the consensus today, finitist
number theory as originally envisaged by Hilbert consists basically of the
quantifier-free theory of primitive recursive functions. It is on the basis of
an identification more or less like this that we can speak of G[[ödel]]’s results
as settling Hilbert’s problems unambiguously.
(4) Girard [1987], p. 34.
Hilbert’s ontology of mathematics distinguished between:
(i) Real (or elementary, finitist ) objects which do exist: they are the
things mathematics is about. Typically the integers, but also the formal
proofs (which can be reduced to integers) are real objects. Real (or elemen-
tary, finitist) properties of real objects are meaningful: typically Fermat’s
last theorem, Goldbach’s conjecture, the four-colour theorem, but also con-
sistency of set theory are elementary in Hilbert’s sense: these properties
Q0
are in fact all 1 , i.e. can be written as ∀x1 . . . ∀xn (t = o) where t is a
primitive recursive function. To these notions of object and property cor-
responds a notion of elementary (or finitist) proof: a finitist proof must use
finitary constructions (i.e. real objects) and elementary properties (and, of
course, the axioms used must be particularly obvious; a candidate for this
is primitive recursive arithmetic PRA).
(ii) Abstract objects which do not actually exist: they are just a façon de
parler. All kinds of infinitary sets, for instance Hilbert spaces, ultrafilters,
are abstract objects. One has also abstracts [[sic]] properties (maybe of
real objects) and non-elementary proof. Hilbert’s idea is that although,
practically speaking, abstract objects are absolutely necessary (they provide
short and elegant proofs), they can (at least in theory) be eliminated from
proofs of elementary properties. (What we should now express by saying
that the abstract methods are conservative w.r.t. elementary methods over
the real properties.) Hilbert’s program is an attempt to prove this fact.
The next set of quotations concerns cut elimination and normalization.


What makes Gentzen’s systems especially interesting is the discovery of a
certain symmetry between the atomic inferences, which may be indicated
by saying that the corresponding introductions and eliminations are in-
verses of each other. The sense in which an elimination, say, is the inverse
of the corresponding introduction is roughly this: the conclusion obtained
by an elimination does not state anything more that what must have al-
ready been obtained if the major premiss of the elimination was inferred
by an introduction. For instance, if the premiss of an & E was inferred by
introduction, then the conclusion of the & E must already occur as one of
the premisses of this introduction. Similarly, if the major premiss A ⊃ B of
an ⊃ E was inferred by an introduction, then a proof of the conclusion B
of the ⊃ E is obtained from the proof of the major premiss of the ⊃ E by
simply replacing its assumption A by the proof of the minor premiss.
(2) Prawitz [1965], p. 8.
A deduction in normal form proceeds from the assumptions of the de-
duction to the conclusion in a direct and rather perspicuous way without
detours; roughly speaking, the assumptions are first broken down into their
parts by successive applications of the elimination rules, and these parts
are then combined to form the conclusion by successive applications of the
introduction rules.
Comment. Cf. the intuitive consideration 20.6 in the tools.
(3) Prawitz [1971], p. 249.
A normal derivation has quite a perspicuous form: it contains two parts,
one analytical part in which the assumptions are broken down in their
components by use of the elimination rules, and one synthetical part in
which the final components obtained in the analytical part are put together
by use of the introduction rules. Between the analytical and the synthetical
part there is a minimum part in which operations on atomic formulas may
occur.
(4) Prawitz [1971], pp. 260 f.
In the way Gentzen presented his analysis, it was technically summed
up in his so-called Hauptsatz, which states that all applications of the so-
called cut-rule can be eliminated from the proofs. This Hauptsatz and our
normal theorem are equivalent in the sense that one can be obtained from
the other by a suitable translation between the two kinds of systems (as
described in some detail in Prawitz (1965) 90–93, where also some other
comments about the relations between the systems are made).
There are certain advantages in carrying out the analysis for natural
deduction besides the fact that the development (including the proofs of
the theorems) flows very natural from the underlying idea. The main ad-
vantage is that the significance of the analysis, as I have tried to describe
it above, seems to become more visible. Further more it has recently been
possible to extend the analysis of first order proofs to the proofs of more
comprehensive systems when they are formulated as systems of natural

deduction [[. . .]], while an analogous analysis with a calculus of sequents for-
mulation does not suggest itself as easily. Finally, the connection between
this Gentzen-type analysis and functional interpretation such as Gödel’s Di-
alectica interpretation becomes very obvious when the former is formulated
for natural deduction [[. . .]].

A proof in a calculus of sequents can be looked upon as an instruction on
how to construct a corresponding natural deduction. This is particularly
evident in the case of intuitionistic or minimal logic. A top-sequent then
corresponds to a natural deduction consisting of just the formula that occur
both in the antecedent and the succedent. As we go downwards in the proof
in the calculus of sequents, we successively enlarge in two directions the cor-
responding natural deductions obtained at the upper levels. When we come
to applications of succedent rules, we enlarge the corresponding natural de-
ductions at the bottom, applying the corresponding I-rules; when we come
to applications of antecedent rules, we usually enlarge the corresponding
natural deductions at the top, applying the corresponding E-rules.
(2) Zucker [1974], p. 6.
The fact that the correspondence fails for disjunction shows that there is
indeed a combinatorial difference between the sequent calculus and natural
deduction, at least with regard to the reduction procedures. This suggests
that it is misleading to think of the sequent calculus merely as a system of
“representations” of natural deduction derivations, showing how they can
be built up. (This is, in any case, clearly false for classical logic [[. . .]].)
(3) Girard [1987], pp. 126 f.
Systems of natural deduction are perfectly equivalent to sequent calculus,
and proof-theorists are still divided on the question as to which formulation
is the best; our choice of presenting sequent calculus in the main part of the
chapter, and natural deduction as an appendix expresses a personal opinion
on the subject: we think that sequents are better, but that it is impossible
to ignore natural deduction.
Advantages of natural deduction: [[. . .]] this formulation is very sim-
ple and elegant; the normalization procedures are easier to write, and, in
general, the proof, viewed as an object is more easily handled by natural
deduction. Another advantage is that, in some sense, natural deduction of-
fers us (at least for the fragment ∀, →, ∧) a primitive notion of proof, the
rules of sequent calculus, appear, in this framework, as derived rules, saying
something about this primitive notion.
Inconveniences of natural deduction: first, natural deduction is more at
ease with intuitionistic systems; since we deal here with classical system,
one has to use a ¬¬-translation [[. . .]], or a special negation rule, both be-
ing unnatural. Secondly, even for intuitionistic logic, the treatment of the
connectives ∃ and ∨ is problematic [[. . .]]. Thirdly, some essential features of
sequent calculus, for instance the cut-rules [[. . .]] are more complex in that
setting.
Advice: the best thing is perhaps to use natural deduction as a heuristic
guide, in problems where this formulation is the simplest [[. . .]], but when
one needs a rigorous proof, it is better to translate everything in terms of
sequents(*).
Comment. (*) refers to a footnote: “The author confesses that he changed his mind
several times on this question.”

The rule about canceling assumptions [[. . .]] should be understood as
follows. When we make the deduction, we are allowed to cancel φ wherever
it occurs as an assumption. But we are not obliged to; we can cancel some
occurrences of φ and not others, or we can leave it completely uncanceled.
The formula φ may not occur as an assumption anyway, in which case we
can forget about canceling it.
Remark 85.14. It might be helpful to point out here that what Martin-Löf [1971]
calls “rules of contraction” is different from the rules of contraction in Gentzen’s
sequential calculi; they are rules for transforming a given proof figure into another
one.
The next set of quotations is meant to give an idea of what is regarded as the
import of the cut elimination theorem.
Quotations 85.15. (1) Schütte [1980], p. 37.

The first important proof theoretical result was the Hauptsatz of first
order predicate calculus by G. Gentzen (1934). According to this theorem,
every deducible formula of first order predicate calculus is deducible only
by inferences whose premises are combined only by parts of the conclusion.
That means that one can eliminate the inferences which Gentzen called
cuts (Schnitte) whose simplest case is the inference from A and A → B to
B. The Hauptsatz of first order predicate calculus is proved in a finitary
way. It gives the possibility to make some conclusions from the structure of
a formula to the deducibility of the formula.
(2) Prawitz [1971], p. 261, footnote 1.
It is an historical fact that Gentzen’s result when technically summed
up in the Hauptsatz has given rise to many misunderstandings. Since the cut
rule may be looked upon as a generalization of modus ponens or as stating
a transitive law of implication, there has been a not to[[o]] infrequent mis-
conception that the significance of the result was that modus ponens or the
transitive law could be eliminated from proofs. As Kreisel remarked in his
lecture at this symposium, one has then to explain what is dubious about
these old respectable principles. Clearly, this belief is not only superficial
but quite mistaken: modus ponens is present also in normal derivations in
the systems of natural deduction (and it is a triviality that the principle of
transitivity holds in these systems although it does not occur as a primi-

tive rule); it is really also present in cut-free derivations in the calculus of
sequents in the form of introduction of ⊃ in the antecedent.
(3) R. M. Smullyan [1968a], p. 560.
The real importance of cut-free proofs is not the elimination of cuts per se,
but rather that such proofs obey the subformula principle.
Comment. Another real so and so. In view of the cut elimination proof in dialec-
tical logic with unrestricted abstraction it seems worthwhile pointing out that the
subformula property is by no means a consequence of the lack of cuts in a proof.
The next couple of quotation concerns three-valued logic as the appropriate

semantics for cut free sequent calculus.
Quotations 85.16. (1) Girard [1987], p. 161.

[[T]]he semantics of the cut-free sequent calculus is necessarily of a different
nature [[than the two valued semantics defined in 22.33 (1) in the tools]]:
we need a concept of model in which the validity of ⊢ A and A ⊢ B does
not entail the validity of ⊢ B (equivalently A and A → B may be valid,
while B is not valid).
The situation is rendered more delicate by the Hauptsatz : since prov-
ability and cut-free provability coincide, it is hopeless to seek a semantic
concept which should discriminate between these two notions! So we shall
look for a semantic interpretation, not of cut-free provability, but of cut-free
derived rules, i.e. cut-free proofs in sequent calculus with extra axioms (for
which we know that the Hauptsatz fails).
Three-valued semantics answer this question: in such models (Schüt-
te valuations) a formula can be true, false, or undetermined. In a valuation
we shall say that A1 , . . . , An ⊢ B1 , . . . , Bm is valid iff the truth of all Ai ’s
implies the non-falsity of some Bj . It is easy to prove the completeness of
the cut-free derived rules w.r.t. Schütte valuations. The original work of
Schütte introduced this semantic notion to give a semantic interpretation
of Takeuti’s conjecture[[.]]
(2) Girard [1987], p. 162.
[[W]]hat makes the interest of three-valued semantics is that it enables us
to understand what we are doing, when trying to seek a new sequent cal-
culus: even with a good experience of sequent calculus, it is not easy to
determine which rules will yield a cut-elimination theorem, and which rules
will not: but, if one translates our rules in three-valued semantics, things
become easier: one can look for those rules which have an interesting three-
valued interpretation, and only after try to give the syntactic proof of cut-
elimination.
The use of three-valued semantics has generally been overlooked, one
of the reasons lying presumably in a certain (justified) contempt for many-
valued logic. This lack of interest is perfectly unjustified, because, in my
opinion, the three-valued semantics of cut-free rules gives surely the key for
a certain number of applications of proof-theory to other parts of logic and

mathematics.
I close this section with some quotations regarding ordinal analysis.
Quotations 85.17. (1) Pohlers [1989], p. 6.

The fact that induction along the wellordering is the only nonfinitistic
means in Gentzen’s proof also suggests using this wellordering as a measure
for the transfinite content of pure number theory. Pursuing this idea one
had defined the proof theoretic ordinal of a formal theory T as the order-
type of the smallest wellordering which is needed for a consistency proof of
T. This definition, however, is somehow vague since it says nothing about
the means used besides the induction along this wellordering (one tacitly
has to assume that these at least have to be formalizable in T). To obtain
a more precise definition one calls an ordinal α provable in T if there is a
primitive recursively definable wellordering ≺ of ordertype α such that the
wellordering of ≺ is provable in T. It is a consequence of Gödel’s second
theorem, that the proof theoretic of T (in the previous sense) cannot be a
provable ordinal of T. Therefore one may define the proof theoretic ordinal
as the least ordinal which is not provable in T. This is the common defi-
nition today. The computation of the proof theoretic ordinal of T is called
the ordinal analysis of T.
(2) Takeuti [1975], pp. 109 f.
One often says that the consistency of PA is proved by transfinite induction
on the ordinals of proofs, as if we were using a general principle of transfinite
induction in order to prove the consistency of mathematical induction.
This is misleading, however. The point is that the consistency proof
uses the notion of accessibility of ε0 , [[. . .]], and otherwise strictly finitist
methods.
Quotations 85.18. (1) Girard [1987], pp. 439 f.

It seems natural to assign two ordinals to current-theories [[. . .]]. For suf-
ficiently big theories, these two ordinals will be the same. Roughly
Q speaking,
ε0 is the ordinal of PA when viewed from the standpoint of 11 -logic (with
the question of well-foundedness heavily emphasized), whereas η0 is the or-
Q1
dinal of PA from the viewpoint of 2 -logic (here the question of direct
limits plays a central role). Which is the best? We do not have enough
information to judge; a moderate answer is: it depends on the context, on
the applications . . .
(2) Kreisel [1987], p. 400.
Formally speaking, normalization is replaced by so-called pruning; in con-
trast to normalization the end result depends on (a sensible choice of) the
order in which reduction steps are applied.
85d. Cut elimination, completeness, and decidability. It wasn’t until

the ‘fifties of the 20th century that it was realized how easily the sequential for-
mulation lends itself to establish completeness results. In the ‘fifties a number of
people seemed to have discovered this way independently of each other, not all of
them with the close link to sequential calculus.
I begin with a set of quotations regarding early considerations of completeness.
Quotations 85.19. (1) Dreben and van Heijenoort in Feferman et al. [1986], p. 52.
[[A]]ccording to Gödel, the only significant difference between Skolem 1923 a
and Gödel 1929–1930 lies in the replacement of an informal notion of “prov-
able” by a formal one[[.]]
Comment. See also Gödel’s remark in his 1929 dissertation, quotation 78.11 (2) in
these materials.
(2) Wang [1974], p. 8; quoting a letter from Gödel.
The completeness theorem, mathematically, is indeed an almost trivial con-
sequence of Skolem 1922. However, the fact is that, at that time, nobody
(including Skolem himself) drew this conclusion (neither from Skolem 1922
nor, as I did, from similar considerations of his own).
(2) Feferman et al. (eds.) [1986], p. 44.
Frege [[. . .]] never saw completeness as a problem, and indeed almost fifty
years elapsed between the publication of Frege 1879 and that of Hilbert and
Ackermann 1928, where the question of the completeness of quantification
theory was raised explicitly for the first time.
To raise the question of semantic completeness the Frege-Russell-White-
head view of logic as all embracing had to be abandoned and Frege’s notion
of a formal system had to become itself an object of mathematical inquiry
and be subjected to the model-theoretic analyses of the algebraists of logic.
Post (1921 ) and Bernays (1926 ) proved for the propositional calculus a
form of syntactic completeness, namely that, if an unprovable formula is
taken as an additional axiom, the resulting system is inconsistent. [[. . .I]]n
September 1928, at the Bologna congress, Hilbert proposed a similar form of
completeness for number theory. [[. . .]] Hilbert [[[1929]]] shifts from this kind
of completeness for number theory to semantic completeness for quantifica-
tion theory (with identity). A system for such a logic could be obtained, he
remarks, by dropping the number-theoretic axioms and introducing an arbi-
trary number of predicate letters. [[. . .]] He distinguishes formulas “that are
not refutable [widerlegbar] through any definite stipulation of the suitable
predicates. These formulas represent the valid logical propositions”. Hilbert
has now adopted the semantic viewpoint and these “not refutable” formu-
las are those for which there is no falsifying interpretation in any domain.
He then comes to semantic completeness: “the question now arises whether
all these formulas are provable through the rules of logical inference, in-
cluding the so-called identity axioms; whether, in other words, the system
of the usual logical rules is complete”. (This was written after Hilbert and
Ackermann 1928 had already been published.)
Hilbert’s goal of a Post-like completeness for number theory is, as we
now know by Gödel’s incompleteness result, unattainable[[.]]
Quotation 85.20. Feferman et al. (eds.) [1986], p. 19.

Much attention had been given in the 1920s and early 1930s to the decision
problem for satisfiability for various classes of formulas in the first-order
predicate calculus (or restricted functional calculus), mainly those grouped
according to the nature of their prefix in prenex normal form. An informal
notion of decidability had been used in the case of positive results, while
some classes of formulas were shown to be reduction classes for the decision
problem for all formulas. (The exact situation was not clarified until Church
[[[1936]]] used the thesis referred to [[as Church’s thesis]] to establish the
undecidability of full quantificational logic, from which it follows that every
reduction class is also undecidable. [[. . .]])
Quotation 85.21. Wang [1960], p. 228.

The treatment of the predicate calculus by Herbrand and Gentzen en-
ables us to get rid of every “Umweg” (cut or modus ponens) so that we
obtain a cut-free calculus in which, roughly speaking, for every proof each
of the steps is no more complex than the conclusion. This naturally suggests
that we can, given any formula in the predicate calculus, examine all the
less complex formulae and decide whether it is provable. The reason why
this does not yield a decision procedure for the whole predicate calculus is
a rule of contraction which enables us to get rid of repetition of the same
formula. As a result, in searching for a proof or a disproof, we may fail in
some case because we can get no proof no matter how many repetitions
we introduce. In such a case, the procedure can never come to an end al-
though we do not know this at any finite stage. While this situation does
not preclude completeness, it does exclude a decision procedure.
Comment. In other words, classical quantificational logic minus contractions is de-
cidable.32

There is nothing in the usual formulation of predicate calculus which makes
one expect that the calculus is complete. In contrast, from the very con-
struction of the calculus of sequents, it is immediately obvious that it is
complete with respect to logical truth. One seems thus to be justified in
saying that this calculus is the natural formulation of a system intended to
generate the first order logical truths.
Comment. Cf. Russell in quotation 86.4 (2) in the next chapter.
32 I am thankful to Hiroakira Ono for having pointed this passage of Wang’s out to me.
(2) Prawitz [1981], pp. 246 f.

Although Gentzen [[[1934]]] arrived at his calculi of sequents by reflecting
upon the systems of natural deduction and, in the case of classical logic,
generalizing the deducibility relation by allowing several sentences taken
disjunctively as conclusions (which to his surprise, as Gentzen noted in his
paper [[[1938]]], removed certain problems connected with classical nega-
tion), there is an equally interesting aspect of the calculus of sequents for
classical logic, namely the close correspondence between its rules and the
clauses of the Tarskian truth definition, which was perhaps first explicitly
noted by Beth [[[1955]]] in terms of his semantic tableaux. More precisely,
the introduction and elimination rules in the system of natural deduction
become formally rules for introducing sentences in the succedent and an-
tecedent, respectively, of a sequent but can also be read as the conditions
for the truth and falsity, respectively, of sentences appearing as members of
an iterated disjunction. In this ways, Gentzen’s Hauptsatz may be looked
upon as a generalization of the fact that the conditions for the truth and
falsity of a sentence cannot be satisfied simultaneously.
Furthermore, systematic application of the rules of the calculus of se-
quents backwards, so to say, yield because of this correspondence all pos-
sible countermodels to the sequent that one started with, if such counter-
models exist. When they do not, i.e. when the procedure of constructing
countermodels breaks down at a finite stage, this will manifest itself in a
demonstration of the fact that no assignment of truth values that refutes
the sequent is consistent in the sense of not assigning both truth and falsity
to the same sentence. Because of a certain property of duality, this demon-
stration constitutes at the same time a proof of the sequent in question. As
a result, we get an immediate proof of Gödel’s completeness theorem for
first order predicate logic, and as a further corollary, we obtain Gentzen’s
Hauptsatz.
Comment. In the tools, this is captured in the so-called principal lemmata, fol-
lowing Schütte’s approach.
85e. Cut elimination and the ω-rule. Cut elimination and semi formal
systems: cut elimination is possible in arithmetics which have a certain rule of
‘infinite induction’, or ‘ω-rule’, instead of the usual rule of induction.
Quotation 85.23. Raggio [1988], p. 95.
In 1951 Schütte [[[1951]]] succe[[e]]ded in extending the Hauptsatz to
arithmetic, but he was obliged to drop the rule of complete induction in-
troducing a rule of infinite induction with an infinite number of premises.
A proof in Schütte’s system, a so-called semiformal system, is no longer a
finite object but it may be described metamathematically in an exact way
and in a more or less constructive fashion. This line of research culminated
in 1964 with the proofs by Schütte and Feferman that there is an ordinal
of the second number class called Γ0 such that in predicative analysis all
inductions to numbers less than Γ0 are provable but the induction up to
Γ0 is not. This is one of the most profound results obtained in proof-theory

and it follows closely the ideas and techniques discovered by Gentzen.
Quotation 85.24. Wang [1974], pp. 50 f.

There is a temptation to cut through the foundational problems by using
the nonconstructive rule of induction (the omega rule) and similar seman-
tic concepts to characterize all true propositions in arithmetic, classical
analysis, and set theory. In this way, of course, e.g. unnatural numbers are
excluded by the basic principles. However, there is not much explaining left
to be done, since what is to be explained is simply taken for granted. With
the infinite rule, more is accepted which is a projection by analogy of the
finite into the infinite. We can never go through infinitely many steps in a
calculation or use infinitely many premisses in a proof unless we have some-
how succeeded in summarizing the infinitely many with a finite schema in
an informative way. Both mathematical induction and transfinite induction
are principles by which we make inferences after we have found by men-
tal experimentations two suitable premisses which summarize together the
infinitely many premisses needed. A very essential purpose of the mathe-
matical activity is to devise methods by which infinity can be handled by
a finite intellect. The postulation of an infinite intellect has little positive
content except perhaps that it would make the whole mathematical activity
unnecessary.
Comment. May I suggest Ramsey: “our inability to write propositions of infinite
length . . . is logically a mere accident”.33
85f. Proof theory and meaning.
Quotations 85.25. (1) Schütte [1956], p. 55; my translation

The importance of sequential logic and related systems for proof theory
consists in the fact that here the essential logical laws are not buried in ax-
ioms, but find expression in the rules of inference, so that the formula to be
proved is successively deducible from basic components. In this construc-
tive deduction the weakening inferences and the structural inferences play
in general a subordinate role. Thus, if weakenings and structural rules are
possibly not used at all, this should yield a tightness of derivations useful
for proof theoretical investigations.
Comment. I shall clearly object to this view of the structural rules as playing a
subordinate role from a dialectical point of view. Cf. quotations 85.28 below.
(2) Sundholm [1986], p. 488.
The introduction and elimination rules must, so to speak, match, not just
in that each connective has introduction and elimination rules, but also
in that they must not interfere with the previous practice. Hence it seems
natural to let one of the (two classes of) rules serve as meaning-giving and
let the other one be chosen in such a way that it(s members) can be justified
according to the meaning-explanation. Such a method of proceeding would

also take care of the ‘paradox’ of inference: one of the two types of rules
would now serve as the direct meaning-given (because meaning-giving!)
way of learning the truth and the other would serve to provide the indirect
means (in conjunction with other justified rules, of course).
Quotations 85.26. (1) Bull and Segerberg [1984], pp. 24 f.

From a philosophical point of view it should be noted that what we
have above [[downward saturated sets and Hintikka systems]] is not yet a
semantics in any but a combinatorial sense of the word. As in the case of
Carnap — there is of course a close connection between state-descriptions
and a downward saturated set — a real semantics is obtained if possible
worlds are postulated and downward saturated sets are identified as partial
descriptions of them.
Comment. No bull. Real semantics for the RealPhilosopher. Like a real wedding in
a real church, with a real priest — there is no fuss like a real fuss.
(2) Bull and Segerberg [1984], p. 28.
It seems fair to say that a deductive treatment congenial to modal logic is
yet to be found, for Hilbert systems are not suited for the purpose of actual
deduction, and in Hintikka/Kripke systems the alternativeness relation in-
troduces an alien element which, moreover, can become quite unmanageable
in special cases.
The situation has given rise to various suggestions. One is that the
Gentzen format, which works so well for truth-functional operators should
not be expected to work for intensional operators, which are far from truth-
functional. (But then Gentzen works well for intuitionistic logic which is not
truth-functional either.) Another suggestion is that the great proliferation
of modal logics is an epidemy from which modal logic ought to be cured:
Gentzen methods work for the important systems, and the other should
be abolished. ‘No wonder natural deduction does not work for unnatural
systems!’.
85g. Structural rules, computing, and the nature of logic. The quota-
tions in this final section touch upon the core of my own enterprise.
Quotation 85.27. Sundholm [1981], p. 162.

The subformula property is defended from the general requirement on
definitions that they be non-impredicative; if the operational rules for the
constant K fail to have the subformula property, the inductive clause for
the extended relation may fail to yield a well-determined concept.
Quotations 85.28. (1) Hacking [1979], p. 294.

The structural rules embody basic facts about deducibility and obtain even
in a language with no logical constants at all.
Comment. I shall later claim that it is the structural rules which constitute classical
logic. It needs to be mentioned, however, that contraction and exchange do not figure
in Hacking’s list of structural rules.
(2) Sundholm [1981], p. 168.
Hacking’s paper may be summed up by a working proof-theorist as: Con-
sider a logic for which cut- and other elimination results are finitistically
provable. As the rules are supposed to have the subformula property, the
well-known Beth-Hintikka-Kanger-Schütte technique for proving the com-
pleteness of cut-free rules may be applied backwards to read off a semantics.
It is hard to see how such a technical groundwork can be made to hold
the grand superstructure Hacking wishes to erect on its basis.
Comment. That doesn’t raise hopes for my project to be communicable to “a work-
ing proof-theorist”. But luckily I know one or the other “working proof-theorist”
myself and this gives me no reason to accept the label “working proof-theorist” as
an excuse for philosophical obtuseness.
(3) Girard [1989], p. 78.
[[I]]t is not too excessive to say that a logic is essentially a set of structural
rules! The three standard structural rules are all of the form
Γ⊢∆
Γ′ ⊢ ∆′ , more precisely:
α) weakening opens the door for fake dependencies: in that case Γ’ and
∆’ are just extensions of the sequences Γ, ∆. Typically, it speaks of causes
without effect, e.g. spending $1 to get nothing —not even smoke—; but it
is an essential too in mathematics (from B deduce A⇒B) since it allows
us not to use all the hypotheses in a deduction. [[. . .]] It is to be remarked
that this rule has been criticized a long time ago by philosophers in the
tradition of Lewis’s “strict implication”, and has led to various “relevance
logics”, which belong to the philosophical side of logic[[. . . .]]
β) contraction is the fingernail of infinity in propositional calculus: it
says that what you have, you will always keep, no matter how you use it.
Comment. The comments on exchange follow on p. 81 of Girard’s paper.
(4) Girard [1989], p. 81.
γ) exchange expresses the commutativity of multiplicatives: Γ’ and ∆’
are obtained from Γ and ∆ by inner permutations of formulas. It is only for
reasons of expressive power that this rule is still present in the main version
of linear logic: a certain amount of commutativity is needed in order to make
a good use of exponentials.
(5) Girard/Lafont/Taylor [1989], p. 30.
[[C]]ontrary to popular belief, these [[structural]] rules are the most important
of the whole calculus, for, without having written a single logical symbol, we
have practically determined the future behaviour of the logical operations.
Yet these rules, if they are obvious from the denotational point of view,
should be examined closely from the operational point of view, especially

the contraction.

The first important methodological contradiction lies in the oppositions
dynamic / static
sense / denotation
finite / infinite
these three oppositions are different aspect[[s]] of the same problem.
Let us start with Frege, who distinguished between sense and denota-
tion: if we take the sentence
Erich von Stroheim is the author of Greed
“Erich von Stroheim” and “the author of Greed ” have the same denota-
tion, i.e. represent the same external object, but have not the same sense
(otherwise it would be pointless to state such a sentence). Denotationally
speaking the two expressions refer to the same thing, whereas one has to
check something (look at a dictionary, make a proof, a computation) to
relate their two distinct senses. This is why
denotation is static, sense is dynamic.
The only extant mathematical semantics for computation are denota-
tional, i.e. static. This is the case for the original semantics of Scott [[. . .]],
which dates back to 1969, and this remains true for the more recent co-
herent semantics of the author [[. . .]]. These semantics interpret proofs as
functions, instead of actions. But computation is a dynamic process, anal-
ogous to — say — mechanics. The denotational approach to computation
is to computer science what statics is to mechanics: a small part of the
subject, but a relevant one. The fact that denotational semantics is kept
constant during a computational process should be compared to the exis-
tence of static invariants like mass in classical mechanics. But the core of
mechanics is dynamics, where other invariants of a dynamical nature, like
energy, impulsion etc. play a dominant role. Trying to modelize programs
as actions is therefore trying to fill the most obvious gap in the theory.
There is no appropriate extant name for what we are aiming at: the name
“operational semantics” has been already widely used to speak of step-by-
step paraphrases of computational processes, while we are clearly aiming
at a less ad hoc description. This is why we propose the name
geometry of interactions
for such a thing.
(2) Girard/Lafont/Taylor [1989], pp. 3 f.
[[I]]t is impossible to say “forget completely the denotation and concentrate
on the sense”, for the simple reason that the sense contains the denotation,
at least implicitly. So it is not a matter of symmetry. In fact there is hardly
any unified syntactic point of view, because we have never been able to
give an operational meaning to this mysterious sense. The only tangible
reality about sense is the way it is written, the formalism; but the formalism
remains an unaccommodating object of study, without true structure, a
piece of soft camembert.
Does this mean that the purely syntactic approach has nothing worth-
while to say? Surely not, and the famous theorem of Gentzen of 1934 shows
that logic possesses some profound symmetries at the syntactical level (ex-
pressed by cut-elimination). However these symmetries are blurred by the
imperfections of syntax. To put it in another way, they are not symmetries
of syntax, but of sense. For want of anything better, we must express them
as properties of syntax, and the result is not very pretty.
So summing up our opinion about this tradition, it is always in search
of its fundamental concepts, which is to say, an operational distinction
between sense and syntax. Or to put these things more concretely, it aims
to find deep geometrical invariants of syntax: therein is to be found the
sense.
The tradition called “syntactic” — for want of a nobler title — never
reached the level of its rival [[algebraic tradition]]. In recent years, during
which the algebraic tradition has flourished, the syntactic tradition was not
of note and would without doubt have disappeared in one or two more
decades, for want of any issue or methodology. The disaster was averted
because of computer science — that great manipulator of syntax — which
posed it some very important theoretical problems.
Some of these problems (such as questions of algorithmic complexity)
seem to require more the letter than the spirit of logic. On the other hand
all the problems concerning correctness and modularity of programs appeal
in a deep way to the syntactic tradition, to proof theory. We are led, then,
to a revision of proof theory, from the fundamental theorem of Herbrand
which dates back to 1930. This revision sheds a new light on those areas
which one had thought were fixed forever, and where routine had prevailed
for a long time.
(3) Martí-Oliet and Meseguer [1989], p. 321.
Girard’s linear logic [[. . .]] presents itself explicitly as a logic of concurrent
interaction in which resources are limited, and are consumed in such in-
teractions. This is of course very similar to Petri nets, where resources
are represented as tokens that are then consumed by transitions. At the
proof-theoretic level, limitation of resources is expressed by forbidding the
structural rules of weakening (which allows new resources to be obtained for
free) and of contraction (which arbitrarily eliminates duplicated resources)
Part E
OUTSIDE THE
TRANSCENDENTAL TRADITION
Das Denken verzweifelnd,
a u s s i ch auch
die Auflösung des Widerspruchs,
in den es sich selbst gesetzt,
leisten zu können,
kehrt zu den Auflösungen
und Beruhigungen zurück,
welche dem Geiste
in andern seiner Weisen und Formen
zu Theil geworden sind.1
1 Hegel [1830], p. 56. A translation can be found in quotation 66.7 (2) in these materials.
PART E
OUTSIDE THE TRANSCENDENTAL TRADITION
The present part E, third part of my background materials for dialectical logic
is less structured than the two foregoing parts; more to the point, it is devoted
to the leftovers, i.e., those topics which were beyond the scope of transcendental
idealism or an historical outline of the philosophy of mathematics, but which still
provide valuable material for my approach.
1201
CHAPTER XXII
ANALYTIC EMPIRICISM
I am in the position of a Jewish chef

preparing ham for a gentile clientele.
Analyticity, essence, and modality
are not my meat.1
For someone like me, who has little sympathy for any sort of philosophy that adorns
itself with the adjective ‘analytic’, and even less for any sort of doctrinal empiricism,
the task of giving an account of some of the topics that prevail in a philosophical
tradition labeled here “analytic empiricism” 2 is aptly described in Quine’s remark
above. But there are these issues — like the analytic-synthetic distinction, problems
with reference and truth, and the distinction of substitutional opacity and referen-
tial transparency, and similar ones — where philosophers in the analytic-empiricist
tradition have created a paradigmatic jumble of ideas — much more illuminating
than the mumbo jumbo of philosophers in the Hegelian tradition ever was — while,
at the same time, they seemed to remain blissfully unaware of the kind of Pando-
ra’s box they had opened. Analytic empiricists can’t get away from the problems
that dominated the more explicitly metaphysical thinking, but the failure of their
attempts is instructive with regard to my endeavour, and if it is only for the pur-
pose of pointing out some endemic misunderstandings (not at all limited to analytic
empiricism).
The issues that interest me in particular are those grouping around the catch
words
reference,
denoting (description), and
identity (substitutivity salva veritate).
§ 86. From logicism to logical empiricism
Although Frege’s attempt at a logical foundation of arithmetic was flawed with

antinomies, it was to gain an enormous influence on philosophy, mainly through
the new assessment of the nature of logic and mathematics provided by Frege’s and
Russell’s investigations.
1 Quine [1977], p. 116.
2 The label analytic empiricism is due to Wang [1985].
1202
§ 86. FROM LOGICISM TO LOGICAL EMPIRICISM 1203
86a. From logicism to logical atomism. This section somewhat continues

from section 75d, taking the issue of denoting as one of the central points.

The subject of denoting is of very great importance, not only in logic
and mathematics, but also in theory of knowledge. For example, we know
that the centre of mass of the solar system at a definite instant is some
definite point, and we can affirm a number of propositions about it; but
we have no immediate acquaintance with this point, which is only known
to us by description. The distinction between acquaintance and knowledge
about is the distinction between the things we have presentations of, and
the things we only reach by means of denoting phrases. [[. . .]] All thinking
has to start from acquaintance; but it succeeds in thinking about many
things with which we have no acquaintance.
Comment. Do I detect a certain similarity to Kant’s formulation? What is missing,
however, is Kant’s insistence that all knowledge has to relate back to Anschauung.
The latter (relating back) seems an endemic omission in British empiricism,3 if not
a characteristic feature of Anglo-Saxon philosophy altogether.
(2) Russell [1905], p. 45.
[[There are]] difficulties which seem unavoidable if we regard denoting phra-
ses as standing for genuine constituents of the propositions in whose verbal
expressions they occur. Of the possible theories which admit such con-
stituents the simplest is that of Meinong. This theory regards any grammat-
ically correct denoting phrase as standing for an object. Thus ‘the present
King of France’, ‘the round square’, etc., are supposed to be genuine ob-
jects. It is admitted that such objects do not subsist, but nevertheless they
are supposed to be objects This is in itself a difficult view; but the chief
objection is that such objects, admittedly, are apt to infringe the law of
contradiction. It is contended, for example, that the existent present King
of France exists, and also does not exist; that the round square is round,
and also not round, etc. But this is intolerable; and if any theory can be
found to avoid this result, it is surely to be preferred.
Comment. Dialetheists were happy to accommodate this difficulty; cf. Priest and
Routley [1983] (On paraconsistency). The passage is also interesting with regard to
Kant, quotation 62.7 (1) in these materials.
(3) Russell [1905], p. 46.
One of the first difficulties that confront us, when we adopt the view
that denoting phrases express a meaning and denote a denotation, concerns
the cases in which the denotation appears to be absent. If we say ‘the King
of England is bald’, that is, it would seem, not a statement about the
complex meaning ‘the King of England’, but about the actual man denoted
by the meaning. But now consider ‘the King of France is bald’. By parity
of form, this also ought to be about the denotation of the phrase ‘the King
3 Cf. my notion of “Hume’s fallacy” in remark 118.12 (1) in the groundworks.
1204 XXII. ANALYTIC EMPIRICISM
of France’. But this phrase, though it has a meaning provided ‘the King
of England’ has a meaning, has certainly no denotation, at least in any
obvious sense. Hence one would suppose that ‘the King of France is bald’
ought to be nonsense; but it is not nonsense, since it is plainly false.
(4) Russell [1905], pp. 47 f.
A logical theory may be tested by its capacity for dealing with puzzles,
and it is a wholesome plan, in thinking about logic, to stock the mind with
as many puzzles as possible, since these serve much the same purpose as
is served by experiments in physical science. I shall therefore state three
puzzles which a theory as to denoting ought to be able to solve; and I shall
show later that my theory solves them.
(1) If a is identical with b, whatever is true of the one is true of the other,
and either may be substituted for the other in any proposition without
altering the truth or falsehood of that proposition. Now George IV wished
to know whether Scott was the author of Waverley; and in fact Scott was
the author of Waverley. Hence we may substitute Scott for the author of
‘Waverley’, and thereby prove that George IV wished to know whether Scott
was Scott. Yet an interest in the law of identity can hardly be attributed
to the first gentleman of Europe.
(2) By the law of excluded middle, either ‘A is B’ or ‘A is not B’ must
be true. Hence either ‘the present King of France is bald’ or ‘the present
King of France is not bald’ must be true. Yet if we enumerated the things
that are bald, and then the things that are not bald, we should not find the
present King of France in either list. Hegelians, who love a synthesis, will
probably conclude that he wears a wig.
(3) Consider the proposition ‘A differs from B’. If this is true, there
is a difference between A and B, which fact may be expressed in the form
‘the difference between A and B subsists’. But if it is false that A differs
from B, then there is no difference between A and B, which fact may be
expressed in the form ‘the difference between A and B does not subsist’.
But how can a non-entity be the subject of a proposition? ‘I think, therefore
I am’ is no more evident than ‘I am the subject of a proposition, therefore
I am’, provided ‘I am’ is taken to assert subsistence or being, not existence.
Hence, it would appear, it must always be self-contradictory to deny the
being of anything; but we have seen, in connexion with Meinong, that to
admit being also sometimes leads to contradictions. Thus if A and B do not
differ, to suppose either that there is, or that there is not, such an object
as ‘the difference between A and B’ seems equally impossible.
Comment. The first half of the sentence before the last reminds me strongly of
Parmenides.

In ‘Scott is the author of Waverley’ the ‘is’, of course, expresses identity,
i.e., the entity whose name is Scott is identical with the author of Waverley.
But, when I say ‘Scott is mortal’ this ‘is’, is the ‘is’ of predication, which
is quite different from the ‘is’ of identity. It is a mistake to interpret ‘Scott

is mortal’ as meaning ‘Scott is identical with one among mortals’, because
(among other reasons) you will not be able to say what ‘mortals’ are except
by means of the propositional function ‘x is mortal’, which brings back the
‘is’ of predication. You cannot reduce the ‘is’ of predication to the other
‘is’. But the ‘is’ in ‘Scott is the author of Waverley is the ‘is’ of identity
and not of predication.*
Comment. There is a footnote to * which I do not want to withhold from the reader.
*The confusion of these two meanings of ‘is’ is essential to the Hegelian
conception of identity-in-difference.
The issue of a confusion of predication and identity is also raised in Russell [1914],
p. 39, fn. 1, where Russell says:
Hegel’s argument in this portion of his “Logic” depends throughout upon
confusing the “is” of predication [[. . .]] with the “is” of identity[[.]]
This confusion business is a relevant issue in the context of my enterprise. I do not
simply dismiss Russell’s confusion theory, but I am more inclined to place it in the
realm of objects and their (exclusive) singletons.4 In chapter XXXI, I start with a
rudimentary form of inclusion as the basic logical constant. On the basis of that
notion of conclusion, adding identity is logically equivalent to adding predication:
each can be defined in terms of the other.
[[M]]any of the stock philosophical arguments about mathematics (derived
in the main from Kant) had been rendered invalid by the progress of math-
ematics in the meanwhile. Non-Euclidean geometry had undermined the
argument of the transcendental aesthetic. Weierstrass had shown that the
differential and integral calculus do not require the conception of the in-
finitesimal, and that, therefore, all that had been said by philosophers on
such subjects as the continuity of space and time and motion must be re-
garded as sheer error. Cantor freed the conception of infinite number from
contradictions, and thus disposed of Kant’s antinomies as well as many of
Hegel’s. Finally Frege showed in detail how arithmetic can be deduced from
pure logic, without the need of any fresh ideas or axioms, thus disproving
Kant’s assertion that ‘7 + 5 = 12’ is synthetic—at least in the obvious
interpretation of that dictum.
Comment. I find something curiously twisted in this passage; I feel like I have been
told only half of the story. Sure, Weierstrass got rid of the infinitesimal, but at
the price of infinite sets; sure, Cantor freed the conception of infinite number from
contradictions, but at the price of new contradictions; in fact, contradictions which
are much closer in character to those of Kant than the contradictions he got rid
of.5 Sure, Frege showed how arithmetic can be deduced from pure logic, but at
4 In view of the specific way in which predication is definable in terms of identity and inclusion,
Žižek’s confusion of predication and inclusion in quotation 95.12 is a prime example of what I have
in mind.
5 Cf. Zermelo’s comment in quotation 68.27 (2) in these materials.
the price of antinomies. In particular the relevance to the philosophical doctrines

mentioned is not as simple and clear as might seem from this passage. Anyway,
Russell continues, as documented in the next quotation.
(2) Russell [1924], p. 325.
Frege’s work was not final, in the first place because it applied only
to arithmetic, not to other branches of mathematics; in the second place
because his premises did not exclude certain contradictions to which all
past systems of formal logic turned out to be liable. Dr. Whitehead and
I in collaboration tried to remedy these two defects, in Principia Math-
ematica, which, however, still falls short of finality in some fundamental
points (notably the axiom of reducibility). But in spite of its shortcomings
I think that no one who reads this book will dispute its main contention,
namely, that from certain ideas and axioms of formal logic, by the help of
the logic of relations, all pure mathematics can be deduced, without any
new undefined idea or unproved propositions.
Comment. Frege did not regard geometry, for instance, as analytic, witness quo-
tation 71.26 (1). In this context, cf. also Fraenkel in quotation 81.19 (1) in these
materials.
Quotations 86.4. (1) Russell [1937], p. ix.

There are three questions in regard to logical constants: First, are there
such things? Second, how are they defined? Third, do they occur in the
propositions of logic? Of these questions, the first and third are highly
ambiguous, but their various meanings can be made clearer by a little dis-
cussion.
First: are there logical constants? There is one sense of this question in
which we can give a perfectly definite affirmative answer: in the linguistic
or symbolic expression of logical propositions, there are words or symbols
which play a constant part, i.e., make the same contribution to the sig-
nificance of propositions wherever they occur. Such are, for example, “or,”
“and,” “not,” “if-then,” “the null-class,” “0,” “1,” “2,” . . .
Comment. For those who want to know my basic line of answer (to be pursued
in the groundworks): the logical constant of negation, for instance, is the set
{h∅, Vi, hV, ∅i}; accordingly the question regarding its “existence” shifts to the ques-
tion regarding the “existence” of this particular set (resp., concept). This will not
settle the question of “existence”, but point to the irrelevance of the question: if it
doesn’t “exist”, make one yourself — just as the manufacturer of the CPU in your
computer has done for you.
(2) Russell [1937], p. xii.
Given a set of logical premisses, we can define logic, in relation to them,
as whatever they enable us to demonstrate. But (1) it is hard to say what
makes a proposition true in virtue of its form; (2) it is difficult to see any
way of proving that the system resulting from a given set of premisses
is complete, in the sense of embracing everything that we would wish to

include among logical propositions.
Comment. As regards the question of completeness, compare Prawitz in quotation
85.22 (1) and (2).
(3) Russell [1919], p. 169; quoted after Wang [1986], p. 112.
Logic is concerned with the real world just as truly as zoology, though with
its more abstract and general features.
Comment. According to Wang, loc. cit., this was Gödel’s favorite sentence.
(4) Russell [1950], p. 370.
[[T]]he scholastics used to say ‘One and Being are convertible terms’. It now
appears that ‘one’ is a predicate of concepts, not of the things to which the
concepts are applicable; the predicate ‘one’ applies to ‘satellite of the earth’
but not to the moon. And for other reasons ‘being’ applies only to certain
descriptions, never to what they describe. These distinctions, it will be
found, put an end to many arguments of metaphysicians from Parmenides
and Plato to the present day.
Comment. My point is that these distinctions, with the problems they cause for
classical logic, vindicate many of the questions Parmenides and Platon asked; in
particular regarding being — it just depends on how you define being.
Quotations 86.5. (1) Wittgenstein [1921], p. 7.

1 The world is all that is the case.
1.1 The world is the totality of facts, not of things.
[[. . .]]
1.2 The world divides into facts.
1.21 Each item can be the case or not the case while everything
remains the same.
Comments. (α) 1.1 sounds to me like saying that the world is a set of relations, not
of first order objects. In section 111a in the groundworks, I shall treat the world
as the ordered pair consisting of both, objects and their relations.
(β) 1.2 sounds to me like saying: a set is partitioned by equivalence classes. Actually,
one such partition is that of true and false sentences, at least according to classical
logic.
(γ) 1.21 reminds me of Cantor’s phrasing in terms of “Vielheiten unverbundener
Dinge”.6 Hegel speaks of a position that views objects “in ihrer Vereinzelung als
selbstständig”,7 translated as severally independent.8
(2) Wittgenstein [1921], p. 11.
2.021 Objects make up the substance of the world.
2.024 Substance is what subsists independently of what is the case.
6 Cf. quotation 73.13 in these materials
7 Hegel [1830], p. 135.
(4) Wittgenstein [1921], pp. 121, 127, 131.

6.1 The propositions of logic are tautologies.
6.11 Therefore the propositions of logic say nothing. (They are the
analytic propositions.)
6.123 Clearly the laws of logic cannot in their turn be subject to the
laws of logic.
6.1261 In logic process and result are equivalent. (Hence the absence
of surprise.)
6.1262 Proof in logic is merely a mechanical expedient to facilitate the
recognition of tautologies in complicated cases.
6.127 [[. . .]]
Every tautology itself shows that it is a tautology.
Comment. This pretty much sums it up, what I regard as the basic insensitivity of
the classical logicist position towards the epistemological relevance of the conceptual
framework. Mind you, it was published fifteen years before Church’s proof of the
undecidability of classical first order logic. I can’t help but thinking of the surprise
paradox. Note also the “merely” (German: “nur”).
(5) Wittgenstein [1921], pp. 133.
6.2 Mathematics is a logical method.
6.21 A proposition of mathematics does not express a thought.
6.232 Frege says that the two expressions [[of a non-trivial equation]]
have the same meaning but different senses.
But the essential point about an equation is that it is not nec-
essary in order to show that the two expressions connected by the
sign of equality have the same meaning since this can be seen from
the two expressions themselves.
Comment. Contrast 6.21 with Frege in quotations 71.25.
4.241 When I use two signs with one and the same meaning [[Bedeu-
tung]], I express this by putting the sign ‘=’ between them.
So ‘s = b’ means that the sign ‘b’ can be substituted for the
sign ‘a’.
[[. . .]]
4.242 Expressions of the form ‘a = b’ are, therefore, mere representa-
tional devices. They state nothing about the meaning of the signs ‘a’
and ‘b’.
Comment. Here is another mere so and so.
5.53 Identity of object I express by identity of sign, and not by using
a sign for identity. Difference of objects I express by difference of signs.
5.5301 It is self-evident that identity is not a relation between objects.
[[. . .]]
[[. . .]]
5.5302 Russell’s definition of ‘=’ is inadequate, because according to it

we cannot say that two objects have all their properties in common.
(Even if this proposition is never correct, it still has sense.)
5.5303 Roughly speaking, to say of two things that they are identical
is nonsense, and to say of one thing that it is identical with itself is
to say nothing at all.
Comment. The notion of identity is a central issue in my approach to dialectical
logic, but I can’t say that Wittgenstein’s assertions make much sense to me. I will
certainly have a sign for identity, and that in accordance with Leibniz’ definition,9
be that primitive or defined. Accordingly, I can express something like a ≡ a.
It is quite impossible for a proposition to state that it itself is true.
Comment. Hence Wittgenstein’s objection to Gödel’s result.10
3.332 No proposition can make a statement about itself, because a
propositional sign cannot be contained in itself (that is the whole of
the ‘theory of types’).
3.333 The reason why a function cannot be its own argument is that
the sign for a function already contains the prototype of its argument,
and it cannot contain itself.
[[. . .]]
That disposes of Russell’s paradox.
5.47
[[. . .]]
One could say that the sole logical constant was what all propo-
sitions, by their very nature, had in common with one another.
But that is the general propositional form.
5.471 The general propositional form is the essences of a proposition.
5.4711 To give the essence of a proposition means to give the essence
of all description, and thus the essence of the world.
5.472 The description of the most general propositional form is the
description of the one and only general primitive sign in logic.
Comment. The least I would have hoped for, in view of Frege’s work, would have
been a consideration regarding the extension of the descriptive phrase “what all
propositions, by their very nature, had in common with one another”; otherwise I
don’t see what distinguishes a phrase like “the sole logical constant” from the phrase
“the will of the people”.11
9 Cf. 70.10 in these materials.

11 I am grateful to A. F. Koch for having attracted my attention to this passage in Wittgen-
stein’s Tractatus.

There are, indeed, things that cannot be put into words. They make
themselves manifest. They are what is mystical.
Comment. The English translation strikes me as pretty barbaric, so I add the Ger-
man original: “Es gibt allerdings Unaussprechliches. Dies zeigt sich, es ist das Mys-
tische.”
What we cannot speak about we must pass over in silence.
My propositions serve as elucidations in the following way: anyone who
understands me eventually recognizes them as nonsensical, when he has
used them—as steps—to climb up beyond them. (He must, so to speak,
throw away the ladder after he has climbed up it.)
Comment. My misgivings about throwing away the ladder: am I not going to need
it anymore? Or am I supposed to have an endless supply of ladders? Dedekind spoke
of the “Treppennatur unseres Verstandes” and as a philosopher who has taken to
mathematics I have come to value the process of iteration. It sounds to me like a
suggestion to throw away the successor operation after we have made the step from
zero to one. It seems to me that Wittgenstein is misjudging the nature of under-
standing in that he seems to think that we can reach a point where the thoughts
that have got us there can be discarded. I regard this as a typical philosophical lim-
itation, a closed mind phenomenon. In short: it strikes me as terribly undialectical.
I want to establish a theory, where Wittgenstein celebrates the mystical.
(4) Russell [1922], p. xx f.12
Mr Wittgenstein’s attitude towards the mystical [[. . .]] grows naturally out
of his doctrine in pure logic, according to which the logical proposition
is a picture (true or false) of the fact, and has in common with the fact
a certain structure. It is this common structure which makes it capable
of being a picture of the fact, but the structure cannot itself be put into
words, since it is a structure of words, as well as of the facts to which
they refer. Everything, therefore, which is involved in the very idea of the
expressiveness of language must remain incapable of being expressed in
language, and is, therefore, inexpressible in a perfectly precise sense. This
inexpressible contains, according to Mr Wittgenstein, the whole of logic
and philosophy. The right method of teaching philosophy, he says, would
be to confine oneself to propositions of the sciences, stated with all possible
clearness and exactness, leaving philosophical assertions to the learner, and
proving to him, whenever he made them, that they are meaningless. It
is true that the fate of Socrates might befall a man who attempted this
method of teaching, but we are not to be deterred by that fear, if it is the
only right method. It is not this that causes some hesitation in accepting
Mr Wittgenstein’s position, in spite of the very powerful arguments which
12 This quotation was instigated by Priest [1995], p. 210.
he brings to its support. What causes hesitation is the fact that, after all,
Mr Wittgenstein manages to say a good deal about what cannot be said,
thus suggesting to the sceptical reader that possibly there may be some
loophole through a hierarchy of languages, or by some other exit.
Comment. Cf. also Carnap, quotation 86.17 (5) below.

I am trying as far as possible [[. . .]] to start with perfectly plain truisms.
My desire and wish is that the things I start with should be so obvious
that you wonder why I spend my time stating them. That is what I aim at,
because the point of philosophy is to start with something so simple as not
to seem worth stating, and to end with something so paradoxical that no
one will believe it.
Comment. Lovely, isn’t it? Usually, it seems the other way round: philosophers
come forward with the most paradoxical assumptions to explain what they consider
obvious. But then there is always the question, what is a perfectly plain truism?
(2) Carnap [1928], p. 108.

[[W]]e have to proceed from that which is epistemically primary, that is to
say, from the “given”, i.e., from experiences themselves in their totality and
undivided unity.
Comment. This strikes me as quite a bit of a metaphysical construction, “experi-

ences themselves”. At some stage I was wondering what it was in the German orig-
inal, but I couldn’t find the energy to spend my time on an issue as silly as that.
Compare what Carnap says about metaphysical statements in quotation 86.17 (1)
below. The question would be whether “experiences in themselves in their totality
and undivided unity” are a fiction or not. Probably Carnap was a bit of a saint and
he could have “experiences in themselves in their totality and undivided unity” —
like Jeanne d’Arc, who could hear God’s voice, or Augustine, who could remember
how it was to learn his first words.13 But, bloody hell, I’m not a saint.
86b. The philosophical significance of logicism for empiricism. About

seventy years ago, a bunch of clever empiricists saw in logicism a chance to get
rid of an old problem for their position: the apriorical nature of mathematics. This
gave rise to what is generally referred to as logical empiricism, or sometimes logical
positivism. This section deals with the period starting in the late twenties, predom-
inantly in Vienna (‘Vienna circle’), around Hans Hahn and Rudolf Carnap until
the middle of the ‘thirties, when after the death of Hans Hahn in 1934 and the
assassination of Moritz Schlick in 1936, the Vienna Circle broke up. By the time of
the “Anschluß”,14 it seems, the “Kreis” was no longer intact.

14 Annexation of Austria by Germany, 11 March 1938.
Quotations 86.8. (1) Hahn [1929], p. 47;

The fundamental thesis of empiricism is that experience is the only source
capable of furnishing us with knowledge of the world, knowledge of facts,
knowledge that has content: all such knowledge originates in what is im-
mediately experienced. The place of mathematics has always presented a
great difficulty for this position; for experience cannot provide us with uni-
versal knowledge, but mathmatical knowledge seems to be universal; all
knowledge originating in experience comes with a coefficient of uncertainty
affixed to it, but in mathematics we notice no uncertainty.
Comment. The point which I find dubious is the inclusion of all knowledge that has
content. The knowledge that Gödel’s sentence is true has content without having
its origin in experience.
(2) Wang [1985], p. 451.
How is empiricism to give an adequate account of the certainty, clarity,
range, and applicability of mathematics?
Comment. Note the close relationship to the question that Kant was concerned
about, certainty and applicability, and keep in mind Frege.
(3) Carnap [1963], p. 47; quoted after Wang [1985], p. 451, in a footnote to the
foregoing quotation .
What was important in this conception from our point of view was the
fact that it became possible for the first time to combine the basic tenet
of empiricism with a satisfactory explanation of the nature of logic and
mathematics.
Quotations 86.9. (1) Hahn [1933], p. 227.

The old conception of logic is approximately as follows: logic is the account
of the most universal properties of things, the account of those properties
which are common to all things; just as ornithology is the science of birds,
zoology the science of all animals, biology the science of all living beings,
so logic is the science of all things, the science of being as such. If this
were the case, it would remain wholly unintelligible whence logic derives
its certainty. For we surely do not know all things. We have not observed
everything and hence we cannot know how everything behaves.
Our thesis, on the contrary, asserts: logic does not by any means treat of
the totality of things, it does not treat of objects at all but only of our way
of speaking about objects; logic is first generated by language. The certainty
and universal validity, or better, the irrefutability of propositions of logic
derives just from the fact that it says nothing about objects of any kind.
(2) Hahn [1931], pp. 136 f; my translation.15
Given any domain of objects, which are in any sort of relations to each
other; let this domain be mapped onto a co-domain in such a way that the
15 This is also translated in McGuiness [1980], p. 33.
objects and relations of the original domain correspond to objects and re-
lations of the co-domain; we can then take the objects and relations of the
co-domain as symbols for the objects and relations of the original domain. If
the mapping thus carried out is not one-one relational but one-many, then
one and the same state of affair in the original domain will correspond to
different complexes of symbols in the co-domain, thus there will be trans-
formations of the symbolism with regard to itself, and the task arises to
give rules for the transformation of one complex of symbols into another
one which maps the same state of affair of the original domain. Such is the
way, according to my view, the language faces reality: the language assigns
complexes of symbols to the states of affairs of the world, and that not in
a one-one relational way (which would be of little use), but in a one-many
relational way; and the logic gives the rules, how a complex of symbols of
the language can be transformed into another one which denotes the same
state of affairs: that is what is called the “tautological” character of logic;
[[. . .]]
Thus logic does not say anything at all about the world, but only relates
to the way I talk about the world, and it will be clear that under this view
the existence of logic as doctrine of the most general properties of objects
is quite compatible.
Quotations 86.10. (1) Hahn [1931], p. 137; my translation.16

The logicist position [[. . .]] claims that there is no difference between mathe-
matics and logic. If this position is feasible, then with the elucidation given
above of the status of logic in the system of our knowledge, the status of
mathematics is also elucidated; like the existence of logic, the existence of
mathematics too is then compatible with the empiricist position.
Comment. Recall that it was Hegel who wanted to have metaphysics dissolved in
logic. In view of this, see paraphrase 108.6 (1) in part F for my speculative twist of
this quotation.
(2) Hahn [1933], p. 232.
The propositions of mathematics are of exactly the same kind as the propo-
sitions of logic: they are tautologous, they say nothing at all about the
objects we want to talk about, but concern only the manner in which we
want to speak of them. The reason why we can assert apodeictically with
universal validity the proposition: 2+3 = 5, why we can say even before any
observations have been made, and can say it with complete certainty, that
it will not turn out that 2 + 3 = 7, is that by “2 + 3” we mean the same as
by “5”—just as we mean the same by “helleborus niger” as by “snow rose.”
For this reason no botanical investigation, however subtle, could disclose
that an instance of the species “snow rose” is not a helleborus niger. We
become aware of meaning the same by “2 + 3” and by “5,” by going back to
the meanings of “2,” “3,” “5,” “+,” and making tautological transformations
16 Cf. McGuiness [1980], p. 34.
until we just see that “2 + 3” means the same as “5.” It is such successive
tautological transformation that is meant by “calculation”[[. . .]]
To be sure, the proof of the tautological character of mathematics is
not yet complete in all details. This is a difficult and arduous task; yet we
have no doubt that the belief in the tautological character of mathematics
is essentially correct.
Comment. Two problems remain: (i) how to establish the meaning of “2,” etc.
and “+”, and (ii) how to establish that all of mathematics can be reduced to the
tautological character of its propositions.
(3) Russell [1946], p. 784.

From Frege’s work it followed that arithmetic, and pure mathematics
generally, is nothing but a prolongation of deductive logic. This disproved
Kant’s theory that arithmetical propositions are ‘synthetic’ and involve a
reference to time. The development of pure mathematics from logic was set
forth in detail in Principia Mathematica, by Whitehead and myself.
Comment. Contrast Whitehead and Russell in quotation 75.17 in these materials

(regarding the axiom of infinity); compare also: Fraenkel in quotation 81.19 (2) in
these materials (regarding the incompleteness of Frege’s work with regard to all
of mathematics), Hilbert in quotation 76.11 (1) (regarding to the non-logical nature
of the axioms of infinity and reducibility), and, more recently, Hodes in quotation
87.11 (3) in the present chapter.

‘logical positivism’ is a name for a method, not for a certain kind of
result. A philosopher is a logical positivist if he holds that there is no special
way of knowing that is peculiar to philosophy, but that questions of fact
can only be decided by the empirical methods of science, while questions
that can be decided without appeal to experience are either mathematical
or linguistic.
(2) Russell [1950], pp. 367 f.
What is distinctive [[for the logical positivist as compared to earlier empiri-
cists]] is attention to mathematics and logic, and emphasis upon linguis-
tic aspects of traditional philosophical problems. British empiricists, from
Locke to John Stuart Mill, were very little influenced by mathematics, and
had even a certain hostility to the outlook engendered by mathematics. On
the other hand, Continental philosophers, down to Kant, regarded mathe-
matics as a pattern to which other knowledge ought to approximate, and
thought that pure mathematics, or a not dissimilar type of reasoning, could
give knowledge as to the actual world. The logical positivists, though they
are as much interested in and influenced by mathematics as Leibniz or
Kant, are complete empiricists, and are enabled to combine mathematics
with empiricism by a new interpretation of mathematical propositions. It
was work on the foundations and on mathematical logic that gave the tech-
nical basis for the school, and without some understanding of this basis it
is impossible to do justice to the grounds for their opinions.
Comment. The “certain hostility” seems to have got the better of later British phi-
losophy again. Unfortunately, the tradition of high esteem for mathematics amongst
Continental philosophers seems to have long gone. Where would there be a Kantian
or Hegelian philosopher today who could claim to have acquired mathematical and
logical literacy? Anyway, I feel very comfortable in Russell’s company, as far as this
quotation goes; the point then is, for me, to take the failure of the “work on the
foundations and on mathematical logic that gave the technical basis” for logical
positivism seriously.
(3) Russell [1950], p. 375.
There is a theory that the meaning of a proposition consists in its
method of verification. It follows (a) that what cannot be verified or falsified
is meaningless, (b) that two propositions verified by the same occurrences
have the same meaning.
I reject both, and I do not think that those who advocate them have
fully realized their implication.
First: practically all the advocates of the above view regard verification
as a social matter. This means that they take up the problem at a late
stage, and are unaware of its earlier stages. Other people’s observations are
not data for me. The hypothesis that nothing exists except what I perceive
and remember is for me identical, in all its verifiable consequences, with the
hypothesis that there are other people who also perceive and remember. If
we are to believe in the existence of these other people—as we must do if
we are to admit testimony—we must reject the identification of meaning
with verification.
86c. Antinomies, the doctrine of types, and logical spheres. The point
is: logicism has failed; so what do logical empiricists do? One dominant aspect is
the dissection of language: object and meta-language.
Quotation 86.12. Carnap [1931a], p. 184.
Ramsey has shown that there are two completely different kinds of
antinomies. [[. . .]]
[[. . .]] Ramsey has shown that antinomies of this second kind cannot be
constructed in the symbolic language of logic and therefore need not be
taken into account in the construction of mathematics from logic.
Quotations 86.13. (1) Carnap [1928], pp. 51 f.
Two objects [[. . .]] are said to be isogenous if there is an argument position in
any propositional function for which the two object names are permissible
arguments. If this is the case, then it holds for any argument position of any
propositional function either that both names are permissible arguments,
or that neither of them is. This is a consequence of the logical theory of
types[[.]]
[[. . .]]
By the sphere of an object we mean the class of all objects which
are isogenous with the given object. This is a consequence of the logical
theory of types [[. . .]]. If two objects are not isogenous, then they are termed
allogenous.
Comment. And, if I may continue, “allogenous” is of course not a property that
could be formulated in that theory of logical types so as to provide a propositional
function; otherwise one would have, for instance, “Hamburg is allogenous relative
to the moon” and “thankfulness is allogenous relative to the moon”, which would
render Hamburg and thankfulness isogenous according to Carnap’s criterion above.
(2) Carnap [1928], pp. 53 f.
Neglect of the difference between concepts of different spheres, we call con-
fusion of spheres.
REFERENCES. There has been no explicit recognition of the in-

dicated type of ambiguity in logic. But it bears a certain resemblance
to the multiplicity of “suppositions” of a word which the Schoolmen
used to distinguish [[. . .]]. It is more closely related to the theory of
types which Russell has developed in order to overcome the logi-
cal paradoxes [[. . .]]. However, Russell has applied his theory only to
formal-logical structures, not to a system of concrete concepts (more
precisely: only to variables and logical constants, not to nonlogical
constants). Our object spheres are Russell’s “types” applied to ex-
tralogical concepts. Thus, the justification for making a distinction
between the various object spheres [[. . .]] is derived from the theory of
types [[. . .]]. Although the theory of types is not generally accepted,
none of its opponents has been able to produce a logical system which
could avoid the contradictions (the so-called paradoxes) from which
the older logic suffers, without using a theory of types.
Comment. In view of the various ‘type free logics’ available since the fifties, the last
point should no longer carry any weight. Cf., however, Scott in quotations 83.7 (4)
and (5) in the last chapter.
86d. Tarski’s concept of truth.
Quotations 86.14. (1) Tarski [1933], pp. 164 f.

A characteristic feature of colloquial language (in contrast to various
scientific languages) is its universality. It would not be in harmony with
the spirit of this language if in some other language a word occurred which
could not be translated into it; it could be claimed that ‘if we can speak
meaningfully about anything at all, we can also speak about it in colloquial
language’. If we are to maintain this universality of everyday language in
connexion with semantical investigations, we must, to be consistent, admit
into the language, in addition to its sentences and other expressions, also
the names of these sentences and expressions, and sentences containing
theses names, as well as such semantic expressions as ‘true sentence’, ‘name’,

‘denote’, etc. But it is presumable just this universality of everyday language
which is the primary source of all semantical antinomies, like the antinomies
of the liar or of the heterological words. These antinomies seem to provide
a proof that every language which is universal in the above sense, and for
which the normal laws of logic hold, must be inconsistent.
(2) Tarski [1933], p. 267.
Philosophers who are not accustomed to use deductive methods in their
daily work are inclined to regard all formalized languages with a certain
disparagement, because they contrast these ‘artificial’ constructions with
the one natural language—the colloquial language. For that reason the fact
that the results obtained concern the formalized languages almost exclu-
sively will greatly diminish the value of the foregoing investigations in the
opinion of many readers. It would be difficult for me to share this view.
In my opinion the considerations of [[essentially the foregoing quotation]]
prove emphatically that the concept of truth (as well as other semantical
concepts) when applied to colloquial language in conjunction with the nor-
mal laws of logic leads inevitable to confusions and contradictions. Whoever
wishes, in spite of all difficulties, to pursue the semantics of colloquial lan-
guage with the help of exact methods will be driven first to undertake the
thankless task of a reform of this language. He will find it necessary to
define its structure, to overcome the ambiguity of the terms which occur
in it, and finally to split the language into a series of languages of greater
and greater extent, each of which stands in the same relation to the next in
which a formalized language stands to its metalanguage. It may, however,
be doubted whether the language of everyday life, after being ‘rationalized’
in this way, would still preserve its naturalness and whether it would not
rather take on the characteristic features of the formalized languages.
Comment. The curious thing for the present enterprise is that Tarski seems to think
he has solved a problem. In contrast, I take it as the formulation of a problem: why
would notions as vacuous/tautological as those considered by Tarski present such
complications as the necessity to distinguish language levels?
86e. Fear of metaphysics, inquisition of language, and worship of sci-

ence. This section concentrates on Carnap and Reichenbach, in particular Carnap’s
idea of analysis of language.
I begin with a quotation from Carnap’s “Überwindung der Metaphysik durch
logische Analyse der Sprache”, a programmatic piece which I find suitable to give
an idea of what I see as wishful thinking and pompous rhetoric leading to the
dishonoured promises characteristic of logical positivism and which is still present
in modern analytic philosophy.17
17 Cf. quotation 91.22 at the end of this chapter.

Quotation 86.15. Carnap [1931b], pp. 219 f.

Durch die Entwicklung der modernen Logik ist es möglich geworden, auf
die Frage nach der Gültigkeit und Berechtigung der Metaphysik eine neue
und schärfere Antwort zu geben. Die Untersuchungen der „angewandten Lo-
gik“ oder „Erkenntnistheorie“, die sich die Aufgabe stellen, durch logische
Analyse den Erkenntnisgehalt der wissenschaftlichen Sätze und damit die
Bedeutung der in den Sätzen auftretenden Wörter („Begriffe“) klarzustellen,
führen zu einem positiven und zu einem negativen Ergebnis. Das positive
Ergebnis wird auf dem Gebiet der empirischen Wissenschaft erarbeitet; die
einzelnen Begriffe der verschiedenen Wissenschaftszweige werden geklärt;
ihr formal-logischer und erkenntnistheoretischer Zusammenhang wird auf-
gewiesen. Auf dem Gebiet der Metaphysik (einschließlich aller Wertphilo-
sophie und Normwissenschaft) führt die logische Analyse zu dem Ergebnis,
daß die vorgeblichen Sätze dieses Gebietes gänzlich sinnlos sind. Damit ist
eine radikale Überwindung der Metaphysik erreicht, die von den früheren
antimetaphysischen Standpunkten aus noch nicht möglich war.
Quotation 86.16. Frege [1979], p. 269 f; translation of Frege [1969], p. 289 (con-
tinuation of quotation 73.7 (3)).
It is difficult to avoid an expression that has universal currency, before you
learn of the mistakes it can give rise to. It is extremely difficult, perhaps
impossible, to test every expression offered us by language to see whether it
is logically innocuous. So a great part of the work of a philosopher consists—
or at least ought to consist—in a struggle against language. But perhaps
only a few people are aware of the need for this.
Quotations 86.17. (1) Carnap [1935], pp. 209 f.

I will call metaphysical all those statements which claim to represent
knowledge about something which is over or beyond all experience, e.g.
about the real Essence of things, about Things in themselves, the Absolute,
and such like. [[. . .]] The sort of statements I wish to denote as metaphys-
ical may most easily be made clear by some examples: “The Essence and
Principle of the world is Water,” said Thales; “Fire,” said Heraclitus; “the
Infinite,” said Anaximander; “Number,” said Pythagoras. “All things are
nothing but shadows of eternal ideas which themselves are in a spaceless
and timeless sphere,” is a doctrine of Plato. From the monists we learn:
“There is only one principle on which all that is, is founded”; but the dual-
ists tell us: “There are two principles.” The materialists say: “All that is, is
in its essence material,“ but the spiritualists say: “All that is, is spiritual.”
To metaphysics (in our sense of the word) belong the principal doctrines
of Spinoza, Schelling, Hegel, and—to give at least one name of the present
time— Bergson.
Now let us examine this kind of statement from the point of view of
verifiability. It is easy to realize that such statements are not verifiable.
From the statement: “The Principle of the world is Water”, we are not able
to deduce any statement asserting any perceptions or feelings or experiences
whatever which may be expected for the future. Therefore the statement,
“The Principle of the world is Water”, asserts nothing at all. [[. . .]] Meta-
physicians cannot avoid making statements nonverifiable, because if they
made them verifiable, the decision about the truth or falsehood of their
doctrines would depend upon experience and therefore belong to the region
of empirical science. This consequence they wish to avoid, because they
pretend to teach knowledge which is of a higher level than that of empirical
science. Thus they are compelled to cut all connection between their state-
ments and experience; and precisely by this procedure they deprive them
of any sense.
Comment. It would have been a nice surprise to find Wittgenstein’s “essence of a
proposition” among these examples of metaphysical nonsense.18 But probably this is
asking for too much in terms of self-awareness. Still there is Carnap’s somewhat half-
hearted engagement with the problem of the self-application of his anti-metaphysical
position in quotation (5) of the present set. In this context, I should also like to
recommend Hegel’s remark in quotation 65.8.
(2) Carnap [1935], p. 215.
[[O]]ur anti-metaphysical thesis [[. . .]] asserts that metaphysical statements—
like lyrical verses—have only an expressive function, but no representative
function. Metaphysical statements are neither true nor false, because they
assert nothing, they contain neither knowledge nor error, they lie completely
outside the field of knowledge, of theory, outside the discussion of truth
or falsehood. But they are, like laughing, lyrics, and music, expressive.
They express not so much temporary feelings as permanent emotional or
volitional dispositions. Thus, for instance, a metaphysical system of monism
may be an expression of an even and harmonious mode of life, a dualistic
system may be an expression of the emotional state of someone who takes
life as an eternal struggle; an ethical system of rigorism may be expressive of
a strong sense of duty or perhaps of a desire to rule severely. Realism is often
a symptom of the type of constitution called by psychologists extroverted,
which is characterized by easily forming connections with men and things;
idealism, of an opposite constitution, the so-called introverted type, which
has a tendency to withdraw from the unfriendly world and to live within
its own thoughts and fancies.
(3) Carnap [1928], p. 234.
Since we consider only factual contents as the criterion for the meaningful-
ness of statements, neither the thesis of realism that the external world is
real, nor that of idealism that the external world is not real can be considered
scientifically meaningful.
(4) Carnap [1935], p. 217.
The only proper task of philosophy is logical analysis.

Comment. I am in favour of logical analysis and I think philosophy could greatly

profit from it; but to declare it the only proper task has a distinctly doctrinal taste
to it which I resent thoroughly.
(5) Carnap [1935], pp. 218 f.
But now it may perhaps be objected: “How about your own statements?
In consequence of your view your own writings, including this book, would
be without sense, for they are neither mathematical nor empirical, that is
verifiable by experience.” What answer can be given to this objection? This
question is decisive for the consistency of the view which has been explained
here.
An answer to the objection is given by Wittgenstein in his book Trac-
tatus Logico-Philosophicus. This author has developed most radically the
view that the statements of metaphysics are shown by logical analysis to be
without sense. How does he reply to the criticism that in that case his own
statements are also without sense? He replies by agreeing with it. He writes:
“The result of philosophy is not a number of ‘philosophical statements,’ but
to make statements clear” (p. 77). “My statements are elucidatory in this
way: he who understands me finally recognizes them as senseless, when he
has climbed out through them, on them, over them. (He must so to speak
throw away the ladder after he has climbed up on it.) He must surmount
these statements; then he sees the world rightly. Whereof one cannot speak,
thereof one must be silent” (p. 189).
I, as well as my friends in the Vienna Circle, owe much to Wittgenstein,
especially as to the analysis of metaphysics. But on the point just mentioned
I cannot agree with him. In the first place, he seems to me to be inconsistent
in what he does. He tells us that one cannot make philosophical statements
and that whereof one cannot speak, thereof one must be silent; and then
instead of keeping silent, he writes a whole philosophical book. Secondly, I
do not agree with his assertion that all his statements are quite as much
without sense as metaphysical statements are. My opinion is that a great
number of his statements (unfortunately not all of them) have in fact sense;
and that the same is true for all statements of logical analysis.
Comment. May I suggest, perhaps, a distinction of good and bad metaphysics, or
benign and malignant ; needless to add, Wittgenstein is doing good metaphysics,
precisely because his metaphysics is directed against bad metaphysics.
Quotations 86.18. (1) Reichenbach [1951], p. vii.

[[P]]hilosophic speculation is a passing stage, occurring when philosophic
problems are raised at a time which does not possess the logical means to
solve them. [[. . .]] [[T]]here is, and always has been, a scientific approach to
philosophy.
(2) Reichenbach [1951], p. 73.
There is no compromise between science and speculative philosophy.
Let us not attempt to reconcile the two in the hope for a higher synthe-
sis. Not all historical developments follow the dialectical law; one line of
thought may die out and leave its place to another that springs from dif-
ferent roots—like a biological species that survives only in fossil form, once
another species, better equipped, has taken over. Speculative philosophy,
after its climax in the system of Kant, has found only mediocre represen-
tatives and is decaying.
Comment. After social Darwinism, cultural Darwinism. The curious thing: those
who evoke these variations of Darwinism always seem to locate themselves — ap-
parently without any trace of doubt — on the line that is not going to die out.
(3) Reichenbach [1951], p. 108.
The rationalism of Leibniz, though inspired by mathematical science, is
speculation in the cloak of logical reasoning and abandons the solid ground
from which modern science has grown—its basis in empirical observation.
His disregard of the empirical component of knowledge led Leibniz to the
belief that all knowledge is logic. Although he saw the analytic nature of
deductive logic, he believed that logic cannot only supply but even replace
empirical knowledge. There are truths of the factual kind, that is, empir-
ical truths, and truths of reason, that is, analytic truths; this distinction,
however, is only a consequence of human ignorance, and if we could have
a perfect knowledge, as God has it, we would see that all that happens
is logically necessary. For instance, God could deduce from the concept of
Alexander that he was a king and conquered the Orient. This analytic in-
terpretation of empirical knowledge is a rationalist blunder which has been
repeatedly committed in the hope to explain mathematical physics.
Comment. I do not know Reichenbach’s sources for his claim that Leibniz believed
that “all knowledge is logic”, but I do have a source for Leibniz’ belief that meta-
physics is hardly different from logic. It can be found on p. 1602 of the ground-
works. I also have quotation 70.17 (1) which I do not find all that compatible with
what Reichenbach says about Leibniz’ belief.
(4) Reichenbach [1951], p. 109.
It may be that we can define a concept Alexander in such a way that
all the history of the man follows analytically from it; but then we could
never know from pure logic whether the observable individual Alexander is
correctly identified by that concept. In other words, the statement that the
observable individual has the properties expressed in the concept would be
synthetic and subject to all the quandaries of empirical knowledge. There is
no way of evading the problems of empiricism by taking refuge in analytic
logic.
Comment. Note the relation to Einstein as quoted in 70.25 (1)–(3) in these mate-
rials.
The distaste for Hegel and dialectics on the one hand and the worship of sci-
ence on the other even made their way into the Trotskyist movement as the next
quotation illustrates.
Quotation 86.19. Burnham [1940], pp. 236 f.

Hegelian dialectics has nothing whatever to do with science [[. . .]]. How the
sciences have influenced the forms of thought no one will ever discover by
spending even a lifetime on the tortuous syntax of the reactionary abso-
lutist, Hegel, but only by studying modern science and mathematics, and
the careful analysis of modern science and mathematics.
Comment. I was told that Burnham started off a Trotzkyist and turned into a
social democrat later which seems to fit well with the superstitious belief in science
displayed in this quotation.
Quotations 86.20. (1) Carnap [1950], p. 20.

Empiricists are in general rather suspicious with respect to any kind of
abstract entities like properties, classes, relations, numbers, propositions,
etc. They usually feel much more in sympathy with nominalists than with
realists (in the medieval sense). As far as possible they try to avoid any
reference to abstract entities and to restrict themselves to what is sometimes
called a nominalistic language, i. e., one not containing such references.
(2) Carnap [1950], pp. 21 f.
Are there properties, classes, numbers, propositions? In order to un-
derstand more clearly the nature of these and related problems, it is above
all necessary to recognize a fundamental distinction between two kinds of
questions concerning the existence or reality of entities. If someone wishes
to speak in his language about a new kind of entities, he has to introduce
a system of new ways of speaking, subject to new rules; we shall call this
procedure the construction of a framework for the new entities in question.
And now we must distinguish two kinds of questions of existence: first,
questions of the existence of certain entities of the new kind within the
framework ; we call them internal questions; and second questions concern-
ing the existence or reality of the framework itself, called external questions.
Internal questions and possible answers to them are formulated with the
help of the new forms of expressions. The answers may be found either by
purely logical methods or by empirical methods, depending upon whether
the framework is a logical or a factual one. An external question is of a
problematic character which is in need of closer examination.
Comment. Would it have crossed Carnap’s mind that his distinction of external
and internal questions may be of even more problematic character than any of the
problems that it is meant to solve?
(3) Carnap [1928], pp. 260.
[[A]] class is not a collection, or the sum, or a bundle of its elements, but a
unified expression for that which the elements have in common.
Comment. This is not stated with regard to some philosophy of mathematics but
to defend a characterization of the “self” in terms of a class. I find it interesting in
view of what Gödel’s says in quotation 78.14, and the quotations in section 83b of
these materials (mostly regarding the iterative concept of set).
§ 87. AFTER EUPHORIA 1223
(4) Carnap [1935], p. 212.

[[I]]t is important to realize that our doctrine is a logical one and has noth-
ing to do with metaphysical theses of the reality or unreality of anything
whatever.
Comment. This sounds like wishful thinking to me.
(5) Carnap [1928], p. 282.
The concept of reality (in the sense of independence from the cognizing
consciousness) does not belong within (rational ) science, but within meta-
physics.
(6) Carnap [1950], p. 40.
Let us grant to those who work in any special field of investigation the
freedom to use any form of expression which seems useful to them; the
work in the field will sooner or later lead to the elimination of those forms
which have no useful function. Let us be cautious in making assertions and
critical in examining them, but tolerant in permitting linguistic forms.
Comment. Sounds beautiful, doesn’t it. But my experience is different. Tell them
that the background of your logical endeavours is Hegel’s Science of Logic and all
tolerance is gone.19 I heard more than once that it wouldn’t be necessary to consider
my logical investigations, because any attempt at making sense of Hegel by means
of modern logic was doomed to failure anyway. There is, however, one remarkable
exception to this pattern and this is a good place to thank Professor Max Drömmer
in Munich for overcoming his dislike of Hegel and actually practicing the tolerance
advocated by Carnap.
Preoccupation with metaphysics, or a straw man version of it, continued. As

late as 1958 a book under the title “Vom Ursprung und Ende der Metaphysik”
(Topitsch) was published. Too silly to be included here.
I close this paragraph with a particular delicacy from the cooking of a Grand
Master.
Quotation 86.21. Popper [1971], p. 1, footnote 1.

I had [[. . .]] (in the winter of 1919–20) formulated and solved the problem of
demarcation between science and non-science and I did not think it worth
publishing.
§ 87. After euphoria
Logical empiricism entered the stage with impressive noises. In this paragraph I shall
focus on some aspects of the steady collapse of high expectations, or, to paraphrase
Reichenbach [1951], p. 73, logical empiricism, “after its climax . . . , has found only
mediocre representatives and is decaying”.20 The point for me, however, is not to
19 Probably, speculative philosophy would not qualify as a “special field of investigation”.
20 See quotation 86.18 (2) above.
gloat over the untenability of a rival approach, but to protect myself against a
similar fate, by trying to learn from the mistakes of others.
87a. The analytic-synthetic distinction. One of the central notions in the

logicist program is that of a tautology. From there it is only a short step to the
analytic-synthetic distinction put forward by Kant.
There is a vast literature on the analytic-synthetic distinction in analytic phi-
losophy, but I haven’t found anything valuable for my project, so I confine myself
here to a few quotations from some of the classic authors.

Kant’s theory of knowledge cannot be disentangled from his belief that
mathematical propositions are both synthetic and a priori. My own work
on the principles of mathematics was to me, at first mainly interesting as
a refutation of the view that mathematical propositions assert more than
can be justified by deductive logic.
(2) Hahn [1933], p. 230.
[[W]]e must distinguish two kinds of statements: those which say something
about facts and those which merely express the way in which the rules
which govern the application of words to facts depend on each other. Let
us call statements of the latter kind tautologies: they say nothing about
objects and are for this very reason certain, universally valid, irrefutable by
observation[[.]]
(3) Reichenbach [1951], p. 304.
The principles of logic and mathematics represent the only domain in which
certainty is attainable; but these principles are analytic and empty. Cer-
tainty is inseparable from emptiness: there is no synthetic a priori.
Comment. And, if I may add, in particular the infinity axiom is analytic and empty.
(It is all just a matter of being a true believer.)
(4) Reichenbach [1951], p. 165.
Kant’s theory of a synthetic a priori knowledge of nature quotes, in addition
to the laws of space and time, the principle of causality as the foremost
instance of such knowledge. Like the development of the problems of space
and time, that of the principle of causality has led, ever since the death of
Kant, to a disintegration of the synthetic a priori.
87b. The three dogmas of logical empiricism. If the title of the present
section somewhat reminds of Quine, this is not just by chance.21

Modern empiricism has been conditioned in large part by two dogmas.
One is a belief in some fundamental cleavage between truths which are ana-
21 If you are wondering what the third dogma is, have a look at quotation 87.11 below.
lytic, or grounded in meanings independently of matters of fact, and truths

which are synthetic, or grounded in fact. The other dogma is reductionism:
the belief that each meaningful statement is equivalent to some logical con-
struct upon terms which refer to immediate experience. Both dogmas, I
shall argue, are ill-founded.
Quotation 87.3. Quine [1953a], pp. 122 f.

Russell ([[. . .]], Principia) had a no-class theory. Notations purporting
to refer to classes were so defined, in context, that all such references would
disappear on expansion. This result was hailed by some, notably Hans
Hahn, as freeing mathematics from platonism, as reconciling mathemat-
ics with an exclusively concrete ontology. But this interpretation is wrong.
Russell’s method eliminates classes, but only by appeal to another realm
of equally abstract or universal classes—so-called propositional functions.
The phrase ‘propositional function’ is used ambiguously in Principia Math-
ematica; sometimes it means an open sentence and sometimes it means
an attribute. Russell’s no-class theory uses propositional functions in this
second sense as values of bound variables; so nothing can be claimed for
the theory beyond a reduction of certain universals to others, classes to
attributes.
As regards the analytic-synthetic distinction the following view seems to have

become somewhat standard.

Kant’s distinction between analytic and synthetic truths is related to Leib-
niz’s distinction between truths of reason and truths of fact, as well as to
Hume’s distinction between relation of idea and matters of fact.
Comment. Observe that there is no support for this view to be found in my quo-
tations of Kant in chapter XV. I shall argue later that it is actually a gross mis-
interpretation of Kant’s distinction. For the moment, however, I leave it at this
comment.
(2) Quine, [1951], p. 20.
Kant’s cleavage between analytic and synthetic truths was foreshad-
owed in Hume’s distinction between relations of ideas and matters of fact,
and in Leibniz’s distinction between truths of reason and truths of fact.
Comment. It is more a shadow, a very dark one indeed, because it is likely to
distract attention to a world of facts, whereas Kant’s distinction is one between two
different sorts of truths of reason: one by analysis and the other one by synthesis
(a combination which is not to be found within the object as it is in itself but has
its origin in the faculties of the mind). Observe that Kant himself had an opinion
on that matter too: see quotation 59.8 (5): it was Locke who came close to making
this distinction.
(3) Putnam, [1962], p. 46.

[[B]]efore the development of non-Euclidean geometry by Riemann and Lo-
bachevsky, the best philosophic minds regarded the principles of geometry
as virtually analytic. The human mind could not conceive their falsity.
Comment. This is a curious piece. Does Kant not qualify as one of the belong to
the “best philosophic minds”? Or is his synthetic a priori subsumed under “virtually
analytic”? Note the criterion of not conceiving the falsity and compare this with
what Poincaré says in quotation 70.23 (3) in these materials.
[[T]]he notion of self-contradictoriness, in the quite broad sense needed for
this definition of analyticity, stands in exactly the same need for clarification
as does the notion of analyticity itself. The two notions are the two sides
of a single dubious coin.
Much of the problem of the analytic-synthetic distinction runs along the lines
of “every bachelor is an unmarried man”.
Quotations 87.6. (1) Quine [1951], p. 24.
Our problem [[. . .]] is analyticity; and here the major difficulty lies not in
the first class of analytic statements, the logical truths, but rather in the
second class, which depends on the notion of synonymy.
(2) Quine [1951], p. 24.
There are those who find it soothing to say that the analytic statements
of the second class reduce to those of the first class, the logical truths, by
definition; ‘bachelor’, for example is defined as ‘unmarried man’. But how
do we find that ‘bachelor is defined as ‘unmarried man’ ? Who defined it
thus, and when? Are we to appeal to the nearest dictionary, and accept the
lexicographer’s formulation as law? Clearly this would be to put the cart
before the horse.
Radical reductionism [[. . .]] set itself the task of specifying a sense-datum
language and showing how to translate the rest of significant discourse,
statement by statement, into it. Carnap embarked on this project in the
Aufbau.
The language which Carnap adopted as his starting point was not a
sense-datum language in the narrowest conceivable sense, for it included
also the notations of logic, up through higher set theory. In effect it included
the whole language of pure mathematics. The ontology implicit in it (that
is, the range of values of its variables) embraced not only sensory events
but classes, classes of classes, and so on. Empiricists there are who would
boggle at such prodigality. Carnap’s starting point is very parsimonious,
however, in its extralogical or sensory part. In a series of constructions
in which he exploits the resources of modern logic with much ingenuity,
Carnap succeeds in defining a wide array of important additional sensory
concepts which, but for his constructions, one would not have dreamed were
definable on so slender a basis. He was the first empiricist who, not content
with asserting the reducibility of science to terms of immediate experience,
took serious steps toward carrying out the reduction.
If Carnap’s starting point is satisfactory, still his constructions were,
as he himself stressed, only a fragment of the full program. The construc-
tion of even the simplest statements about the physical world was left in
a sketchy state. Carnap’s suggestions on this subject were, despite their
sketchiness, very suggestive. He explained spatio-temporal point-instants
as quadruples of real numbers and envisaged assignment of sense qualities
to point-instants according to certain canons. [[. . .]]
Carnap did not seem to recognize, however, that his treatment of phys-
ical objects fell short of reduction not merely through sketchiness, but in
principle. Statements of the form ‘Quality q is at point-instant x; y; z; t’
were, according to his canons, to be apportioned truth values in such a way
as to maximize and minimize certain over-all features, and with growth in
experience the truth values were to be progressively revised in the same
spirit. I think this is a good schematization (deliberately oversimplified, to
be sure) of what science really does; but it provides no indication, not even
the sketchiest, of how a statement of the form ‘Quality q is at x; y; z; t’ could
ever be translated into Carnap’s initial language of sense data and logic.
The connective ‘is at’ remains an added undefined connective; the canons
counsel us in its use but not in its elimination.
Carnap seems to have appreciated this point afterward; for in his later
writings he abandoned all notion of the translatability of statements about
the physical world into statements about immediate experience. Reduction-
ism in its radical form has long since ceased to figure in Carnap’s philosophy.
Quotation 87.8. Adjukiewicz [1934], p. 259.

Die Hauptthese des gewöhnlichen Konventionalismus, wie er etwa von Poin-
caré vertreten wird, besteht in der Behauptung, daß es Probleme gibt, die
durch die Erfahrung so lange nicht lösbar sind, als nicht eine willkürliche
Konvention eingeführt wird, welche erst dann, im Verein mit den Erfah-
rungsdaten, das Problem zu lösen gestattet. Die Urteile, in welchen diese
Lösung besteht, sind uns also nicht durch die Erfahrungsdaten allein aufge-
zwungen, ihre Annahme hängt teilweise von unserer Willkür ab, da wir die
angenommene Konvention, welche die Lösung des Problems mitbestimmt,
willkürlich ändern können, wodurch wir dann zu anderen Urteilen kommen.
In dieser Abhandlung beabsichtigen wir diese These des gewöhnlichen
Konventionalismus zu verallgemeinern und zu radikalisieren. Wir wollen
hier nämlich die Behauptung aufstellen und begründen, daß nicht nur eini-
ge, sondern alle Urteile, die wir annehmen, und die unser ganzes Weltbild
ausmachen, durch die Erfahrungsdaten noch nicht eindeutig bestimmt sind,
sondern von der Wahl der Begriffsapparatur abhängen, durch die wir die
Erfahrungsdaten abbilden. Diese Begriffsapparatur können wir aber so oder
ganz anders wählen, wodurch unser ganzes Weltbild geändert wird.
Quotation 87.9. Cohen and Nagel [1934], p. 173.

If in an inference the conclusion is not contained in the premise, it cannot be
valid; and if the conclusion is not different from the premises, it is useless;
but the conclusion cannot be contained in the premises and also possess
novelty; hence inferences cannot be both valid and useful.
87c. Further down the track.
Quotation 87.10. Carnap [1961] pp. v f.

I still agree with the philosophical orientation which stands behind this
book [[The Logical Structure of the World ]]. This holds especially for the
problems that are posed, and for the essential features of the method which
was employed. The main problem concerns the possibility of rational recon-
struction of the concepts of all fields of knowledge on the basis of concepts
that refer to the immediately given. By rational reconstruction is here meant
the searching out of new definitions for old concepts. The old concepts did
not ordinarily originate by way of deliberate formulation, but in more or
less unreflected and spontaneous development. The new definitions should
be superior to the old in clarity and exactness, and, above all, should fit
into a systematic structure of concepts. Such a clarification of concepts,
nowadays frequently called “explication,” still seems to me one of the most
important tasks of philosophy, especially if it is concerned with the main
categories of human thought.
For a long time, philosophers of various persuasions have held the view
that all concepts and judgments result from the coöperation of experience
and reason. Basically, empiricists and rationalists agree in this view, even
though both sides give a different estimation of the relative importance
of the two factors, and obscure the essential agreement by carrying their
viewpoints to extremes. The thesis which they have in common is frequently
stated in the following simplified version: The senses provide the material
of cognition, reason synthesizes the material so as to produce an organized
system of knowledge. There arises then the problem of finding a synthesis
of traditional empiricism and traditional rationalism. Traditional empiri-
cism rightly emphasized the contribution of the senses, but did not realize
the importance and peculiarity of logical and mathematical forms. Ratio-
nalism was aware of this importance, but believed that reason could not
only provide the form, but could by itself (a priori) produce new content.
Through the influence of Gottlob Frege, under whom I studied in Jena, but
who was not recognized as an outstanding logician until after his death,
and through the study of Bertrand Russell’s work, I had realized, on the
one hand, the fundamental importance of mathematics for the formation
of a system of knowledge and, on the other hand, its purely logical, for-
mal character to which it owes its independence from the contingencies
of the real world. These insights formed the basis of my book. Later on,
through conversations in Schlick’s circle in Vienna and through the influ-

ence of Wittgenstein’s ideas they developed into the mode of thought which
characterized the “Vienna Circle.”
Quotation 87.11. (1) Wang [1985], p. 452.

Carnap [[. . .]]:
(C1) Logicism: mathematics is reducible to logic.
(C2) Analyticity: logical truths are analytic.
(C3) Conventionalism: analytical propositions are true by convention
and, therefore, void of content.
(2) Wang [1985], p. 453; partly quoting Gödel.
Of the various senses of the term ‘analytic’, Gödel singles out two sig-
nificant ones [[. . .]]:
(2a) Tautological. This term has been used equivocally. If it is to have
a sufficiently definite and broad sense it has to involve in some manner defi-
nitions plus axioms and deductions. When applied to set theory or number
theory, etc., it has the “purely formal sense that terms occurring can be de-
fined (either explicitly or by rules for eliminating them from sentences con-
taining them) in such a way that the axioms and theorems become special
cases of the law of identity and disprovable propositions become negations
of the law. In this sense [i.e., if ‘analytic’ is taken to mean ‘tautological’ in
this sense] even the theory of integers is demonstrable nonanalytic, provided
that one requires of the rules of elimination that they allow one actually
to carry out the elimination in a finite number of steps in each case.” (To
remove the restriction on being finite would beg the question of giving an
account of mathematics and would be circular in proving, e.g., the axiom
of choice or the axiom of infinity, tautological. Note that even first-order
logic is not tautological in the specified sense.)
(2b) Analytic. A “proposition is called analytic if it holds, ‘owing to the
meaning of the concepts occurring in it,’ where this meaning may perhaps
be undefinable (i.e., irreducible to anything more fundamental).” In this
sense the axioms and theorems of mathematics, set theory, and logic all are
analytic, but need not, as a result, be “void of content.”
Comment. One of the particular points of interest for me: not even first-order logic
is tautological in the specified sense.
(3) Hodes [1995], p. 448, first column.
As a defence of derivational logicism, Principia was flawed by virtue of its
reliance on three axioms, a version of the Axiom of Choice, and the axioms
of Reducibility and Infinity, whose truth were controversial. Reducibility
could be avoided by eliminating the ramification of the logic (as suggested
by Ramsey). But even then, even the arithmetic of the natural numbers re-
quired use of Infinity, which in effect asserted that there are infinitely many
individuals (i.e., entities of type 0). Though Infinity was “purely logical,”
i.e. contained only logical expressions, in his Introduction to Mathemati-

cal Philosophy (p. 141) Russell admits that it “cannot be asserted by logic
to be true.” Russell then (pp. 194–95) forgets this: “If there are still those
who do not admit the identity of logic and mathematics, we may challenge
them to indicate at what point in the successive definitions and deductions
of Principia Mathematica they consider that logic ends and mathematics
begins. It will then be obvious that any answer is arbitrary.” The answer,
“Section 120, in which Infinity is first assumed!,” is not arbitrary.
§ 88. The rise of common sense and ordinary language, and the decline
of logical empiricism
Somewhere along the line, logicism got lost and the paradigm of mathematics was
replaced by ordinary language. It is this development which I am mainly thinking
of when I speak of “the decline of logical empiricism”.
88a. Psychology and the rule of objects. The days of Frege’s struggle
against psychology in foundational issues are over; philosophers in a somewhat em-
piricist tradition have found back to their favourite playground: mostly driveling on
about how children learn to do this or that, or how some primitive people in some
primordial past discovered how to do this or that.

We are prone to talk and think of objects. Physical objects are the
obvious illustration when the illustrative mood is on us, but there are also
all the abstract objects, or so there purport to be: the states and qualities,
numbers, attributes, classes. We persist in breaking reality down somehow
into a multiplicity of identifiable and discriminable objects, to be referred
to by singular and general terms. We talk so inveterately of objects that
to say we do so seems almost to say nothing at all; for how else is there to
talk?
Comment. Cf. Frege in quotation 71.15 (2) and 71.17.
Quotations 88.2. (1) Quine [1953], quoted after [1966], p. 226.

None of us learned his words quite like anyone else. But we use them in
sufficient systematic agreement for communication — which is no accident,
since language is subject to the law of survival of the fittest.
(2) Quine [1969], p. 81.
A child learns his first words and sentences by hearing and using them
in the presence of appropriate stimuli. These must be external stimuli,
for they must act both on the child and on the speaker from whom he
is learning. Language is socially inculcated and controlled; the inculcation
and control turn strictly on the keying of sentences to shared stimulation.
Internal factors may vary ad libitum without prejudice to communication
as long as the keying of language to external stimuli is undisturbed. Surely
§ 88. COMMON SENSE AND THE DECLINE OF LOGICAL EMPIRICISM 1231
one has no choice but to be an empiricist so far as one’s theory of linguistic

meaning is concerned.
Comment. Note the similarity to Augustine in the next quotation.
(3) Wittgenstein [1953], p. 2e ; translation of Augustine.
When they (my elders) named some object, and accordingly moved towards
something, I saw this and I grasped that the thing was called by the sound
they uttered when they meant to point it out. Their intention was shewn by
their bodily movements, as it were the natural language of all peoples: the
expression of the face, the play of the eyes, the movement of other parts
of the body, and the tone of voice which expresses our state of mind in
seeking, having, rejecting, or avoiding something. Thus, as I heard words
repeatedly used in their proper places in various sentences, I gradually
learnt to understand what object they signified; and after I had trained my
mouth to form these signs, I used them to express my own desires.
Comment. That’s how simple it is! You just have to remember how you learnt
talking yourself. But perhaps it needs a superior mind, at least superior to mine, to
remember how learning my first words as a child was.22 Anyway, Wittgenstein can
make sense of it, as the next quotation indicates.
These words, it seems to me, give us a particular picture of the essence
of human language. It is this: the individual words in language name objects
— sentences are combinations of such names. — In this picture of the
language we find the roots of the following idea: Every word has a meaning.
This meaning is correlated with the word. It is the object for which the word
stands.
(5) Quine [1969], p. 83.
Epistemology in its new setting, conversely, is contained in natural science,
as a chapter of psychology. But the old containment remains valid too, in
its way. We are studying how the human subject of our study posits bodies
and projects his physics from his data, and we appreciate that our position
in the world is just like his.
Comment. This reminds me of Hume’s words “the science of man is the only solid
foundation for the other sciences”.23 The point of an explanation of universally valid
a priori sciences seems to have lost its place. We are back to what Kant, talking
about Locke, called “a certain physiology of the human understanding”.
Quotation 88.3. Ramsey [1927], p. 138.
Suppose I am at this moment judging that Cæsar was murdered then it is
natural to distinguish in this fact on the one side either my mind, or my
present mental state, or words or images in my mind, which we will call
the mental factor or factors, and on the other side either Cæsar, or Cæsar’s
22 In section 6 of the first book of his Confessiones, Augustinus states that he could not
remember early childhood; so may be he only learned speaking in late childhood.

23 Hume [1739], p. xx; see quotation 58.17 (2) in these materials.
murder, or Cæsar and murder, or the proposition Cæsar was murdered, or

the fact that Cæsar was murdered, which we will call the objective factor
or factors; and to suppose that the fact that I am judging that Cæsar was
murdered consists in the holding of some relation or relations between these
mental and objective factors. The questions that arise are in regard to the
nature between them, the fundamental distinction between these elements
being hardly open to question.
88b. Common sense and intuition. In view of the notion of intuition in

Kant’s philosophy it is worth while emphasizing that the sort of intuition which is
at issue in this section is more a sort of vague common sense.
Quotations 88.4. (1) Warren [1967], p. 182.

[[W]]hatever the merits of intuition as a source of knowledge, intuition as a
certification of knowledge is another matter. Certification is not gained by
intuition.
(2) Van Benthem [1984b], p. 309.
Indeed, the ill repute of the term ‘intuition’ may be partly due to a
misapplication. It is highly unlikely that intuition would settle such specific
issues as the validity of concrete inference schemata. An appeal to intuitions
in discussions of the latter type often amounts to a refusal to argue about
the evidence. On the other hand, the proper place for intuition would seem
to be at the level of the general structure of our concepts — in the spirit of
Kant’s philosophy. To paraphrase this great philosopher, we have certain a
priori intuitions concerning the basic logical notions, and no human mind
is entirely without them.
Comment. Poor spirit of Kant’s philosophy; but a somewhat predictable result of
the translation of “Anschauung” as “intuition”.
Quotation 88.5. Jech [1977], p. 350.

After all, mathematics abounds in “counterintuitive” examples. Just look
at Weierstrass’ construction of a continuous nondifferentiable function.
Quotation 88.6. Wang [1974], p. 192.

In any event, even if we agree that our intuition did once lead to con-
tradictions, that fact does not justify the view that we run a high risk of
self-contradiction whenever we use our intuition. [[. . .]] The iterative concept
of set is an intuitive concept and this concept has led to no contradictions.
88c. Ordinary language and linguistic analysis. In what follows ordinary

language is, of course, English.24
Ordinary language philosophy seems to have developed largely in blissful igno-
rance of Frege’s analysis, his struggle and his failure to provide a logical foundation
of arithmetic.
24 To consider any other language, it seems, would have to count as unnatural, at least, if not
insane.
The point of this section is not so much to provide useful material, as to sketch
aspects of the regression of mainly English philosophy into an impotent state of
preoccupation “with the different ways in which silly people can say silly things”.25
It is in this ineptness that English (even more so Oxford)26 philosophy is on a par
with the Continental tradition, despite all the differences that seem to prevail. I
hate to think that this is a problem of philosophy altogether, independent of the
different forms it has taken in the last half century.
The attitude is aptly described in Flew’s introduction to a collection of essays
on Logic and Language from which I take the first quotation.
Quotation 88.7. Flew [1951], pp. 2 f.

We have tried to compile a collection of articles which would as far as
possible satisfy a multiple set of criteria. Firstly, they had to be immediately
readable by and intelligible to the layman. All symbolism was therefore
excluded. This was easily achieved, for the protagonists of this movement
in England, unlike those of the similar and parallel philosophic tendencies
of Logical Positivism and Semanticism on the Continent and in the United
States, strive to write in plain untechnical and unsymbolic English. And
in this they stand squarely in the British tradition of Thomas Hobbes and
John Locke, of Bishop Berkeley and David Hume.
Quotations 88.8. (1) Austin [1956], pp. 383 f.27

First, words are our tools, and, as a minimum, we should use clean
tools: we should know what we mean and what we do not, and we must
forearm ourselves against the traps that language sets us. Secondly, words
are not (except in their own little corner) facts or things: we need therefore
to prise them off the world, to hold them apart from and against it, so that
we can realize their inadequacies and arbitrariness, and can relook at the
world without blinkers. Thirdly, and more hopefully, our common stock of
words embodies all the distinctions men have found worth drawing, and
the connexions they have found worth marking, in the lifetimes of many
generations: these surely are likely to be more numerous, more sound, since
they have stood up to the long test of the survival of the fittest, and more
subtle, at least in all ordinary and reasonably practical matters, than any
that you or I are likely to think up in our arm-chairs of an afternoon—the
most favoured alternative method.
Comment. The emphasis, I suggest, lies on the phrase “in all ordinary and rea-
sonably practical matters”. I suppose what Austin had in mind was very ordinary,
indeed, and nothing like a theory of relativity, for instance, was ever to disturb this
ordinariness. Note the point of words as “clean tools”: “we should know what we
mean”; contrast this with Hilbert’s proof theory as a way of making the mathemati-
25 Russell [1959], p. 250; quoted after Rée [1993], p. 10.

26 Cf. Russell in quotation 88.8 (3) below.
27 This quotation was instigated by Rée [1993], p. 10.
cian’s tools the object of investigation, and Hegel’s comment that to “know what
one says is much rarer than one thinks”.28
(2) Rée in Urmson and Rée [1989], p. xi.
[[M]]any would regard as the Golden Age of twentieth-century English phi-
losophy — the “linguistic” movement centered in Oxford in the 1950s, which
was inspired by the later Wittgenstein, and advocated by Austin, Hare,
Strawson, and above all Ryle.
(3) Russell [1956], pp. 321 f; the editor Marsh quoting Russell.
Bad philosophy has always been an Oxford specialty[[.]]
Comment. I am inclined to agree. There seems to be something distinctly bovine
in Oxford philosophy.
Quotation 88.9. Ammerman [1965], p. 2 f.

The word “analysis” when used in philosophy bears obvious affinities to the
word’s use in a science such as chemistry. To analyze, we may say roughly,
is to take apart in order to gain a better understanding of what is being
analyzed. The chemist is concerned with the analysis of complex physical
substances into their constituent parts. The philosopher, on the other hand,
is interested in analyzing linguistic or conceptual units. He is concerned, in
general, with coming to understand the structure of language by a careful
study of the elements and their interrelations.
We will use the word “analysis” (or “analytic philosophy”), then, to
refer to any philosophy which places its greatest emphasis upon the study
of language and its complexities. [[. . .]]
[[. . .]]
Although all analytic philosophers would agree that the study of lan-
guage is of the greatest importance, there is no general agreement about
which language can most fruitfully be studied by the philosopher. Indeed, it
is just at this point that a fundamental cleavage has occurred between the
various philosophers who practice analysis. Some of them have concluded
that philosophical analysis ought consist primarily in the construction of
new, artificial language systems (sometimes called calculi, because of their
affinity to mathematical systems). The rules of these artificially constructed
languages are intended to be clearer, more complete, and more precise than
the rules that govern our use of language in ordinary discourse. Just as
science had to create its own technical vocabulary and introduce concepts
(e.g., force, mass, atom) that are more precise than those supplied by com-
mon sense, so also, these philosophers argue, philosophy must develop its
own vocabulary and set of concepts in order to resolve its problems.
Other analysts have disagreed with this argument. They contend that
such artificial languages are of little help in resolving philosophical prob-
lems. It is their view that philosophical problems can best be approached
by a careful analysis of the ordinary, natural language we all use to commu-

nicate with each other. For this reason, these philosophers are sometimes
(but not accurately) referred to “ordinary language” philosophers. A more
accurate way of distinguishing these two main “schools” of analysis is to
refer to the proponents of artificial language analysis as Logical Positivists
and the philosophers interested in analyzing ordinary language as Linguistic
Analysts.
Comment. Note the phrase “the ordinary, natural language we all use to commu-
nicate with each other”. Hasn’t Frege warned us of using the definite article in
phrases like “the will of the people”; or is “the ordinary, natural language we all use”
a different issue?
The following set of quotations is meant to give an impression of the usurpation

and distortion of Frege’s idea in analytic empiricism.
Quotations 88.10. (1) Searle [1971], p. 2.

Frege’s most important single discovery in the philosophy of language
was the distinction between sense and reference. He elucidated this distinc-
tion in terms of the following puzzle about identity statements: How is it
possible for a true statement of the form a is identical with b, to contain
any more factual information than a statement of the form, a is identical
with a?
Comment. In view of what Frege says in the introduction to his Grundgesetze, this
is a curiously lopsided way of putting things, in particular speaking of a discovery.
Frege introduces his distinction of sense and reference for the purpose, among others,
of elucidating his abolishment of the three stroke identity sign,29 and it plays a
further role in the introduction of the spiritus lenis for the purpose of communicating
the course of value (Wertverlauf) and establishing a notion of equality for the
extensions of concepts.
(2) Searle [1971], p. 4.
Frege asks: how do proper names and definite descriptions refer to things?
His answer is: in virtue of their sense.
Comment. I would be curious to know where Frege actually asked and answered
this question in this way.
(3) Searle [1971], p. 7.
Instead of seeing the relations between words and the world as something
existing in vacuo, one now sees them as involving intentional actions by
speakers, employing conventional devices (words, sentences) in accordance
with extremely abstract set of rules for the use of those devices. For example,
the real strength of Frege’s theory of sense and reference as opposed to
Russell’s theory of definite descriptions emerges in Strawson’s conception
of reference as a speech act. Once one sees referring as an action that is
performed in the utterance of an expression with a particular sense provided
29 See Frege [1893], p. IX, quotation 71.20 (1) in these materials.
by the rules for the use of the expression, then it is easier to see that it is
subject to the sorts of error that plague actions generally (one can fail to
refer to a king of France for the same reason that one can fail to hit a king
of France: there is no such person), and on this account there is much less
motivation for trying to identify referring (one kind of speech act), with
asserting an existential proposition (quite another kind of speech act), as
Russell in effect does.
Comment. It needs a RealPhilosopher to discover real strength. Frege was still
caught in that narrow perspective of a logical enterprise characterized in quotation
71.13 (1) in these materials where there was no room for “the interaction of speaker
and listener”.
(4) Searle [1971], p. 39.
In a typical speech situation involving a speaker, a hearer, and an ut-
terance by the speaker, there are many kinds of acts associated with the
speaker’s utterance. The speaker will characteristically have moved his jaw
and tongue and made noises. In addition, he will characteristically have
performed some acts within the class which includes informing or irritating
or boring his hearers[[.]]
Comment. Not to mention the speaker’s “fine Kantian beard” 30 . Anyway, I don’t
think that Frege would have been impressed;31 but I would want to argue that the
colour of the speaker’s underpants accounts for the différance in speech-acts. In
other words, I wish to assert that Searle’s revolutionary ideas, together with those
of Drucilla Cornell’s friend Jacques, occupy an important place in the landscape of
philosophy — stultifera navis.32
Quotations 88.11. (1) Strawson [1950], p. 316

I think it is true to say that Russell’s Theory of Descriptions [[. . .]] is
still widely accepted among logicians as giving a correct account of the use
of such expressions in ordinary language.
(2) Russell [1957], p. 338
My theory of descriptions was never intended as an analysis of the state
of mind of those who utter sentences containing descriptions.
(3) Russell [1957], p. 337.
[[There is]] a fundamental divergence between myself and many philoso-
phers with whom Mr. Strawson appears to be in general agreement. They
are persuaded that common speech is good enough, not only for daily life,
but also for philosophy. I, on the contrary, am persuaded that common
speech is full of vagueness and inaccuracy, and that any attempt to be pre-
cise and accurate requires modification of common speech both as regards
vocabulary and as regards syntax.
30 I gladly acknowledge Michèle Le Dœuff’s authorship of this beautiful phrase.
31 I take quotation 71.13 (1) in these materials as a clear expression by Frege that he did
not intend to deal with anything like a “speech act” in his theory.
32 Thank you, Michel Foucault.
§ 89. TRUTH AND REFERENCE 1237
Quotation 88.12. Ammerman [1965], p. 12.

Philosophers can never again, except at their peril, ignore the importance
of language when attempting to resolve philosophical problems. This is a
minimal, but lasting, accomplishment of analysis. Whether further philo-
sophical investigation of language will in the end lead to the dissolution of
all philosophical puzzles, as Wittgenstein seems to have believed, or whether
it will issue in a new, linguistically oriented metaphysics, as Strawson sug-
gests, is in many ways the most important question confronting philosophy
today.
Comment. Of course, it all depends on what is meant by language, but what I would
want to know is at whose peril philosophers can ignore the lack of importance of
linguistically oriented metaphysics.
§ 89. Truth and reference
89a. The semantic notion of truth. While Tarski’s concept of truth in

formalized languages is a generally recognized contribution to metamathematics, it
also has a so-called philosophical dimension which seems widely acclaimed in the
tradition of analytic philosophy but otherwise it is far from being uncontroversial.
The following quotations are meant to characterize the position of the more analytic
tradition.

The present article [[The Concept of Truth in Formalized Languages]] is
almost wholly devoted to a single problem—the definition of truth. Its task
is to construct—with reference to a given language—a materially adequate
and formally correct definition of the term ‘true sentence’. This problem,
which belongs to the classical questions of philosophy, raises considerable
difficulties. For although the meaning of the term ‘true sentence’ in col-
loquial language seems to be quite clear and intelligible, all attempts to
define this meaning more precisely have hitherto been fruitless, and many
investigations in which this term has been used and which started with ap-
parently evident premisses have often led to paradoxes and antinomies (for
which, however, a more or less satisfactory solution has been found). The
concept of truth shares in this respect the fate of other analogous concepts
in the domain of the semantics of language.
(2) Tarski [1935], p. 153.
A thorough analysis of the meaning current in everyday life of the term
‘true’ is not intended here. Every reader possesses in greater or less degree
an intuitive knowledge of the concept of truth and he can find detailed
discussions on it in works on the theory of knowledge. I would only mention
that throughout this work I shall be concerned exclusively with grasping
the intentions which are contained in the so-called classical conception of
truth (‘true—corresponding with reality’) in contrast, for example, with the
utilitarian conception (‘true—in a certain respect useful’).
Comment. There are so many things that I have been told, implicitly, that I would
have; a belief in some god or other higher being, a moral feeling for what is good
and bad, a ‘nature’ as a male, an understanding of reality as independent from
my consciousness, in short, all that rubbish that people tell you when they try to
dominate you with their ideas. Now Tarski is telling me that I would have, “in greater
or less degree an intuitive knowledge of the concept of truth”. Maybe this was just
meant to be a nice gesture, but to me it comes across as an arrogant imposition. I
am not surprised about the fierce opposition Tarski’s semantic concept of truth has
met with in the transcendental-speculative tradition. No matter how incompetent
philosophers in the transcendental-speculative tradition may be in technical respect,
they still recognize philosophical bluff, even if it is in the introduction to a formal
logical work.
(3) Tarski [1944], pp. 53 f.
We should like our definition to do justice to the intuitions which adhere
to the classical Aristotelian conception of truth—intuitions which find their
expression in the well-known words of Aristotle’s metaphysics
To say of what is that it is not, or of what is not that it is, is false, while
to say of what is that it is, or of what is not that it is not, is true.
If we wished to adapt ourselves to modern philosophical terminology,
we could perhaps express this conception by means of the familiar formula:
The truth of a sentence consists in its agreement with (or correspondence
to) reality.
(For a theory of truth which is to be based upon the latter formulation the
term “correspondence theory” has been suggested.)
Comment. Cf. quotations 57.21 and 57.22, respectively, as regards the citation from
Aristoteles and the version in “modern philosophical terminology”.

Durch die formale Fassung der Antinomie [[des Lügners]] wird zunächst
einmal zur vollen Deutlichkeit gebracht, daß die Antinomie mit der Frage
der sachlichen Wahrheit, im Sinne eines erkenntnistheoretischen Problems
nichts zu tun hat, daß vielmehr von dem Begriff des Zutreffens (der Wahr-
heit) einer Aussage für das Zustandekommen des Widerspruchs nur dasjeni-
ge gefordert wird, was [[. . .]] sich in der Umgangssprache etwa so formulieren
läßt „Ein Satz der Form ‚die Aussage A trifft zu‘ ist seinerseits eine Aussa-
ge; aus dieser Aussage kann die Aussage A gefolgert werden und umgekehrt
auch aus der Aussage A jener Satz.“
Bemerkung. Der Ausdruck „Wahrheitsdefinition“ darf nicht dazu verlei-
ten, von einer solchen Definition eine philosophische Aufklärung über den
Wahrheitsbegriff zu erwarten. Es handelt sich vielmehr zumeist nur um ei-
ne Präzisierung derjenigen Deutung der Formeln, welche ohnehin bei der
üblichen Anwendung des Formalismus zugrunde gelegt wird, und die Auf-
gabe für die Definition besteht darin, daß diese Deutung auf generelle Art,
in Abhängigkeit von der Gestalt der jeweiligen Formel, zum Ausdruck zu

bringen ist.

Personally, I should not feel hurt if a future world congress of the “theo-
reticians of truth” should decide—by a majority of votes—to reserve the
word “true” for one of the non-classical conceptions, and should suggest an-
other word, say, “frue,” considered here. But I cannot imagine that anybody
could present cogent arguments to the effect that the semantic conception
is “wrong” and should be entirely abandoned.
(2) Tarski [1944], pp. 70 f.
I have heard it remarked that the formal definition of truth has nothing
to do with “the philosophical problem of truth.” However, nobody has ever
pointed out to me in an intelligible way just what this problem is. I have
been informed in this connection that my definition, though it states neces-
sary and sufficient conditions for a sentence to be true, does not really grasp
the “essence” of this concept. Since I have never been able to understand
what the “essence” of a concept is, I must be excused from discussing this
point any longer.
In general, I do not believe that there is such a thing as “the philosoph-
ical problem of truth.”
Comment. Admittedly, philosophers hardly ever seem to say clearly what they think
the philosophical problem connected with the notion of truth consists in.33 Since
it is actually a relevant problem in the context of my enterprise, I do not want to
let Tarski’s remarks stand unanswered. If, as Tarski says, “The truth of a sentence
consists in its agreement with (or correspondence to) reality”, then the philosophical
problem lies in the notion of reality. In its most immediate form, the question is:
how is this reality given to our discourse? Problems here can be avoided if the reality
concerned is a formalized theory as in the case of §50 in the tools, where a truth
definition is given for a formalized theory of first order arithmetic.34 But this does
not apply to Tarski’s intuitive formulation. If we assume, with Locke, for instance,
that colour is a secondary property, how then can a proposition of the form “snow
is white” be true in the sense of Tarski’s formulation? In this respect, what I regard
as the philosophical problem of truth is intimately linked to the question of how
language relates to the world.35 This, however, is a problem that Tarski passes over
in silence. Differently put, it is the role of concepts in knowledge, that is problematic.
If this had been recognized, I am inclined to speculate, the correspondence theory
of truth might well have led to the question of the ‘truth’ of concepts.
(3) Tarski [1944], p. 71.
[[T]]he semantic definition of truth implies nothing regarding the conditions
under which a sentence like (1):
33 Tugendhat [1960] can serve as a paradigm case.
34 Cf. also quotation 89.2 (2) above.
35 Cf. Goodstein in quotation 89.7 (1) below.
(1) snow is white

can be asserted. It implies only that, whenever we assert or reject this
sentence, we must be ready to assert or reject the correlated sentence (2):
(2) the sentence “snow is white” is true.
Thus we may accept the semantic conception of truth without giving
up any epistemological attitude we may have had; we may remain naïve re-
alists, critical realists or idealists, empiricists or metaphysicians—whatever
we were before. The semantic conception is completely neutral toward all
these issues.
Comment. As a basic rule, I am suspicious against claims of complete neutrality.
According to my experience it is only fools and/or salesmen who try to convince
you that something is “completely neutral”. More specifically, as a philosopher in
the tradition of Hegel, I am suspicious towards a notion of truth which doesn’t go
beyond what Hegel (is reported as having) called “die formelle Wahrheit, die bloße
Richtigkeit” (“formal truth, bare correctness”).36
(4) Tarski [1956], p. 402; quoted after Suppes [1961], p. 11, footnote *.
The main source of the difficulties met with seems to lie in the following: it
has not always been kept in mind that the semantical concepts have a rel-
ative character, that they must always be related to a particular language.
People have not been aware that the language about which we speak need
by no means coincide with the language in which we speak. They have car-
ried out the semantics of a language in that language itself and, generally
speaking, they have proceeded as though there was only one language in the
world. The analysis of the antinomies mentioned shows, on the contrary,
that the semantical concepts simply have no place in the language to which
they relate, that the language which contains its own semantics, and within
which the usual logical laws hold, must inevitably be inconsistent.
Comment. The point at stake in the present study is, of course, the somewhat silent
premise not to question that the “usual logical laws hold”.
A quotation regarding a more technical aspects of Tarski’s approach to the
notion of truth in formalized theories can be found in quotation 79.14 in these
materials.
Quotations 89.4. (1) Putnam [1978], p. 9.
The nature of truth is a very ancient problem in philosophy; but not until
the present century did philosophers and logicians attempt to separate this
problem from the problems of the nature of knowledge and the nature of
warranted belief. That one could have a theory of truth which is neutral
with respect to epistemological questions, and even with respect to the
great metaphysical issue of realism versus idealism, would have seemed
preposterous to a nineteenth-century philosopher. Yet that is just what the
most prestigious theory of truth, Tarski’s theory, claims to be.
Comment. Does it come as a surprise that I strongly disagree, in particular with the
claim of epistemological neutrality? The adherence to classical logic places Tarski’s
theory in principle within realism. I do, however, acknowledge that Tarski’s ap-
proach enables a sharpening of the question regarding the nature of truth.
(2) Putnam [1978], p. 2.
What Tarski does is show how, in the context of a formalized language, one
can define ‘true’ (or a predicate which can be used in the place of ‘true’)
using only the notions of the objects language and notions of pure mathe-
matics. In particular, no semantical notion — no such notion as ‘designates’,
or ‘stands for’, or ‘refers to’ — is taken as primitive by Tarski (although
‘refers to’ gets defined — defined in terms of non-semantical notions — in
the course of his work). Thus anyone who accepts the notions of whatever
object language is in question — and this can be chosen arbitrarily — can
also understand ‘true’ as defined by Tarski for that object language. ‘True’
is just as legitimate as any notion of first order science.
Concerning semantics vs. syntax. In analytic philosophy, formal systems are

usually considered as representing an intuitive theory. Hence, semantics play an
essential role in analytic philosophy. The following quotation concerns modal logic.
Quotations 89.5. (1) Hintikka [1969], p. 23.

[[Syntactic]] methods are not always the best to create philosophical illumi-
nation in logic. The methods best suited to increase conceptual clarity are
here, as in many other areas of logic, the semantical ones (in the sense of
the term in which it has been applied to Carnap’s and Tarski’s studies). It
is not very helpful merely to put one’s intuition into the form of a deductive
system, as happens on the syntactical method. They are rarely sharpened
in the process.
(2) Leblanc [1973], p. 241.
Some like their logic mixed with a lot of ontology. To them first-order
logic, for example, adjudicates on basic matters of existence, and any se-
mantic account of it should come with things, sets of things, and relations
between things. They likewise think of modal logic as adjudicating on basic
questions of possibility, and to them any semantic account of it should be
complete with possible worlds, possible individuals, and so on.
Others prefer their logic straight: They view it as a handbook (of a
highly sophisticated kind, to be sure) for drawing inferences. Of course
there are things, and if one is to discourse about them, he must — sooner
or later — think of them as belonging to sets, bearing relations to one
another, and so on. But first-order logic can be, and hence is perhaps best,
explicated without recourse to “models”: What is dispensable simply is not
of the essence. And those who so wish may of course ponder over possible
worlds, possible individuals, ways in which the latter preserve their identity
from one possible world to another, and so on. But modal logic can be, and
hence is perhaps best, explicated without recourse to (Leibniz’ and) Kripke’s

musings.
89b. Truth, correspondence, and the problem of reference. I begin

with the problem of correspondence.
Quotation 89.6. O’Connor [1975], p. 70.

The sense in which two things may be said to correspond, match, fit or
accord with one another obviously depends on the nature of the things
that are being related. The word means different things in different cases.
Portraits correspond with their originals, maps with their terrain, the score
of a sonata with a performance, the Greek text of the Odyssey with a
translation, a structural formula in chemistry with the organization of the
molecule — the list could be extended indefinitely without bringing much
insight into the problem.
Comment. O’Connor is right: a list of this sort could be extended indefinitely with-
out bringing any insight into the problem. This is, indeed, the battlefield of phi-
losophy. It seems worthwhile, therefore, to point out my strategy. In O’Connor’s
example the two things which correspond are in some way significantly different. In
my approach in the groundworks, I investigate ways in which one and the same
thing can be related to itself — and not match.
As regards the role of symbols; also the problem of distinctions such as sign
and object, use and mention. The following quotations are intended to give an
impression of the diversity of possible positions.

Problems concerning the nature of signs and the relation of language to
reality find expression in such questions as: ‘Is language no more than a
system of signs? Has language a content, or does it float above reality like a
bubble above the earth? Can language point to something outside itself, has
it roots in some actuality or are the truths of language independent of all
experience? If language is a medium of communication (between human be-
ings) then what is it that is communicated, and how is this communication
effected?’
Comment. Notice the phrase “Has language a content ”. See variations 100.7 in the
groundworks for other congenial formulations.
(2) French [1967], p. 635.
What the later Wittgenstein did was to free people from the assumption
that language, in order to be meaningful, must denote entities existing in
one realm or another. According to Wittgenstein, the word ‘meaning’ is
being used illicitly if it is used to signify a thing that in one way or another
is said to correspond to the word. The meaning(s) of a word is not a denoted
object, but the way(s) the word is used in the language.
(3) Barwise [1980], p. 95.

Walker Percy writes: “The word is that by which the thing is conceived
or known.” Carl Jung puts another light on it: “Because there are innumer-
able things beyond the range of human understanding, we constantly use
symbolic terms to represent concepts we cannot fully understand.” [[. . .]]
[[. . .]] Man understands reality through the symbols he uses to represent
it[[.]]
Comment. If this is all you can expect when a mathematical logician engages in
philosophical ideas, then I don’t see much help coming from that corner.

The fundamental-seeming philosophical question, How much of our sci-
ence is merely contributed by language and how much is a genuine reflection
of reality? is perhaps a spurious question which itself arises wholly from a
particular type of language. Certainly we are in a predicament if we try
to answer the question; for to answer the question we must talk about the
world as well as about the language, and to talk about the world we must
already impose upon the world some conceptual scheme peculiar to our own
language.
Comment. What is shining up here is the problem addressed in quotation 65.2 (1)
and (2),37 and quotation 71.27 (3):38 talking about the world without employing
language. It may be worthwhile pointing out again that this is one of the central
issues of my enterprise. My aim is to get at the objective character of what Quine
dubs a ‘predicament’.
(2) Quine [1950], p. 79.
We can improve our conceptual scheme, our philosophy, bit by bit while
continuing to depend on it for support; but we cannot detach ourselves from
it and compare it objectively with an unconceptualized reality. Hence it is
meaningless, I suggest, to inquire into the absolute correctness of a concep-
tual scheme as a mirror of reality. Our standard for appraising basic changes
of conceptual scheme must be, not a realistic standard of correspondence to
reality, but a pragmatic standard. Concepts are language, and the purpose
of concepts and of language is efficacy in communication and in prediction.
Such is the ultimate duty of language, science, and philosophy, and it is in
relation to that duty that a conceptual scheme has finally to be appraised.
Comment. Good old pragmatism; advocating meaninglessness and ultimate duties;
pledging an open mind but any time ready to escape into higher spheres. The
problem is the “global indefiniteness” 39 of Quine’s “conceptual scheme” which shows
up in the phrase “as a mirror of reality”. It is not meaningless, I suggest, to inquire
into the “absolute correctness” of a conceptual scheme such as arithmetic, as a pure
form of thought, and it is possible to outline a somewhat immanent development à
37 Hegel regarding things in themselves.
38 Frege regarding objective.
39 Cf. Wang in quotation 95.13 in these materials; also quotations 95.5.
la Gödel: adding true but unprovable statements like that of consistency. It would
also not be meaningless, I suggest, to inquire into the “absolute correctness” of
what naively is called “reality”, and that as a form of thought, not as a mirror
of anything. In other words, it is not meaningless, I suggest, to inquire into the
dialectic of concepts.
§ 90. Logic, ontology, and flight from intension
The phrase “flight from intension” in the title of this paragraph may sound familiar40
and the reader is right in expecting that it is to some considerable extent dedicated
to a presentation of Quine’s concerns about the handling of intensional attitudes
and the discussion it evoked. It seems to tell of some kind of primordial fear of
intension.
As an opening to this paragraph, I include a quotation which, to my mind, dis-
plays beautifully a fin de siècle-complacency, equalling Wood in quotation 69.9 (2)
in these materials.
Quotation 90.1. D. Bell [1990], p. 87.

For us, now, in the last quarter of the twentieth century, the nature of logic
— that is, of the discipline of formal, deductive logic — is very largely un-
problematic: it is a pure science devoted to the investigation and codification
of relations of deductive consequence holding between sentences, or perhaps
between the thoughts or propositions they express. And in this connection
we understand the need to distinguish (proof theoretic) relations of syntac-
tic consequence and (model theoretic) relations of semantic consequence;
there is a general consensus as to how issues concerning, say formal sche-
mata, calculi, interpretation, truth, validity, consistency, and completeness
are related one to the other; and today we can be clearer than ever before
about how, if at all, the subject matter of logic is related to that of other
disciplines like psychology, mathematics, set theory, ontology, epistemology,
or linguistics.
90a. Quantification, variables, and ontology. Recall that in section 86d

I quoted Carnap regarding his aversion to questions of the existence or reality of
the framework itself.

A curious thing about the ontological problem is its simplicity. It can
be put in three Anglo-Saxon monosyllables: ‘What is there?’ It can be
answered, moreover, in a word—‘Everything’—and everyone will accept
this answer as true.
Comment. Sure, everyone will accept this answer as true; just as sure as “all the
world would regard” it more advantageous to “modify the laws of optics and suppose
that light does not travel rigorously in a straight line” than to “renounce Euclidean
40 If not, cf. chapter VI in Quine [1960].
§ 90. LOGIC, ONTOLOGY, AND FLIGHT FROM INTENSION 1245
geometry”.41 I regard this passage as a paragon of analytic mindlessness. For a

start there is the conflation of being and existence. If the question is “what can be
predicated?”, then the answer may well be “everything”, because we can’t speak of
what cannot be predicated (Parmenides); but if the question is “what exists?”, then
we may well say “not everything; unicorns, for example, do not exist”.
(2) Quine [1948], pp. 1 f.
Suppose McX maintains there is something which I maintain there is not.
McX can, quite consistently with his own point of view, describe our dif-
ference of opinion by saying that I refuse to recognize certain entities. [[. . .]]
When I try to formulate our difference of opinion, on the other hand,
I seem to be in a predicament. I cannot admit that there are some things
which McX countenances and I do not, for in admitting that there are such
things I should be contradicting my own rejection of them.
[[. . .]]
This is the old Platonic riddle of nonbeing. Nonbeing must in some
sense be, otherwise what is it that there is not? This tangled doctrine might
be nick-named Plato’s beard ; historically it has proved tough, frequently
dulling the edge of Occam’s razor.
Comment. It is as if Frege had never (been translated as having) said that affirma-
tion of existence “is in fact nothing but denial of the number nought” and “existence
is a property of concepts”. But then, Quine probably could not have said that a
concept is empty either.
Quine establishes a link between what he calls the “ontological problem” and
the various positions in the foundations of mathematics.
Quotation 90.3. Quine [1948], pp. 14 f.

[[T]]he modern philosophical mathematicians have not on the whole recog-
nized that they were debating the same old problem of universals in a newly
clarified form. But the fundamental cleavages among modern points of view
on foundations of mathematics do come down pretty explicitly to disagree-
ments as to the range of entities to which the bound variables should be
permitted to refer.
The three main mediaeval points of view regarding universals are des-
ignated by historian as realism, conceptualism, and nominalism. Essentially
these same three doctrines reappear in twentieth-century surveys of the phi-
losophy of mathematics under the new names logicism, intuitionism, and
formalism.
Realism, as the word is used in connection with the mediaeval contro-
versy over universals, is the Platonic doctrine that universals or abstract
entities have being independently of the mind; the mind may discover them
but cannot create them. Logicism, represented by Frege, Russell, White-
head, Church and Carnap, condones the use of bound variables to refer to
41 Poincaré [1913], p. 81; see quotation 70.23 (3) in these materials.
abstract entities known and unknown, specifiable and unspecifiable, indis-

criminately.
Conceptualism holds that there are universals but they are mind-made.
Intuitionism, espoused in modern times in one form or another by Poincaré,
Brouwer, Weyl, and others, countenances the use of bound variables to refer
to abstract entities only when those entities are capable of being cooked up
individually from ingredients specified in advance. As Fraenkel has put it,
logicism holds that classes are discovered while intuitionism holds that they
are invented—a fair statement indeed of the old opposition between realism
and conceptualism. [[. . .]]
Formalism, associated with the name of Hilbert, echoes intuitionism in
deploring the logicist’s unbridled recourse to universals. But formalism also
finds intuitionism unsatisfactory. This could happen for either of two oppo-
site reasons. The formalist might, like the logicist, object to the crippling
of classical mathematics; or he might, like the nominalists of old, object to
admitting abstract entities at all, even in the restrained sense of mind-made
entities. The upshot is the same: the formalist keeps classical mathematics
as a play of insignificant notations.
Quotations 90.4. (1) Kahn [1973], p. 1.

[[T]]he very term “ontology” is derived from ὤν/ὄντος, the present participle
of the Greek verb be. Etymologically speaking, ontology is the theory of
being or what is. At the same time, it is not entirely clear how far a theory
of the verb belongs to linguistic, how far it belongs to philosophy proper.
Comment. The difference between Quine and Kahn: the specification there.
(2) Van Fraassen [1976], p. 102.
Being [[. . .]] belongs to any subject of discourse, existence or non-existent,
possible or impossible, real or imaginary or unimaginable or inconceivable.
(3) Benveniste [1971], p. 164; pinched from Derrida [1971], p. 201.
What matters is to see clearly that there is no connection, either by nature
or by necessity, between the verbal notion of ‘to exist, really to be there’
and the function of the ‘copula.’ One need not ask how it happens that the
verb ‘to be’ can be lacking or omitted. This is to reason in reverse. The real
question should be the opposite: how is it that there is a verb ‘to be’ which
gives verbal expression and lexical consistency to a logical relationship in
an assertive utterance?
I continue with a few quotations regarding the notion of the variable.
Quotation 90.5. Shoenfield [1967], p. 7.

Unlike a name, which has only one meaning, a variable has many meanings.
In an analysis text, a variable may mean any real number; or, as we shall
say, a variable varies through the real numbers. However, a variable keeps
the same meaning throughout any one context.
Comment. This is a point where dialectical logic may be said to deviate from clas-
sical logic: in one way of looking at it the stipulation that a variable can always
keep its meaning in the context of a deduction will be rejected. More to the point:
the truth value of a sentence variable may not be the same after this variable has
been engaged in an inference.

The variable qua variable, the variable an und für sich and par excellence,
is the bindable objectual variable. It is the essence of ontological idiom, the
essence of the referential idiom.
(2) Quine [1976], pp. 274.
Functional abstraction or class abstraction can be taken as basic, and quan-
tification can be defined in terms of either of them. Church took the one
line in his lambda calculus, and I the other in my logic based on inclusion
and abstraction. Even Peano had a full-blown class abstraction on which
he based his existential quantification, though his universal quantification
took another line. However, reduction to quantification is generally to be
preferred to these alternative reductions, because quantification is wanted
not only in set theory but also in elementary theories where there is no call
for classed or functions; and moreover the logic of quantification, unlike set
theory, is complete, compact, and convenient.
Such, then is a standard theory, in Tarski’s phrase; it is simply quantifi-
cation logic, or the predicate calculus, with one or another fixed lexicon of
predicates appropriate to one or another particular subject matter. The on-
tology of a theory, thus standardized, is the range of values of the variables
of quantification; for the variables are the variables of quantification. And
of course it is clear from the readings of objectual quantification in ordinary
language that the ontology consists of those values; for the quantifiers mean
‘everything is such that’ and ‘something is such that’.
(2) Quine [1976], pp. 276 f.
Peano [[. . .]] introduced the inverted epsilon for the words ‘such that’, or
for the equivalent in his three romance languages [[. . .]]. But he introduced
no functor for class abstraction. He saw his inverted epsilon as already
class abstraction; here was his confusion. He did not distinguish between
the general term, or predicate, and the class name, a singular term.
The same conflation may be seen in Peano’s upright epsilon. For the ep-
silon that is now standard in set theory comes from Peano; and he adopted
it as his copula of predication, the initial of the Greek ἐστί. He inverted it
for his ‘such that’ because this is the inverse of predication; the two cancel.
Peano was thus sensitive to the relation between predication and relative
clauses: ‘a is a thing x such that F x’ reduces to ‘F a’. He was indeed sensitive
to the role of the variable as relative pronoun; he was explicit on this. But
he must be convicted of the conflation, on two counts. He provides explic-
itly that his ‘such that’ expressions designate classes; ‘(x p) ǫ Cls’. And,
what is more to the point, he quantifies over them with bound variables;
whereas a relative clause or ‘such that’ clause properly conceived is rather

a predicate, susceptible at best of substitution for an unbindable schematic
predicate letter. [[. . .]] With all his sensitivity to grammar he was insensitive
to the distinction between general and abstract singular; insensitive to the
ontological import of values of variables.
90b. Possible worlds, referential transparency and extensional opac-

ity.42 According to a naive view, the terms of a language refer in one way or another
to objects. But of what kind are these objects? This question becomes particularly
urgent when modal contexts are concerned. The relevance of this issue for my enter-
prise has to be seen in the failure of substitutivity of equal terms, i.e., extensional
opacity, in modal contexts.
The issue had already been raised by Frege who stressed the extensional char-
acter of logic.43
Quotation 90.7. Geach [1972], p. 226.

Frege held a purely extensional view of concepts. He adopted the mathe-
matician’s attitude toward definitions—that it does not much matter which
definition you choose, so long as the same objects come under it; and he
expressly says that, though proper identity holds only between objects, the
analogue of identity for concepts holds if and only if concepts are coexten-
sive, i.e., have the same objects falling under them.
Comment. But Frege also considers the case of the ‘ungerade Bedeutung’ in quo-
After Frege’s analysis of sense and reference, there came the idea of possible
worlds.
Quotation 90.8. Kripke [1959], p. 2.

The basis of the informal analysis which motivated these definitions [[of
validity in a model]] is that a proposition is necessary if and only if it is
true in all “possible worlds.” (It is not necessary for our present purposes
to analyze the concept of a “possible world” any further.)
Comment. This is how the notion of possible worlds entered the philosophical stage
and set the standard of discussion.
The next quotation marks the opposite extreme; apart from that it strikes me
as a perfect summary of its author’s position, in content as well as form.

Intensions are creatures of darkness, and I shall rejoice with the reader
when they are exorcised[[.]]
42 My warmest thanks go to Matthias Kaiser in Oslo for having proffered heaps of ideas and
materials regarding Quine’s position on quantified modal logic.

43 Cf. quotations 71.29 (1) and (2) in these materials.
Comment. Exorcism is the word. May I also suggest Inquisition, auto-da-fé, and
the stake? Anyway, inasmuch as I am a reader, I resent being incorporated in this
way into the eerie dream world of an orthodox fanatic.
Remark 90.10. Unfortunately, one of the most original and challenging positions,
that of Ruth Barcan Marcus, is not represented in this section. For the moment
I can do no more than refer readers to Barcan Marcus [1960], [1961], [1962], and
[1981].
Modal distinction collapses if equivalent wffs can be substituted mutually.44
From this Quine concludes that the scope of is extensionally opaque. No substi-
tution within the scope of . In view of this, Quine goes on to propose to give up the
modalities, at least in the form of sentential operators which allow in-quantification.
There is, however, another option: to give up the purely referential interpretation
of variables.
One of Quine’s points concerning the modalities has been that quantified modal
logic commits one to essentialism, i.e., the view that an object, of itself and by
whatever name or none, must be seen as having some of its traits necessarily and
others contingently, despite the fact that the latter traits follow just as analytically
from some ways of specifying the objects as the former do from other ways of
specifying it.
Quotation 90.11. Goodman [1947], pp. 122 f.
Suppose, for example, that all I had in my pocket on V-E day was a group
of silver coins. Now we would not under normal circumstances affirm of a
given penny P
If P had been in my pocket on V-E day, P would have been silver,
even though from
P was in my Pocket on V-E day
we can infer the consequent by means of the general statement
Everything in my pocket on V-E day was silver.
Quotations 90.12. (1) Quine [1953c], pp. 158 f.

There are three different degrees to which we may allow logic, or seman-
tics, to embrace the idea of necessity. The first or least degree of acceptance
is this: necessity is expressed by a semantical predicate attributable to state-
ments as notational forms—hence attachable to names of statements. [[. . .]]
A second and more drastic degree in which the notion of necessity may
be adopted is in the form of a statement operator. [[. . .]]
Finally the third and gravest degree is expression of necessity by a sen-
tence operator. This is an extension of the second degree, and goes beyond
it in allowing the attachment of ‘nec’ not only to statements but also to
open sentences, such as ‘x > 5’, preparatory to the ultimate attachment of
quantifiers[[.]]
44 See section 30d in the tools.
Comment. Perhaps it is worthwhile noting here for the casual reader who doesn’t
quite know why such a quotation has found its way into these materials: in §134
in the groundworks, I shall do just what Quine labels the ‘third and gravest
degree’, viz. introduce a sentence operator of necessity — but I shall do so by
explicit definition, and that in a system of abstraction and inclusion; in other words,
it doesn’t amount to an extension.
(2) Quine [1960], p. 196.
‘Implies’ and ‘is analytic’ are best viewed as general terms, to be pred-
icated of sentences by predicative attachment to names (e.g. quotations) of
sentences. In this they contrast with ‘not’, ‘and’ and ‘if-then’, which are not
terms but operators attachable to the sentences themselves. Whitehead and
Russell, careless of the distinction between use and mention of expressions,
wrote ‘p implies q’ (in the material sense) interchangeably with ‘If p then q’
(in the material sense). Lewis followed suit, thus writing ‘p strictly implies
q’ and explaining it as ‘Necessarily not (p and not q)’.
(3) Quine [1977], p. 113.
The predicate is more comfortable than the sentence functor, for it occasions
no departure from extensional logic.
(4) Quine [1962], p. 323.
Lewis founded modern modal logic, but Russell provoked him to it. For
whereas there is much to be said for the material conditional as version
a of ‘if-then’, there is nothing to be said for it as a version of ‘implies’;
and Russell called it implication, thus apparently leaving no place open
for genuine deductive connections between sentences. Lewis moved to save
the connections. But his way was not, as one could have wished, to sort
out Russell’s confusion of ‘implies’ with ‘if-then’. Instead, preserving that
confusion, he propounded a strict conditional and called it implication.
I contrast the foregoing quotations with one from Hilbert and Bernays which
I find a philosophically satisfactory formulation of the problem of the so-called
‘paradoxes of material implication’.
Quotation 90.13. Hilbert and Bernays [1934/68], p. 47, footnote 1.

Die sprachlichen Aussagen-Verknüpfungen sind Verknüpfungen der Aussa-
gen selbst, nicht ihrer Wahrheitswerte. Dieser Unterschied macht sich al-
lerdings für die Konjunktion und die Negation nicht fühlbar, wohl aber für
die disjunktive und die hypothetische Verknüpfung. Die Bedeutung einer
Aussage „A oder B“ ist, daß A, B Möglichkeiten sind, die zusammen den
Bereich der Möglichkeiten in einer Hinsicht erschöpfen. Eine Aussage „wenn
A, so B“ bringt einen Zusammenhang zum Ausdruck, zufolge dessen das
Zutreffen von A ein Erkenntnisgrund für das Zutreffen von B ist. In beiden
Fällen läßt sich der Inhalt der Aussage nicht einfach als eine Beziehung
zwischen den Wahrheitswerten von A und B ausdrücken.
Comment. Bernays speaks of “Erkenntnisgrund”. Note the proximity to Frege’s
“Erkenntniswert”.
Quotations 90.14. (1) Quine [1953c], pp. 175 f.

Aristotelian essentialism. This is the doctrine that some of the attributes of
a thing (quite independently of the language in which the thing is referred
to, if at all) may be essential to the thing, and others accidental. E.g., a
man, or talking animal or featherless biped (for they are in fact all the same
things), is essentially rational and accidentally two-legged and talkative, not
merely qua man but qua itself.
Comment. So Locke, with his distinction of primary and secondary qualities, was
an Aristotelian essentialist.
(2) Quine [1953c], p. 176.
Necessity as semantical predicate reflects a non-Aristotelian view of
necessity: necessity resides in the way in which we say things, and not
in the things we talk about. Necessity as statement operator is capable
[[. . .]] of being reconstrued in terms of necessity as a semantical predicate,
but has, nevertheless, its special dangers; it makes for an excessive and
idle elaboration of laws of iterated modality, and tempts one to a final
plunge into quantified modality. This last complicates the logic of singular
terms; worse, it leads us back into the metaphysical jungle of Aristotelian
essentialism.
Quotations 90.15. (1) Føllesdal [1969a], pp. 147 f.

According to Quine, it is difficult to make any sense of such sentences,
in particular it is unclear what the objects are, if any, that are referred to
in modal contexts. Consider, for example,
(1) (∃x)N(x > 7)
“Would 9, that is, the number of planets, be one of the numbers neces-
sarily greater than 7?” Quine has asked, pointing out that such an affirma-
tion would be true in the form
(2) N(9 > 7)
and false in the form
(3) N(the number of planets > 7)
The difficulty, according to Quine, is due to the modal context’s being
referentially opaque; ‘the number of planets’ cannot be substituted for ‘9’
in (2) salva veritate although the two expressions refer to the same object.
In order to make sense of quantification, Quine concludes, the positions of
the variables have to be referential in the sentence following the quanti-
fier, names which refer to the same object must be interchangeable in these
positions[[.]]
(2) Føllesdal [1969a]. pp. 155 f.
[[G]]iven Quine’s thesis, if one wants to quantify into modal contexts without
having modal distinctions collapse, then these contexts have to be referen-
tially transparent and extensionally opaque. And essentialism is just this
combination of referential transparency and extensional opacity: whatever

is true of an object is true of it regardless of how it is referred to (referen-
tial transparency), and among the predicates true of an object, some are
necessarily true of it, others only accidentally (extensional opacity).
(3) Hintikka [1975], p. 103 (giving an account of Quine’s position).
Q. Modern logic was conceived in sin, the sin of confusing use and mention.
It originated from C. I. Lewis’ notion of strict implication, which was an
unwitting and illegitimate attempt to say in a language something about
that language. Hence one can never hope to make decent interpretational
sense of modal logic, and the same goes a fortiori for quantified modal
logic.
Comment. Note the phrase “to say in a language something about that language”;
compare Findlay in quotation 69.25 (3) in these materials.
(4) Hintikka [1975], pp. 105 f (giving an account of Quine’s position).
[[T]]he problem of quantified modal logic turns on the semantics of singu-
lar terms and other referring expressions. How are they related to their
objectual targets? It ought to be a tautology to say that the task of such
referring expressions as proper names is to refer to individuals, to pick out
members of one‘s universe of discourse. But the striking thing about modal
contexts is that this not what our referring expressions do there, or is not
at all that they do. If it were, substitutivity of identity (SI) and existential
generalization (EG) would apply in these contexts. [[. . .]]
Thus the notorious failure of these laws in modal contexts shows that
in these contexts our referring expressions do not just refer. Something else
is involved here besides the references of our terms. The truth of modal
statements does not depend only on the individuals referred to, but also on
the referring expressions themselves. [[. . .]]
[[. . .]] For what sense would it make to ask whether there exists an
individual satisfying an open sentence F (x) if an individual‘s satisfying
this open sentence does not depend on it alone but also on the way it is
referred to?
[[. . .]]
[[. . .]] SI and EG [[. . .]] are our bridges to the notions of individual and
ontology. If they are destroyed we do not know where to turn for our individ-
uals. It is for this reason that the failure of the SI and of EG is absolutely
crucial interpretationally. It shows that in modal contexts it is not clear
what the individuals are that we are really speaking of, for in such con-
texts we just are not any longer speaking of ordinary individuals alone in
an ordinary way.
(5) Montague [1963], p. 294.
[[I]]f necessity is to be treated syntactically, that is, as a predicate of sen-
tences, as Carnap and Quine have urged, then virtually all of modal logic
[[. . .]] must be sacrificed.
Comment. Montague’s result can be found as lemma 48.33v in the tools.
(6) Hintikka [1975], p. 104.

C. [[. . .]] What you are willing to grant to modal logicians they cannot
have in any case, and the real defence of their position remains unrecognized
by you . . .
Q. So much the worse for modal logic. What you are averring is simply
that even the lower degrees of modal involvement cannot be defended in the
teeth of unproblematic elementary arithmetic. What could vindicate more
completely my criticism of modal logic?
Comment. The allusion made is to Montague’s antinomy.45
Quotation 90.16. Føllesdal [1969a], p. 148.
As observed by Quine, the difficulties connected with quantification into
modal contexts would vanish if one were to exclude from one’s universe of
discourse all objects that can be uniquely specified in ways which are not
necessarily equivalent to one another and retain only objects such that any
two conditions uniquely determining x are necessarily equivalent, i.e. such
that:
(5) (y)(F y ≡ . y = x).(y)(Gy ≡ . y = x). ⊃ N(y)(F y ≡ Gy)
However, as Quine points out in From a Logical Point of View (1st ed.,
pp. 152–153), (5) has consequences which some modal logicians might be
reluctant to accept, for example:
(x)(y)(x = y ⊃ N(x = y))
that is, all identities are necessary.
In Word and Object, Quine draws a further, disastrous consequence
from (5), viz. the consequence that every true sentence is necessarily true
or
p ⊃ Np
Since the converse holds, too, this means that modal distinctions collapse;
the whole point of the modalities vanishes.
Comment. Such are the costing of deisimilitude. There is no room for intensionali-
ties in the heavens.
I close this section with a quotation from Hintikka which may be regarded as a
sufficient explanation, at least by some, of the difficulties pointed out by Quine.
Quotation 90.17. Hintikka [1962], pp. 139 f.
[[T]]he failure of referential transparency in epistemic contexts is due to the
possibility that two names or other singular terms which de facto refer
to one and the same object (or person) are not known (or believed) by
someone to do so and that they will therefore refer to different objects (or
persons) in some of the “possible worlds” we have to discuss (explicitly or
implicitly) in order to discuss what he knows or believes and what he does
45 As regards Montague’s antinomy, see remark 48.26 in the tools.
not. The referential opacity is not due here to anything strange happening
to the ways in which our singular terms refer to objects nor to anything
unusual about the objects to which they purport to refer. It is simply and
solely due to the fact that we have to consider more than one way in which
they could refer (or fail to refer) to objects. What we have to deal with
here is therefore not so much a failure of referentiality as a kind of multiple
referentiality.
Comment. That comes a bit closer to my view (systemic ambiguity) than anything
else. The point is, what constitutes multiple referentiality?
90c. Description, identity, and substitutivity. One way to get rid of the
problems linked to the failure of substitution is to treat definite description contex-
tually. This possibility seems to have been first pointed out in Church [1942] and
was discussed at greater detail by A. F. Smullyan.
Quotations 90.18. (1) A. F. Smullyan [1948], p. 35.
[[T]]he modal paradoxes arise not out of any intrinsic absurdity in the use
of the modal operators but rather out of the assumption that descriptive
phrases are names.
(2) A. F. Smullyan [1948], p. 37.
It is not, essentially, the unrestricted use of modal operators which violates
Leibniz’s Law. It is rather that the modal paradoxes arise out of neglect of
the circumstance that in modal contexts the scopes of incomplete symbols,
such as abstracts or descriptions, affect the truth value of those contexts.
(3) A. F. Smullyan [1948], p. 35.
[[T]]he logical modalities need not involve paradox when they are referred to
a system in which descriptions and class abstracts are contextually defined.
Comment. In other words, a no-class theory.46
Quotations 90.19. (1) Montague and Kalish [1959], pp. 85 f.
We [[. . .]] impose on Leibniz’ principle the restriction that the replaced oc-
currence of a name be proper. [[. . .]]
There are some difficulties [[. . .]] for which the remedy is not obvious.
Perhaps the most conspicuous among these are the cases involving the con-
junction ‘that’. It is to such cases that the present discussion is directed.
We shall further restrict our attention to a single rule, Leibniz’ principle in
its revised form. Our remarks, however will apply with slight modification
to a number of other logical principles.
We begin with three examples:
(3) The number of planets = 9.
Kepler was unaware that the number of planets > 6
Therefore Kepler was unaware that 9 > 6.
(4) 9 = the number of planets.
It is necessary that 9 = 9.
46 Cf. Russell in quotation 75.10 (6) in these materials.
Therefore it is necessary that the number of planets = 9

(5) 9 = the number of planets.
It is provable in arithmetic that 9 is not prime.
Therefore it is provable in arithmetic that the number of planets
is not prime.
In each of these examples, the premisses are ordinarily regarded as true and
the conclusion false, and each argument seems to be an instance of Leibniz’
principle. The common characteristic is the occurrence of the subordinate
conjunction ‘that’. (Similar paradoxes can be generated with the help of
‘whether’ or the interrogatives ‘what’, ‘where’, ‘why’, ‘when’, ‘who’, ‘how’,
in their use as subordinate conjunctions. [[. . .]])
Comment. Note the proximity to Aristoteles’ categories and compare also Frege in
quotation 71.29 (2) in these materials. This forms one of the cornerstones of my
approach to a foundation of categories.
(2) Kaplan [1969], p. 194.
[[W]]e should restrict our attention to a smaller class of names; names which
are so intimately connected with what they name that they could not but
name it. I shall say that such a name necessarily denotes its object [[. . .]].
Such a relation is available; based on the notion of a standard name.
A standard name is one whose denotation is fixed on logical, or perhaps
I should say linguistic, grounds alone. Numerals and quotation names are
prominent among the standard names. Such names do, in the appropriate
sense, necessarily denote their denotations.
Comment. The problem, I suggest, lies hidden in the phrase “in the appropriate
sense”; it will be inappropriate, I suppose, to ask what the code of Gödel’s sentence
(with regard to some assignment of natural numbers to the expressions of a formal
language) denotes. There is no assignment which would allow us to express that
predicate within the theory itself.
(3) Van Fraassen [1976], p. 127.
The usual pattern of predicate modification at the center of the discus-
sions by Reichenbach, Davidson, and [[R. H.]] Thomason, is this: A subject is
qualified by a number of expressions, each of which qualifies this predicate
in some particular respect. These qualifiers answer such questions as:
Where?
How?
With what?
When?
To whom?
While what else happened?
How much?
Comment. Cf. Aristoteles’ categories in quotations 57.15.

[[I]]f identity is taken strictly as the relation that every entity bears to itself
only, [[Hume]] is at a loss to see what is relational about it, and how it differs
from the mere attribute of existing. Now the root of the trouble is confusion
of sign and object. What makes identity a relation, and ‘=’ a relative term,
is that ‘=’ goes between distinct occurrences of singular terms, same or
distinct, and not that it relates distinct objects.
Similar confusion of sign and object is evident in Leibniz where he
explains identity as a relation between the signs, rather than between the
named object and itself: “Eadem sunt quorum unum potest substitui alteri,
salva veritate.” Frege at one time took a similar line. This confusion is
curiously doubled in Korzybski, when he argues that ‘1 = 1’ must be false
because the two sides of the equation are spatially distinct.
Identity evidently invites confusion between sign and object in men
who would not make the confusion in other contexts. Those involved in-
clude most of the mathematicians who have liked to look upon equations
as relating numbers that are somehow equal but distinct. Whitehead once
defended the view, writing e.g. that “2 + 3 and 3 + 2 are not identical; the
order of the symbols is different in the two combinations, and this difference
of order directs different processes of thought.” It is debatable how much
this defense depends on confusion of sign and object, and how much on a
special doctrine that numbers are thought processes. Wittgenstein’s mis-
take is more clearly recognizable, when he objects to the notion of identity
that “to say of two things that they are identical is nonsense, and to say
of one thing that it is identical with itself is to say nothing.” Actually of
course the statements of identity that are true and not idle consist of unlike
singular terms that refer to the same thing.
Comment. This quotation has to be seen in the context of Frege’s claim that an
opposition to his non-distinction of equality and identity “will very likely rest on
an inadequate distinction between sign and thing signified.” 47 The issue is of major
relevance in the context of my enterprise and readers might find it interesting to
learn that my own approach in the third book is necessarily based on a distinction of
equality and identity, simply because even a weak form of extensionality is refutable
in the presence of unrestricted abstraction with remarkably modest logical means.48
The point with regard to Quine is his “divinisimilitude”: he seems satisfied with
establishing that identity is a relation that everything bears to itself; but he doesn’t
consider the process that may be necessary to establish that we are dealing with
one and the same thing. The problem is the underlying referential semantics. My
point is not so much to criticize this approach as to provide an alternative. The
basic line of my criticism can be found in chapter XXX.
47 Cf. Frege [1893], p. IX; see quotation 71.20 (1) in these materials.
48 Cf. U. Petersen [2000], p. 376.
Quotation 90.21. Caton [1976], p. 173.

In connection with the relation of identity, Leibniz is frequently credited
with having enunciated a principle of substitution. This principle is taken
to be expressed by his dictum „eadem sunt, quorum unum potest substitui
alteri salva veritate“, a principle which appears on its face to speak of things
themselves being substituted one for the other but which has often in re-
cent times, been taken to be a principle of the substitutivity of expressions
which designate identical things. We may refer to these respectively as the
principle of substitutivity of identicals and the principle of substitutivity of
designations of identicals.
Comment. It may be worthwhile having a look at the context of the Leibniz “dic-
tum”, quotation 70.10 in these materials to see that Leibniz is explicitly speaking
of propositions and that it is formulated as a definition, not a principle.
I close this section with a quotation regarding the relevance of these issues for
empiricism.
Quotations 90.22. (1) Føllesdal [1965], p. 264.

[[T]]he usefulness of the notions of physical necessity, contrary-to-fact con-
ditionals, etc., depends largely on whether one can quantify into them, as
illustrated e.g. in Quine’s example [[. . .]] concerning solubility in water. If
one cannot find a quantifying into these contexts, then these notions, to-
gether with a wealth of related notions, will probably have to be abandoned,
to the extent that they cannot be made clear without appeal to physical
necessity, e.g. in the case of dispositions by conceiving them as enduring
structural traits of objects.
However, to seems to me that arguments analogous to those of Quine
against quantification into logical modalities apply to the contexts of phys-
ical necessity and related contexts as well.
(2) Føllesdal [1965], p. 270.
[[I]]f we want to make sense of quantification into causal contexts, we appar-
ently have to require that no identity statement which is true in our actual
world is false in any physically possible world.
What this amounts to, is that in order to make sense of quantification,
identity should be universally substitutive. Tampering with the substitutiv-
ity of identity prevents the derivation of undesirable results, but only at the
cost of making both identity and quantification unintelligible.
Comment. I do not want to contest the claim of unintelligibility, I only want to
ask for whom; i.e., what, precisely, are the ideas that constitute a framework in
which identity (or, rather, equality) and quantification are so irrevocably tied to
each other; and if it is only to be able to distance myself from these ideas.
90d. Intentionality. This section contains only two quotations, and that es-
sentially from a single paper — I just didn’t get any further.
Quotations 90.23. (1) Brentano, quoted after Føllesdal [1972] , p. 418.

Every mental phenomenon is characterized by what the scholastics in
the Middle Ages called the intentional (and also mental) inexistence of an
object, and what we could call, although in not entirely unambiguous terms,
the reference to a content, a direction upon an object.
(2) Føllesdal [1972], p. 418.
Brentano held that even in these cases our mental activity, our thinking or
our sensing, is directed towards some object. The directedness has nothing
to do with the reality of the object, Brentano held. The object is itself
contained in our mental activity, ‘intentionally’ contained in it.
§ 91. The return of metaphysics 49

Although there might not really be an agreement as to what metaphysics is, after all,
philosophy in the analytic tradition has seen a comeback of questions and problems
which would have been dismissed as metaphysical in the early decades of the XXth
century.
91a. Notions of realism. I begin with a few quotations regarding various
notions of realism.
Quotations 91.1. (1) Weitz [1966], p. 1.

Realism as formulated in the twentieth century by G. E. Moore and
Bertrand Russell, its founders and leading exponents, is the dualistic doc-
trine that mind and matter and universals and particulars are ultimate.
(2) Popper [1982], p. 2.
[[R]]ealism. That is to say, the reality of the physical world we live in: the
fact that this world exists independently of ourselves; that it existed before
life existed, according to our best hypotheses; and that it will continue to
exist, for all we know, long after we have all been swept away.
(3) Horwich [1982], p. 181.
The debate surrounding realism is hampered by an aversion to explicit for-
mulation of the doctrine. The literature is certainly replete with resounding
one-liners: ‘There are objective facts’, ‘Truth is correspondence with real-
ity’, ‘Reality is mind-independent’, ‘Statements are determinately either
true or false’, ‘Truth may transcend our capacity to recognize it’. But such
slogans are rarely elaborated on. All too often the argument, for or against,
will proceed as though the nature of realism were so well-understood that
no careful statements of the position is required. Consequently, several dis-
tinct and independent positions have at various times been identified with
realism and the debate is marked by confusion, equivocation and arguments
at cross-purposes to one another.
49 I did not say “the return of the rag queen”; and even less did I add “to the lumpenphiloso-
phy”. But I would be lying if I said it hadn’t crossed my mind.

§ 91. THE RETURN OF METAPHYSICS 1259
Comment. This sounds promising; but what follows is as weak as everything else
in this field.
Quotation 91.2. Lachs [1967], p. 285.
[[T]]here are two forms of realism we must distinguish. One is a theory about
universals and their relation to thought; the other is a theory about particles
and their relation to sense. The first maintains that universals are neither
mental in nature, nor dependent on thought for their being or existence.
The second contends that spatio-temporal particulars are non-mental, and
exist independently or our perception of them. The theories are similar in
making the essentialist realist claim that consciousness does not create its
objects: they differ only in the type of consciousness, and hence the type of
object, concerning which their claims are made.
Quotations 91.3. (1) Hirst [1967], p. 78.
Direct realism is the general view that perception is a direct aware-
ness, a straightforward confrontation (or being in touch, contact) with the
external object.
(2) Hirst [1967], p. 82.
A closer examination is required not only of the concepts of datum and
reference but also of the general relation of mind and body presupposed
in perception and of the nature of mental contents; above all the theory
must take full account of the numerous quasi-interpretative activities which
modern psychology has found in perception.
(3) Dummett [1982], pp. 110.
A naive realist about the physical world supposes that, in perception, we are
in direct contact with physical objects: we know them as they really are.
When, under normal conditions, I perceive an object, a Cartesian doubt
is impossible, according to the naive realist: it would be senseless, given
my perceptual state, to suppose that the object was not present or was
otherwise than I perceive it to be; mistakes occur only because perception
does not always take place under normal conditions.
Quotations 91.4. (1) Dummett [1982], p. 55.
[[R]]ealism is a view about a certain class of statements [[. . . W]]e may regard
a realistic view as consisting in a certain interpretation of statements in
some class, which I shall call ‘the given class’.
So construed, realism is a semantic thesis, a thesis about what, in
general, renders a statement in the given class true when it is true. The
very minimum that realism can be held to involve is that statements in the
given class relate to some reality that exists independently of our knowledge
of it, in such a way that that reality renders each statement in the class
determinately true or false, again independently of whether we know, or
are even able to discover, its truth-value. Thus realism involves acceptance,
for statements of the given class, of the principle of bivalence, the principle
that every statement is determinately either true or false.
(2) Dummett [1982], p. 56.

Rejection of the principle of bivalence for statements of some given
class always involves a repudiation of a realistic interpretation of them; and
adoption of an anti-realistic view often turns critically upon such a rejection
of bivalence.
Comment. This strikes me as interesting in view of Priest’s dialetheism which not
only seems to combine realism and a rejection of bivalence (in the sense of allowing
truth value gluts), but actually claims that realism cannot escape dialetheism.50
(3) Dummett [1982], p. 62.
It is natural to say that whether or not a sentence is true depends both on
its meaning and on the way the world is, on the constitution of external re-
ality; and, since the semantic values of the component expressions together
determine whether or not the sentence is true, it is plain that, in associ-
ating particular semantic values with these expressions, we have already
taken the contribution of external reality into account. But, given the way
the world is, whether a sentence is or is not true depends upon its meaning;
so, given the way the world is, the semantic value of an expression depends
only on its meaning. It follows that a grasp of the meaning of a specific ex-
pression must be something which, taken together with the way the world
is, determines the particular semantic value that is [[sic]] has. Thus meaning
must be something that determines semantic value: in Frege’s terminology,
sense determines reference.
Comment. Note the awkward phrase “given the way the world is” which is finally
dropped. I can see no reason, and it is not explained, how it is justified to be
suddenly dropped. I find this a thoroughly misleading way of putting things.
(4) Putnam [1983], p. 298.
Dummett uses the term “realism” for a view which:
(1) Assumes a world consisting of a definite totality of discourse-inde-
pendent objects and properties;
(2) Assumes “strong bivalence” — that is, that an object either deter-
minately has or determinately lacks any property P which may significantly
be predicated of that object; and
(3) Assumes the correspondence theory of truth in a strong realist sense
of “correspondence”: i.e., a predicate corresponds to a unique state of af-
fairs, involving the properties and objects mentioned in (1), and is true if
that state of affairs obtains and false if it does not obtain.
(5) Putnam [1982], p. 141.
What the metaphysical realist holds is that we can think and talk about
things as they are independently of our minds, and that we can do this by
virtue of a “correspondence” relation between the terms in our language and
some sorts of mind-independent entities. [[. . .]] Today material objects are
taken to be paradigm mind-independent entities, and the “correspondence”
50 Cf. quotations 91.7 below.
is taken to be some sort of causal relation. [[. . .]] On this view, it is no puzzle
that we can refer to physical things, but reference to numbers, sets, moral
values, or anything not “physical” is widely held to be problematical if not
actually impossible.

In one way of conceiving it, realism is an empirical theory. One of the facts
that this theory explains is the fact that scientific theories tend to ‘converge’
in the sense that theories are, very often, limiting cases of later theories
(which is why it is possible to regard theoretical terms as preserving their
reference across most changes of theory). Another of the facts it explains
is the more mundane fact that language-using contributes to getting our
goals, achieving satisfaction, or what have you.
The realist explanation, in a nutshell, is not that language mirrors the
world but that speakers mirror the world — i.e. their environment — in
the sense of constructing a symbolic representation of that environment.
(2) Putnam [1978], p. 125.
the world is supposed to be independent of any particular representation
we have of it — indeed, it is held that we might be unable to represent
the world correctly at all (e.g. we might all be ‘brains in a vat’, the
metaphysical realist tells us).
(3) Putnam [1978], p. 113.
The realist believes that the truth or falsity of our sentences depends (usu-
ally) on something external — i.e. on fact which are not (logically equivalent
to) experiential facts, facts which are not about sensations (or about lan-
guage rules, etc.). But realism, if true, must be expressible in some language
or other. [[. . .]] Suppose LR is a language adequate for the expression of the-
sis of realism. E.g. in LR we might want to give the sort of account I have
been sketching, of how there is a correspondence between sentences and
extra-linguistic (and non-phenomenal) facts the existence of which explain
the contribution language using makes to the success of over-all behaviour.
If conclusive verificationism is correct, there must be phenomenal truth
conditions for every sentence in every intelligible language. [[. . .]] So, in par-
ticular, the sentences of LR which purport to describe the allegedly non-
phenomenal states of affairs corresponding to the sentences of, say, natural;
language, actually have phenomenal truth conditions — i.e. these states of
affairs are not really non-phenomenal after all. So realism is either false or
(worse) inexpressible.
(4) Putnam [1980], p. 477.
The problem of realism in the philosophy of science — of empirical real-
ism, not metaphysical realism — is to show that scientific theories can be
regarded as better and better representations of an objective world with
which we are interacting.
Quotations 91.6. (1) Hacking [1983], p. 21.

Scientific realism says that the entities, states and processes described by
correct theories really do exist. Protons, photons, fields of force, and black
holes are as real as toe-nails, turbines, eddies in a stream, and volcanoes.
The weak interaction of small particle physics are as real as falling in love.
Theories about the structure of molecules that carry genetic codes are either
true or false, and a genuinely correct theory would be a true one. [[. . .]]
Anti-realism says the opposite: there are no such things as electrons.
Certainly there are phenomena of electricity and of inheritance but we
construct theories about tiny states, processes and entities only in order
to predict and produce events that interest us. The electrons are fictions.
Theories about them are tools to thinking. Theories are adequate or useful
or warranted or applicable, but no matter how much we admire the specu-
lative and technological triumphs of natural science, we should not regard
even its most telling theories as true.
(2) J. Young [1987], p. 641.
Realism and anti-realism are semantic theses about what makes true
sentences true. Realism is the thesis that reality determines the truth values
of sentences. It does so independently of their possible knowledge. According
to realists, that is, sentences have “objective truth conditions.” Anti-realism,
on the other hand, is the thesis that truth does not transcend what speakers
can be warranted in asserting. Sentences have warranting conditions as their
truth conditions. Warranting conditions are conditions which speakers are
able to recognise as counting in favour of the truth of a sentence.
Quotations 91.7. (1) Priest [1991], p. 361.

A realist [[in one sense of that word]] holds that there are things which exist,
and which are what they are, quite independently of thought. They would
exist even if they were not thought about. It must be possible, therefore,
for there to be things that are not conceived of, indeed, inconceivable.
Comment. If I understand Priest correctly, what he is suggesting is that realism

commits you to dialetheism. But is there anything special about ‘existence’ in that
argument? Is there something hidden in the phrase “which are what they are“, or
would the point extend to ‘being thought’ ? If such a real thing is being thought,
then its being thought is quite independent of its being thought, i.e., it would be
thought even if it were not being thought.
(2) Priest [1995], p. 76.

One can conceive of an object that is not conceived (inconceivable), in the
appropriate sense; I do, and so can you.
Comment. Quotation 92.11 (2) reveals how you can do so, at least according to
Priest.
91b. Reference and realism. The notion of reference contributed to a dis-

cussion about realism.
Remark 91.8. Kripke’s view of names as ‘rigid designators’ is lacking in this section;
but apart from the fact that I find this stuff too painful to engage with, the whole
thing bears very little on my project.
Quotation 91.9. Dummett [1982], p. 57.

Reference is a relation between a singular term, of a kind that can oc-
cur within statements of the given class, and some one object within the
domain.

Is reference just a relation between one thing, which happens to be a word
(say, ‘Mond’) or a particular event of that word’s being uttered, and another
thing (say, the moon)? If so, is it a relation which is just as much part of
the natural-causal order as the relation ‘is chemically bonded to’ ?
(2) Putnam [1978], p. 17.
[[I]]f I am referring to X that means [[. . .]] that I stand in a certain ‘physi-
calistic’ relation to X, tentatively the relation of being connected to Y, or
some property X has, by a certain kind of causal chain (to be specified!).
Is it possible to distinguish between those properties which an object has only

by virtue of its particular description, and those properties which it has “in itself”?
The following quotation characterizes a view ascribed to Feyerabend (with re-
gard to the Bohr-Rutherford description of an electron).

If nothing fits the exact Bohr-Rutherford description of an electron, then
‘electron’ in the sense in which Bohr-Rutherford used it does not refer.
Moreover, if the theoretical description of an electron is different in two
theories, then the term ‘electron’ has a different sense [[. . .]] in the two
theories.
(2) Putnam [1978], pp. 23 f.
This line of reasoning can be blocked by arguing (as I have in various
places, and as Saul Kripke has) that scientific terms are not synonymous
with descriptions. Moreover, it is an essential principle of semantic me-
thodology that when speakers specify a referent for a term they use by
a description and, because of mistaken factual beliefs that these speak-
ers have, that description fails to refer, we should assume that they would
accept reasonable reformulations of their description (in cases where it is
clear, given our knowledge, how their description should be reformulated
so as to refer, and there is no ambiguity about how to do it in the practical
context). (This is, roughly, the principle of benefit of the doubt [[. . .]])
To give an example: there is nothing in the world which exactly fits the
Bohr-Rutherford description of an electron. But there are particles which
approximately fit Bohr’s description: they have the right charge, the right
mass, and they are responsible for key effects which Bohr-Rutherford ex-
plained in terms of ‘electron’; for example, electric current in a wire is flow
of these particles. The principle of benefit of the doubt dictates that we
treat Bohr as referring to these particles.
Comment. The principle of science fiction as applied to scientific description. What-
ever its value for natural science, it doesn’t seem to have any for mathematics or
higher order logic. But is it possible to extend the kind of reasoning (from the first
of the two quotations) to mathematics? It seems to me that Putnam [1980] is doing
essentially that.
Quotation 91.12. Sayers [1985], p. 121.

The account of realism for which I am arguing rejects this account of
reference. Reference should not be seen in such exclusive either/or terms.
There are no theories which entirely and absolutely lack reference, nor any
whose reference is perfect. Reference is, rather, a matter of degree. Our
terms and theories approximate more or less to objective reality; they refer
more or less adequately to it.
Comment. This is a nice example of what I call the shopping mentality in philos-
ophy: “I have a kilo of best mince, thanks.” Typically, philosophers seem to think
that it is enough to know what they want ; “I have a notion of reference that allows
fuzziness, thanks.” If it is possible, and how it can be realized under actual condi-
tions are questions, it seems, that are below the dignity of a philosopher — or, in
some cases, just considered defeatist. It seems to me that Frege’s remarks on the
(lack of) power of definitions in quotations 71.24 are of a similar kind.

One of the puzzling things about the metaphysical realist picture is that
it makes it unintelligible how there can be a priori truths, even contextual
ones, even as a (possibly unreachable) limit. An a priori truth would have
to be the product of a kind of direct ‘intuition’ of the things themselves.
Even verbal truth is hard to understand.
(2) Putnam [1978], p. 137 f.
Suppose we include a sentence S in the ideal theory T1 just because it
is a feature we want the ideal theory to have that it contain S. (Suppose
we even hold S ‘immune’ from revision, as a behaviouristic fact about us.)
Assuming S doesn’t make T1 inconsistent, T1 still has a model. And since
the model isn’t fixed independently of the theory, T1 will be true — true
in the model (from the point of view of meta-T1 ); true in all admissible
models, from the point of view of a theory in which the terms of T1 do
not determinately refer to begin with. So S will be true! ‘S’ is ‘analytic’ —
but it is an ‘analyticity’ that resembles Kant’s account of the synthetic a
priori more than it resembles his account of the analytic. For the ‘analytic’
sentence is, so to speak, part of the ‘form of the representation’ and not ‘the
content of the representation’. It can’t be false of the world [[. . .]] because
the world is not describable independently of our description.
(3) Putnam [1980], p. 464.
[[I]]n many different areas there are three main positions on reference and
truth: there is the extreme Platonist position, which posits nonnatural men-
tal powers of directly “grasping” forms (it is characteristic of this position
that “understanding” or “grasping” is itself an irreducible and unexplicated
notion); there is the verificationist position which replaces the classical no-
tion of truth with the notion of verification or proof, at least when it comes
to describing how the language is understood; and there is the moderate
realist position which seeks to preserve the centrality of the classical notions
of truth and reference without postulation nonnatural mental powers.
Quotations 91.14. (1) Hintikka and Sandu [1994], p. 278.

Frege has one magnificent achievement to this credit, viz. the creation
of modern formal logic. As a philosopher and as a theoretical logician,
he was nevertheless as parochial as he was, geographically speaking. (He
travelled barely more than Immanuel Kant.) Hence Frege’s concepts and
problems offer singularly unfortunate starting points for constructive work
in the foundations of logic and mathematics. Even if he is right in some of
his views, they depend on severely restrictive assumptions that have to be
noticed and (if possible) eliminated. These restrictive assumptions have not
been sufficiently acknowledged in most of the recent discussions of Frege’s
ideas. This has given rise to wild overstatements of Frege’s achievement.
(2) Hintikka and Sandu [1994], pp. 282 f.
Frege is said to have created the logical theory of quantification, but
this claim can be defended on the level of syntactical theory only. That is,
in his logical system, he had symbols expressing first-order quantifiers, as
well as a rule of universal generalization, a rule of substitution and a rule
of alphabetic change of bound variables.
However, on the semantical level, Frege arguably misunderstood the
semantics of quantifiers by construing them as higher-order predicates (sec-
ond-level functions, more exactly). On such a view what e.g. the existential
quantifier does in a sentence like
(2) (∃x) S[x]
is to say that the complex or simple predicate S[x] is not empty.
[[. . .]]
This way of looking at quantifiers should be contrasted with another
one which was emphasized by Skolem and Hilbert in the twenties. On this
view quantifiers codify suitable universal choice functions. The connection
between quantifier-dependence and choice function is one of the basic ideas
behind Skolem functions and Hilbert’s epsilon-calculus. [[. . .]] It is precisely
this view on quantifiers, and not the Fregean one, which made mathemati-
cians in the twenties aware of the logical power of quantification theory.
[[. . .]]
The general point underlying this difference in emphasis is that the

idea of quantifiers as higher-order predicates sits happily only with quan-
tifiers not dependent on others. Once we begin to consider dependencies
between quantifiers, the Fregean interpretation of quantifiers loses much of
its appeal. It is singularly an unhelpful tool in understanding such crucial
phenomena as quantifier ordering, scope, quantifier dependence and inde-
pendence, etc. It also hides the remarkable parallelism between the universal
quantifier and conjunction as well as between the existential quantifier and
disjunction.
Comment. The crucial point for me is the role of contraction hidden in the ε-
operator. In the jargon of Hintikka and Sandu: the ε-operator is singularly an un-
helpful tool in focusing on the ambiguity inherent in quantification because it lumps
them together in one choice function.51
Quotations 91.15. (1) Ellis [1988], p. 420.

[[T]]arski’s theory of truth leads us naturally to ask this important question:
can every truth be expressed in some language whose primitive terms denote
only real entities, and whose primitive predicates are satisfied, if at all, only
by such entities or sequences of them? That is, can every truth be expressed
in a realistically interpretable extensional language?
Comment. As opposed to this: the point of the present study is whether in a re-
alistically interpretable extensional language there isn’t too much expressible as
truth.
(2) Hintikka and Sandu [1994], p. 281.
The principle of compositionality is one of the main pillars on which
Tarski-type truth-definitions rest. In such a truth-definition, the concept of
truth and satisfaction for a complex expression are defined step by step in
terms of the truth and satisfaction of certain simpler expressions. In brief,
a Tarski-type truth-definition works its way from inside out, and hence
cannot accommodate any real semantical context-dependence.
Quotations 91.16. (1) Wettstein [1986], p. 185

GOTTLOB FREGE motivates his famous distinction between sense and
reference by formulating what amounts to a condition of adequacy for a
semantic account of singular terms. Frege’s idea is that any such account
must provide an answer to a crucial question concerning the cognitive sig-
nificance of language: the question of how identity sentences in which proper
names flank the identity sign can both state truths and be informative.
Comment. This sounds like Frege’s sense reference distinction needed motivation.
Under my reading it is the sense reference distinction which Frege employed to
motivate his identification of identity and quality in arithmetic.52
51 As regards this last point, cf. Hazen [1987a], where this issue is beautifully expounded.
52 Cf. Frege in quotation 71.20 (1) in these materials.
(2) Wettstein [1986], p. 203.

It was natural for Frege, given his conception of the semantic enter-
prise to require that an adequate semantic account yield the epistemolog-
ical riches in question. Remarks like Dummett’s “a theory of meaning is
a theory of understanding,” John Searle’s “the philosophy of language is a
branch of the philosophy of mind,” and Stephen Schiffer’s “the basis of a
theory of reference must . . . be a theory of the thought in the mind of a
person using a singular term,” all issue from such a perspective. The seeds
of a radically differently conception of semantics can be found in the work of
the new theorists. An account of linguistic meaning is no longer to be seen
as an account of anything like what the competent speaker understands by
his terms, but rather as an account of the practices he has mastered.
Comment. Here it is again, the same old subjective treadmill that Frege had tried
so hard to overcome. Perhaps it is worthwhile reminding readers that Frege’s alleged
“semantic enterprise” was first and foremost an attempt at providing a justification
for the “supposition that the reference of a sentence is its truth value”.53 This, in
turn, was to serve as the basis for his notion of the ‘Werthverlauf 54 of a function’
which he then employed to extend the notion of equality from objects to concepts,
i.e., as “an equality between ranges of values”,55 in order to be able to formulate his
“Grundgesetz V”. This background must be forgotten, before Frege can be placed in
the vicinity of something like a “theory of the thought in the mind of a person using
a singular term”. In this context, I also recommend Russell’s response to Strawson
in quotations 88.11 (2) and (3) in these materials.
(3) Wettstein [1986], p. 201
[[M]]any of the new theorist’s arguments against Frege take the form of
pointing out that the sense-reference model, no matter what its apparent
advantages with regard to problems of cognitive significance, just cannot
be correct since it is incompatible with actual linguistic practice.
Comment. This line of argument sounds to me like saying that Einstein’s theory of
the relativity of the order of events with regard to the observer cannot be correct
since it is incompatible with actual intuitive experience.
Quotation 91.17. Thiel [1972], p. 42.
[[D]]as Auftreten der von Russell gefundenen Antinomie zeigt, daß ein sinn-
loser Ausdruck verwendet worden ist und diese Verwendung keineswegs
harmlos war. Damit wird man die Frage stellen dürfen, weshalb wir denn
überhaupt sinnlose Ausdrücke zulassen sollen, statt Freges Forderungen an
die korrekte Bildung von Ausrücken zu verschärfen und sinnlose Ausdrücke
ein für allemal zu verbieten. Dies hat man nun in der Tat versucht und
die sog. prädikativen oder konstruktiven Systeme geschaffen, deren Wider-
spruchsfreiheit sich beweisen läßt. In ihnen sind weder falsche Aussagen wie
54 Translated as “course-of-values” in Furth [1964] and “value-range” in Geach and Black
[1952].
Russells Antinomie ableitbar noch solche Pseudo-Aussagen, die zwar nicht

falsch, aber doch sinnlos sind.
[[. . .]] Auf diese Weise ist Freges Ziel der Begründung einer widerspruchs-
freien Analysis erreicht worden — auf einem von ihm nicht vorhergesehenen
Weg.
Comment. The German paradigm: “ein für allemal verbieten” — jawohl! The ques-
tion that Thiel does not ask in his zeal to exclude meaningless expressions is where
the criterion of meaninglessness comes from and if it can itself be formulated mean-
ingfully — a question that drove Wittgenstein into mysticism as quotations 86.6 in
this chapter illustrate. It may also be worthwhile mentioning that what has been
achieved by the so-called predicative systems can be equally achieved by type-free
(impredicative) systems,56 and whatever it is, it does not reach Frege’s goal.
91c. Epistemology.
Quotations 91.18. (1) Hamlyn [1967], pp. 8 f.

Epistemology, or the theory of knowledge, is that branch of philosophy
which is concerned with the nature and scope of knowledge, its presuppo-
sitions and basis, and the general reliability of claims of knowledge.
(2) Ameriks [1982], p. 126.
[[Epistemology]] has above all to explain the basic fact that we know about
the empirical world through judgments.
Comment. This is where I can’t suppress my suspicion against anything that comes
as a ‘basic fact’.
Quotations 91.19. (1) Rorty [1980], p. 3.

To know is to represent accurately what is outside the mind; so to under-
stand the possibility and nature of knowledge is to understand the way in
which the mind is able to construct such representations.
(2) Hirst [1967], p. 80.
“Represent” is usually interpreted in accordance with the doctrine of pri-
mary and secondary qualities — that is, the sensa resemble the object in
spatiotemporal properties but not insofar as colors, sounds, smells, and
other secondary qualities are concerned. Modern analogies of “represent-
ing” are the relation between a map or radar screen and the region they
cover or between television or movies and the studio events reproduced.
Comment. Can radar screen and television distract from the lack of content? I hope
not. Without calling on fancy technology, there is a precise notion of representa-
tion, however: numeralwise expressibility and representability;57 but probably this
doesn’t appeal to a philosophical mind.
56 Cf. Schütte [1960], pp. 337 ff, for an informative survey of the different possibilities for a
foundation of analysis.
57 See definitions 13.37 in the tools.
Quotation 91.20. C. Parsons [1971], p. 236.

The ability to get at “the same object” from different points of view —
different individual minds, different places and times, different characteri-
zations by language — is one of the essentials of objective knowledge. If this
is lacking, then the entities involved should be denied objective existence.
Quotation 91.21. French [1967], p. 623.

I wish to argue that [[Kant’s constitutive-regulative]] distinction is important
historically, as it marks one of the first glimmerings of a problem which has
been pivotal in much contemporary philosophy, especially on the analytic
side.
The problem to which I am referring is simply this: how are we to regard
certain kinds of statements which are, one now sees, significantly different
both from the ordinary observational statements that one encounters in the
more mundane affairs of life, and from ordinary scientific statements? What
are we to say of the sentence which contains words like ‘substance’, ‘God’,
‘evil’, or ‘mind’ ? Or should one say, with the early Wittgenstein, “The
correct method in philosophy would really be the following: to say nothing
except what can be said, i.e. propositions of natural science—i.e. something
that has nothing to do with philosophy—and then, whenever someone else
wanted to say something metaphysical, to demonstrate to him that he had
failed to give a meaning to certain signs in his propositions . . . What we
cannot speak about we must pass over in silence”?
I close this chapter with an up-to-date example of that disarming rhetoric of

uninhibited conceit that is so characteristic of analytic empiricism ever since Carnap
formulated his thesis of the meaninglessness of metaphysics.
Quotation 91.22. Magill [1993], p. 63.

Promotional literature for the [[First European Congress of Analytic
Philosophy, Aix-en-Provence, 23–26 April 1993]] announced that ‘the tra-
dition of contrasting “Analytic” and “Continental” philosophy . . . is inade-
quate, for the values of analytic philosophy are universal[[’.]]
CHAPTER XXIII
PARADIGMS OF METHOD AND CONFIRMATION
Denn eben wo Begriffe fehlen,

Da stellt ein Wort zur rechten Zeit sich ein1
There are some philosophers

who are capable of discoursing after dinner
on various results in modern science
and their philosophical significance,
but manage to preserve a basic opacity
while creating an illusion of understanding
and not committing factual mistakes.2
One of the main problems for a theory of dialectic is to find a paradigm; in view
of the somewhat self-referential nature of dialectical knowledge, it would be un-
satisfactory to simply go around and state how things are; things just aren’t that
simple for dialectic philosophy. Analytic philosophers may be happy with natural
language as a paradigm; dialectic philosophy has to find one which accounts for
a certain interconnectedness of the subject and object of knowledge. Now, it will
probably be clear to the reader that in the end my paradigm will be mathematical
logic, more precisely, something like a system of type free lambda calculus endorsed
with a non-classical logic; but I don’t want to impose this end result on the reader
without looking, at least cursorily, at various options. In other words, the present
chapter is concerned with aspects of method and confirmation. What is to count
for or against a theory?
I’m afraid the whole thing is quite messy but I have grown tired of sifting
through materials.
§ 92. General considerations
Dialectic is not simply a theory that could be expressed like a “theory” of how
bees make baby-bees. Some philosophers tried to talk about a “rational core” of
1 Goethe, Faust, 1. Teil, V. 1993, “Studierzimmer”. Partly also quoted in Marx, Kapital I,
p.83, footnote 24 (continued from p. 82).

2 Wang [1974], p. 16.
1270
dialectic, but to my mind, most of them just made fools of themselves when they
actually began to say what they took that to be. So the first thing that has to
be acknowledged is that in dialectic we are dealing with a rather mind-boggling
phenomenon: contradictions, paradoxes, and antinomies. This requires an awareness
of method that is not commonly found among philosopher, despite, or perhaps
because of, their big claims.
92a. Text, context, pretext. Etymology seems one of the all-time favorites
of people who can’t handle the subject itself. To my mind, Heidegger is the uncon-
tested master of the art of reading a whole philosophy out of a single word.
Quotation 92.1. Krell [1978], p. 273; translating Heidegger [1962]

How do we explain the mathematical if not by mathematics? In such ques-
tions we do well to keep to the word itself. Of course, the issue is not always
there where the word occurs. But with the Greeks, from whom the word
stems, we may safely make this assumption.
Quotations 92.2. (1) Heidegger [1965], p. 17.

Das Andenken an den Anfang der Geschichte, in der sich das Sein im Den-
ken der Griechen enthüllt, kann zeigen, daß die Griechen von früh an das
Sein des Seienden als die Anwesenheit des Anwesenden erfuhren. Wenn wir
εἶναι durch „sein“ übersetzen, dann ist die Übersetzung sprachlich richtig.
Wir ersetzen jedoch nur einen Wortlaut durch einen anderen. Prüfen wir
uns, dann stellt sich alsbald heraus, daß wir weder εἶναι griechisch den-
ken, noch eine entsprechend klare und eindeutige Bestimmung von „sein“
denken. Was sagen wir also, wenn wir statt εἶναι „sein“ und statt „sein“
εἶναι und esse sagen? Wir sagen nichts. Das griechische und das lateinische
und das deutsche Wort bleiben in der gleichen Weise stumpf. Wir verraten
uns bei dem gewohnten Gebrauch lediglich als Schrittmacher der größten
Gedankenlosigkeit, die je innerhalb des Denkens aufgekommen und bis zur
Stunde in der Herrschaft geblieben ist. Jenes εἶναι aber sagt: anwesen. Das
Wesen dieses Anwesens ist tief geborgen in den anfänglichen Namen des
Seins. Für uns aber sagt εἶναι und οὐσία als παρ- und ἀπουσία zuvor dies:
im Anwesen waltet ungedacht und verborgen Gegenwart und Andauern,
west Zeit. Sein als solches ist demnach unverborgen aus Zeit. So verweist
Zeit auf die Unverborgenheit, d.h. die Wahrheit von Sein.
Comment. Das gespreizte Anwesen im Ansinnen des Heideggerschen Andenkens
kann so recht nur in deutscher Sprache zum Anwalten kommen.
(2) Heidegger [1953], pp. 54 f.
Die ganze Abwandlungsmannigfaltigkeit des Verbum »sein« ist durch drei
verschiedene Stämme bestimmt.
Die beiden zuerst zu nennenden Stämme sind indogermanisch und kom-
men auch im griechischen und lateinischen Wort für »sein« vor.
1. Das älteste und eigentliche Stammwort ist »es«, sanskrit »asus«,
das Leben, das Lebende, das, was von ihm selbst her in sich steht und
geht und ruht: das Eigenständige. Hierzu gehören im Sanskrit die verbalen
1272 XXIII. PARADIGMS OF METHOD AND CONFIRMATION
Bildungen esmi, esi, esti, asmi. Dem entsprechen im Griechischen εἰμί und
εἶναι , im Lateinischen esum und esse. Zusammen gehören: sunt, sind und
sein. Bemerkenswert bleibt, daß sich in allen indogermanischen Sprachen
von Anfang an das »ist« (ἔστιν, est . . . .) durchhält.
2. Der andere indogermanische Stamm lautet bhû, bheu. Zu ihm gehö-
ren das griechische φύω , aufgehen, walten, von ihm selbst her zu Stande
kommen und im Stand bleiben. [[. . .]]
Desselben Stammes ist das lateinische Perfekt fui, fuo; ebenso unser
deutsches »bin«, »bist«, wir »birn«, ihr »birt« (im 14.Jahrh. erloschen).
[[. . .]]
3. Der dritte Stamm kommt nur im Flexionsbereich des germanischen
Verbum »sein« vor: wes a.ind.: vasami; germ.: wesan, wohnen, verweilen,
sich aufhalten; zu vest gehören ̥ǫστ ά, ̥άστ υ, Vesta, vestibulum. Hieraus
bildet sich im Deutschen: »gewesen«; ferner: was, war, es west, wesen. Das
Particip »wesend« ist noch in an-wesend, ab-wesend erhalten. Das Sub-
stantivum »Wesen« bedeutet ursprünglich nicht das Was-sein, die quiddi-
tas, sondern das Währen als Gegenwart, An- und Ab-wesen. Das »sens« im
Lateinischen prae-sens und ab-sens ist verlorengegangen. [[. . .]]
Aus den drei Stämmen entnehmen wir die drei anfänglichen anschaulich
bestimmten Bedeutungen: leben, aufgehen, verweilen.
Comment. This sounds quite different to the “is” as copula.
(3) Heidegger [1953], p. 49.
ἐξίσατασθαι – »Existenz«, »existieren« bedeutet für die Griechen gerade:
nicht-sein. Die Gedankenlosigkeit und Verblasenheit, in der man das Wort
»Existenz« und »existieren« zur Bezeichnung des Seins gebraucht, belegt
erneut die Entfremdung gegenüber dem Sein und einer ursprünglich mäch-
tigen und bestimmten Auslegung seiner.
I continue with a quotation which seems directly related to Marx and, perhaps,
let us recall, also to the problem of commodity fetishism, but, let us not be fooled,
it is actually relating to nothing else but Derrida himself.
Quotations 92.3. (1) Derrida [1994], pp. 155 f.3

The commodity table, the headstrong dog, the wooden head faces up, we
recall, to all other commodities. The market is a front, a front among fronts,
a confrontation. Commodities have business with other commodities, these
hardheaded specters have commerce among themselves. And not only in
tête-à-tête. That is what makes them dance. So it appears. But if the “mys-
tical character” of the commodity, if the “enigmatic character” of the prod-
uct of labor as commodity is born of the “social form” of labor, one must
still analyze what is mysterious or secret about this process, and what the
secret of the commodity form is (das Geheimnisvolle der Warenform). This
secret has to do with a “quid pro quo.” The term is Marx’s. It takes us
3 I am grateful to Valerie Kerruish for having brought this particularly intriguing specimen
of philosophical writing to my attention.

back once again to some theatrical intrigue: mechanical ruse (mekhane) or

mistaking a person,repetition upon the perverse intervention of a prompter
[souffleur ], parole souffleé, substitution of actors or characters. Here the
theatrical quid pro quo stems from an abnormal play of mirrors. There is a
mirror, and the commodity form is also this mirror, but since all of a sud-
den it no longer plays its role, since it does not reflect back the expected
image, those who are looking for themselves can no longer find themselves
in it. Men no longer recognize in it the social character of their own labor.
It is as if they were becoming ghosts in their turn. The “proper” feature
of specters, like vampires, is that they are deprived of a specular image, of
the true, right specular image (but who is not so deprived?). How do you
recognize a ghost? By the fact that it does not recognize itself in a mirror.
Now that is what happens with the commerce of the commodities among
themselves.
Comment. The ramble goes on, of course, but I think this is sufficient to give an
impression of Derrida’s way of using commodity fetishism as a pretext for his own
performance.
(2) Derrida [1978], p. 84.
[[W]]e will be incoherent, but without systematically resigning ourselves to
incoherence.
Comment. The methodological credo of Derridadaism.
92b. Common sense, ordinary language, and introspection. Since this

is written in English, a remark seems appropriate: “common sense” is not just an
English word, I also take it to stand for a very English notion, like that of Common
Law, or Commonwealth, or House of Commons. The French have good sense (“le bon
sens”), the Germans sane human understanding (“der gesunde Menschenverstand” 4).
Hegel also speaks of the mean understanding (“der gemeine Verstand”).5
Quotation 92.4. Karl Kraus [1955], p. 183.

Der gesunde Menschenverstand sagt, daß er mit einem Künstler bis zu
einem bestimmten Punkt »noch mitgeht«. Der Künstler sollte auch bis
dorthin die Begleitung ablehnen.
Quotation 92.5. Jaggar [1983], p. 29, for the most part quoting Descartes [1637],
p. 81.
Good sense is of all things in the world the most equally distributed,
for everybody thinks himself so abundantly provided with it, that
even those most difficult to please in other matters do not commonly
desire more of it than they already possess.
The apparent irony of this formulation is mitigated by what follows it:
4 As it has manifested itself in the rejection of “entartete Kunst”, for instance; cf. also quo-
tation 92.4 below. Also: das “gesunde Volksempfinden”.

5 I would be happy to settle for “mean sense”.
It is unlikely that this is error on their part; it seems rather to be

evidence in support of the view that the power of living a good life and
of distinguishing the true from the false, which is properly speaking
what is called Good Sense or Reason, is by nature equal in all men.
p. 259.
[[T]]o satisfy the conditions of the problem, the opponents of [[Hume]] should
have penetrated very deeply into the nature of reason, so far as it is con-
cerned with pure thought—a task which did not suit them. They found
a more convenient method of being defiant without any insight, viz., the
appeal to common sense. It is indeed a great gift of heaven to possess right
or (as they now call it) plain common sense. But this common sense must
be shown in deeds well-considered and reasonable thoughts and words, not
by appealing to it as an oracle when no rational justification of oneself can
be advanced. To appeal to common sense when insight and science fail, and
no sooner—this is one of the subtle discoveries of modern times, by means
of which the most superficial ranter can safely enter the list with the most
thorough thinker and hold his own. But as long as a particle of insight
remains, no one would think of having recourse to this subterfuge.
Comment. In view of my introductory remarks: “plain common sense” is the trans-
lation of “geraden” or “schlichten Menschenverstand”. Otherwise, Kant speaks of
the “gemeinen Menschenverstand”, translated as “common sense”.
(2) Carus and Ellington [1977], p. 109; translating Kant [1783], pp. 369 f.
The appeal to common sense is even more absurd, when concepts and
principles are said to be valid, not insofar as they hold with regard to ex-
perience, but outside the conditions of experience. For what is common
sense? It is normal good sense, so far as it judges rightly. What is normal
good sense? It is the faculty of the knowledge and use of rules in concreto,
as distinguished from the speculative understanding, which is a faculty of
knowing rules in abstracto. Common sense can hardly understand the rule
that every event is determined by means of its cause and can never compre-
hend it thus generally. It therefore demands an example from experience;
and when it hears that this rule means nothing but what it always thought
when a pane was broken or a kitchen-utensil missing, it then understands
the principle and grants it. Common sense, therefore, is only of use so far
as it can see its rules (though they actually are a priori) confirmed by ex-
perience; consequently, to comprehend them a priori, or independently of
experience, belongs to the speculative understanding and lies quite beyond
the horizon of common sense. But the province of metaphysics is entirely
confined to the latter kind of knowledge, and it is certainly a bad sign of
common sense to appeal to it as a witness, for it cannot here form any
opinion whatever, and men look down upon it with contempt until they are
in trouble and can find in their speculation neither advice nor help.

[[C]]ommon sense is more fallible than it likes to believe.
(2) Russell [1946], p. 586.
[[Locke]] is always sensible, and always willing to sacrifice logic rather than
become paradoxical. He enunciates general principles which, as the reader
can hardly fail to perceive, are capable of leading to strange consequences;
but whenever the strange consequences seem about to appear, Locke blandly
refrains from drawing them. To a logician this is irritating; to a practical
man, it is a proof of sound judgment.
Comment. I find it curious to read these lines written by a man who is famous
for having pushed the logical idea of the extension of concepts to its paradoxical
conclusion. Compare this with what Hegel says in quotation 66.23 (1); the differ-
ence between thinkers such as Locke and Hegel is perfectly expressed in these two
quotations.
(3) Russell [1946], p. 586.
Since the world is what it is, it is clear that valid reasoning from sound
principles cannot lead to error; but a principle may be so nearly true as
to deserve theoretical respect, and yet may lead to practical consequences
which we feel to be absurd. There is therefore a justification for common
sense in philosophy, but only as showing that our theoretical principles
cannot be quite correct so long as their consequences are condemned by an
appeal to common sense which we feel to be irresistible. The theorist may
retort that common sense is no more infallible than logic. But this retort,
though made by Berkeley and Hume, would have been wholly foreign to
Locke’s intellectual temper.
Quotation 92.8. Aune [1967], p. 43

[[C]]ommon experience is entirely adequate to show that clear-headed men
never accept a claim merely because it is made, without regard to the
peculiarities of the agent and of the conditions under which it is produced.
For such men, the acceptability of every claim is always determined by
inference.
Comment. Hurray, here is a generalized statement that can be established by com-
mon experience!6
All things ordinary having extraordinary relevance in ordinary thinking.
Quotations 92.9. (1) T. Parsons [1982], p. 367.

I am suggesting that the test be applied to real live uses of English sentences.
The advantage of this is that the test depends on what people mean when
they speak in ordinary language about ordinary matters[[.]]
Comment. Gosh, “real live uses of English sentences”; what a bold and courageous
philosophical idea!
6 Sorry, but I didn’t cite this because I wanted it to be taken seriously.
(2) McKeon [1975], p. 14.

Polemo, the third head of the Academy, said that we should exercise our-
selves with facts and not with dialectical speculations which may gain us
praise for skill in asking questions but be a hindrance to us in ordering our
lives.
Comment. Sounds like philosophy had already reached a low.
Quotation 92.10. Tugendhat [1960], p. 197.

Sätze wie [[der Lügner]], die ihre eigene Falschheit aussagen, müssen
offenbar vermieden werden, und wir vermeiden sie auch faktisch in der
natürlichen Sprache aus ebenso prinzipiellen Gründen wie den Widerspruch
überhaupt.
Comment. A standard case of horror contradictionis. Notice that it is apparently no
problem for Tugendhat to avoid propositions which express their own falsity, prob-
ably because a superior mind like that of a philosopher recognizes them without
difficulty and avoids them in a natural way. But, of course, no one can reasonably
expect of philosophers to look after logicians and tell them how to avoid contradic-
tions in higher order logic.
Quotations 92.11. Myers [1986], p. 199.

Philosophers and psychologists typically use “introspection” to mean a form
of observation. Their concept of introspection resembles Locke’s notion of
reflection, which he defines as the “perception of the operations of our mind
within us”, deserving the name “internal sense” and being further charac-
terized as “that notice which the mind takes of its own operations”.
(2) Priest [1991], p. 363.
[[A]] simple thought experiment. I ask you to consider an object which is not
being conceived. You do so. Indeed, in the relevant sense you can conceive
of anything that is referred to by a simple grammatical noun-phrase of
English, just because you understand it. By mentally rehearsing the phrase
you bring the object it refers to before the mind in the required sense.
Comment. You may try Transcendental Meditation if you find it difficult to men-
tally rehearse the phrase, but be cautious in using hyperventilation, as you might
find the object hitting your mind stronger than in the required sense. Once you’ve
mastered the stage of conceiving the inconceivable you may go on to conceive God
(in the relevant sense, of course).
Quotation 92.12. Wittgenstein [1956], p. 114; quoted after Wang [1974], p. 162.
Why are the Newtonian laws not axioms of mathematics? Because we could
quite well imagine things being otherwise. But — I want to say — this only
assigns a certain role to those propositions in contrast to another one. I.e.:
to say of a proposition: ‘This could be imagined otherwise’ or ‘We can
imagine the opposite too,’ ascribes the role of an empirical proposition to
it.
Comment. My point regarding such considerations is always the question “who can
imagine otherwise?”; to a good extent, I say, this is a matter of which end of the
cane you are on. What I suggest is that it is the coercive “we” which is at the bottom
of this kind of philosophy.
Quotation 92.13. Alston [1989], p. 107.

I can form an intelligible conception of someone’s failing to believe that
p, where p seems obviously true. Perhaps this person has been rendered
unduly skeptical by overexposure to the logical paradoxes.
Comment. Intelligible it may be, at least to some, but how intelligent ?
Quotation 92.14. Wang [1954], p. 266.

[[M]]ost people, when confronted with Cantor’s indenumerability arguments,
presumable have some uneasy feeling and suspect the presence of some
hidden fallacy. Undoubtedly ordinary mathematicians would consider such
arguments as extraordinarily uncommon.
Comment. For once, at least, I find myself on the side of most people.
And, of course, there are always ‘naive intuitions’. They lie, as it were, at the
bottom of the issue, the lowest level.
Quotation 92.15. Herzberger [1982], p. 485.

[[A]] number of people have proposed antinomic logics or impossible worlds
in an effort to represent paradoxical statements as being both true and
false.
Whatever the merits of these various approaches, I don’t think any of
them begins to provide an adequate representation of our naive intuitions[[.]]
Comment. One more mention of ‘our naive intuitions’ and I am going to scream.
So I think it’s best just to close this paragraph and try to get rid of that foul taste.
I close this section with one of these lines that you get from a calender, usu-
ally attributed to someone famous, in this case to Einstein, but never properly
referenced: Common sense is the collection of prejudices acquired by age eighteen.
92c. Authority, law, and custom. The simplest answer to the question
“Why do we claim A” is “Because it is the case”. And if you don’t accept that . . .
we’ve got means to teach you.

[[A]]ny argument against realism which is based on modern atomic theory
— on quantum mechanics — ought to be silenced by the memory of the
reality of the events of Hiroshima and Nagasaki.
(2) Perutz [1986], p. 49; my translation
When I asked Popper what made him assume that biochemistry cannot
be reduced to chemistry, I received the schoolmasterish answer, I would find
out myself if I were to think about it for a night.
Quotation 92.17. Augustine, quoted after Kline [1967], p. 1.

The good Christian should beware of mathematicians and all those who
make empty prophecies. The danger already exists that the mathematicians
have made a covenant with the devil to darken the spirit and to confine man
in the bonds of Hell.
Comment. This would have been written roughly at the time a Christian mob
lynched Hypatia.
Quotation 92.18. Thomas Aquinas, quoted after Fried [1940], p. 204.

By faith alone do we hold, and by no demonstration can it be proved, that
the world did not always exist[[.]]
Comment. Amen. However, cf. quotation 92.24 below for possible complications.
I include two particularly silly remarks just for the records.
Quotations 92.19. (1) Land [1993], p. 101.7

As for its ‘own’ or ‘inner’ law, logic has never been anything other than
the distillation of juridical procedure, the abstract form of inclusion or
non-inclusion of a case under a law (species under genus), which has been
predominantly thematised as judgment, although a language of propositions
has more recently risen to prominence.
(2) Finocchiaro [1980], p. 302.
[[S]]ince Galileo is regarded by almost everybody as a model scientist, and
since logicians want very much to be scientific, it will be very important for
them to examine his work to see to what extent it can be used to justify
their own techniques of formalization and abstraction.
Quotation 92.20. Fichte, Werke, ii, p. 454, quoted from Wallace [1873], p. 297.
The old woman who frequents the church—for whom by the way I cherish
all possible respect—finds a sermon very intelligible and very edifying which
contains lots of texts and verses of hymns she knows by rote and can repeat.
In the same way readers who fancy themselves far superior to her, find a
work very instructive and clear which tells what they already know, and
proofs very stringent which demonstrate what they already believe. The
pleasure the reader takes in the writer is a concealed pleasure in himself.
What a great man! (he says to himself); it is as if I heard or read myself.
Quotations 92.21. (1) Lama Anagarika Govinda, Foundations of Tibetan Mysti-

cism (Rider, London 1973), p. 93; quoted after Capra [1975], p. 147.
The Buddhist does not believe in an independent or separately existing
external world, into whose dynamic forces he could insert himself. The
external world and his inner world are for him only two sides of the same
7 I am grateful to Valerie Kerruish for drawing my attention to this particularly impressive
specimen of legal reasoning.

fabric in which the threads of all forces and of all events, of all forms of
consciousness and of their objects are woven into an inseparable net of
endless, mutually conditioned relations.
Comment. Modern physics seeking for a paradigm in Eastern wisdom.
(2) Murti [1955], p. 138; quoted after Capra [1975], p. 142.
Things derive their being and nature by mutual dependence and are nothing
in themselves.
Comment. Fine. But what exactly does this dependency look like?
92d. Performative inconsistency, psychiatry, and science fiction. I be-

gin with an interesting attempt to employ self-reference in the context of metaphys-
ical reasoning. Although it is not sufficiently elaborated either, I find it a pleasant
change from the usual dogmatism.
Quotations 92.22. (1) Boyle [1972], p. 25.

I will analyze a form of philosophical argument widely used by metaphysi-
cians[[. . .]].
The argument-form that I will examine is that in which a position is
criticized as undercutting itself. A position attacked by such an argument
allegedly denies or cannot account for some condition that is required for
it to make sense or be true. This argument-form has been called the argu-
ment from self-referential consistency. It seeks to show that a position of
theory which refers to itself, that is to say, includes itself in its subject mat-
ter, cannot account for itself. Such a theory can be called self-referentially
inconsistent or self-refuting.
(2) Boyle [1972], p. 27
The peculiar character of performative inconsistency can be revealed
by comparing it with other kinds of inconsistency. Performative inconsis-
tencies are not purely formal inconsistencies like those of elementary logic.
Nor are they merely empirical inconsistencies like the discrepancies that can
arise between statements and the facts they purportedly describe. Finally,
these arguments neither point out nor depend upon semantical meaning-
lessness. In short these arguments and the inconsistencies they reveal are
not simply matters of syntax or semantics, nor are they simply empirical
arguments based only on an appeal to contingent facts. Since the inconsis-
tency here is between what a sentence expresses and its actual utterance,
this inconsistency might be called a matter of pragmatics. [[. . .]]
The irreducibility of performative inconsistency to the syntactical con-
tradictions between elements of a formal system is easily shown. The incon-
sistency of performatively inconsistent statements is removed if their scope
is limited so that the self-referential instance does not arise.
(3) Boyle [1972], p. 30.
This recognition of the irreducibility of performative inconsistency ei-
ther to the contradictions of logical theory or to the nonsense of the seman-
tical paradoxes undercuts the most common objection to the use of argu-
ments that exploit performative inconsistencies. The objection—formulated
by Bertrand Russell and others—assumes that all self-reference is vicious
and consequently that all statements that refer to themselves are meaning-
less.
(4) Boyle [1972], p. 31
This undercutting of a common criticism of self-referential arguments,
however, is not enough to establish that these arguments are in fact valid. To
establish this validity, the root of the inconsistency must be revealed. [[. . .]]
I submit that the “inconsistency”—if this is not a completely misleading
word in this context—is a case of the discrepancy that obtains between
a statement and a state of affairs that falsifies it. [[. . .]] In other words, a
statement is performatively inconsistent when it is falsified by some aspect
of its utterance.
(5) Boyle [1972], p. 40.
I call this type of argument peculiarly metaphysical, first, because it
can terminate in statements that are certain, and secondly, because it can
be used to make statements about the whole of reality or about “being as
such”.
Comment. The language is vague, e.g., “syntactical contradictions between elements
of a formal system”, but the idea might nevertheless look attractive. Still, it is not
quite what I am aiming at.
Quotations 92.23. (1) Laing [1970], p. 13.

it hurts Jack
to think
that Jill thinks he is hurting her
by (him) being hurt
to think
that she thinks he is hurting her
by making her feel guilt
at hurting him
by (her) thinking
that he is hurting her
by (his) being hurt
to think
that she thinks he is hurting her
by the fact that
da capo sine fine
(2) Laing [1970], p. 21.
JILL You put me in the wrong
JACK I am not putting you in the wrong
JILL You put me in the wrong for thinking you put me in the wrong.
(3) Laing [1970], p. 30.

JACK You are a pain in the neck
To stop you giving me a pain in the neck
I protect my neck by tightening my neck muscles,
which gives me the pain in the neck
you are.
JILL My head aches through trying to stop you
giving me a headache.
(4) Adams [1979], p. 73. Marvin speaking.
‘I’m not getting you down at all am I?’
Quotation 92.24. Adams [1979], p. 50; quoting from The Hitch Hiker’s Guide to
the Galaxy.
[[‘I ]]f you stick a Babel fish in your ear you can instantly understand any-
thing said to you in any form of language. The speech patterns you actually
hear decode the brainwave matrix which has been fed into your mind by your
Babel fish.
‘Now it is such a bizarrely improbably coincidence that anything so
mindbogglingly useful could have evolved purely by chance that some thinkers
have chosen to see it as a final and clinching proof of the non-existence of
God.
‘The argument goes something like this: “I refuse to prove that I exist,”
says God, “for proof denies faith, and without faith I am nothing.”
‘“But,” says Man, “the Babel fish is a dead giveaway isn’t it? It could
not have evolved by chance. It proves you exist, and so therefore, by your
own arguments, you don’t.’QED.”
‘“Oh, dear,” says God, “I hadn’t thought of that,” and promptly vanishes
in a puff of logic.
‘“Oh, that was easy,“ says Man, and for an encore goes on to prove that
black is white and gets himself killed on the next zebra crossing.
[[. . .]]
‘Meanwhile, the poor Babel fish, by effectively removing all barriers of
communication between different races and cultures, has caused more and
bloodier wars than anything else in the history of creation.’
Comment. Recently, dialetheism has provided the means for a full rehabilitation of
God in His Existence by introducing the notion of an illicit use of the disjunctive
syllogism. Surely, if anyone should have the privilege of Being and Non-Being, it is
God.
Quotation 92.25. Hofstadter [1979], p. 489.
Achilles: Fabulous! Let me just focus down onto the tip of those flames [[of
Magrittes’s painting The Fair Captive]], where they meet the picture
frame . . . Its such a funny feeling to be able to instantaneously “copy”
anything in the room—anything I want—onto that screen. I merely
need to point the camera at it, and it pops like magic onto the screen.
Crab: Anything in the room, Achilles?

Achilles: Anything in sight, yes. That’s obvious.
Crab: What happens, then, if you point the camera at the flames on the
TV screen?
(Achilles shifts the camera so that it points directly at that part of the
television screen on which the flames are—or were—displayed.)
Achilles: Hey, that’s funny! That very act makes the flames disappear
from the screen! Where did they go?
Crab: You can’t keep an image still on the screen and move the camera
at the same time.
Comment. My emphasis is on: “that very act”. Or, differently put, you can’t have
your cake and eat it too.8
92e. Reductionism, scientific revolution, and academic reality.
Quotations 92.26. (1) Burks [1986], p. 38

[[T]]he idea of reducing one system to a more basic, underlying system is
fundamental to metaphysics. The modern materialist says not only that
mind is reducible to physiology, but physiology is reducible to evolutionary
biology, biology to chemistry, chemistry to physics, and — he keeps going
— physics to astronomy — ultimately we all came from the big bang. All
sciences, the materialist says, can be arranged in this kind of hierarchical
structure of many levels, each science dependent on the ones below it.
Comment. This business about reducibility is very vague. The suggestion of a re-
duction of physics to astronomy with an allusion to the big bang sounds hilarious
to me and I wonder if it has ever been seriously suggested. But philosophers seem
to love to discuss questions of this sort; particularly, if they lack competence in the
fields covered. Be this as it may, a particular sort of reduction will play an eminent
role in this study: logicism, the (attempt at a) reduction of arithmetic to logic.
What is challenging for me regarding such reductions is their actual realization,
not philosophical discussions about their possible realization. The reason is a very
practical one which will be clear to anybody who’s ever tried to do something and
not just talk about how it can be done.
(2) Wang [1974], p. 298.
A commonly accepted belief today among biologists is that all manifesta-
tions of life can ultimately be explained by the laws governing inanimate
matter. This belief is often labeled mechanism or materialism. According
to this view, the ultimate goal of the life sciences is to account for the ori-
gin and properties of life (and mind) by means of the principles of physics.
Psychology is reducible to physiology (the machinery of the brain), physi-
ology (and biology) is reducible to chemistry and physics (the machinery of
life), chemistry is reducible to physics. The complexity of the phenomena of
8 Cf. footnote 53 on p. 1462 in the groundworks.
life (and mind) comes from the complex organization of the large number
of objects involved and not from the complexity of the fundamental laws
governing the basic objects.
One working hypothesis is that, whatever uncertainties there are with
regard to the foundations of physics and however they will be resolved,
they will not affect seriously the superstructure that makes up biology (and
psychology). It seems unquestionable that for a long time to come results
in biology will be sufficiently stable (or imprecise) so that they will not be
affected by alternative theories of elementary particles. It is, however, not
equally clear that resolutions of the unresolved difficulties in fundamental
physics will not affect the ultimate program of a complete account of life in
terms of physics.
(3) Burks [1986], p. 39
The kinetic theory of gases constitutes a model reduction. Consider a
gas held in a container. The gas is a system with mass and temperature,
exerting pressure on its container. The ideal gas law states that these three
measurable quantities are related in a simple way: pressure × volume =
constant × temperature. Thus a gas is a system with various properties
(pressure, volume, temperature) related by a certain law, the gas law.
But the gas is also a bunch of particles, septillions of them each with
its weight, position, and velocity, and these particles are governed by their
laws, the laws of mechanics, which tell how they move and bump against
one another and the walls of the container.
The simplest analogy to this dual nature of the gas is a forest and its
trees, the forest being made of trees as the gas is made of particles. As the
old saying goes, sometimes we cannot see the forest for the trees. But in
the kinetic theory of gases we look at both the forest and the trees, both
the gas as a unitary system and the gas as a system of many particles,
for reduction involves a relation between these two systems. Moreover, we
must look also at the laws of each system, for though these are different,
Maxwell and Boltzmann were able to reduce the law of the gas-system to
the laws of the particle-system. That is, Maxwell and Boltzmann deduced
the law “pressure × volume = constant × temperature” from the laws of
mechanics governing the individual particles.
(4) Dummett [1980], p. 66
Reductionism, properly so called, is the thesis that there exists a transla-
tion of statements of the given class into those of some other class, which
I shall call the reductive class. This translation is proposed, not merely
as preserving truth-values, but as part of an account of the meanings of
statements of the given class: it is integral to the reductionist thesis that
it is by an implicit grasp of the schema of translation that we understand
those statement. The most celebrated example of a reductionist thesis is
that embodied in classical phenomenalism: the given class here consists of
statements about material objects, and the reductive class of statements
about sense-data.
Reduction is closely related to deduction. I include a quotation regarding the

undue emphasis of deduction as a source of knowledge from someone who is famous
for his work on the reduction of mathematics to logic.

The Greeks in general attached more importance to deduction as a source of
knowledge than modern philosophers do. In this respect, Aristotle was less
at fault than Plato; he repeatedly admitted the importance of induction,
and he devoted considerable attention to the question: how do we know the
first premisses from which deduction must start? Nevertheless, he, like other
Greeks, gave undue prominence to deduction in his theory of knowledge.
The following set of quotations is taken from Kuhn’s Structure of Scientific

Revolutions.
Quotations 92.28. (1) Kuhn [1962], p. 6.

[[T]]he major turning points in scientific development associated with the
names of Copernicus, Newton, Lavoisier, and Einstein [[display]] [[m]]ore
clearly than most other episodes in the history of at least the physical
sciences [[. . .]] what all scientific revolutions are about. Each of them neces-
sitated the community’s rejection of one time-honored scientific theory in
favor of another incompatible with it. Each produced a consequent shift in
the problems available for scientific scrutiny and in the standards by which
the profession determined what should count as an admissible problem or
as a legitimate problem-solution.
(2) Kuhn [1962], p. 37.
[[O]]ne of the things a scientific community acquires with a paradigm is
a criterion for choosing problems that, while the paradigm is taken for
granted, can be assumed to have solutions. To a great extent these are
the only problems that the community will admit as scientific or encour-
age its members to undertake. Other problem, including many that had
previously been standard, are rejected as metaphysical, as the concern of
another discipline, or sometimes as just too problematic to be worth the
time. A paradigm can, for that matter, even insulate the community from
those socially important problems that are not reducible to the puzzle form,
because they cannot be stated in terms of the conceptual and instrumental
tools the paradigm supplies.
(3) Kuhn [1962], p. 73.
Consider [[. . .]] the late nineteenth century crisis in physics that prepared
the way for the emergence of relativity theory. One root of that crisis can be
traced to the late seventeenth century when a number of natural philoso-
phers, most notably Leibniz, criticized Newton’s retention of an updated
version of the classic conception of absolute space.[[*]] They were very nearly,
though never quite, able to show that absolute positions and absolute mo-
tions were without any function at all in Newton’s system; and they did
succeed in hinting at the considerable aesthetic appeal a fully relativis-

tic conception of space and motion would later come to display. But their
critique was purely logical. Like the early Copernicans who criticized Aris-
totle’s proof of the earth’s stability, they did not dream that transition to a
relativistic system could have observational consequences. At no point did
they relate their views to any problems that arose when applying Newtonian
theory to nature.
Comment. The footnote [[*]] refers the reader to Jammer [1954], pp. 114–24, part of
which is contained in quotation 70.17 (4) in these materials.
(4) Kuhn [1962], p. 158.
The man who embraces a new paradigm at an early stage must often do so
in defiance of the evidence provided by problem-solving. He must, that is,
have faith that the new paradigm will succeed with the many large problems
that confront it, knowing only that the older paradigm has failed with a
few. A decision of that kind can only be made on faith.
At this point, a look at a concrete example taken from academic reality seems
appropriate.
Quotation 92.29. Kreiser [1973], p. X.

Bei Frege reichte die „eigentliche“, „nützliche“ Tätigkeit in den Augen der
Behörden gerade noch aus, um 1896 die Zustimmung zur Ernennung zum
ordentlichen Honorarprofessor zu geben. In einem Schreiben vom 3.6.1908
an die „durchlauchtigsten Erhalter“ der Universität zu Jena teilte der Uni-
versitätskurator Dr. von Eggeling mit, er könne Frege „zu keiner Auszeich-
nung vorschlagen, da seine Lehrtätigkeit untergeordneter Art und für die
Universität ohne besonderen Vorteil ist.“
Comment. How unfortunate that Frege’s work was not assessed by someone like
Scott who does not want to be reminded of a bad review of Beethoven.9 Or could
the problem possibly be that a Beethoven is not always easily recognizable for
MostPeople? And when someone has emerged as Beethoven, that MostPeople who
has written a bad review of him has long been replaced by another MostPeople.
Have you ever heard of one university curator Eggeling before?
Quotation 92.30. Analysis 45 (1985), p. 69 (“Rejoinder”).

[[X]] in his reply declares that of course [[Y]] is the case and that he is at loss
to understand why I should think otherwise. I am at a loss to understand
why he should think that he has met my challenge, since he has produced
no arguments, merely an example of [[Y]] together with a reiteration of the
disputed claim.
Comment. The titanic struggle in full swing; the heroes are wielding their /swords10
with that deadly wanking precision that only philosophers are capable of.
10 I, too, can write under erasure; and it is actually not all that difficult: you just use \hskip-
\labelsep and put the whole thing into \mbox; at least in LATEX.
I leave the final word in this paragraph to Heidegger. Honi soit qui mal y pense.
Quotations 92.31. (1) Krell [1978], p. 272; translating Heidegger [1962].

The greatness and superiority of natural science during the sixteenth
and seventeenth centuries rests in the fact that all the scientists were
philosophers. They understood that there are no mere facts, but that a
fact is only what it is in the light of the fundamental conception, and
always depends upon how far that conception reaches. The characteris-
tic of positivism—which is where we have been for decades, today more
than ever—by way of contrast is that it thinks it can manage sufficiently
with facts, or other and new facts, while concepts are merely expedients
that one somehow needs but should not get too involved with, since that
would be philosophy. Furthermore, the comedy—or rather the tragedy—of
the present situation of science is that one thinks to overcome positivism
through positivism. To be sure, this attitude prevails only where average
and supplemental work is done. Where genuine and discovering research
is done the situation is no different from that of three hundred years ago.
That age also had its indolence, just as conversely, the present leaders of
atomic physics, Niels Bohr and Heisenberg, think in a thoroughly philo-
sophical way, and only therefore create new ways of posing questions and,
above all, hold out in the questionable.
Comment. Heil Bohr and Heisenberg, our leaders of atomic physics.
(2) Krell [1978], p. 94; translating Heidegger [1929], p. 25.
No particular way of treating objects of inquiry dominates the others. Math-
ematical knowledge is no more rigorous than philological-historical knowl-
edge. It merely has the character of “exactness,” which does not coincide
with rigor. To demand exactness in the study of history is to violate the
idea of the specific rigor of the humanities.
Comment. Another “merely”. But in principle I agree. Just as mathematics is rigor-
ously (“streng”) exact, the humanities are rigorous in their own way: strictly coffee-
house.
(3) Krell [1978], p. 434; translating Heidegger [1969], p. 64.
No prophecy is necessary to recognize that the sciences now establishing
themselves will soon be determined and regulated by the new fundamental
science that is called cybernetics.
§ 93. The paradigm of physics
As I said before, my paradigm is mathematics, but I have always felt that physics
is very close to what I am aiming at, the second choice so to speak. Admittedly,
my engagement with physics never got beyond a basically dilettante state;11 this
is something which I ask the reader to keep in mind when looking at the following
quotations.
11 Whereas I call myself an advanced dilettante as regards mathematical logic.
§ 93. THE PARADIGM OF PHYSICS 1287
A point of particular interest is the abandoning of a classical theory. Kant’s

claims about the nature of space and time may serve as a link.
93a. Galileo Galilei. I begin with Galileo Galilei because of his pioneering
role in the employment of mathematical tools in the formulation and deduction of
physical laws. It is only a few scattered remarks about him which I collected on my
way, as it were, because they struck me as quite telling, not just about Galilei, but
also about those who made them.
Quotations 93.1. (1) Mason [1962]. pp. 155 f.

Galileo was of the view that suitably chosen mathematical demonstra-
tion could be applied to the investigation of any problem involving qualities
which were measurable, beyond the spatial measures of length, areas, an
volumes which had been the traditional subject-matter of geometry. Inves-
tigating the scale effect, he studies amounts of matter, later called masses,
in this way; then he enquired into dynamical problems involving the mea-
sures of time and velocity in a similar fashion. Here the central problem
for Galileo was that of the fall of bodies under gravitational force. First of
all he argued has way out of the Aristotelian view that heavy objects fell
faster than lighter ones. What would happen, he asked, if a heavy and a
light body were tied together and then dropped from a height? From the
Aristotelian point of view it could be maintained that the time occupied by
their fall would be either the mean of the times the two bodies would have
taken to fall separately, or the time taken by a body with their combined
weights to fall from the same height. ‘The incompatibility of these results’,
Galileo wrote, ‘showed Aristotle to be wrong.’ To find out what actually
did happen in the gravitational fall of bodies Galileo made the experiment
of measuring the time taken by smooth metallic spheres to roll down given
lengths of a graduated inclined plane. The free fall of an object under grav-
ity was too rapid to be observed directly, and so Galileo ‘diluted gravity’,
using the device of the inclined plane, in order that his metallic spheres
should move downwards under gravity with measurable speeds.
(2) Mason [1962]. pp. 156 f.
Thence Galileo proceeded to show the value of mathematical demon-
stration in science by developing the theory of the path traced out by the
flight of a projectile. He considered the movement of a sphere rolling across
a table with uniform speed until it came to the edge, when it traced a
curved path to the floor. At any point on this path the sphere would have
two velocities; one horizontal, remaining constant owing to the principle of
inertia, the other vertical, increasing with time because of gravity. In the
horizontal direction the sphere would sweep out equal distances in equal
times, but vertically, the distances it covered would be proportional to the
square of the time. Such relations determine the form of the path described,
namely a semi-parabola. The path of a projectile shot from a cannon would
then be a full parabola, giving a maximum range when the gun was ele-
vated at an angle of forty-five degrees. Thus what Tartaglia had observed
as a fact, Galileo deduced theoretically from the results of his inclined plane
experiments. Galileo wrote in this connection:
‘The knowledge of a single fact acquired through a discovery of its
causes, prepares the mind to understand and ascertain other facts, without
need of recourse to experiments, precisely as in the present case, where by
argumentation alone the author proves with certainty that the maximum
range occurs when the elevation is forty-five degrees.’
Quotations 93.2. (1) Krell [1978], p. 289 f; translating Heidegger [1962],

Galileo did his experiment at the leaning tower of Pisa, where he was profes-
sor of mathematics, in order to prove his statement. In it bodies of different
weights did not arrive at precisely the same time after having fallen from
the tower, but the difference in time was slight. In spite of these differences
and therefore really against the evidence of experience, Galileo upheld his
proposition. The witnesses to this experiment, however, became really per-
plexed by the experiment and Galileo’s upholding his view. By reason of
this experiment the opposition toward Galileo increased to such an extent
that he had to give up his professorship and leave Pisa.
Comment. This strikes me as a nice example of Heidegger’s cottage perspective.
Compare what Galilei himself had to say at the end of the following quotation.
(2) S. Drake [1980], p. 25.
Historians generally have doubted the story about Galileo and the Leaning
Tower of Pisa, first told after Galileo’s death by a protégé who was not born
until long after the incident. According to this story, the demonstration was
performed in the presence of Galileo’s students and some professors. It is
probable that his students, who had been taught Aristotle’s rules by their
professors of philosophy, would argue against him that weight must affect
speed of fall. Galileo’s Leaning Tower demonstration would then have been
not just to show the students, but to convince the professors that Aristotle’s
physics must be revised, as he was already arguing.
In a later dispute (1612) with Galileo, a professor of philosophy at
Pisa conducted experiments from the Leaning Tower to support Aristotle,
observing that bodies of the same material and different weights do not hit
the ground exactly together. The basic difference between his approach and
Galileo’s is illustrated in Galileo’s last book:
Aristotle says that a hundred-pound ball falling from a height of a hundred
cubits hits the ground before a one-pound ball has fallen one cubit. I say they
arrive at the same time. You find, on making the test, that the larger ball beats
the smaller one by two inches. Now, behind these two inches you want to hide
Aristotle’s ninety-nine cubits and speaking only of my tiny error remain silent
about his enormous mistake.
Comment. May I suggest that the difference between ninety-nine cubits and two
inches is meaningless to philosophers.12 It’s all just numbers to them and can be left
12 One cubit is approximately 20.6 inches, according to my encyclopedia.
to ‘uncultured button-pushers and knob-twiddlers’.13 What counts for philosophers

is the fact that they don’t arrive at exactly the same time, not the accuracy of
prediction.
(4) Drake [1980], p. vi.
When Galileo began making physical measurements, he put philosophy
aside temporarily; and when measurement led him to laws he lost inter-
est in causes, indefinitely postponing his return to philosophy.
Quotations 93.3. (1) S. Drake [1980], p. 70; quoting Galilei.

I think that tastes, odours, and so on are no more than mere names
so far as the objects in which we locate them are concerned, and that
they reside only in consciousness. If living creatures were removed, all these
qualities would be wiped out and annihilated.
Comment. Compare Berkeley in quotation 58.11 (4); also Parmenides in quotations
57.6 for a stronger position (including change of place, for instance).
(2) S. Drake [1980], p. 71; quoting Galilei.
If [[the philosophers’]] opinions and their voices have the power to call
into existence the things they name, then I beg them to do me the favour
of naming a lot of old hardware I have about my house ‘gold’.
93b. Absolute space and time: Newton’s mechanics. My interest in

Newton’s mechanics is mainly motivated by his notions of space and time.
Quotations 93.4. (1) Capra [1975], p. 56.

The stage of the Newtonian universe, on which all physical phenomena
took place, was the three-dimensional space of classical Euclidean geometry.
It was an absolute space, always at rest and unchangeable. In Newton’s own
words, ‘Absolute space, in its own nature, without regard to anything ex-
ternal, remains always similar and immovable’. All changes in the physical
world were described in terms of a separate dimension, called time, which
again was absolute, having no connection with the material world and flow-
ing smoothly from the past through the present to the future. ‘Absolute,
true, and mathematical time’, said Newton, ‘of itself and by its own nature,
flows uniformly, without regard to anything external.’
The elements of the Newtonian world which moved in this absolute
space and absolute time were material particles, in the mathematical equa-
tions they were treated as ‘mass points’ and Newton saw them as small,
solid, and indestructible objects out of which all matter was made.
Comment. One may recognize here the pattern of Kant’s cosmological antinomies.
(2) Einstein [1933], pp. 145 f.
Newton, the first creator of a comprehensive and workable system of
theoretical physics, still believed that the basic concepts and laws of his
13 According to Marsh [1956], p. 321, “the phrase, as one might expect, comes from the Oxford
Holy of Holies, All Souls’ ”.

system could be derived from experience; his phrase ‘hypotheses non fingo’
can only be interpreted in this sense. In fact, at that time it seemed that
there was no problematical element in the concepts, Space and Time. The
concepts of mass, acceleration and force, and the laws connecting them,
appeared to be directly borrowed from experience. But if this basis is as-
sumed, the expression for the force of gravity seems to be derivable from
experience; and the same derivability was to be anticipated for the other
forces.
One can see from the way he formulated his views that Newton felt by
no means comfortable about the concept of absolute space, which embodied
that of absolute rest; for he was alive to the fact that nothing in experience
seemed to correspond to this latter concept. He also felt uneasy about the
introduction of action at a distance. But the enormous practical success of
his theory may well have prevented him and the physicists of the eighteenth
and nineteenth centuries from recognizing the fictitious character of the
principles of his system.
On the contrary, the scientists of those times were for the most part
convinced that the basic concepts and laws of physics were not in a log-
ical sense free inventions of the human mind, but rather that they were
derivable by abstraction, i.e. by a logical process, from experiments. It was
the general Theory of Relativity which showed in a convincing manner the
incorrectness of this view. For this theory revealed that it was possible for
us, using basic principles very far removed from those of Newton, to do
justice to the entire range of the data of experience in a manner even more
complete and satisfactory than was possible with Newton’s principles. But
quite apart from the question of comparative merits, the fictitious charac-
ter of the principles is made quite obvious by the fact that it is possible
to exhibit two essentially different bases, each of which in its consequences
leads to a large measure of agreement with experience. This indicates that
any attempt logically to derive the basic concepts and laws of mechanics
from the ultimate data of experience is doomed to failure.
If, then, it is the case that the axiomatic basis of theoretical physics
cannot be an inference from experience, but must be free invention, have
we any right to hope that we shall find the correct way? Still more – does
this correct approach exist at all, save in our imagination? Have we any
right to hope that experience will guide us aright, when there are theo-
ries (like classical mechanics) which agree with experience to a very great
extent, even without comprehending the subject in its depths? To this I
answer with complete assurance, that in my opinion there is the correct
path and moreover, that it is in our power to find it. Our experience up to
date justifies us in feeling sure that in Nature is actualized conviction that
pure mathematical construction enables us to discover the concepts and
the laws connecting them which give us the key to the understanding of the
phenomena of Nature. Experience can of course guide us in our choice of
serviceable mathematical concepts; it cannot possibly be the source form
which they are derived; experience of course remains the sole criterion of
the serviceability of a mathematical construction for physics, but the truly
creative principle resides in mathematics. In a certain sense, therefore, I
hold it to be true that pure thought is competent to comprehend the real,
as the ancients dreamed.
Quotation 93.5. Maxwell [1873], quoted from Coley and Hall [1980], p. 86.
An atom is a body which cannot be cut in two. A molecule is the smallest
possible portion of a particular substance. No one has ever seen or handled
a single molecule. Molecule science, therefore, is one of those branches of
study which deal with things invisible and imperceptible by our senses, and
which cannot be subjected to direct experiment.
93c. Relativity. Einstein’s theory of relativity is of special interest in my

approach to dialectic, mainly because of the way in which he got rid of some absolute
notions like those of space and time, and that of force acting at a distance. Central
question: simultaneity.
Quotations 93.6. (1) Einstein [1949a], p. 53, quoted after Wang [1974], p. 12.
A paradox upon which I had already hit at the age of sixteen: If I pursue
a beam of light with the velocity c (velocity of light in a vacuum), I should
observe such a beam of light as a spatially oscillatory electro-magnetic field
at rest.
(2) Einstein [1949b], p. 679, quoted after Wang [1974], p. 12.
Today everybody knows, of course, that all attempts to clarify this paradox
satisfactorily were condemned to failure as long as the axiom of the abso-
lute character of time, viz., of simultaneity, unrecognizedly was anchored in
the unconscious. Clearly to recognize this axiom and its arbitrary charac-
ter really implies already the solution of the problem. The type of critical
reasoning which was required for the discovery of this central point was
decisively furnished, in my case, especially by the reading of David Hume’s
and Ernst Mach’s philosophical writings.
(3) Einstein [1920], p. 22.
The concept [[simultaneity]] does not exist for the physicist until he has the
possibility of discovering whether or not it is fulfilled in an actual case. We
thus require a definition of simultaneity such that this definition supplies
us with the method by means of which, in the present case, he can decide
by experiment whether or not both the lightning strokes occurred simulta-
neously. As long as this requirement is not satisfied, I allow myself to be
deceived as a physicist (and of course the same applies if I am not a physi-
cist), when I imagine that I am able to attach a meaning to the statement
of simultaneity.
Comment. To counter with Frege: The concept exists, of course; that’s not for a
physicist to decide; that’s logic. What the physicist deals with is the possibility of
a mapping to observables, or realization. And there it may happen that no such
mapping is possible, in which case the concept is empty as regards its application
in physics.
(4) Einstein [1949b], p. 674; quoted after Wang [1987], p. 153.
[[My theoretical attitude]] is distinct from that of Kant only by the fact that
we do not conceive of the “categories” as unalterable (conditioned by the
nature of the understanding) but as (in the logical sense) free conventions.
They appear to be apriori only insofar as thinking without the positing of
categories and of concepts in general would be as impossible as breathing
in a vacuum.
Comment. Notice: the view of categories as, firstly, not unalterable, secondly, free
conventions.14
(5) Einstein [1952], p. vi.
[[S]]pace-time is not necessarily something to which one can ascribe a sepa-
rate existence, independently of the actual objects of physical reality. Phys-
ical objects are not in space, but these objects are spatially extended. In
this way the concept “empty space” loses its meaning.
Comment. Contrast what Kant says in quotation 61.19 (3): “We can never represent
to ourselves the absence of space, though we can quite well think it as empty of
objects.”
Quotation 93.7. Einstein [1920], p. 63.

“If we pick up a stone and then let it go, why does it fall to the ground?”
The usual answer to this question is: “Because it is attracted by the earth.”
Modern physics formulates the answer rather differently for the following
reason. As a result of the more careful study of electromagnetic phenomena,
we have come to regard action at a distance as a process impossible without
the intervention of some intermediate medium. If, for instance, a magnet
attracts a piece of iron, we cannot be content to regard this as meaning
that the magnet acts directly on the iron through the intermediate empty
space, but we are constrained to imagine—after the manner of Faraday—
that the magnet always calls into being something physically real in the
space around it, that something being what we call a “magnetic field.” In
its turn this magnetic field operates on the piece of iron, so that the latter
strives to move towards the magnet.
Quotation 93.8. Einstein [1920], p. 71.

[[C]]lassical mechanics starts out from the following law: Material parti-
cles sufficiently far removed from other material particles continue to move
uniformly in a straight line or continue in a state of rest. [[. . .]] [[T]]his fun-
damental law can only be valid for bodies of reference K which possess
certain unique states of motion, and which are in uniform translational mo-
tion relative to each other. Relative to other reference-bodies K the law is
not valid. Both in classical mechanics and in the special theory of relativity
14 As regards the second aspect, cf. Poincaré in quotation 70.23 (2).
we therefore differentiate between reference-bodies K relative to which the

recognised “laws of nature” can be said to hold, and reference-bodies K
relative to which these laws do not hold.
But no person whose mode of thought is logical can rest satisfied with
this condition of things. He asks: “How does it come that certain reference-
bodies (or their states of motion) are given priority over other reference-
bodies (or their states of motion)?

One of the most interesting aspects of relativity theory for the philo-
sophical-minded consists in the fact that it gave new and surprising insights
into the nature of time [[. . .]]
[[. . .]] [[I]]t seems that one obtains an unequivocal proof for the view of
those philosophers who, like Parmenides, Kant, and the modern idealists,
deny the objectivity of change and consider change as an illusion or an
appearance due to our special mode of perception.
Comment. There is a footnote to the last sentence which quotes Kant [1787], p. 54.15
(2) Wang [1974], pp. 12 f.
Gödel points out that the fruitfulness of the positivistic point of view
in [[Einstein’s]] case is due to a very exceptional circumstance, namely the
fact that the basic concept to be clarified, i.e. simultaneity, is directly ob-
servable, while generally basic entities (such as elementary particles, the
forces between them, etc.) are not. Hence the positivistic requirement that
everything has to be reduced to observations is justified in this case. That,
generally speaking, positivism is not fruitful even in physics seems to fol-
low from the fact that, since it has been more or less adopted in quantum
physics (i.e. about 40 years ago) no substantial progress has been achieved
in the basic laws of physics[[.]]
[[W]]hat Kepler, in a simple and sublime manner, articulated in the form of
laws of celestial motion, Newton converted into the nonconceptual, reflec-
tive form of the force of gravity. The whole manner of the “proof” presents
in general a confused tissue of lines of merely geometrical construction to
which a physical meaning of independent forces is given, of the empty con-
cepts of the understanding of a force of acceleration, of particles of time, at
whose beginning those forces always play a renewed role, and of a force of
inertia, which presumably continues its previous effect, and so on. A ratio-
nal proof of the quantitative determinations of free motion can only rest on
the determinations of the concepts of space and time, the moments whose
relation is motion.
Comment. According to Wang [1987], p. 254: “in the early 1930s [[Gödel]], says
Menger ‘showed me a passage [in a book of Hegel] which appeared to completely
anticipate general relativity theory’.”
Quotation 93.10. Capra [1975], p. 186.

Black holes are among the most mysterious and most fascinating ob-
jects investigated by modern astrophysics and illustrate the effects of rela-
tivity theory in a most spectacular way. The strong curvature of space-time
around them prevents not only all their light from reaching us, but has an
equally striking effect on time. If a clock, flashing its signals to us, were at-
tached to the surface of the collapsing star, we would observe these signals
to slow down as the star approached the event horizon, and once the star
had become a black hole, no clock signals would reach us any more. To an
outside observer, the flow of time on the star’s surface slows down as the
star collapses and it stops altogether at the event horizon. Therefore, the
complete collapse of the star takes an infinite time. The star itself, however,
experiences nothing peculiar when it collapses beyond the event horizon.
Time continues to flow normally and the collapse is completed after a finite
period of time, when the star has contracted to a point of infinite density.
So how long does the collapse really take, a finite time or an infinite time?
In the world of relativity theory, such a question does not make sense. The
lifetime of a collapsing star, like all other time spans, is relative and depends
on the frame of reference of the observer.
Comment. There seems to be some similarity to Achilles and the tortoise.
93d. Uncertainty and complementarity. The philosophical background of

quantum mechanics is often tied to an idea of interaction, viz., that between the
process of measurement, and that of a pre-existing situation. This once was very
attractive to me, but I have given it up for a number of reasons. One reason is
that even for quantum mechanics, interaction between observer and observed is not
a convincing model. There is the EPR-experiment which depends crucially on the
condition “without in any way disturbing a system”, and which occasioned Bohr
to push the problem of the influence more into the direction of the description.16
On the level of description, however, interaction is not an attractive model simply
because it is not clear who or what is acting (except, perhaps, for what the average
naive realist thinks that a loony idealist would have to think, i.e., the description
is acting).
The following is an early document concerning the situation (contribution) of
the observer.
Quotations 93.11. (1) Mason [1962], p. 296.

Laplace in 1812 put forward his famous conception of a Divine Calculator
who, knowing the velocities and positions of all the particles in the world at
a particular instant could calculate all that had happened in the past, and
all that would happen in the future. [[. . .]] The Divine Calculator was the
mathematical physicist writ large, a being who appeared to be outside of
the system he was investigating, and who contemplated the world as though
16 Cf. Niels Bohr. A Centenary Volume (A. P. French and P. J. Kennedy eds.). Harvard
University Press: Cambridge MA and London 1985, p. 142 f.

it were a play, deducing from the events of the moment the preceding and
the subsequent action. [[. . .]]
The biologist, Bonnet, 1720–90, somewhat earlier had put forward a
similar conception in the sphere of psychology. He suggested that if an
Intelligence could have analysed the workings of all the fibres in Homer’s
brain, he would have been able to picture the Iliad as it was conceived of by
the poet. However, the mathematician, Condorcet, 1743–94, perceived that
there were drawbacks to the idea when it was extended to the human sphere.
The physical, organic, and human worlds would be identical in principle,
he wrote in 1782,
‘for a being who, as a stranger to our race, studied human society as we
study the beaver and the bee . . . But here the observer is part of the society
he observes, and truth can only be judged, imprisoned, or bribed.’
Comment. Compare Lenin’s words in quotation 95.29 (Wang) below.
(2) Mason [1962], p. 297.

By the end of the eighteenth century it was appreciated that observers as
well as instruments had their own errors, and techniques were devised for
obtaining the most accurate measurement by averaging several observa-
tions.
(3) Bohm [1980], p. 49.
Of course, modern physics states that actual streams (e.g., of water) are
composed of atoms, which are in turn composed of ‘elementary particles’,
such as electrons, protons, neutrons, etc. For a long time it was thought
that these latter are the ‘ultimate substance’ of the whole of reality, and
that all flowing movements, such as those of streams, must reduce to forms
abstracted from the motions through space of collections of interacting par-
ticles. However, it has been found that even the ‘elementary particles’ can
be created, annihilated and transformed, and this indicates they too are
relatively constant forms, abstracted from some deeper level of movement.
One may suppose that this deeper level of movement may be analysable
into yet finer particles which will perhaps turn out to be the ultimate sub-
stance of the whole of reality. However, the notion that all is flux, into which
we are inquiring here, denies such a supposition. Rather it implies that any
describable event, object, entity, etc., is an abstraction from an unknown
and undefinable totality of flowing movement. This means that no matter
how far our knowledge of the laws of physics may go, the content of these
laws will still deal with such abstractions, having only a relative indepen-
dence of existence and independence of behaviour. So one will not be led to
suppose that all properties of collections of objects, events, etc., will have
to be explainable in terms of some knowable set of ultimate substances. At
any stage, further properties of such collections may arise, whose ultimate
ground is to be regarded as the unknown totality of the universal flux.
Quotation 93.12. Heisenberg [1930], p. 47.

Die Welt der aus der täglichen Erfahrung stammenden Begriffe ist zum
ersten Male verlassen worden in der Einsteinschen Relativitätstheorie. Dort
stellte sich heraus, daß man die gewöhnlichen Begriffe nur anwenden kann
auf Vorgänge, in denen die Geschwindigkeit der Lichtfortpflanzung als prak-
tisch unendlich angesehen werden kann. Das durch die moderne Experi-
mentalphysik verfeinerte Erfahrungsmaterial zwang also zu einer Revision
der überkommenen und zur Ausbildung neuer Begriffe; aber unser Denken
vermag sich nur langsam dem erweiterten Erfahrungsbereich und seiner
Begriffswelt anzupassen und daher erschien die Relativitätstheorie anfangs
abstrakt und fremd. [[D]]ie Erfahrungen aus der Welt der Atome [[zwingen]]
zu einem noch viel weitergehenden Verzicht auf bisher gewohnte Begriffe.
In der Tat beruht unsere gewöhnliche Naturbeschreibung und insbesondere
der Gedanke einer strengen Gesetzmäßigkeit in den Vorgängen der Natur
auf der Annahme, daß es möglich sei, Phänomene zu beobachten, ohne sie
merklich zu beeinflussen. Einer bestimmten Wirkung eine bestimmte Ursa-
che zuzuordnen, hat nur dann einen Sinn, wenn wir Wirkung und Ursache
beobachten können, ohne gleichzeitig in den Vorgang störend einzugreifen.
Das Kausalgesetz in seiner klassischen Form kann also seinem Wesen nach
nur für abgeschlossene Systeme definiert werden. In der Atomphysik ist aber
im allgemeinen mit jeder Beobachtung eine endliche, bis zu gewissem Grade
unkontrollierbare Störung verknüpft, wie dies in der Physik der prinzipiell
kleinsten Einheiten auch von vornherein zu erwarten war. Da andererseits
jede raum-zeitliche Beschreibung eines physikalischen Vorganges durch die
Beobachtung des Vorganges bedingt ist, so folgt daß die raum-zeitliche Be-
schreibung von Vorgängen einerseits und das klassische Kausalgesetz an-
dererseits komplementäre, einander ausschließende Züge des physikalischen
Geschehens darstellen.
I turn to the problem of interpretation: the classical versus the quantum world.

Heisenberg tried to explain the limitations which his interpretation im-
posed on all possible measurements by pointing out that if we measure an
elementary particle, we disturb it or interfere with it.
This early interpretation of Heisenberg’s implied that the particle had
a sharp position and momentum; but owing to our interfering with it, we
could never measure both sharply.
Quotations 93.14. (1) Heisenberg [1958], p. 58.

If one wants to give an accurate description of the elementary particle —
and here the emphasis is on the word “accurate” — the only thing that can
be written down as a description is a probability function. But then one
sees that not even the quality of being (if that may be called a “quality”)
belongs to what is described. It is a possibility for being or a tendency for
being.
(2) Heisenberg [1958], p. 69.

Natural science does not simply describe and explain nature; it is a part of
the interplay between nature and ourselves; it describes nature as exposed
to our method of questioning. This [[. . .]] makes a sharp separation between
the world and the I impossible.
(3) Heisenberg [1958], p. 70.
[[A]]ctually the position of classical physics is that of dogmatic realism. It
is only through quantum theory that we have learned that exact science is
possible without the basis of dogmatic realism.
Quotation 93.15. Heisenberg [1958], pp. 61 f.

After this comparison of the modern views in atomic physics with Greek
philosophy we have to add a warning, that this comparison should not be
misunderstood. It may seem at first that the Greek philosophers have by
some kind of ingenious intuition come to the same or very similar conclu-
sions as we have in modern times only after several centuries of hard labor
with experiments and mathematics. This interpretation of our comparison
would, however, be a complete misunderstanding. There is an enormous
difference between modern science and Greek philosophy[[.]]
Quotation 93.16. Bohm [1951/79], p. 624.

Classical concepts are characterized by three assumptions concerning the
properties of matter:
(1) The world can be analyzed into distinct elements.
(2) The state of each element can be described in terms of dynamical
variables that are specifiable with arbitrarily high precision.
(3) The interrelationship between parts of a system can be described
with the aid of exact causal laws that define the changes of the above
dynamical variables with time in terms of their initial values. The behavior
of the system as a whole can be regarded as the result of the interaction of
all of its parts.
Quotations 93.17. (1) Bohm [1980], p. 57.

[[T]]hought proper begins in this way with thought, conscious of itself
through its distinguishing itself from nonthought.
Comment. Has Bohm read Fichte?
(2) Bohm [1980], p. 64.
We have to be very alert and careful [[. . .]], for we tend to try to fix
the essential content of our discussion in a particular concept or image, and
talk about this as if it were a separate ‘thing’ that would be independent
of our thought about it. We fail to notice that in fact this ‘thing’ has by
now become only an image, a form in the overall process of thought, i.e.,
response of memory, which is a residue of past perception through the mind
(either someone else’s or one’s own). Thus, in a very subtle way, we may
[[. . .]] be trapped in a movement in which we treat something originating
in our own thought as if it were a reality originating independently of this

thought.
Comment. Note the similarity in the first part to Kant’s explanation of how reason
falls into antinomies: treating appearances as independent of representation.
(3) A. Petersen [1985], p. 302

Traditional philosophy has accustomed us to regard language as some-
thing secondary, and reality as something primary. Bohr considered this
attitude toward the relation between language and reality inappropriate.
When one said to him that it cannot be language which is fundamental,
but that it must be reality which, so to speak, lies beneath language, and of
which language is a picture, he would reply, “We are suspended in language
in such a way that we cannot say what is up and what is down. The word
‘reality’ is also a word, a word which we must learn to use correctly.”
(4) Penrose [1994], p. 309.
One of the main issues concerns the ‘reality’ of the quantum formal-
ism—or even of the quantum-level world itself. In this connection, I cannot
resist quoting a remark that was made to me by Professor Bob Wald, of
the University of Chicago, at a dinner-party some years ago:
If you really believe in quantum mechanics, then you can’t take it

seriously.
It seems to me that this expresses something profound about quantum the-

ory and about people’s attitude to it. Those who are most vehement about
accepting the theory as being in no way in need of modification tend not to
think that it represents the actual behaviour of a ‘real’ quantum-level world.
Niels Bohr, who was a leading figure in the development and interpretation
of quantum theory, was one of the most extreme in this respect. He seems to
have regarded the state vector as no more than a convenience, useful only
for calculating probabilities for the results of ‘measurements’ that might be
performed on a system. The state vector itself was not to be thought of as
providing an objective description of any kind of quantum-level reality, but
as representing merely ‘our knowledge’ of the system. Indeed, it was to be
regarded as doubtful that the very concept of ‘reality’ applied meaningfully
at the quantum level. Bohr was certainly someone who ‘really believed in
quantum mechanics’, and his view of the state vector seemed, indeed, to
be that it should not be ‘taken seriously’ as the description of a physical
reality at the quantum level.

Measurement only determines what is already the case: it does not bring
into existence the observable measured, or cause it to ‘take on a sharp value’
which it did not already possess.
(2) Putnam [1981], p. 194.

In the two-slit experiment, photons or electrons or any other quantum
mechanical particles are released from a point source. Between the particles
(say, photons) and the detector (say, a photographic plate) a barrier is
placed in the form of a wall with two fine slits. The uncertainty in the
position of the photon — more precisely, the fact that the photon is in a
superposition of various position states — permits each photon to interact
with both slits, so that what one gets on the photographic plate is not
a simple sum of the patterns that one would obtain by just performing
the experiment with the right slit open. Rather, it is as if, in the case of
each photon that gets through, half the photon went through each slit and
the two halves then intermingled and interfered (in the manner of waves
— in fact, this phenomenon constitutes the celebrated ‘wave aspect’ of
the photon). The final result is a system of visible interference fringes in
the photographic picture. Yet, in spite of all this wave-like behavior, each
individual photon strike like the emulsion at one and only one definite point.
(2) Dalla Chiara [1977], p. 331.
From a logical point of view the measurement problem of quantum me-
chanics (QM) can be described as a characteristic question of ‘semantical
closure’ of a theory: to what extent can a consistent theory (in this case
QM) be closed with respects to the objects and the concepts which are
described and expressed in its metatheory? As is well known, the limita-
tive theorem of logic and the paradoxes of set theory teach us that there
are some definite limits to the semantical closure of any consistent theory
(which satisfies some standard formal requirements). In particular, such
theories can never express and prove all that is expressed and proved in
their metatheories; further, they cannot generally describe (up to certain
limitations) their universe as their own object. From an intuitive point of
view one can recall a number of arguments which justify why a well-behave
scientific theory cannot be logically self-sufficient. However, we cannot help
finding a much more disagreeable situation when a given physical theory
(QM), owing to purely logical reasons, turns out to be subject to some limi-
tations concerning its capacity of describing and expressing certain specific
physical objects and concepts. This is the curious situation which arises
with the measurement problem of QM.
Comment. This would still be my aspiration, to develop a unified theory of the
fixed point property and the measurement problem.
Quotation 93.19. Mermin [1981], p. 397.

We often discussed his notions on objective reality. I recall that during
one walk, Einstein suddenly stopped, turned to me and asked whether I
really believed that the moon exists only when I look at it.
A. Pais
As O. Stern said recently, one should no more rack one’s brain about the
problem of whether something one cannot know anything about exists all
the same, than about the ancient question of how man angels are able to
sit on the point of a needle. But it seems to me that Einstein’s questions
are ultimately always of this kind.
W. Pauli
Pauli and Einstein were both wrong. The questions with which Einstein
attacked the quantum theory do have answers; but they are not the an-
swers Einstein expected them to have. We now know that the moon is
demonstrably not there when nobody looks.
Quotations 93.20. (1) Edwards [1979], p. 4.

We will assume that each experiment that one is interested in performing
has a finite formal description E, and that performing E yields a result
in a finite set RE . The experiments can be thought of as being on a fixed
‘system’ which may be in various internal states α ∈ S depending upon how
we prepared the system. The breakup of the description of an experiment
into the pair (α,E), consisting of an initial ‘preparation’ followed by an
actual ‘observation’, though somewhat conventional, is very important for
the development of our formalism. On the other hand, the notion of the
‘system itself’ is often quite opaque and misleading. It plays a role similar
to Kant’s ding an sich. There is little trouble when one is talking about
billiard balls, tables, and the like, where one can simply ‘point’ to the
system itself which one is trying to study. But how does one ‘point’ to
an electron? Our experiments should really be thought of as ‘modes of
perceiving’, ‘ways of looking’, and the results of our experiments should
be thought of as ‘what was seen’. Careless use of the notion of the ‘system
itself’ inevitably leads to the notorious measurement problem which itself is
reminiscent of the classical problem of the relationship between phenomena
and numina. In our approach one deals strictly with phenomena and never
even supposes the existence of numina. Of course, by invoking Berkeley’s
universal observer one can connect our approach with that of the traditional
realist position.
(2) Mermin [1981], p. 405.
[[T]]he prevailing interpretation of the quantum theory [[. . .]] emphatically
denied the existence of instruction sets, insisting that certain physical prop-
erties (said to be complementary) had no meaning independent of the ex-
perimental procedure by which they were measured. Such measurement, far
from revealing the value of a preexisting property, had to be regarded as
an inseparable part of the very attribute they were designed to measure.
Properties of this kind have no independent reality outside the context of a
specific experiment arranged to observe them: the moon is not there when
nobody looks.
(3) Edwards [1979], pp. 37 f.
[[C]]lassical systems are characterized by having all of their (theoretical)
observables compatible, i.e., there is a single theoretical perspective from
which one can see all there is to be seen. In order to obtain a truly non-
classical theory, one must assume that the deep logic is non-Boolean and
that its Boolean subalgebras interlock in a complex fashion. That quantum
mechanics contain incompatible observables, such as the position and mo-
mentum observables, Q and P, was first realized by Dirac and Heisenberg
in late 1926.
(4) Bohm [1951/79], p. 625.
[[I]]t is only at the classical level that definite results for an experiment
can be obtained, in the form of distinct events which are associated in a
one-to-one correspondence with the various possible values of the physical
quantity that is being measured. This means that without an appeal to
a classical level, quantum theory would have no meaning. We conclude
then that quantum theory presupposes the classical level and the general
correctness of classical concepts in describing this level; it does not deduce
classical concepts as limiting cases of quantum concepts.
The next quotation concerns the idea of ultimate substance; it should be seen,
at least partly, in the context of Kant’s second antinomy.
Quotations 93.21. (1) Capra [1975], pp. 78 f.
In the history of man’s penetration into this submicroscopic world, a stage
was reached in the early 1930s when scientists thought they had now fi-
nally discovered the ‘basic building blocks’ of matter. It was known that
all matter consisted of atoms and that all atoms consisted of protons, neu-
trons and electrons. These so-called ‘elementary particles’ were seen as the
ultimate indestructible units of matter: atoms in the Democritean sense.
Although quantum theory implies, as mentioned previously, that we can-
not decompose the world into independently existing small units, this was
not generally perceived at that time. The classical habits of thought were
still so persistent that most physicists tried to understand matter in terms
of its ‘basic building blocks’, and this trend of thought is, in fact, quite
strong even today.
(2) Capra [1975], p. 82.
[[T]]he basic question was whether one could divide matter again and again,
or whether one would finally arrive at some smallest indivisible units. After
Dirac’s discovery, the whole question of the division of matter appeared in
a new light. When two particles collide with high energies, they generally
break into pieces, but these pieces are not smaller than the original particles.
They are again particles of the same kind and are created out of the energy
of motion (‘kinetic energy’) involved in the collision process. The whole
problem of dividing matter is thus resolved in an unexpected sense. The
only way to divide subatomic particles further is to bang them together in
collision processes involving high energies. This way, we can divide matter
again and again, but we never obtain smaller pieces because we just create
particles our of the energy involved in the process. The subatomic particles
are thus destructible and indestructible at the same time.
Comment. Cf. Kant’s second antinomy, Kant [1781] (A), pp. 434 ff.
Quotations 93.22. (1) Capra [1975], pp. 303 f.

To discover the ultimate fundamental laws of nature remained the aim of
natural scientists for the three centuries following Newton.
In modern physics, a very different attitude has now developed. Physi-
cists have come to see that all their theories of natural phenomena, includ-
ing the ‘laws’ they describe, are creations of the human mind; properties of
our conceptual map of reality, rather than of reality itself. This conceptual
scheme is necessarily limited and approximate, as are all the scientific the-
ories and ‘laws of nature’ it contains. All natural phenomena are ultimately
interconnected, and in order to explain any one of them we need to under-
stand all the others, which is obviously impossible. What makes science so
successful is the discovery that approximations are possible. If one is sat-
isfied with an approximate ‘understanding’ of nature, one can describe se-
lected groups of phenomena in this way, neglecting other phenomena which
are less relevant. Thus one can explain many phenomena in terms of a few,
and consequently understand different aspects of nature in an approximate
way without having to understand everything at once. This is the scientific
method; all scientific theories and models are approximations to the true
nature of thing, but the error involved in the approximations is often small
enough to make such an approach meaningful.
(2) Popper [1982], p. 3.
We owe to Kant the first great attempt to combine a realistic interpre-
tation of natural science with the insight that our scientific theories are not
simply the result of a description of nature—of ‘reading the book of nature’
without prejudice’—but that they are, rather, the products of the human
mind: ‘Our intellect does not draw its laws from nature, but it imposes
its laws upon nature.’ I have attempted to improve this excellent Kantian
formulation as follows: ‘Our intellect does not draw its laws from nature,
but it tries—with varying success—to impose upon nature laws which it
freely invents.
Quotation 93.23. Laing and Cooper [1964], p. 40.

The only valid theory of knowledge today is one which is founded on that
principle of microphysics which asserts that the experimenter is part of the
experimental system.
Quotation 93.24. Kojève [1947], p. 177.

The isolated Object is but an abstraction, and that is why it has no fixed
and stable continuity (Bestehen) and is perpetually deformed or perturbed.
Therefore it cannot serve as a basis for a Truth, which by definition is
universally and eternally valid. And the same goes for the “object” of vulgar
psychology, gnoseology, and philosophy, which is the Subject artificially
isolated from the Object—i.e., yet another abstraction.
Comment. The last sentence comes with a footnote which is actually the raison
d’être n for the whole quotation:
This interpretation of science, on which Hegel insisted very much, is
currently admitted by science itself. In quantum physics it is expressed in
mathematical form by Heisenberg’s relations of uncertainty. These relations
show on the one hand that the experience of physics is never perfect, be-
cause it cannot achieve a description of the “physical real” that is both
complete and adequate (precise). On the other hand, the famous principle
of “complementary notions” follows from it, formulated by Bohr: that of
the wave and the particle, for example. This means that the (verbal) phys-
ical description of the Real necessarily implies contradictions: the “physical
real” is simultaneously a wave filling all of space and a particle localized in
one point, and so on. By its own admission, Physics can never attain Truth
in the strong sense of the term.
I take this opportunity to remind readers of the second motto to this chapter, the
one proffered by Wang.
Quotation 93.25. Quine [1948], pp. 18 f

The analogy between the myth of mathematics and the myth of physics
is [[. . .]] strikingly close. Consider, for example, the crises which was pre-
cipitated in the foundations of mathematics, at the turn of the century,
by the discovery of Russell’s paradox and other antinomies of set theory.
These contradictions had to be obviated by unintuitive, ad hoc devices;
our mathematical myth-making became deliberate and evident to all. But
what of physics? An antinomy arose between the undular and the corpus-
cular accounts of light; and if this was not as out-and-out a contradiction as
Russell’s paradox, I suspect the reason is that physics is not as out-and-out
as mathematics. Again, the second great modern crisis in the foundations
of mathematics—precipitated in 1931 by G"odel’s proof [[. . .]] that there
are bound to be undecidable statements in arithmetic—has its companion
piece in physics in Heisenberg’s indeterminacy principle.
Comment. So the ‘myth’ is already there; and where is the ‘narrative’ ?
A last couple of quotations is intended as a warning of missiles of mysticism

and mythology launched from a physicist’s basis.
Quotations 93.26. (1) Davies [1989], p. 3.

[[T]]he idea that time does not stretch back for all eternity but was created
with the universe was anticipated in the fifth century by St Augustine.
There is thus a scientific counterpart to the creation ex nihilo of Christian
tradition.
(2) Davies [1989], p. 9.
The unexpected success of simple mathematical laws in physics bolsters the
belief that science is tapping into an already existing external reality.
§ 94. The foundations of mathematics as a philosophical challenge
The precision in the studies on the foundations of mathematics has attracted many
philosophers. Though very often their dealing with the subject did not go beyond the
level of kids playing pilot in a cardboard box, the project of a logicist foundation
of mathematics — along with its failure, the antinomies — initiated some quite
remarkable efforts by mathematically minded philosophers. Since I am going to use
certain paradigms developed in the foundations of logic and mathematics later on
anyway, I shall only deal superficially with this topic in the present paragraph.
94a. The mathematical method: discovery vs. invention. In quotation
85.19 (2), Hilbert (Über das Unendliche) speaks of “the ingenious method of ideal
elements”. Something of this kind will form an integral part in my theory of dialectic.
The present section is devoted to a few quotations to illustrate the situation in
mathematics.
I begin with a general remark concerning the development of mathematics in
the last two hundred years.

As mathematics passed the year 1800 and entered the recent period, there
was a steady trend toward increasing abstractness and generality. By the
middle of the nineteenth century, the spirit of mathematics had changed so
profoundly that even the leading mathematicians of the eighteenth century,
could they have witnessed the outcome of half a century’s progress, would
scarcely have recognized it as mathematics. The older point of view of
course persisted, but it was no longer that of the men who were creating
new mathematics. Another quarter of a century, and it had become almost
a disgrace for a first-rank mathematician to attack a special problem of
the kind that would have engaged Euler in much of his work. Abstractness
and generality, directed to the creation of universal methods and inclusive
theories, became the order of the day.
(2) E. T. Bell [1940/45], p. 170.
Most of the great mathematicians of the eighteenth and early nineteenth
centuries were more like engineers than modern mathematicians in their
thinking; a formula revealed in a flash of intuition, or hastily inferred from
loose reasoning, was as good as any other provided it worked. Their formulas
worked admirably. Gauss (1777–1855) was the first great mathematician to
rebel successfully against intuition in analysis. Lagrange (1736–1813) had
tried and failed.
(3) Russell [1950], p. 369.
Let us enumerate a few of the errors that infected mathematics in the
time of Hegel. There was no defensible definition of irrational numbers,
and consequently no ground for the Cartesian assumption that the position
of any point in space could be defined by three numerical co-ordinates.
There was no definition of continuity, and no known method of dealing
with the paradoxes of infinite number. The accepted proofs of fundamental
§ 94. FOUNDATIONS OF MATHEMATICS AS A PHILOSOPHICAL CHALLENGE 1305
propositions in the differential and integral calculus were all fallacious, and
were supposed, not only by Leibniz, but by many later mathematicians, to
demand the admission of actual infinitesimals.
The next set of quotations is meant to give an illustration of the way in which
new objects are added in mathematics, i.e., how the mathematical realm of discourse
has been enlarged. This is not just in view of Hilbert’s “method of ideal elements” 17
but also in view of Hegel’s “thought-forms”.

Until positive rational fractions and negative numbers were invented
by mathematicians (or ‘discovered,’ if the inventors happened to be Pla-
tonic realists), a quadratic equation with rational integer coefficients had
precisely one root, or precisely two, or precisely none. A Babylonian of a
sufficiently remote century who gave 4 as the root of x2 = x + 12 had solved
his equation completely, because –3, which we now say is the other root, did
not exist for him. Negative numbers were not in his number system. The
successive enlargements of the number system necessary to provide all alge-
braic equations with roots equal in number to the respective degrees of the
equation was one of the outstanding landmarks in mathematical progress,
and it took about four thousand years of civilized mathematics to establish
it.
Comment. ‘Ontology’.
(2) E. T. Bell [1940/45], p. 172.
The earliest extensions of the system of natural numbers were the Baby-
lonian and Egyptian fractions. These illustrate one prolific method of gen-
erating new numbers from those already accepted as understood, namely,
inversion. To solve the problem ‘by what must 6 be multiplied to produce
2?’, a new kind of ‘number,’ the fraction 31 , must be invented. Here the
direct operation is multiplication, and the inverse, division. The other pairs
of elementary inverses are addition and subtraction; raising to powers and
extracting roots.
All of these elementary operations were known to the ancients. The
inverses, division and subtraction, of the rational operations, multiplication
and addition, necessitated the invention of common fractions and negative
numbers; the operation inverse to powering was in part responsible for the
invention of irrationals, including the pure imaginaries and the ordinary
complex numbers. The solution of an algebraic equation, or of a system
of such equations in several unknowns, can be restated as a problem in
inversion with respect to iterations of addition and multiplication. Up to
about 1840, algebraic equations were probably the most prolific source of
extensions of the natural numbers.

(3) E. T. Bell [1940/45], pp. 173 f.

Diophantus, in the third century a.d., encountering −4 as the formal solu-
tion of a linear equation, rejected it as absurd. In the first of the seventh
century, Brahmagupta is said to have stated the rules of signs in multiplica-
tion; he discarded a negative root of a quadratic. The rules of signs became
common in India after their restatement by Mahavira in the ninth century.
Al-Khowarizmi, of about the same time, made no advance except that he
appears to have exhibited a positive and a negative root for a quadratic
without explicitly rejecting the negative.
Of the Europeans, Fibonacci in the early thirteenth century rejected
negative roots, but took a step forward when he interpreted a negative
number in a problem concerning money as a loss instead of a gain. It has
been claimed that the Indians did likewise. L. Pacioli (1445 ?–1514, Tuscan)
in the second half of the fifteenth century is credited with a knowledge of the
rule of signs on such evidence as (7−4)(4−2) = 3×2 = 6. M. Stifel (1487 ?–
1567, German) a fine algebraist for his time, called negative numbers absurd
in the middle of the sixteenth century. Cardan, in his Ars magna (1545),
stated the rule ‘minus times minus gives plus’ as an independent position;
he also is said to have recognized negative numbers as ‘existent,’ but on
evidence which seems doubtful. In fact, he called negatives ‘fictitious.’
Quotations 94.3. (1) Ariew and Garber [1989], p. 236.

As for myself, I cherished mathematics only because I found in it the traces
of the art of invention in general [[.]]
(2) Ariew and Garber [1989], p. 237.
[[. . .]] I have recognized that metaphysics is scarcely different from the true
logic, that is, from the art of invention in general; for, in fact, metaphysics
is natural theology, and the same God who is the source of all goods is also
the principle of all knowledge. This is because the idea of God contains
within it absolute being, that is, what is simple in our thought, from which
everything that we think draws its origin.
Comment. While the first part may sound quite modern, the second seems to carry
a distinct undertone of a bygone age. But I wish to remind readers that our idea
of God as some kind of patriarchal being that has chosen to be represented on this
planet by old men rallying against contraception and denying women to take part
in their colourful performances, is a very idiosyncratic one indeed; certainly not
anything that could be called a principle.
I add a few quotations on arithmetization as a tool.
Quotations 94.4. (1) P. P. Wiener [1951], p. 15; translating Leibniz C 155–156;

instigated by Burks [1986], pp. 2 f.
[[I]]f we could find characters or signs appropriate for expressing all our
thoughts as definitely and exactly as arithmetics expresses numbers or geo-
metric analyses expresses lines, we could in all subjects in so far as they
are amendable to reasoning accomplish what is done in Arithmetic and

Geometry.
For all inquiries which depend on reasoning would be performed by
the transposition of characters and by a kind of calculus, which would
immediately facilitate the discovery of beautiful results. For we should not
have to break our heads as much as is necessary today, and yet we should
be sure of accomplishing everything the given facts allow.
Moreover we should be able to convince the world what we should have
found or concluded, since it would be easy to verify the calculation either
by doing it over or by trying tests similar to that of casting our nines in
arithmetic. And if someone would doubt my results, I should say to him:
“Let us calculate, Sir,” and thus by taking to pen and ink, we should soon
settle the question.
I still add in so far as the reasoning allows on the given facts. For
although certain experiments are always necessary to serve as a basis for
reasoning, nevertheless, once these experiments are given, we should derive
from them everything which anyone at all could possible derive; and we
should even discover what experiments remain to be done for the clarifica-
tion of all further doubts.
(2) Burks [1986], pp. 2 f, commenting essentially on the foregoing quotation.
To accomplish this, Leibniz planned to develop a universal language in
which all terms would be represented by positive integers. Atomic terms
would be represented by prime numbers, and compound terms by products
of primes. For example, if the term “animal” is expressed by 2 and “rational”
by 3, the term man is expressed by 2 time 3, or 6. The statement “all men
are rational” then reduces to the arithmetic truth “6 is divisible by 3.”
Here are the seeds of modern computer data processing. Leibniz is
proposing to code alphabetic symbols as numerals, so that alphabetic in-
formation can be processed numerically!
(3) Shankar [1994], pp. 1 f, footnote 1; quoting Leibniz.
What must be achieved is in fact this: that every paralogism be recognized
as an error of calculation, and every sophism when expressed in this new
kind of notation, appear as a solecism or barbarism, to be corrected easily
by the laws of this philosophical grammar.
Once this is done, then when a controversy arises, disputation will no
more be needed between two philosophers than between two computers. It
will suffice that, pen in hand, they sit down to their abacus and (calling in
a friend, if they so wish) say to each other: let us calculate.
(4) Odifreddi [1989], pp. 87.
The first attempt to find number-like connections between propositions of
various sorts probably goes back to Lullus’ Ars Magna, but it was Leibniz
[1666] who dreamt of arithmetization as a general method to replace rea-
soning in natural language by arithmetical propositions, with the goal of
substituting arguments with computations. [[. . .]]
[[. . .]]
Contrary to Leibniz’s dreams, the effect of arithmetization was, ironi-
cally, not to shield the language against its oddities, but rather to spring a
leak in Arithmetic, through which the linguistic paradoxes poured only to
reveal the inadequacies of formalism.
Comment. Now here is a notion of irony for you; or, actually for me. Just contrast
the empty rhetoric in Flay [1990], quotation 69.4 (11) in these materials.
Quotations 94.5. (1) Hazen [1983] p. 333.

Abstraction leads to mathematical power, but also to the risk of mean-
inglessness and contradiction. The infinitesimals of 18th-century analysis
are a typical example: with their aid, classical mathematical physics was
developed, but their definitions were too contradictory to define anything,
and, whatever reconstructions our logically more self-conscious age has
managed to provide, it was far from clear to contemporaries what consistent
principle would permit the separation of the useful from the absurd conse-
quences of their postulation. The 19th century reacted against them, and
attempted to give a rigorous foundation for analysis. The new foundation,
however, again postulated and depended on generalizations about abstract
objects as far removed from any sort geometrical or physical intuition as
the supposed infinitesimals had been. Real numbers were constructed as
limits of arbitrary converging sequences of rationals (or, in Dedekind’s ver-
sion, of arbitrary bounded sets of rationals), without regard to how these
sequences (sets) were to be defined, and functions as arbitrary sets of or-
dered pairs of these real numbers (subject to the sole condition that every
real be a first element of a pair in the set, and be paired only with one
second element), again without regard to how these sets were to be defined.
When pursued beyond the immediate needs of the foundations of analysis,
the principles postulated for these sets allowed the definition of an infinite
hierarchy of ever increasing infinite ‘numbers’. By the end of the century
another reaction had set in, with one school of mathematicians, led by
Poincaré, rejecting the set theoretic foundations as resting on unjustified
assumptions. They were delighted when, at around the turn of the century,
the fundamental principles of the new theory, implicit in the work of oth-
ers and formulated to new standards of logical explicitness by Frege, were
discovered to lead to explicit contradictions.
(2) U. Petersen [2002], p. 1308, translating Kowalewsky [1918], p. 731.
Before the discovery of set theory one thought to be justified without a
second thought to deny space and time transcendental reality because they
are infinite.
This was because in mathematics one only knew one infinity: the po-
tential infinity of a variable magnitude which grows in such a way as to
transcend any given limit. Because one didn’t have any other concept of in-
finity, one thought that the infinite as such would have to be indeterminate
(unbestimmt), incomplete (unabgeschlossen), and thus inferred the ideality
of space as follows: space is infinite, i.e. at least regarding extension inde-

terminate, incomplete, and something that is indeterminate cannot exist in
itself, hence space cannot exist in itself. Now that we have set theory we
know that the concept of infinity does not necessarily include that of the
indeterminate-incomplete being.
Quotations 94.6. (1) E. E. Harris [1987], p. 32.

Mathematicians have extended the concept of number to include neg-
ative, irrational, and imaginary numbers, and they interpret the symbols
for these higher types of numbers so that, in calculation with them, the
fundamental algebraical laws are not violated. These are the laws of com-
mutation, association, and distribution. Such laws obviously hold only if
the symbols employed stand for entities of the kind defined by Frege; that
is, sets of unitary particulars mutually in external relation.
Comment. Compare Hegel’s “severally independent”-phrase in quotation 63.18 (5)
of these materials and also variations 108.11 in the groundworks. The quotation
is of further interest in view of the structural rules in the sequential calculi of logic:
assumptions in classical logic are entities which stand to each other in external
relation; hence their manipulation is guided, in a certain sense, by fundamental
algebraical laws. Cf. the comment to the next quotation.
(2) Harris [1987], p. 33.
If [[. . .]] the units that made up a collection were internally related so
that they affected one another in certain ways or constituted one another
by their mutual relations, if, in short, we were dealing with wholes and
not with mere collections, the order in which the elements were aggregated
would not be indifferent and the algebraic laws would no longer hold.
Logicians have been particularly careful to ensure that their manip-
ulation of the symbols which they employ should conform to the three
fundamental algebraic laws, and we now see that the condition for this is
that these symbols should represent entities that are externally related or
are composed of externally related elements. That [[. . .]] was precisely how
Frege viewed concepts and the objects which fall under them; these are
what logic is about, and the metaphysical presupposition of a pluralistic
atomism is [[. . .]] apparent.
Comment. I suggest that this be seen in connection with Cantor’s “Vielheiten un-
verbundener Dinge” in quotation 73.13. Compare also the immediately preceding
quotation by the same author and relate the structural rules to the condition of
externally related elements; for instance: exchange to commutation and contraction
to distributivity (at least partly). The point of what I am doing in part G in the
groundworks is to start logic without making any such assumption, i.e., without
any structural rules.
You may call it mischief, but I just have great pleasure in closing this section
with a sample of Heidegger’s thoughts regarding the nature of mathematics.
Quotation 94.7. Krell [1978], p. 276; translating Heidegger [1962].

We see three chairs and say that there are three. What “three” is the three
chairs do not tell us, nor three apples, three cats, nor any other three things.
Rather, we can count three things only if we already know “three.” In thus
grasping the number three as such, we only expressly recognized something
which, in some way, we already have. This recognition is genuine learning.
The number is something in the proper sense learnable, a mathēma, i.e.,
something mathematical.
94b. The issue of foundation. I begin with a quotation from Descartes,

taken from the Cambridge Dictionary of Philosophy, regarding the paradigm of
geometry (more geometrico).
Quotation 94.8. Cambridge Dictionary of Philosophy, p.193, second column, on

Descartes.
Descartes saw mathematics as a kind of paradigm for all human under-
standing: “those long chains composed of very simple and easy reasonings,
which geometers customarily use to arrive at their most difficult demonstra-
tions, gave me occasion to suppose that all the things which fall within the
scope of human knowledge are interconnected in the same way” (Discourse
on the Method, Part II).
Comment. Cf. Dedekind’s remark regarding our “Treppenverstand” in quotation
71.2 (3) in these materials.
Quotations 94.9. (1) Shoenfield [1962], p. 1.

For what reason do we accept the axioms? We might try to use ob-
servation here: but this is not very practical and is hardly in the spirit of
mathematics. We therefore attempt to select as axioms certain laws which
we feel are evident from the nature of the concepts involved.
(2) Shoenfield [1962], p. 132.
One of the principal objectives of the foundations of mathematics is to “ex-
plain”, as far as possible, the basic concepts of mathematics. In the simplest
case, concepts are explained in terms of other concepts. Thus the concept
of rational number is explained in terms of the simpler concepts of integer
and ordered pair. The object is not to reduce all problems about ratio-
nal numbers to problems about integers and ordered pairs, but to be able
to say that we understand the notion of a rational number as fully as we
understand the notions of integer and ordered pair.
In the case of such fundamental concepts as natural number or set, such
a procedure is impossible; we cannot hope to explain these concepts fully
in terms of other concepts. However, we can hope to explain the concept of
set in stages. Thus we might start with certain concepts which we regard as
being thoroughly understood; these might include some particularly simple
sets which we feel offer no problem to the understanding. In terms of these,
we may explain certain further sets.
Comment. Note the quotation marks around the first use of explain.

At first sight, it is not easy to understand why so much attention is paid
to the thesis that mathematics is reducible to logic. One important reason
is undoubtedly its relation to the contention of logical empiricists that all
a priori propositions are analytic, since it is relatively easy to concede that
logic is analytic and since mathematics is conspicuously a priori but not
easily seen to be analytic.
(2) Wang [1974], pp. 23 f; partly quoting Blanshard.
[[O]]nce mathematics, which has been traditionally the main stumbling block
to empiricism, has been taken care of by reducibility, it becomes easier for
ardent empiricists to stipulate that no nonanalytic knowledge can be a pri-
ori. Now why is this analyticity thesis of all a priori knowledge regarded
as so important by philosophers? The reason is that for a long time the
majority of philosophers had regarded the chief concern of philosophy as
the study of a priori knowledge about the world. From this point of view,
the analytic or verbal doctrine of the a priori, if true, would trivialize or
wipe out philosophy as a profession. To quote Blanshard [[[1962], p. 259]]:
This account of a priori knowledge is ingenious, plausible, and ex-
tremely important. It is important because, if true, it effectively dis-
credits the main instrument of speculative philosophy and theology,
and also many of what were supposed to be their self-evident conclu-
sions. . . . In the positivist conception of reason, the necessity of an
insight is no guarantee that it applies to nature or reveals anything
outside the confines of meaning. . . . That a new theory would find
us with our occupation gone is no argument against it.
Quotation 94.11. Stegmüller [1976], pp. 1 f.

The development of the philosophy of mathematics into an exact science
called metamathematics was prompted by the foundational crisis in mathe-
matics. And since this crisis was spawned by the discovery of the antinomies
in set theory, it is often thought to represent a tragic event in modern math-
ematics. Judging from the consequences, however, one might sooner arrive
at the opposite conclusion: the discovery of antinomies was a most fortunate
event, for it forcefully urged the formalization and precise articulation of
the objects with which the philosophy of mathematics deals. Intuitive ideas
about mathematical thinking were replaced by objects capable of being
exactly described, and the philosophy of mathematics evolved into mathe-
matical foundational studies which in all branches led to disciplines which
today are considered a part of mathematics.
Unfortunately, empirical sciences have not undergone a foundational
crisis of this sort, and thus there has been no compelling reason to transform
conceptual and theoretical constructions in these disciplines into objects
suitable for exact metatheoretical study. One simply went ahead as if such
objects exist. [[. . .]]
Thus, there remained only a philosophy-of-the-as-if patterned after
the big brother metamathematics. What proved so extraordinarily fruitful
there, must prove fruitful here too.
Comment. It will probably be clear that this is the point where I take transcen-
dental idealism to be enjoying a prima facie advantage: Kant came forward with
an antinomy of pure reason more than hundred years before set theoretical antin-
omies were discovered. But then, Kant’s antinomy is hardly convincing. So the set
theoretical antinomies were a godsend.
Quotation 94.12. Wang [1987], p. 288, quoting Gödel.

‘Concrete intuition,’ ‘concretely intuitive’ are used as translations of ‘An-
schauung,’ ‘anschaulich.’ [[. . .]] What Hilbert means by ’Anschauung’ is sub-
stantially Kant’s space-time intuition confined, however, to configurations
of a finite number of discrete objects. Note that it is Hilbert’s insistence on
concrete knowledge that makes finitary mathematics so surprisingly weak
and excludes many things that are just as incontrovertibly evident to ev-
erybody as finitary number theory. E.g., while any primitive recursive defi-
nition is finitary, the general principle of primitive recursive function is not
a finitary proposition, because it contains the abstract concept of function.
94c. Philosophical quibbles. Hegel’s dogma that it is only “the subordi-

nate form of scientific method which can be employed in mathematics” 18 implies
that mathematics cannot provide a method for so superior a science as philosophy.
This attitude is still fashionable amongst Continental philosophers of the twentieth
century, sometimes only as a poorly concealed fear of mathematics.
Quotations 94.13. (1) Tugendhat [1960], p. 191.

Was [[. . .]] in der Mathematik für den Aufbau deduktiver Systeme legitim
und notwendig ist, dürfte sich bei unkritischer Übertragung auf philosophi-
sche Grundbegriffe als methodische Naivität erweisen. Philosophische Be-
griffe stehen für komplexe Sachzusammenhänge, deren Explikation verbaut
wird, wenn sie durch eine exakte Definition auf einen bestimmten Teila-
spekt festgelegt werden. Die Isolierung eines Teilphänomens ist freilich legi-
tim, wenn sie ausdrücklich geschieht und das Verhältnis des thematisierten
Aspekts zum Ganzen angegeben wird.
Comment. Das Dürftige eines solchen “dürfte”-Räsonnements zeigt sich in der Spra-
che; deshalb lasse ich dieses schöne Exemplar deutscher Katheder Philosophie im
Original stehen. Notice: an exact definition limits us to a partial aspect. What that
means is most likely that an exact definition deprives philosophers of the kind of
Tugendhat of their playground.
18 Miller [1969], p. 53; translating Hegel [1812], p. 50: “das Untergeordnete der Wissenschaft-
lichkeit, die in der Mathematik Statt finden kann”. Cf. the phrase “Mangelhaftigkeit dieses Erken-
nens”, Hegel [1807], p. 41; quotation 65.10 (6): “defective cognition”.
(2) Henrich [1975], p. 225.

Es kann wohl kaum bestritten werden, daß Hegels Theorie in keinem Falle
eine deduktive Form im strikten Sinne hätte annehmen können. Immer
müßte sie auch auf eine Folgeordnung von Grundbegriffen Bezug nehmen,
die sich faktisch einstellen und die sich auf die Weise semiotischer Prozesse
fortbestimmen.
Comment. Despite all its vagueness and conceptual clumsiness, this passage seems
to get close to an important issue: the dynamical structure aimed at in Hegel’s
“theory” is not easily compatible with the character of a deductive theory in the strict
sense. But then, not even the theory of classical arithmetic can take the form of a
deductive theory in the strict sense. Every strict formalization of arithmetic contains
true but unprovable sentences, a fact which Gödel has called ‘the incompletability’ of
mathematics.19 From this there emerged in the foundational studies of mathematics
the idea of a progression of deductive theories,20 in the sense of adding certain true
but unprovable wffs, most notably consistency statements. If this idea is to be
transferred to Hegel’s “theory”, then the question is what drives a progression of
theories from a dialectical position? And this is where the weakness of Henrich’s
argumentation shows up: a vague reference to ‘semiotic processes’ just won’t do,
at least not if a theoretical standard is aimed at that somewhat matches that of
mathematical logic.21
Quotations 94.14. (1) Foucault [1969], pp. 188 f.

There is perhaps only one science for which one can neither distinguish
these different thresholds [[positivity, epistemologization, scientificity, and
formalization]], nor describe a similar set of shifts: mathematics, the only
discursive practice to have crossed at one and the same time the thresholds
of positivity, epistemologization, scientificity, and formalisation. The very
possibility of its existence implied that which, in all other sciences, remains
dispersed throughout history, should be given at the outset: its original
positivity was to constitute an already formalized discursive practice (even
if other formalizations were to be used later). Hence the fact that their
establishment is both enigmatic (so little accessible to analysis, so confined
within the form of the absolute beginning) and so valid (since it is valid both
as an origin and as a foundation); hence the fact that in the first gesture of
the first mathematician one saw the constitution of an ideality that has been
deployed throughout history, and has been questioned only to be repeated
and purified; hence the fact that the beginning of mathematics is questioned
not so much as a historical event as for its validity as a principal of history:
and hence the fact that, for all the other sciences the description of its
historical genesis, its gropings and failures, its late emergence is related
to the meta-historical model of a geometry emerging suddenly, once and
for all, from the trivial practices of land-measuring. But if one takes the
19 Cf. quotation 95.23 (2) in this chapter.
20 Cf. Turing [1939] and Feferman [1962], for instance.
21 This question will be of particular relevance in §120 in the groundworks.
establishment of mathematical discourse as a prototype for the birth and

development of all the other sciences, one runs the risk of homogenizing
all the unique forms of historicity, of reducing to the authority of a single
rupture all the different thresholds that a discursive practice may cross,
and reproduce endlessly, at every moment in time, the problem of origin:
the rights of the historico-transcendental analysis would thus be reinstated.
Mathematics has certainly served as a model for most scientific discourses
in their efforts to attain formal rigour and demonstrativity; but for the
historian who questions the actual development of the sciences, it is a bad
example, an example at least from which one cannot generalize.
Comment. Is this the standard practice of the historian, to take an example and
generalize? So a bad example is one which doesn’t lend itself to generalization. I
call this bad practice.
(2) Derrida [1971b], p. 227.
Outside the mathematical text—which it is difficult to conceive as providing
metaphors in the strict sense, since it is attached to no determined ontic
region and has no empirical sensory content—all the regional discourses,
to the extent that they are not purely formal, procure for philosophical
discourse metaphorical contents of the sensory type.
Quotations 94.15. (1) Stegmüller [1979], pp. 10 f.

Worte wie »Fundierung« oder »Begründung«, üben auf junge Philosophie-
rende eine eigentümliche Faszination aus. Hat man einmal etwas ›begrün-
det‹, so scheint man es auf einen Felsengrund gestellt zu haben. Diese Me-
tapher von der ›Begründung‹ ist nichts wert, und hinter ihr steckt gar
nichts. Alles, was wir tun, erfinden und unternehmen, uns ausdenken oder
konstruieren, kann sich als fehlerhaft erweisen. In diesem Sinn bejahe ich
vollkommen die These des Popperschen Fallibilismus. »Aber«, so könnte
einer einwenden, »du hast doch selbst im Titel deines ersten wissenschafts-
theoretischen Bandes das Wort ›Begründung‹ verwendet!« Das ist richtig.
Ich habe es aber nicht in diesem metaphorischen, sondern in einem ganz ge-
wöhnlichen, normalen Sinn gebraucht. Das, was in diesem Sinn ›begründet‹
werden kann, sind Erwartungen, Überzeugungen, Hoffnungen.
Comment. The title referred to by Stegmüller in which the word “Begründung”
(foundation, grounds) occurs is “Wissenschaftliche Erklärung und Begründung”. In
view of Stegmüller’s elaboration above, this would have to be read: “Wissenschaft-
liche Erklärung und Begründung von Erwartungen, Überzeugungen, Hoffnungen”
(Scientific explanation and foundation/grounds of expectations, convictions, hopes).
(2) Stegmüller [1979], p. 11.
»Eine Wissenschaft begründen« ist [[. . .]] zunächst nichts anderes als eine
unsinnige Wortzusammenstellung. Wobei ich keineswegs leugne, daß man
dieser Wendung im nachhinein eine harmlosere Deutung geben kann (die
man aber besser schon in der Formulierung zum Ausdruck hätte bringen
können, um nicht etwas vorzutäuschen, was nicht möglich ist). So etwa kann
§ 95. THE PARADIGM OF LOGIC 1315
es eine Abkürzung sein für: »diese Wissenschaft auf der Grundlage weni-
ger Axiome und weniger Grundbegriffe systematisch aufzubauen«; oder für
»diese Wissenschaft in einer für den Lernenden möglichst zweckmäßigen
Weise darstellen«; oder für »einen methodischen Aufbau dieser Wissen-
schaft liefern, wobei man für die benützten Methoden möglichst plausible
Gründe anzuführen versucht«.
(3) Stegmüller [1979], p. 13.
Wer glaubt, er könne auf irgendeinem Wege eine Garantie für etwas erzielen,
der ist nicht mehr von dem Wunsch nach wissenschaftlicher Erkenntnis
beseelt, sondern eher beherrscht von dem Bestreben, sich in eine unfehlbare
Gottheit zu verwandeln.
§ 95. The paradigm of logic
95a. The resilience of the Aristotelian paradigm. The view of Aristote-

les’ syllogisms as incorporating the idea of logic seems to be still dominant in large
parts of philosophy. It certainly was the view when the idea of dialectic that I am
dealing with was formed.
[[T]]he most merited and most important aspect of the disfavour into which
syllogistic doctrine has fallen is that this doctrine is such a long-drawn out,
notionless occupation with a subject matter whose sole content is the No-
tion itself. The numerous syllogistic rules remind one of the procedure of
arithmeticians who similarly give a host of rules about arithmetical oper-
ations, all of which rules presuppose that one has not the Notion of the
operation. But numbers are a notionless material and the operations of
arithmetic are an external combining or separating of them, a mechanical
procedure—indeed, calculating machines have been invented which perform
these operations; whereas it is the harshest and most glaring of contradic-
tions when the form determinations of the syllogism [[Schlu"s]], which are
Notions, are treated as a notionless material.
Comment. Miller translates “Schluß” as “syllogism” which doesn’t allow to keep the
German “Schluß” and “Syllogismus” apart; I would prefer “inference”. “Syllogistik”
is translated as “syllogistic doctrine”.
(2) U. Petersen [2002], p. 1315; translating Hegel [1817], p. 146; partly revising Miller
[1969], p. 685; continuation of the foregoing quotation.
The extreme of this notionless handling of the Notion determinations
of the Inference must be that Leibniz (Opp. Tom. II. P. I.) subjected the
Inference to the combinatory calculus [[. . . .]] Leibniz makes a big deal of the
usefulness of combinatory analysis, not only in order to find the forms of
Inference, but also to find the connections of other Notions. The operation
by which this is found is the same as that by which it is calculated how
many connections of letters are allowed by an alphabet, how many kinds of
throws are possible in a game of dice, or plays with an ombre card, and so
on. Thus one finds here the determinations of the Inference placed in one
class with the points of a dice and the ombre card, the reasonable taken as
something dead and notionless, and what is left aside is the peculiarity of the
notion and its determinations, to relate themselves as spiritual essences and
through this relating to sublate their immediate determination. [[. . .]] This
was linked to a pet thought of Leibniz, grasped in his youth and in spite of
its immaturity and shallowness not relinquished by him even in later life, of
a universal characteristic of Notions, — a script-language [Schriftsprache],
in which every notion is to be represented how it is a relation of others,
or related to other notions — as if in the connection of reason, which is
essentially dialectical, a content would keep the same determinations which
it has when it is fixed for itself.
Comment. Frege developed his “Begriffsschrift” (notion-script) roughly sixty years
after these lines had been written; and Gödel obtained his results roughly another
fifty years later. Together, it would seem, they could have done much to bring
Hegel’s narrow perspective regarding the shallowness of Leibniz’ idea of a script-
language, and the failure of Frege’s logic together. But this would have required
some engagement with mathematical-logical methods on the part of the Hegelian
philosophers which to assume appears almost counterfactual “from a late twentieth-
century perspective”. Hegel’s view of the shallowness of Leibniz’ idea is related to
Cornford’s view of the emptiness of tautologies; two forms of intellectual arrogance
of philosophers. Together they seem to have divided philosophy amongst them, both
squarely in the tradition of Berkeley’s horror of precision.22
Quotations 95.2. (1) Heidegger [1978], pp. 18 f.

What does logic have to do with the freedom of existence? How does the
basic question of being belong here? Logic does not treat being directly but
deals with thinking. “Thinking” is of course an activity and comportment
of humans, but still only one activity among others. The investigation into
thinking as a form of human activity would then fall under the science
of man, under anthropology. The latter is, of course, not philosophically
central, but only reports how things look when man thinks, how primitive
peoples “think” differently than we do and follow different laws. These an-
thropological and psychological questions about forms and types of thinking
are certainly not philosophical. But it remains open whether these are the
only questions and even the only radical questions.
If thinking is a mode of Dasein’s comportment and if it is not abandoned
to arbitrariness but stands under laws, then the question must be asked:
What are the fundamental laws belonging to thinking as such? What is, in
general, the character of this lawfulness and regulation? We can obtain an
answer only by way of a concrete interpretation of the basic laws of thinking
which belong to its essence in general.
22 Cf. Berkeley in quotation 58.12 (2) in these materials and Heidegger in quotation 95.2 (3)
below.
What is meant by “basic principles,” and what is their essence? What

principles are there? The tradition gives us the principle of identity, the
principle of non-contradiction, the principle of excluded middle, the princi-
ple of sufficient reason, principium identitatis, principium contradictionis,
principium exclusi tertii, principium rationis sufficientis. Are these all? In
what order do they stand? What intrinsic connection do they have? Where
do they find their foundation and their necessity? Are we dealing here with
laws of nature, with psychological or moral laws? Or of what sort are they
that Dasein is subject to them?
But the account of the laws governing thinking pushes us back into the
question of the conditions of their possibility. How must that being which
is subject to such laws, Dasein itself, be constituted so as to be able to be
thus governed by laws? How “is” Dasein according to its essence so that
such an obligation as that of being governed by logical laws can arise in
and for Dasein?
Comment. There is an interesting strategy here, which can be found in Heidegger
throughout: first he throws up a whole lot of questions; then he says something like:
“tradition gives us . . . ”; and then he continues as if he had given a sufficient answer.
Heidegger has not yet given an account of the laws governing thinking, but he is
already throwing up more questions.
(2) Heidegger [1978], pp. 104 f.
Every science, including metaphysics, and every form of prescientific
thinking uses, as thinking, the formal rules of thought. Using the rules
of thought in the thinking process is inevitable. [[. . .]] But does it follow
from the inevitability of using rules in the thought process that science
has its basis in logic? Not at all. For the inescapability of rule usage does
not in itself immediately imply the inescapability of logic. Using rules does
not necessarily require a science of the rules of thought and certainly not
a reasoned knowledge of these rules in the sense of traditional logic. For
otherwise the thoughtful justification of logic itself would be intrinsically
impossible, or superfluous. A fully developed logic would have to exist then
already insofar as there was thought.
Even if one wished to concede that the inevitability of using rules in
scientific thought implies the inevitability of logic for science and for meta-
physics, this would not in turn mean logic is the foundation of science as
such. The procedure of thinking and the use of rules are, among others,
requirements which flow from the essence of science, but the essence of
science does not have its intrinsic possibility in the procedure of thinking
and the use of rules. Moreover, the converse is true. The unavoidability of
rule usage can be established only from the intrinsic essential possibility of
science, can only be justified metaphysically. Not only rule usage, but the
rules themselves, need metaphysics for their justification. [[. . .]]
[[. . .]]
Therefore: 1) Logic is not even the operational condition for thinking
but only a science of the rules as such preserved in a tradition. 2) But if
it is a science of rules, then obviously logic, in its traditional form, can

neither clarify nor justify these very rules in their essence. 3) Insofar as the
intrinsic possibility of something that provides a foundation, the explication
of the intrinsic possibility of thinking, as such, is the presupposition of
“logic” as a science of the rules of thinking. 4) The entire problem of the
possible precedence of metaphysics over logic or logic over metaphysics
cannot be posed, discussed, and solved by logic, unless logic is conceived as
the metaphysics of truth.
Comment. I would have thought that Heidegger’s thinking was a striking counter
example to the view that thinking proceeds according to the formal rules of thought,
whatever they may be. The paradigm of “using rules” is extremely resilient when it
comes to understanding logic. Contrast the paradigm of the priority of a concept
which, through its being applied, is filled, or fills itself, with content. This is the
reason, I suggest, why Heidegger never understood Hegel; or, differently put, why
the idea of Hegel’s encyclopedia and the role of logic in it, is wasted with regard to
Heidegger and those who follow his line of thinking in ‘chicken and egg’ terms.
(3) Heidegger [1978], p. 106.
Leibniz time and again comes close to founding metaphysics on logic, prob-
ably most clearly and emphatically in the treatise [[. . .]] Primae veritates (C.
581–23)[L.267]. Contemporary logic shows a new distortion of the problem.
Not only is metaphysics reduced to logic, but logic is itself reduced to math-
ematics. Contemporary logic is symbolic, mathematical logic, and thus a
logic which follows the mathematical method.
Comment. Two issues seem to be confounded here, that of reducing one science
to another one and that of employing in one science the methods of another one.
Physics, for instance, successfully employs mathematical methods, without having
been reduced, so far at least, to mathematics. As regards logic and mathematics,
there was an (unsuccessful) attempt to “reduce” mathematics to logic, so-called
logicism; and apart from that there is a successful employment of mathematical
methods in the metatheory of logic.
95b. The use of logical tools. Whitehead and Russell used the vicious circle
principle to deal with the philosophical problem of the self-refuting sceptic.
Quotations 95.3. (1) Whitehead and Russell [1910], p.40.

[[T]]he imaginary sceptic, who asserts that he knows nothing, and is refuted
by being asked if he knows that he knows nothing, has asserted nonsense,
and has been fallaciously refuted by an argument which involves a vicious-
circle fallacy. In order that the sceptic’s assertion may become significant,
it is necessary to place some limitation upon the things of which he is
asserting his ignorance, because the things of which it is possible to be
ignorant form an illegitimate totality. But as soon as a suitable limitation
has been placed by him upon the collection of propositions of which he is
asserting his ignorance, the proposition that he is ignorant of every member
of this collection must not itself be one of the collection.
(2) French [1967], p. 628.

Like Kant, Ryle rejects the belief that minds are things, but different sorts of
things from bodies. To think that the word ‘mind’ denotes a distinct entity
(of whatever kind) is to make what Ryle calls “the Cartesian category-
mistake.” To make a category-mistake is to allocate a concept to a logical
type to which it does not belong. The word ‘mind’ is not of the same logical
type as those nouns that can be used ostensively.
Comment. Now here is someone who has heard of ‘logical types’ and has a use for
them. We don’t know yet what constitutes a category, but we have already ‘category
mistakes’.
Possible worlds semantic and the question of “how can language be sexist?” or:
how can language favor a particular ontology.
Quotations 95.4. (1) J. Hintikka [1981], p.212.

[[T]]he tableau method, or some variant thereof, is already a partial formal-
ization of the dialectical method, or in any case of one of its models. It
is not yet a satisfactory reconstruction of the dialectical method, however,
for the method presumably applies much more widely than in the realm of
deductive first-order inferences, instructive though this special case is his-
torically. We have to widen this reconstruction so as to take into account
also nonanalytic steps to new information.
(2) M. B. Hintikka and J. Hintikka [1983], p. 146.
Western philosophical thought has been overemphasizing such ontological
models as postulate a given fixed supply of discrete individuals, individu-
ated by their intrinsic or essential (non-relational) properties. These models
are unfavorably disposed towards cross-identification by means of functional
or other relational considerations. Is it to go too far to suspect a bias here?
It seems to us that a bias is unmistakable in recent philosophical semantics
and ontology. There we find almost everyone postulating a given domain
of discrete individuals whose identity from one model (world) to another is
unproblematic. An especially blatant example of this trend is Kripke’s no-
tion of a rigid designator, which becomes virtually useless as soon as cross-
identification is recognized as a problem. (No wonder Kripke has been led
to argue that the re-identification of temporally persistent physical objects
must be taken for granted.)
Recall that the paradigm of logicism was taken up by the empiricists at the
beginning of the twentieth century.
Quotations 95.5. (1) Wang [1974], p. ix.

Directly and indirectly logic plays an important part in much of contem-
porary Anglo-American academic philosophy. Sociologically, this is com-
fortable for those of us who are interested in both logic and philosophy.
But I have for years had two interrelated misgivings. The way logic is com-
monly used in philosophy seems to me to do less than justice to the full
richness of logic as a study of the foundations of mathematics; and the ex-

cessive emphasis on the importance of logic, seems to me to have led to a
far from balanced view of philosophy, especially as it is understood in the
traditional sense. Moreover, the much publicized juxtaposition of logic with
positivism, (or empiricism or ‘analytic’ philosophy) has burdened logic with
a guilt by association, resulting in a surprising ignorance of logic on the part
of philosophers of other persuasions. This has the unfortunate consequence
that not only is logic misused by those who are guided by misplaced preci-
sion, but the other philosophers also are not making what is a proper and
fruitful use of logic, namely, to use it not so much explicitly but more as a
way of acquiring the habit of precise thinking.
(2) Wang [1985], p. 459.
Both Carnap and Quine pay a good deal of attention to logic and math-
ematics; yet, in my opinion, both of them fail to take logic and mathematics
seriously as clear representatives of our conceptual knowledge. This is not
surprising, since the commandment of empiricism is dominant for both of
them and, when it conflicts with the other commandment of the centrality
of logic, they are forced to adapt and modify the latter to suit the former.
As a result, we see in both of them an ambivalent attitude towards logic
(and mathematics): logic is one of the two pillars of their philosophies; yet
it is not permitted to possess any measure of autonomy for fear of im-
pinging on the monopoly of fact and content by (one type of) experience,
which ultimately rests exclusively on sense experience of a kind that, though
indeterminate, must in no way be contaminated by contact with any con-
ceptual intuition. Indeed the empiricists’s stipulative use of the terms ‘fact’
and ‘content’ is a conspicuous petitio principii: by definition there can be
no mathematical (or conceptual) facts and contents. This preemptive us-
age has the defect of reinforcing a preconception and concealing the glaring
conflict of mathematics with empiricism.

Instead of adhering to an uncritical imitation of physical methods and tech-
niques, philosophers are to find different ways of thinking and communica-
tion which are appropriate to foundational studies.
Comment. This is all very well. The point is, which are they?
(2) Wang [1987], p. 2.
A question raised by G[[ödel]] on a public occasion in 1972, which l had
also asked myself before, is the feasibility of a computer ‘knowing’ its own
program. One difficulty is to give the idea a less elusive meaning, perhaps
by relating it to ideas that lack its seeming simplicity. Recently it occurred
to me that one could conceivably approach the problem by attempting to
find an analogue of the hidden ‘rational core’ in Hegel’s Phenomenology
of the Spirit, namely, in terms of a journey toward self-knowledge within
a world consisting of computers. Such a project ought to do something to
reconcile the antagonism between the humanists and the technologists.
Comment. Sounds nice to me, but I seriously doubt that this would do anything
to “reconcile the antagonism between humanists and technologists”, simply because
the humanists will not be able to overcome their horror mathematicae.
Quotations 95.7. (1) Rehder [1983], p. 227.
Imagine the following ingenious method for solving philosophical prob-
lems. Let P denote the problem under investigation. Assume that P admits
a logical representation in a sufficiently rich language L, state a set AP
of axioms plus some rules of transformation and inference, including the
modus ponens. Call the triplet S = (P, L, AP ) our formalized philosophi-
cal problem. Now! Let f(S) be a function of S whose value T = f(S) is the
(or one) solution of S. Properties of the “solving function” f are studied in
the rest of our imagined paper. Also some generalizations are suggested,
and the philosophical impact of f and T is discussed and related to recent
publications on the same subject matter.
A caricature? Of course it is. But we’ll see that this caricature has some
striking, if exaggerated, similarities with a very popular modern approach
to philosophy. This approach is adorned with a respectable name: “Philo-
sophical Logic”, although quite a few of its aspects should rather remind us
of a variation of a famous Wittgensteinian theme, the disastrous invasion
of philosophy by logic.
(2) Rehder [1983], p. 237.
The fads and fallacies in fashionable philosophical logic [[. . .]] arise from
one common source: their use of attractive tools and machinery of logic is
premature and deceptively straightforward. The employment of logic su-
perimposes structures upon the philosophical problems which are too often
alien to them. At the very least, more care is needed. And more knowledge.
If you want to enjoy the elegance of logic, you had better prepare yourself
to take the pains of the ensuing mathematics as well. Otherwise the air
might get very thin, and the elaborate construction very void.
Quotation 95.8. Wang [1974], p. 56.
One must not forget that formal systems are tools and only tools in
the study of philosophy. Like other tools, they are useful only for certain
purposes. They are not the philosopher’s stone which can solve all prob-
lems for us. When applied indiscriminately to all questions, they cause at
best waste and at worst disaster. The use of formal systems in studying
the philosophy of mathematics has proved to be successful, so much so that
nowadays nobody can hope to become a serious mathematical philosopher
unless he possesses considerable skill in the manipulation of formal sys-
tems. On the other hand, the application of these tools to the treatment
of problems of inductive logic, meaning, time, causality, has met with little
success. It is indeed hard to estimate whether these commendable efforts
to expand the sphere of influence of mathematical logic have done mere
good or more harm to philosophy. One is tempted to wonder whether these
applied logicians have not committed the ‘fallacy of too many digits’: viz.
the fallacy, emphasized by Ramsey, of working out to (say) seven places of

decimals a result only valid to two.
Quotation 95.9. Flew [1979], p. 296.

The truth or falsity of a quantified statement (that is, one containing
a quantifier) cannot be assessed unless one knows what totality of objects
is under discussion, or where the values of the variables may come from.
For example, ‘All five-year-olds go to school’ may be true if one is just
talking about English children, but false if one is talking more generally
about European children.
Comment. Only years of logical training in analytic philosophy and a total ab-
sorption of the doctrine of objectual quantification can get you to the point of
understanding the significance of this remark, I suppose.
95c. Employing logical and mathematical results and/or concepts.
Quotation 95.10. Ariew and Garber [1989], p. 217 (33.); translating Leibniz, Mon-
adology.
There are [[. . .]] two kinds of truths, those of reasoning and those of fact.
The truths of reasoning are necessary and their opposite is impossible; the
truths of fact are contingent, and their opposite is possible. When a truth
is necessary, its reason can be found by analysis, resolving it into simpler
ideas and simpler truths until we reach the primitives.
Quotation 95.11. Hilbert [1931], p. 485 (quoted after Carnap [1931b], p. 241).
In einem neueren philosophischen Vortrage finde ich den Satz: ‚Das Nichts
ist die schlechthinnige Verneinung der Allheit des Seienden‘. Dieser Satz ist
deshalb lehrreich, weil er trotz seiner Kürze alle hauptsächlichen Verstöße
gegen die in meiner Beweistheorie aufgestellten Grundgesetze illustriert.
Comment. Heidegger springs to mind, of course, but I wasn’t able to find exactly the
sentence in question, only: „Das Nichts ist die vollständige Verneinung der Allheit
des Seienden“,23 which I find close enough.
Quotation 95.12. Žižek [1991], pp. 42 f.

This triad, this ternary structure in which the Universal, confronted with
its particular content, redoubles into positive and negative, encompassing
and exclusive, “pacifying” and “destructive” – in other words: in which the
initial position, confronted (“mediated”) with the multitude of its particu-
lar negations, is retroactively trans-coded into pure, self-relating negativity
– furnishes the elementary matrix of the dialectical process. Such a self-
referring logical space where the universal genus encounters itself in the
form of its opposite within its own species [[. . .]] – that is to say, where a set
comes across itself within its own elements – is based on the possibility of
reducing the structure of the set to a limit-case:
23 Heidegger [1965], p. 29.
that of a set with one sole element: the element has to differ only from the
empty set, from the set which is nothing but the lack of the element itself
(or from its place as such, or from the mark of its place – which amounts
to saying that it is split). The element has to come out for the set to exist,
it has to exclude itself, to except itself, to occur as deficient or in surplus.
Within this logical space, the specific difference no longer functions as the
difference between the elements against the background of the neutral-
universal set: it coincides with the difference between the universal set itself
and its particular element – the set is positioned at the same level as its
elements, it operates as one of its own elements, as the paradoxical element
which “is” the absence itself, the element-lack (that is, as one knows from
the fundamentals of set theory, each set comprises as one of its elements
the empty set). This paradox is founded in the differential character of the
signifier’s set: as soon as one is dealing with a differential set, one has to
comprise in the network of differences the difference between an element
and its own absence. in other words, one has to consider as a part of the
signifier its own absence – one has to posit the existence of a signifier which
positivizes, “represents”, “gives body to” the very lack of the signifier – that
is to say, coincides with the place of inscription of the signifier. This differ-
ence is in a way “self-reflective”: the paradoxical, “impossible” yet necessary
point at which the signifier differs not only from another (positive) signifier
but from itself as signifier.
Comment. The quote is identified as: Jacques-Alain Miller, “Matrice”, Ornicar? 4,
Paris 1975, p. 6. As regards the claim that “each set comprises as one of its own
elements the empty set”, I am torn between admiration and doubt. The admiration
is for the capacity to say virtually the silliest thing possible in set theory, somewhat
adding to the claim of analytic philosophers (or also Russell) that Hegel(ians) tend
to confuse the different uses of “is” in natural language, in the present case the “is”
of being an element and that of being subsumed as a subset. The doubt is whether
this might not be due to the poetry of the translator, as in the case of Derrida in
quotation 95.17 (2) below. Or, perhaps, “the Giant of Ljubljana” just doesn’t live
up to Wang’s image of the philosopher who is “capable of discoursing after dinner”
on results in set theory without “committing factual mistakes”.
Quotations 95.13. Wang [1985], p. 450.

One common feature of Carnap’s and Quine’s philosophical practice is
their preoccupation with local (and mostly formal) precision, accompanied
by a surprising willingness to tolerate or even celebrate global indefiniteness.
[[. . .]] Behind their practice is a shared wish to adhere to the physical and
other more concrete or tangible objects and experiences (such as linguistic
expressions and observation sentences).
What is fundamental in Carnap and Quine is, I think, not the two
dogmas of analyticity and reductionism which separate them, but rather
the denial of any autonomy to conceptual knowledge which unites them.

We well know that quantification theory, which is so much more com-
plex than the Boolean predicate calculus, has its serious motivation in
polyadic predicates. When we move to polyadic logic, the bound variable
quits the wings and gets into the act. The basic job of the bound variable
is cross-reference to various places in a sentence where objective reference
occurs; and whereas monadic logic calls for this service only in the prepara-
tions, polyadic logic calls for it also within the ongoing algorithm, in order
to keep track of permutations and identifications of arguments of polyadic
predicates. It is in such permutations and identifications that the bound
variable enters essentially into the algorithm, and here it is, by the way,
that the decision procedures cease to be generally available.
There is evidence of a connection. Polyadic logic remains decidable, like
monadic logic, as long as there is no crossing up of argument places. [[. . .]]
The variable, then, it seems, is the focus of indecision.
Comment. According to Kripke [1962], monadic quantificational logic becomes un-
decidable when modality is added; but I don’t know if there is a connection.

It is ridiculous when philosophers base their work on false or ill-digested
mathematical ‘facts’. An example [[. . .]] is the impression, still widespread
among philosophers though diminishing, that the paradoxes have reduced
set theorists to comparing a large number of differing systems embodying
different, completely unintuitive, approaches to the paradoxes. In fact, es-
sentially one system, Zermelo-Fraenkel set theory, is seriously studied by
working set theorists; and most set theorists today regard it as based upon
an intuitive idea. Yet many philosophical morals have been drawn from the
alleged plethora of competing and unintuitive approaches.
Comment. I agree with Kripke as regards philosophers, but who would not agree
that it is generally ridiculous to base one’s work on “false or ill-digested” facts, be
they mathematical or whatever. In that respect Kripke’s remark is rather vacuous;
but I also find his example rather silly. Kripke confounds what “most set theorists
today” regard as “intuitive” with what might be called a mathematical fact. The
point here seems to me that the problem raised by the set theoretical paradoxes is
not ‘solved’ to the satisfaction of some philosophers, but to the satisfaction of almost
all set theorists. Whether the solution is intuitive is not a mathematical fact. The
situation is symptomatic. Another example is Gödel’s incompleteness theorem. It is
simply hair-raising what some philosophers say about it, but I find it hard to take
when mathematicians/logicians try to sell their interpretations as the actual result.
Mathematicians, no matter how well they think they have understood a result, hold
no monopoly on its interpretation.
Quotations 95.16. (1) Suszko [1968], p. 210.

[[C]]an formal logic tell us something about the development of knowledge?
I think it can. I think, to use a suggestive terminology, that we can build
something like a diachronic formal logic as opposed to the synchronic formal

logic today. It must be added that diachronic logic can only be simply an
application of notions, theorems and methods used in synchronic logic to
the problem of change in human knowledge.
(2) Suszko [1968], p. 211.
When considering in formal logic the phenomenon of knowledge we take
into account the ordered pair (S, M ) called epistemological opposition or,
shortly, E-oppositions. The first member S is called here the subject or the
mind and the second member M is called the object or the world for the
mind in the given epistemological opposition.
95d. Incompleteness, undecidability, and the interconnectedness of

everything — after dinner.24 The incompleteness and undecidability results are
a special issue in that they seem to invite philosophical muddling.
Quotations 95.17. (1) Fogarasi [1954], p. 68; my translation.

The results of Skolem, Gödel, Church and other researchers have shown in
different respects that the axiomatic theory by way of closed systems is not
capable to eliminate, to solve the contradictions which arise in mathematics.
The deepest reason for this can of course only be comprehended by the
materialist-dialectical logic which reveals the relation of mathematics and
reality.
Comment. Hurray for dialectical logic!
(2) Derrida [1972], p. 189.
An undecidable proposition, as Gödel demonstrated in 1931, is a proposi-
tion which, given a system of axioms governing a multiplicity, is neither an
analytical nor deductive consequence of those axioms, nor in contradiction
with them, neither true nor false with respect to those axioms. Tertium
datur, without synthesis.
Comments. (α) There is a sentence in the text immediately preceding this quotation
which might be relevant: “Allusion, or “suggestion” as Mallarmé says elsewhere, is
indeed that operation we are here by analogy calling undecidable.” This makes at
least clear that Derrida doesn’t commit himself to a strict mathematical-logical
notion of “undecidable”. But who would have expected that anyway?
(β) Readers who feel uncomfortable with this rendering of Gödel’s result might feel
consoled when reading what Derrida has said in his own words:25
Une proposition indécidable, Gödel en a démontré la possibilité en 1931,
est une proposition qui, . . .
I.e., “An undecidable proposition, the possibility of which Gödel has demonstrated
in 1931, is a proposition which, . . . ”. I regard this as a fine example of the poetry of
24 Cf. Wang in the second motto to this chapter; although I don’t want to say the following
quotations avoid “making factual mistakes”.

25 La dissémination, Éditions du Seuil, Paris 1972, pp. 248 f.
the translator; but then, does it matter much for philosophers? Gasché quotes the
English translation with no apparent discomfort.
(γ) What justifies the last sentence of this quotation from Derrida’s dissemination
of Gödel’s result? It strikes me as a philosopher’s quick jump.
(3) Ryan [1982], p. 17.
To use Gödel’s terminology, the system is necessarily “undecidable,” because
it generates elements that can be proved to belong to the system and not to
belong to it at the same time. [[. . .]] A formal system of axioms is undecidable
if it is incomplete, and, according to Gödel, all such systems are undecidable.
Comment. Poor Gödel; now he has been turned into a paraconsistent logician.
Notice also the new theorem: “A formal system of axioms is undecidable if it incom-
plete”.26
(4) Gasché [1986], p. 240.
Gödel demonstrates that metalogical statements concerning the com-
pleteness and consistency of systems any more complex than logical systems
of the first order cannot be demonstrated within these systems.
(5) Bochenski [1980], p. 14.
According to [[Gödel’s first theorem]], the impossibility of all all-em-
bracing philosophical systems — like that of Hegel — has been shown once
and for all[[.]]
Comment. Cf. what Girard says in quotation 95.22 below regarding formulations
of the meaning of Gödel’s theorem.
If this wasn’t enough to explain why philosophers have acquired such a bad
reputation, there is more to come, but I don’t blame anyone for skipping the next
two sets of quotations.
Quotations 95.18. (1) Derrida [1972], p. 189.

“Undecidability” is not caused here by some enigmatic equivocality, some
inexhaustible ambivalence of a word in a “natural” language, and still less
by some “Gegensinn der Urworte” (Abel). In dealing here with hymen, it is
not a matter of repeating what Hegel undertook to do with German words
like Aufhebung, Urteil, Meinen, Beispiel, etc., marveling over that lucky
accident that installs a natural language within the element of speculative
dialectics[[.]] What counts here is not the lexical richness, the semantic in-
finiteness of a word or concept, its depth or breadth, the sedimentation
that has produced inside it two contradictory layers of signification (con-
tinuity and discontinuity, inside and outside, identity and difference, etc.).
26 The problem here seems to be that the undecidability of a theory and that of a sentence
are confounded. An undecidable theory may well be complete, witness classical first order logic (cf.
quotation 79.8), and a decidable theory may well be incomplete, i.e., have undecidable sentences,
witness a monadic non-classical first order logics like the intuitionistic one: its decidability is proved
like that of classical monadic first order logic and its incompleteness follows from the unprovability
of tertium non datur.
What counts here is the formal or syntactical praxis that composes and
decomposes it.
(2) Gasché [1986], p. 241.
Derrida emphasized that infrastructures were to be called undecidable only
by analogy. The notion of the undecidable, he remarks in his Introduction
to the Origin of Geometry, in its very negativity, “has such a sense by
some irreducible reference to the ideal of decidability.” Its revolutionary
and disconcerting sense “remains essentially and intrinsically haunted in
its sense of origin by the telos of decidability—whose disruption it marks”
(O, p. 53). Yet what is being thought under the title of the infrastructures
transcends the project of definiteness itself. Therefore, undecidable must be
understood to refer not only to essential incompleteness and inconsistency,
bearing in mind their distinction from ambiguity, but also to indicate a level
vaster than that which is encompassed by the opposition between what is
decidable and undecidable.
(3) Derrida [1962], p. 53 f.
In its very negativity, the notion of the un-decidable—apart from the fact
that it only has such a sense by some irreducible reference to the ideal of
decidability*—also retains a mathematical value derived from some unique
source of value vaster than the project of definiteness itself. This whole
debate is only understandable within something like the geometrical or
mathematical science, whose unity is still to come on the basis of what is
announced in its origin. Whatever may be the responses contributed by
the epistemologist or by the activity of the scientific investigator to these
important intra-mathematical questions of definiteness and completeness,
they can only be integrated into this unity of the mathematical tradition
which is questioned in the Origin. And they will never concern, in the
“objective” thematic sphere of science where they must exclusively remain,
anything but the determined nature of the axiomatic systems and of the
deductive interconnections that they do or do not authorize. But the ob-
jective thematic field of mathematics must already be constituted in its
mathematical sense, in order for the values of consequence and inconsis-
tency to be rendered problematic, and in order to be able to say, against
the classic affirmations of Husserl, “tertium datur.”
Comment. There is a footnote marked here by *, part of which I find worth quoting,
if only to give an even better idea of the unfathomable depth of Derrida’s insight:
Thus, undecidability has a revolutionary and disconcerting sense, it is itself
only if it remains essentially and intrinsically haunted in its sense of origin
by the telos of decidability—whose disruption it marks.
Quotations 95.19. (1) Derrida [1982], p. 9.
The sign is usually said to be put in the place of the thing itself, the present
thing, “thing” here standing equally for meaning or referent. The sign rep-
resents the present in its absence. It takes the place of the present. When
we cannot grasp or show the thing, state the present, the being-present,
when the present cannot be presented, we signify, we go through the detour

of the sign. We take or give signs. We signal. The sign, in this sense, is
deferred presence. Whether we are concerned with the verbal or the writ-
ten sign, with the monetary sign, or with electoral delegation and political
representation, the circulation of signs defers the moment in which we can
encounter the thing itself, make it ours, consume or expend it, touch it,
see it, intuit its presence. [[. . .]] According to this classical semiology, the
substitution of the sign for the thing itself is both secondary and provi-
sional: secondary due to an original and lost presence from which the sign
thus derives; provisional as concerns this final and missing presence toward
which the sign in this sense is a movement of mediation.
Comment. What seems to be completely absent in this view is the point that Leibniz
insists on: the signs amongst themselves can have the same structure as the objects
for which they stand.27 This is also the point which I shall focus on in the ground-
works.
(2) Ryan [1982], p. 20.
[[B]]ecause all language is metaphoric (a sign substituted for a thing), no
metametaphoric description of language is possible that escapes infinite
regress. [[Derrida]] is criticizing the transcendental impulse in general, the
desire to construct truths through a language supposedly so formal that
it renders the truth of the thing itself in its presence without any rep-
resentational mediation. Any absolute knowledge, conveyed in necessarily
metaphoric language, would, therefore, also have to provide an account of
all metaphors, but such a metametaphorics is impossible without recourse
to yet one more metaphor, the language of the formal description itself, a
metaphor that would be undecidable because it both participates in and
remains outside of the system it helps to enunciate. To account for the
language of the account would require another account—and a potentially
interminable repetition of the problem.
Comment. This view, whether it represents Derrida’s position or not, is stuck in
a somewhat Kantian version of the problem: it only sees an infinite regress. The
point of my approach is that we are not just confronted with the possibility of an
infinite regress but also a loop. This loop is a source of systemic ambiguity; and
this systemic ambiguity is the source of the categories. What Gödel has shown,
i.e., as a result (not as something like a definition, as (the translation of) Derrida
p. 189 seems to suggest), is that arithmetic already does what Ryan refers to in
the sentence starting “Any absolute knowledge . . . ”. There is nothing intrinsically
“absolute” about (formalized) arithmetic. This is one point. What is more important
for me is the conclusion: “a potentially interminable repetition . . . ”. If this is indeed
representing Derrida correctly, then this just shows that Derrida is still caught in
a framework of classical (binary, two-valued) logic which got him bogged down in
the (classical) interpretation of Gödel’s result in the sense of a limitation and an
infinite progression of “accounts”. Hegel had the label “bad infinity” for that. The
point of the present study is not to remove a bad infinity, but to achieve a sort of
totalization that can provide for a new category.
(3) Derrida [1981], p. 35.
[[I]]t always will be impossible, and for essential reasons, to reduce absolutely
the natural languages and nonmathematical notation. We must also be wary
of the “naive” side of formalism and mathematism, one of whose secondary
functions in metaphysics, let us not forget, has been to complete and confirm
the logocentric theology which they otherwise could contest. Thus in Leibniz
the project of a universal, mathematical, and nonphonetic characteristic is
inseparable from a metaphysics of the simple, and hence from the existence
of divine understanding, the divine logos.
The effective progress of mathematical notation thus goes along with
the deconstruction of metaphysics, with the profound renewal of mathe-
matics itself, and the concept of science for which mathematics has always
been the model.

Meine Aufgabe ist es nicht, über den Gödelschen Beweis, zum Beispiel, zu
reden, sondern an ihm vorbeizureden.
Comment. In the light of this remark I am inclined to say that Wittgenstein has
perfectly achieved his goal. But then, it seems to me, he has generally an allem
vorbeigeredet. I am not sure how far the irony here was intended by Wittgenstein.
Bernays [1957], p. 120, remarks after quoting this particular passage that Wittgen-
stein doesn’t lack the “Spaßhafte im Ausdruck”, and often loves “sich als Schelm zu
gebärden”. The phrase “an etwas vorbeireden”: “to talk past something”, mainly in
the sense of “to fail to address an issue”. Anscombe’s translation, “My task is, not to
talk about (e.g.) Gödel’s proof but to pass it by” (Meine Aufgabe ist es, nicht über
den Gödelschen Beweis, zum Beispiel, zu reden, sondern ihn zu umgehen), strikes
me as an attempt to cover up the “Spaßhafte im Ausdruck”, or, as I would put it,
the folly of Wittgenstein’s remark. Boos [1987], p. 21, translates correctly, to my
mind: “My task is not to talk about the Gödelian proof, for example; but to talk
past it”.28
(2) Wang [1987], p. 49, quoting from Gödel’s draft of a reply to Menger, dated 12
April 1972.
As far as my theorem about undecidable propositions is concerned, it is
indeed clear from the passages you cite [[Wittgenstein [1956], pp. 50–55 and
176 ff.]] that Wittgenstein did not understand it (or pretended not to under-
stand it). He interprets it as a kind of logical paradox, while in fact it is just
the opposite, namely a mathematical theorem within an absolutely uncon-
troversial part of mathematics (finitary number theory or combinatorics).
Incidentally, the whole passage [[mentioned]] seems nonsense to me.
28 Boos translated from a different edition (Suhrkamp, 1974), which accounts for the semicolon
after “example”.
The first of the following couple of quotations is meant as an example of what I

consider a perfectly sound way of talking philosophically about mathematical results
and Hegel’s dialectic without exposing oneself to Wang’s or Kripke’s criticism.
Quotations 95.21. (1) Findlay [1963], p. 6 f.

Two of the greatest logico-mathematical discoveries of fairly recent
times may in fact be cited as excellent and beautiful examples of Hegelian
dialectic: I refer to Cantor’s generation of transfinite numbers, and to
Goedel’s theorem concerning undecidable sentences. In the case of Cantor
we first work out the logic of the indefinitely extending series of inductive,
natural numbers, none of which transcends finitude or is the last in the
series. We now pass to contemplate this series from without, as it were,
and raise the new question as to how many of these finite, natural numbers
we have. To answer this we must form the concept of the first transfinite
number, the number which is the number of all these finite numbers, but
is nowhere found in them or among then, which exists, to use Hegelian lan-
guage, an sich in the inductive finite numbers, but becomes für sich only
for higher-order comment. And Cantor’s generation of the other transfi-
nite numbers, into whose validity I shall not here enter, are all of exactly
the same dialectical type. Goedel’s theorem is also through and through
dialectical, though not normally recognized as being so. It establishes in a
mathematicized mirror of a certain syntax-language that a sentence declar-
ing itself, through a devious mathematicized circuit, to be unproveable in a
certain language system, is itself unproveable in that system, thereby set-
ting strange bounds to the power of logical analysis and transformation.
But the unproveable sentence at the same time soars out of this logico-
mathematical tangle since the proof of its unproveability in one language
is itself a proof of the same sentence in another language of higher level, a
situation than which it is not possible to imagine anything more Hegelian.
(2) Kosok [1966], pp. 263 f.
From the perspective of a principle of Non-Identity, dialectic logic can
be taken as a way of generalizing Goedel’s theorem, and instead of regarding
it merely as a limitation to the expression of consistent systems in ordinary
logical structures, it now becomes the starting point for a dialectic logic,
which regards these limitations as the essence of its structure. According to
Goedel’s theorem, our present mathematical-logic system is so constituted
that consistency and completeness appear as contraries: given one, the other
cannot necessarily be shown. But as we have seen, the principle of Non-
Identity equally presents us with that contrariness. Indeed, it is significant
that the formula giving rise to Goedel’s result is similar in form to the
principle of Non-Identity: i.e. it is possible to construct a true formula G
such that its demonstration is implied by and implies the demonstration of
its negation. This is symbolized by Dem(G)↔Dem(-G), which results in the
conclusion that while true the formula G cannot consistently be expressed
within the given formal system without expansion, which would then only
produce higher order incompleteness, namely another formula G′ having a

fate similar to G.
The above situation appears as a limitation of expression only if we
view formal structures merely from the perspective of the law of Identity,
wherein we regard the essence of a given term as already fixed and formed,
independent of the activity of reflection. Reflection, however, opens up to
any given X to an indeterminate -X, placing the given thus in a new context,
within which both the given X and the -X become transformed due to the
mutually limiting nature of the coupling relation expressing the co-existing
of X and -X.
Comment. I find this hard to take; it certainly doesn’t sound very competent to
me.
I close this section with a quotation from Girard regarding “the fortune of
Gödel’s theorem” which strikes me as very fitting, and that not just in view of the
Derridadaist nonsense in quotations 95.18 above.
Quotation 95.22. Girard [1987], pp. 75 ff.

Gödel’s theorems are very popular; this popularity comes from the fact
that, undoubtedly, the theorems have a deep philosophical meaning. The
natural temptation is to extrapolate, exactly to abstract the theorems from
their precise mathematical context, in order to get less precise, but more
general statements. The usual experience of the proof-theorist with Gödel’s
theorems is that, in a first step one gets so struck that one tries to refor-
mulate one’s personal view of the world to fit the contents of the theorems.
Later one, with a reasonable experiences of these results, this kind of «dra-
matic» consequences appear as ridiculous extrapolations. The situation is
quite different with outsiders (philosophers, poets, musicians, journalists...)
who have usually no opportunity to correct their first superficial impressions
concerning the theorems through mathematical practice.
The problem is precisely that these people have wide access to mass me-
dia, therefore their misconceptions concerning Gödel’s theorems are widely
spread. Of course, this makes the name of Gödel very famous, and this is
not so bad (similarly a bad movie on, say, Mahler can have the positive
effect of making Mahler’s music more popular...).
The current misunderstanding of Gödel’s theorem can be expressed by
«a theory cannot think about itself ». This is a metaphorical approach to
the second theorem, the most popular of both theorems. This is a very bad
expression of the meaning of the second theorem:
— If one means by this that the facts concerning [[PA]] cannot be proved
in [[PA]], this is a nonsense, since the second theorem says that Con(PA)
is a formula of the language, but not provable.
— If the meaning is that facts concerning [[PA]] cannot be proved in
[[PA]], this is more correct, but trifling! Observe that most important facts
concerning PA (except, of course results like Con(PA)) can be proved in
PA itself[[.]] [[. . .]]

A correct statement of this misunderstanding (if misunderstandings can
be stated correctly) is: «the reductionist approach fails». This means that
each new (consistent) theory is in some sense «transcendental» over the
previous ones, and cannot be reduced to them. This follows from the first
incompleteness theorem: the new theory proves new elementary formulas.
At a more general level, we have people who say bravely that «by
Gödel’s theorem, science cannot think about itself»; at this level of general-
ity, where formal theories are replaced by science, or – why not? – thought,
Gödel’s results are used in a way which makes them say approximately the
apposite [[sic!]] of the original meaning! Gödel’s results say nothing about
science nor thought in general; if Gödel’s results say something in this
area, this concerns only the limitations of formal thought, or mechanical
thought ; if one wants to reformulate Gödel’s result in one general sentence
it is something like: «there are scientific facts which cannot be mechanically
approached», of which must be said: one can forget everything in Gödel’s
proofs, except the fact that provability is mechanical (in the technical sense:
recursive).
[[. . .]]
In fact what happens with Gödel’s theorem is that some people use
it as a scientific alibi for their fantasies; for instance, there is an idea of
transgression (of the limits of a theory) in Gödel’s theorems; but ideas of
transgression are most fashionable among painters and musicians of the
XXth century, and these people can refer to Gödel’s theorems. But at the
turn of the century, they would have referred to Mallarmé for instance, who
expressed these ideas of transgression, in very beautiful and cryptic poems;
the difference between Mallarmé and Gödel, is that Mallarmé’s poems are
perfectly adapted to this situation, whereas it is dishonest to use Gödel’s
results in a completely metaphorical (and usually incorrect) sense.
95e. Self-reference and diagonalization. The next set of quotations pro-

vides more examples of using the diagonal method for philosophical argumentation.

In a generous universe there are more things than can be named even with
an infinitude of names. For, let us recall again the twin discrepancy [[. . .]]
between sets and open sentences. We saw that some of the sets are not
determined by any of the sentences. But these sets will lack names; for if a
set has a name, say ‘a’, then the set is determined by the sentence ‘x ∈ a’.
Comment. Quine is talking about names here. This is concerning the jettisoning of
the purely referential interpretation of quantification: Marcus. Quine’s argument in
[1970] against substitutional quantification.
(2) Wang [1974], p. 324; quoting from Gödel’s Gibbs lecture 1951.
The human mind is incapable of formulating (or mechanizing) all its math-
ematical intuitions. I.e.: If it has succeeded in formulating some of them,
this very fact yields new intuitive knowledge, e.g. the consistency of this
formalism. This fact may be called ‘the incompletability’ of mathematics.
(3) Wang [1987], p. 171.
The unsolvable problems and G[[ödel]]’s undecidable propositions can be
viewed as a part of the surprising proofs √ of impossibility in the mathemat-
ical tradition: the Greek discovery that 2 is not a rational number, the
nineteenth-century discoveries that it is not possible to solve an equation of
the fifth degree by means of radicals, that e and π are not algebraic num-
bers, that it is not possible to trisect every angle by ruler and compass, that
the parallel axiom is not deducible from the other axioms of Euclid, that
the reals numbers are not countable, etc. At the same time, the concepts
of arbitrary mechanical procedures and formal systems are more general
and give more a feeling of finality in the sense that we do not have defi-
nite conceptions of going beyond that are, for example, comparable with
the extension of the domain of numbers from the rational to the algebraic.
Indeed, the generality of these concepts reminds one of Kant’s concept of
all possible experience and his speculations about its limits.
(4) Grim [1988], p. 356.
There is no set of all truths.
For suppose there were a set T of all truths, and consider all subsets of
T, elements of the power set PT.
To each element of this power set will correspond a truth. To each
set of the power set, for example, a particular truth T , either will or will
not belong as a member. In either case, we will have a truth: that T , is a
member of that set, or that it is not.
There will then be at least as many truths as there are elements of the
power set PT. But by Cantor’s power set theorem the power set of any set
will be larger than the original. There will then be more truths than there
are members of T, and for any set of truths T there will be some truth left
out.
There can be no set of all truths.
(5) Penrose [1994], p. 193.
[[T]]he human understanding of mathematical truth cannot be entirely re-
duced to something that is computationally checkable. [[. . . T]]he human
notion of perceiving the unassailable truth of Π1 -sentences cannot be made
into any computational system, precise or otherwise. There is no paradox
here, though the conclusion may be found disturbing. It is in the nature
of any argument by reductio ad absurdum that one arrives at a contradic-
tory conclusion, but such seeming paradox serves only to rule out the very
hypothesis that was being previously entertained.
Comment. The “unassailable truth” of something like Gödel’s sentence is precisely
what is at stake in the present study; i.e., in less Academic Routine Jargon (ARJ),
it is one of the crucial points to be questioned in my approach to dialectical logic
through reassessing Cantor’s diagonal method. The point is to enquire into the
nature of reductio ad absurdum and not just call on it.
(6) Penrose [1994], p. 195.
As a means to trying to circumvent the limitations imposed by Gödel’s
theorem, one might try to devise a robot that is somehow able to ‘jump out
of the system’ whenever its controlling algorithm gets caught in a compu-
tational loop. It is, after all, the continued application of Gödel’s theorem
that keeps leading us into difficulties with the hypothesis that mathematical
understanding can be explained in terms of computational procedures[[.]]
The master himself seems to have had somewhat different hopes as the following
quotation suggests.

Tarski has stressed in his lecture (and I think justly) the great impor-
tance of the concept of general recursiveness (or Turing’s computability).
It seems to me that this importance is largely due to the fact that with
this concept one has for the first time succeeded in giving an absolute def-
inition of an interesting epistemological notion, i.e., one not depending on
the formalism chosen. In all other cases treated previously, such as demon-
strability or definability, one has been able to define them only relative to
a given language, and for each individual language it is clear that the one
thus obtained is not the one looked for. For the concept of computability
however, although it is merely a special kind of demonstrability or decid-
ability, the situation is different. By a kind of miracle it is not necessary
to distinguish orders, and the diagonal procedure does not lead outside the
defined notion. This, I think, should encourage one to expect the same thing
to be possible also in other cases (such as demonstrability or definability).
Comment. An “absolute definition of an interesting epistemological notion” is quite
a challenge for a dialectician of my kind, but I haven’t done enough work on the
notion of general recursive functions to take it up.
I continue with a bit of Kripke’s motivation for introducing a partial truth

predicate.

Consider the ordinary statement, made by Jones:
(1) Most (i.e., a majority) of Nixon’s assertions about Watergate are
false.
Clearly, nothing is intrinsically wrong with (1), nor is it ill-formed. Ordinar-
ily the truth value of (1) will be ascertainable through an enumeration of
Nixon’s Watergate-related assertions, and an assessment of each for truth
or falsity. Suppose, however, that Nixon’s assertions about Watergate are
evenly balanced between the true and the false, except for one problematic
case,
(2) Everything Jones says about Watergate is true.
Suppose, in addition, that (1) is Jones’s sole assertion about Watergate, or

alternatively, that all his Watergate-related assertions except perhaps (1)
are true. then it requires little expertise to show that (1) and (2) are both
paradoxical: they are true if and only if they are false.
Quotations 95.26. (1) Slezak [1982], p. 51.

If we are Turing machines, an interesting question to consider is: What
are the manifestations of the inevitable Gödel limitations? The intriguing
speculation which suggests itself here is that there may be an intimate con-
nection between the notorious peculiarities of ‘diagonalisations’ and the
perplexities of introspective knowledge. Presumably any system of internal
representation or ‘language of thought’ would share the fundamental prop-
erties of formal systems. The introspective perplexities of the self could be
the embodiments or realisation of the formal self-referential schemata —
that is, abstract ‘competence’ models of inherent, shared intuitions. Far
from refuting mechanism, Gödel’s theorem may even provide the most per-
suasive support for it.
(2) Boos [1983b], p. 157.
Contemporary metamathematical arguments reflect dilemmas of rationalist
metaphysics; not only do they not resolve them, as the logical positivists
seemed to hope: they seem to represent them, in formal miniature.
(3) Slezak [1984], p. 34.
The suggestive reverberations of the diagonal motif are perhaps most in-
terestingly seen in Descartes’s classical ‘Cogito’ argument: Just as the Liar
sentence says of itself that it is untrue, and the Gödel sentence says of it-
self that it is unprovable, so Descartes’s argument seems to be saying that
doubt is indubitable.

[[I]]ssues in the philosophy of science having to do with reference of theo-
retical terms and issues in the philosophy of mathematics having to with
the problem of singling out a unique “intended model” for set theory are
both connected with the Löwenheim-Skolem Theorem and its near relative,
the Gödel Completeness Theorem. Issues having to do with reference also
arise in philosophy in connection with sense data and material objects and,
once again, these connect with the model-theoretic problems we have been
discussing. (In some way it really seems that the Skolem Paradox underlies
the characteristic problems of 20th century philosophy.)
(2) Putnam [1980], p. 478.
It is a striking fact that this entire problem does not arise to the standpoint
of mathematical intuitionism. This would not be a surprise to Skolem: it was
precisely his conclusion that “most mathematicians want mathematics to
deal, ultimately, with performable computing operations and not to consist
of formal propositions about objects called this or that.”
(3) Putnam [1980], p. 481.

What Skolem really pointed out is this: no interesting theory (in the sense
of first-order theory) can, in and of itself determine its own objects up
to isomorphism. Skolem’s argument can be extended [[. . .]] to show that
if theoretical constraints do not determine reference, then the addition of
operational constraints will not do it either. It is at this point that reference
itself begins to seem “occult”; that it begins to seem that one cannot be any
kind of a realist without being a believer in nonnatural mental powers.

The claim Gödel makes is that “V = L” is false “in reality”. But what
on earth can this mean? [[. . .]]
[[T]]he realist standpoint is that there is a fact of the matter—a fact
independent of our legislation—as to whether V = L or not. [[. . .]]
[[. . .]] If I am right, then the “relativity of set-theoretic notions” extends
to a relativity of “V = L” (and, by similar arguments, of the axiom of
choice and the continuum hypothesis as well).
(2) Putnam [1980], p. 471.
It may well be the case that the idea that statements have their truth values
independent of embedding theory is so deeply built into our ways of talking
that there is simply no “ordinary language” word or short phrase which
refers to the theory-dependence of meaning and truth.
(3) Putnam [1980], p. 472.
[[T]]he realist—the hard core metaphysical realist—holds that our intentions
single out “the” model, and that our beliefs are then either true or false in
“the” model whether we can find out their truth value or not.
(4) Putnam [1980], p. 474.
[[A]]n appeal to mysterious powers of the mind is made by some. Chisholm
(following the tradition of Brentano) contends that the mind has a faculty
of referring to external objects (or perhaps to external properties) which he
calls by the good old name “intentionality”. Once again most naturalistically
minded philosophers (and, of course psychologists), find the postulation of
unexplained mental faculties unhelpful epistemology and almost certainly
bad science as well.
(5) Putnam [1980], p. 482.
[[T]]he world does not pick models or interpret languages. We interpret our
languages or nothing does.
We need, therefore, a standpoint which links use and reference in just
the way that the metaphysical realist standpoint refuses to do. The stand-
point of “nonrealist semantics” is precisely that standpoint. From that
standpoint, it is trivial to say that a model in which, as it might be, the set
of cats and the set of dogs are permuted [[. . .]] is “unintended” even if corre-
sponding adjustments in the extensions of all the other predicates make it
end up that the operational and theoretical constraints of total science or
total belief are all “preserved”. [[. . .]] [[F]]rom the viewpoint of “nonrealist”
semantics, the metalanguage is completely understood, and so is the object
language. So we can say and understand, “ ‘cat’ refers to cats”. Even though
the model referred to satisfies the theory, etc., it is “unintended”; we rec-
ognize that it is unintended from the description through which it is given
(as in the intuitionist case). Models are not lost noumenal waifs looking for
someone to name them; they are constructions within our theory itself, and
they have names from birth.
Quotation 95.29. Wang [1981], pp. 16 f.

[[W]]hen we contrast formal thinking with dialectical thinking, we have in
mind something much broader than obtaining results which can be writ-
ten out in given formal systems [[. . .]]. Rather it corresponds more or less
to abstract thinking or, more broadly, it is taken to be any thinking that
does not capture the real situation in its full concreteness. If we take formal
thinking in such a broad sense, as one does sometimes, we might even say
paradoxically that at each moment we are only capable of formal thinking
and the essence of dialectics is to realize this and strive for better approxi-
mations to the whole real situation all the time. In fact, this would appear
to be one possible interpretation of the following famous passage:
“We cannot imagine, express, measure, depict movement without inter-
rupting continuity, without simplifying, coarsening, dismembering, stran-
gling that which is living. The representation of movement by means of
thought always makes coarse, kills, . . . and not only by means of thought,
but also by sense-perception, and not only of movement, but of every con-
cept.
And in that lies the essence of dialectics.
And precisely this essence is expressed by the formula: the unity, iden-
tity of opposites.”
Comment. In a footnote, the passage is identified as “V. I. Lenin, Philosophical
notebooks (Collected works, vol. 38), pp. 259–260.”
I close this chapter with a fine example of poor philosophical argumentation, if

only to reiterate a main theme.
Quotation 95.30. Burbidge [1993], pp. 89 f.

Russell challenged the legitimacy of self-reference — of having thought think
about itself. Examples of sentences that refer to themselves (where someone
says he is currently lying, for example), lead to contradictions that should
be avoided. So Russell jumps to a theory of types that decrees that no
logical expression can ever refer to itself.
CHAPTER XXIV
TRUTH, PARADOXES, AND NON-CLASSICAL

LOGICS
Non-classical logics commonly employ the same symbols as classical logic which has
led to some confusion, given that the connectives are governed by different axioms;
or, as some may prefer to put it, a certain confusion seems to have been the starting
point for non-classical logics: we are actually dealing with different connectives; it
is only the symbols which are the same.
The aim of the present chapter is not to discuss or assess this claim, but to
provide materials for a later discussion. My own position on these matters is not
separable from my view on non-classical logic and, as a result, not easily communi-
cated in the present context. Suffice it here to say that according to my view, the
non-classical logics presented in the present study are indeed dealing with the same
connectives. The point is that already in sentential logic the realm for which these
connectives were originally defined is left, but this remains hidden to a classical, or
orthodox position, such as Quine’s, for instance. In view of this I regard Quine’s
comments on the issue in quotation 96.20 (1) as simply missing the point.
The issue of non-classical logics is a particularly unpleasant one because the
standard of the discussion seems to be reduced to the lowest common denominator of
the disputers in both fields. Given that the philosophical standard of mathematical
logicians matches the mathematical-logical standard of philosophers in terms of
incompetence, the outcome is cruel. It is as if you were to put Johan van Benthem
and Jacques Derrida together to discuss the relation between undecidability and
dialectic.1
As regards my approach to dialectical logic, these considerations will be dealt
with in more detail in §115. In this chapter I shall only present selected quotations
from the philosophical background of a few of those non-classical logics which have
some relevance for my view of dialectical logic; formal aspects were treated in the
tools.
§ 96. General aspects
There exists quite a variety of non-classical logics, none of which, however, managed
so far to supersede classical logic in its universal acknowledgment. In this first
1 For those who want to get a taste, I recommend a look at Van Benthem [1979b] and Derrida
[1980].
1338
§ 96. GENERAL ASPECTS 1339
paragraph I shall have a cursory glance at some of the reasons for trying to overcome
classical logic.
96a. Bivalence and dissatisfaction with classical logic. I open this sec-
tion with a quotation from a classical orthodox position.

Bivalence is a basic trait of our classical theories of nature. It has us
positing a true-false dichotomy across all the statements that we can ex-
press in our theoretical vocabulary, irrespective of our knowing how to de-
cide them. In keeping with our theories of nature we have viewed all such
sentences as having factual content, however remote from observation.
The problem with bivalence is that it is not clear which statements actually
qualify and in that respect the quotation from Quine gives no clue.
The issue of bivalence was already discussed by Aristoteles and quotations 57.17,
57.20, and 57.21 in these materials were meant to give an ideas of that.
In view of quotations 57.20, it will be obvious that “it is raining” is not a propo-
sition; at least not as long as it has not been specified where and when it is raining.
Although classical sentential logic makes no such claim as to which sentences are
in fact either true or false, this question forms an important part of the intuitive
background of classical logic. At first sight, natural language seems to provide ap-
propriate examples of propositions in the above sense. Particularly in the beginning
it is quite tempting to illustrate the aim of logic by examples of the following sort:
Snow is grey.
Ravens are white.
Jack is married.
Jack and Jill are married.
Introductory texts in logic provide ample supply of such examples. But apart from
the fact that it is rarely made clear that they do not form a genuine part of the theory
of classical logic, they are quite misleading. First of all, it doesn’t need much wit
to show that snow is sometimes grey, but sometimes not. All it needs is the right
times and places to look.2 It is possible, however, to find perfectly truth-definite
propositions; for example:
4 is a prime.3
7 added to 3 equals 4.
There are infinitely many prime numbers.
Still, there are a lot of people who are not satisfied with the fact that classical logic
has a perfect realm of application, but are concerned about those sentences which
are not true or false.
2 It is possible to find yellow snow too; and it wasn’t necessarily Santa’s reindeer.
3 The predicate “prime” is actually a good example, because it is decidable whether or not a
natural number is a prime.
1340 XXIV. TRUTH, PARADOXES, AND NON-CLASSICAL LOGICS
Quotation 96.2. Haack [1978], p. 162.

A good deal of informal discourse is, in some degree, vague. And so the
question arises whether, and if so, how, logicians should take account of
this fact.
Comment. I do not regard this as a question of logic, but one of morals (obligation).
Some people love to deviate to them. Suffice it to say that I have no intention to
join them.
A common strategy is to complain about the lack of application of classical

logic in some particular field of interest, usually something like ordinary language,
intuitive reasoning, just the way MostPeople thinks — or doesn’t.
Quotations 96.3. (1) Finocchiaro [1980], p. 297.

At a memorable symposium Bar-Hillel lamented the fact that formal logi-
cians have devoted so little effort to the study of argumentation in natural
language. Calling the situation “one of the greatest scandals in human ex-
istence”, he challenged “anybody here to show me a serious piece of argu-
mentation in natural language that has been successfully evaluated as to
its validity with the help of formal logic”.
(2) Finocchiaro [1980], p. 301.
[[T]]he value of formal logic ultimately depends on its contributions to the
improvement of actual reasoning[[.]]
Quotations 96.4. (1) Haack [1978], p. 163.

Formal logical systems are supposed to be relevant to the assessment of
informal arguments; but the classical logical systems, in which every wff
is either true or else false, seem inappropriate for the assessment of infor-
mal arguments with premises and/or conclusions which, because of their
vagueness, one hesitates to call either definitely true or definitely false.
(2) Haack [1978], p. 163.
[[I]]t is clear that vague sentences can occur in informal arguments without
threatening their validity [[. . .]] a vague sentence can play a genuine role in
an argument (‘John likes capable girls; Mary is capable and intelligent; so
John will like Mary’); and so logicians must take vagueness more seriously.
(3) Haack [1978], p. 164.
The idea that increase of precision may not be an unmixed blessing is not
new; Duhem pointed out (1904 pp. 178–9) that the statements of theoret-
ical physics, just because they are more precise, are less certain, harder to
confirm, than the vaguer statements of common sense. Popper (1961, 1976)
has also suggested that precision may be a ‘false ideal’.
Comment. Adorno could contribute to this line of thought. Cf. also Berkeley in
quotation 58.12 (2).
Remark 96.5. In the context of the present study it is important to stress that none
of these attempts can claim to have overcome classical logic, since, as I have pointed
out above, classical logic has a paradigmatic field of application: the metatheory of
formalized theories. Only when classical logic is starting to lead into trouble here,
will I start to reconsider classical logic. My point is, of course, that it already has;
but this will require further elaboration for which the present chapter is not the
place.

Generally speaking, the weaker the logical means we use, the less is the
danger of inconsistency—the danger that a contradiction is derivable. So
intuitionistic logic can also be looked upon as an attempt to make more
certain that our arguments are consistent and that we do not get into hidden
inconsistencies or paradoxes or antinomies. How safe such a weakened logic
is, as such, is a question into which I do not want to enter now; but obviously
it is at least a little safer that the full classical logic.
Comment. I include this here as a warning against the “obviously . . . ”. Gödel has
shown in his paper [1933] that classical logic can be regarded as a subsystem of
intuitionistic logic.4 In other words, if there is an inconsistency in a theory based
on classical logic, then there is also one in that theory based on intuitionistic logic.
Quotation 96.7. Myhill [1984], pp. 130 f.

[[T]]he Fregean concept of property is inconsistent with classical logic. So if
we want to take Frege’s principle seriously, we must begin to look at some
kind of nonclassical logic.
I close this introductory section with a quotation from Takeuti which I find
attractive in its simplicity, but I don’t think can be upheld for more than the
purpose of illustration.
Quotation 96.8. Takeuti [1980], p. 167.

The classical logic is the logic of the absolute. The intuitionistic logic is the
logic of the mind. The quantum logic is the logic of the particles.
96b. Truth, paradoxes, and partial recursive functions. The notion of

truth occupies a central place in the discussion of the paradoxes. Kripke’s paper
[1975] seems to mark the tip of an iceberg of attempts to come to grips with the
semantical paradoxes. Although this is not congenial to the approach taken here, I
still want to include a few quotations, mainly to be able to say what my approach
is not about.
Quotations 96.9. (1) Kripke [1975], p. 694, footnote 9.

By an “orthodox approach”, I mean any approach that works within clas-
sical quantification theory and requires all predicates to be totally defined
4 Cf. theorem 24.28 in the tools.
on the range of the variables. Various writers speak as if the “hierarchy of

languages” or Tarskian approach prohibited one from forming, for example,
languages with certain kinds of self-reference, or languages containing their
own truth predicates. On my interpretation, there are no prohibitions; there
are only theorems on what can and cannot be done within the framework of
ordinary classical quantification theory. Thus Gödel showed that a classical
language can talk about its own syntax; using restricted truth definitions
and other devices, such a language can say a great deal about its own se-
mantics. On the other hand, Tarski proved that a classical language cannot
contain its own truth predicate, and a higher order language can define
a truth predicate for a language of lower order. None of this came from
any a priori restrictions on self-reference other than those deriving from
the restriction to a classical language, all of whose predicates are totally
defined.
Comment. Short: “there are no prohibitions . . . other than those deriving from
the restriction to a classical language, all of whose predicates are totally defined.”
In view of what I am doing in this study it might be worthwhile pointing out
that these are the prohibitions of the Tarskian approach which are at stake in the
present enterprise and I see no point in trying to sweep them under the carpet by
way of interpretation. In other words, there are prohibitions and there is an a priori
restriction 5 despite Kripke’s attempt at wrapping them in euphemisms and passing
them off as consequences.
(2) Kripke [1975], pp. 694.
Philosophers have been suspicious of the orthodox approach as an anal-
ysis of our intuitions. Surely our language contains just one word ‘true’, not
a sequence of distinct phrases ptruen q, applying to sentences of higher and
higher levels.
(3) Kripke [1975], p. 698.
Almost all the extensive recent literature seeking alternatives to the
orthodox approach—I would mention especially the writings of Bas van
Fraassen and Robert L. Martin—agrees on a single basic idea: there is to be
only one truth predicate, applicable to sentences containing the predicate
itself; but paradox is to be avoided by allowing truth-value gaps and by
declaring that paradoxical sentences in particular suffer from such a gap.
(4) Kripke [1975], pp. 700, footnote 18.
I have been amazed to hear my use of the Kleene valuation compared
occasionally to the proposals of those who favor abandoning standard logic
“for quantum mechanics,” or positing extra truth values beyond truth and
falsity, etc. Such a reaction surprised me as much as it would presumably
Kleene, who intended (as I do here) to write a work of standard mathemat-
ical results, provable in conventional mathematics. “Undefined” is not an
5 This use of a priori hurts my soul as a transcendental philosopher. It must feel worse for
a theoretical physicist when a psychologist talks about a field theory of thought, for instance. At
least the notion of a priori never was precise.
extra truth value, any more than—in Kleene’s book—u is an extra number
in sec. 63. Nor should it be said that “classical logic” does not generally hold,
any more than (in Kleene) the use of partially defined functions invalidates
the commutative law of addition. [[. . .]] Mere conventions for handling terms
that do not designate numbers should not be called changes in arithmetic;
conventions for handling sentences that do not express propositions are not
in any philosophically significant sense “changes in logic.“
Comment. This seems clear enough to me; but it is obviously not enough to protect
his approach from being subsumed under the label of a deviant logic. There are
enough people around who think they know it better anyway. Notice, however, that
in view of Frege’s original doctrine, admitting partially defined functions would have
to count as a change of logic,6 at least of Frege’s logic.
Quotations 96.10. (1) Skyrms [1984], p. 119.

For partial recursive functions, the diagonal procedure does not lead outside
the category, the reason being that the functions are undefined in the crucial
place. Why not model a theory of truth which admits self-reference along
these lines? I think that it is fair to say that the idea is already really
there in Gödel (1944) and (1946). It has come to fruition in contemporary
truth-value-gap treatments of
self-reference, especially in the work of Kripke.
(2) Gupta [1982], pp. 175 f.
The Liar paradox raises two distinct though related problems about the
concept of truth. The first is the descriptive problem of explaining our use
of the word ‘true’, and, in particular, of giving the meaning of sentences
containing ‘true’. It is a fact that although we do not have clear intuitions
about the meaning of some sentences containing ‘true’, e.g., the paradoxical
sentences, yet we do manage to use successfully various other sentences that
contain the word ‘true and we do have fairly clear intuitions about what
these sentences mean. [[. . .]]
The second problem that the Liar paradox raises about the concept
of truth is the normative one of discovering the changes (if any) that the
paradox dictates in our conception and use of ‘true’.
Quotations 96.11. (1) Herzberger [1982], p. 479.

ONE lesson I am inclined to draw from the recent history of philosophical
struggles with semantic paradoxes, is that new techniques for suppressing
them are unlikely to advance our understanding of the basic problem. No
sooner has one semantic paradox been defeated than some new paradox has
arisen to take its place.
Comment. I don’t know whose understanding is behind “our understanding”, but
my experience doesn’t leave much hope that it is anything other than some brand
of naive ethnocentricity.
6 Cf. Frege in quotation 71.16 (6) in these materials.
(2) Herzberger [1982], p. 481.

It is no use inquiring into the properties of “semantically closed” languages,
because Tarski’s definition leaves too little room for any such languages.
This has been remarked, for example, by Kripke, and it is in effect Tarski’s
indefinability theorem: expressively rich languages of the standard sort can-
not contain their own truth predicate in Tarski’s sense. For this reason, one
doesn’t want to postulate Tarski’s schema T as a basic assertion of naive
semantics. That would render naive semantics itself an inconsistent and
thereby false theory.
Comment. I’m glad that “one doesn’t want” to do that in naive semantics. That
keeps this variety of naivety at a healthy distance from me.
(3) Herzberger [1982], p. 487.
[[A]] paradoxical statement is a statement whose value will be true at one
stage of evaluation, false at some later stage, true again at some still later
stage, and so on ad infinitum. Every attempt to evaluate it by the ordinary
semantic rules, sooner or later will have its verdict reversed. This doesn’t
mean that it’s neither true nor false; nor does it mean that it’s both true
and false. Its fundamental semantic characteristic is neither a truth value
nor the absence of a truth value, but a valuational pattern.
Comment. One further step, namely acknowledgement of some epistemological rel-
evance of different ‘valuational patterns’ for a foundation of categories (theoretical
constants), and we’re almost there: in speculative philosophy.
Gluts or gaps?
Quotations 96.12. (1) Dowden [1984], p. 125.
It is [[. . .]] interesting to see the similarity in structure between this seman-
tics of so-called truth-value “gluts” and the semantics of truth-value “gaps”
proposed by Kripke (1975). However, at the more practical level of choosing
among competing solutions to the paradoxes, the cost of Priest’s proposal
is at least as great as Kripke’s. Neither theory treats the Strengthened Liar
sentences (“This sentence is not true”) the way it treats the ordinary Liar.
(2) Priest [1984], pp. 158 f.
[[W]]e use the notion of exclusion negation, i.e., a functor ∗ such that p∗Aq
is true precisely if pAq is not true and false otherwise. The extended liar is
then simply the sentence ‘∗This sentence is true’ (α). Now a contradiction
is quickly forthcoming. For if (α) is within the set of true sentences, it is
in the complement of this set. And if it is without the set of true sentences
then (α) is true, and so within the set of true sentences. The extended
contradiction obviously shows that the gap theorist has gained little by
his/her manoeuverings. S/he has still ended up inconsistent. (Of course, it
is always possible for the gap theorist to maintain that exclusion negation
is not expressible in the language in question. However, this is obviously
unsatisfactory. For consistency is purchased only at the price of castrating
the language, so that something which can be said cannot be said in it.
If the language does not contain an exclusion negation there can be no

objection to simply adding on to it, i.e., extending it by a new function
and stipulating this to heave the appropriate truth condition.) What is not
clear is why it should be thought that it sinks the glut theorist too. After
all, the aim of the glut theorist is not to avoid contradictions, but precisely
to allow for them in a way that does not lead to disaster.
Comment. This sounds like saying: precisely because the aim of the glut theorist
is not to avoid contradictions, contradictions can’t do any harm. If this sounds
hopelessly naive, this is precisely what I think it is. What sinks the glut theorist
is the direct fixed point property for terms7 which, as a consequence, Priest [1997]
struggles to ward off as “unpalatable”.
(3) Priest [1984], pp. 159 f.
A final desperate measure is to take logical consequences to be defined
proof-theoretically and simply specify that A ∧ ∗A/B is an acceptable rule
of inference. However, the answer to this is the same as that given by the
doctor to the patient who came to see her saying “Doctor, it hurts when I
do this” . . . .
Comment. I hate to think of the answer. But I find it unfair that Priest is evoking
a woman doctor to make his point. In any case, I just hope I will not get that
response, if I find myself in the situation of having to say “Doctor, it hurts when I
pee.”
96c. Many-valued logics. The origins of many-valued logics go back to Łu-

kasiewicz and Post.
Quotations 96.13. (1) Łukasiewicz [1910], pp. 507 f.

[[T]]he possibility is by no means excluded that constructions which count
today as free of contradiction nevertheless contain a deeply hidden contra-
diction which we have not yet been able to discover. [[. . .]] A newly discov-
ered contradiction by B. Russell, which touches on the logical foundations
of mathematics, demonstrates that we encounter completely unexpected
and unexplained difficulties with such constructions.
(b) Actual objects and reconstructive abstractions, insofar as they cor-
respond to reality, appear to be placed beyond contradiction. In fact there
is known to us no single case of a contradiction existing in reality. Indeed
it is generally impossible to suppose that we might meet a contradiction in
perception; the negation which inheres in contradictions is not at all percep-
tible [wahrnehmbar ]. Actually existing contradictions could only be inferred
[erschlossen].
(2) Bolc and Borowik [1992], p. 23.
The origins of multivalued logics can be traced as deep as the treatises
of Aristotle. He was the first to object to rigid bivalence of statements.
His doubts concerned the so-called Law of the Excluded Middle (p ∨ ∼ p),
7 Cf. theorem 130.8 in the groundworks and/or theorem 42.21 in the tools.
considered as undeniable truth in classical logic. The idea of accepting sen-

tences which (in a given instance) fail to be either absolutely true of ab-
solutely false aroused contention between the Epicureans, on the one side,
and the Stoics (including Chrysippus) on the other. The latter represented
the standpoint of extreme determinism, admitted the possibility that nei-
ther of two statements, one of which negated the other, must necessarily be
true; in particular, when the statements involved events that were to come.
Today, it seems, many valued logic is associated with the name of Łukasiewicz,
partly because of the popularity of his infinite-valued logic which allows unrestricted
abstraction.
I quote from Łukasiewicz’ early investigation into the principle of contradiction
in Aristotle and his farewell lecture.
Quotations 96.14. (1) Łukasiewicz [1910], p. 486.

Just as in the course of the nineteenth century a more exact examination of
the Euclidean parallel line postulate has led to new, non-Euclidean systems
of geometry, so the conjecture would not be entirely out of order that a
fundamental revision of basic laws (Grundgesetze) of Aristotle’s logic might
perhaps lead to new non-Aristotelian systems of logic.
(2) Łukasiewicz [1910], pp. 506.
19. Every proof of the principle of contradiction must take into ac-
count the fact that there are contradictory objects (e.g., the greatest prime
number). In the most general formulation: “the same characteristic cannot
belong and not belong to an object at the same time” is in terms of the
principle of contradiction most certainly false.[[∗ ]] It could only be true, and
then it would also be proven formally, if the word “object” is to designate
only objects which are free from contradiction.
Comment. There is a footnote to ∗ which I include:
So far as I know, Meinong first put this proposition forward. At the
occasion of certain critical observations of B. Russell’s, Meinong expressed
himself in the following way (Über die Stellung der Gegenstandstheorie im
System der Wissenschaft, [Leipzig, 1907], p. 16): “B. Russell lays the real
emphasis on the fact that by recognizing such (scil. impossible) objects the
principle of contradiction would lose its unlimited validity. Naturally I can
in no way avoid this consequence. . . . Indeed the principle of contradiction
is directed by no one at anything other than the real and the possible.”
(3) Łukasiewicz [1918], p. 86.
Logical coercion is most strongly manifested in a priori sciences. Here
the contest was to the strongest. In 1910 I published a book on the principle
of contradiction in Aristotle’s work, in which I strove to demonstrate that
that principle is not so self-evident as it is believed to be. Even then I strove
to construct non-Aristotelian logic, but in vain.
Now I believe to have succeeded in this. My path was indicated to
me by the antinomies, which prove that there is a gap in Aristotle’s logic.
Filling that gap led me to a transformation of the traditional principles of

logic.
[[. . .]] I have proved that in addition to true and false propositions there
are possible propositions, to which objective possibility corresponds as a
third in addition to being and non-being.
This gave rise to a system of three-valued logic, which I worked out
in detail in last summer. That system is as coherent and self-consistent as
Aristotle’s logic, and is much richer in laws and formulae.
That new logic, by introducing the concept of objective possibility, de-
stroys the former concept of science, based on necessity. Possible phenomena
have no causes, although they themselves can be the beginning of a causal
sequence. An act of a creative individual can be free and at the same time
affect the course of the world.
(4) Łukasiewicz [1920], p. 87.
Three-valued logic is a system of non-Aristotelian logic, since it assumes
that in addition to true and false propositions there also are propositions
that are neither true nor false, and hence, that there exists a third logical
value. That third logical value may be interpreted as “possibility” and may
be symbolized by 21 .
(5) Scott [1976], p. 64.
[[H]]ow did Łukasiewicz come to propose his many-valued truth tables? As
far as I can make out he said very little in print about the genesis of the
idea. And once an idea like this is put forward, it has a life with no one
ever asking what right it has to travel so far or where its intellectual visa
is. To go far it needs two principal qualities: first, it must be rather simple-
minded so as not to tax the patiences of the people it will meet; and in the
second place it must stand in opposition to something ‘classical’ so we can
all have the thrill of the break-through or revolution. None of us wants to
be remembered for his bad reviews of a Beethoven. Speculation in ideas is
as risky as speculation in land, but unfortunately the punishment does not
come as quickly, and bad shares are still thick on the market.
Comment. It is not just that bad shares are still thick on the market, it is also that
they are hard to identify; or, in other words, trying to sort out the good shares from
the bad ones, is just what speculation is all about.
96d. Questions of method. In an area as sensitive as that of a non-orthodox

approach, I naturally have to dedicate considerable care to the question of method.
I begin with quotations which are intended to represent what was a classic line for
a considerable time.

If we [[. . .]] analyze the [[essential]] assumptions which lead to the antinomy
of the liar, we notice the following:
(I) We have implicitly assumed that the language in which the antinomy
is constructed contains, in addition to its expressions, also the names of
these expressions, as well as semantic terms such as the term “true” referring
to sentences of this language; we have also assumed that all sentences which
determine the adequate usage of this term can be asserted in the language.
A language with these properties will be called “semantically closed.”
(II) We have assumed that in this language the ordinary laws of logic
hold.
[[. . .]] Since every language which satisfies both of these assumptions is
inconsistent, we must reject at least one of them.
It would be superfluous to stress here the consequences of rejecting the
assumption (II), that is, of changing our logic (supposing this were possible)
even in its more elementary and fundamental parts. We thus consider only
the possibility of rejecting the assumption (I). Accordingly, we decide not
to use any language which is semantically closed in the sense given.
Comment. Note the “it would be superfluous”-phrase here. There is some ironical
aspect to this: Łukasiewicz’ infinite-valued logic which is indeed a very successful
way of “changing our logic” was published in a joint communication with Tarski in
1930 and can be found in English translation in the collection of Tarski’s papers
Logic, Semantics, Metamathematics.
(2) Stegmüller [1957], p. 38 f.
Wir stehen also vor der Alternative: Preisgabe der semantischen Geschlos-
senheit der Sprache oder Ersetzung logischer Grundregeln durch neue. Das
letztere würde eine wissenschaftliche Katastrophe darstellen; denn es ist
nicht einzusehen, wie bei Verwerfung jener einfachen logischen Prinzipien
und Deduktionsregeln, die bei der Konstruktion der semantischen Antino-
mien verwendet wurden, auch nur ein geringer Bestandteil des als „Wissen-
schaft“ bezeichneten Forschungsbetriebes aufrechterhalten werden könnte.
Comment. This is essentially Tarski’s “it would be superfluous”-claim with some
fantastic embellishment (“scientific catastrophe”). The point for my enterprise is,
firstly, how can the “simple logical principles and rules of deductions” be restricted
so as to avoid the fatal conclusion from the paradoxes and, secondly, how can
something like arithmetic be done on the basis of the remaining logical tools?

[[I]]t can easily be shown that if one were to accept contradictions then
one would have to give up any kind of scientific activity: it would mean
a complete break-down of science. This can be shown by proving that if
two contradictory sentences are admitted, any sentence whatsoever must
be admitted.
(2) Popper [1940], p. 410.
[[F]]rom two contradictory premisses, we can logically deduce anything, and
its negation as well. We therefore convey with such a contradictory theory—
nothing. A theory which involves a contradiction is entirely useless, because
it does not convey any sort of information.
From this, we see the real significance of the so-called “law of contra-
diction”. This logical rule, which forbids contradictions, thereby inducing
us never to accept any contradiction, secures the possibility of conveying
something with the help of a deductive system. Once a contradiction were
admitted, all science would collapse.
Comment. Silly Cantor didn’t know that and used an inconsistent set to prove the
well-ordering theorem. But luckily there is a Popper to set things right.
(3) Popper [1943], p. 50.
There is little hope for Hegelian dialectics to find support in even the
weakest of logics. . . .
Comment. This is probably the best opportunity to point out that the crux of
Gentzen’s approach to proof theory is to show that something like the ‘law of con-
tradiction’, or in different forms ex falso quodlibet, modus ponens etc., is redundant
in logic. In terms of the so-called “ideal calculus” this means that dropping the cut
rule disables the ill effects of the logical paradoxes, while a notion of number can
still be obtained.8
Quotations 96.17. (1) Maydole [1975], p. 269.

It is a common misconception that the paradoxes of Naive Set Theory
(NST) can be dodged by simply giving up the Principle of Bivalence, and
by shifting thereby from the classical two-valued logic to a many-valued
logic.
(2) R. L. Martin [1984], pp. 1 ff.
(T) Any sentence is true, if and only if, what it says is the case.
[[. . .]]
[[. . .]] The principle of bivalence states that every sentence is true or false.
[[. . .]]
[[T]]he ordinary Liar [[. . .]] is [[. . .]] independent of the principle of bivalence[[.
. . . W]]e can show, without any use of the principle of bivalence, that [[. . .]]:
(S0 ) There is a sentence that says of itself only that is false.
is [[. . .]] incompatible with (T). The argument [[. . .]] does rely on some other
semantic principles, besides (T); but no appeal is made to the principle of
bivalence. Here is the argument:
Let s0 be the ordinary Liar. First we show that s0 is not false, as follows:
suppose s0 is false; then, since that is what it says, it is true, and hence
not false. (Principle: no sentence is both true and false.) Therefore, s0 is
not false. But now we can see that s0 is false, since s0 says something, the
negation of which (s0 is not false) is true. (Principle: a sentence is false if
its negation is true.) Thus a contradiction.
Comment. My point in a nutshell: no use is made of the principle of bivalence itself,
8 Cf. §41 in the tools.
only of a principle which is equivalent to the principle of bivalence (on the basis of
classical logic). In symbols:
1. A ↔ tru(pAq) (Tarski’s truth definition T)
2. ¬(tru(pAq) ∧ fls(pAq)) (No sentence is true and false)
3. tru(p¬Aq) → fls(pAq) (a sentence is false if it is not true)
This last ‘principle’ can be seen in a few steps to be equivalent — on the basis of
classical sentential logic — with Martin’s ‘principle of bivalence’:
4. (tru(p¬Aq) → fls(pAq)) ↔ ¬tru(p¬Aq) ∨ fls(pAq) (classical sentential logic)
5. ¬tru(p¬Aq) ↔ ¬¬A (instantiation of T)
6. ¬¬A ↔ A (classical sentential logic)
7. ¬tru(p¬Aq) ↔ tru(pAq) (from 5., 6., and 1. by transitivity of →)
8. (tru(p¬Aq) → fls(pAq)) ↔ tru(pAq) ∨ fls(pAq) (from 4. and 7.
by substitutivity)
The right side of this last bi-implication is a symbolic formulation of Martin’s ‘prin-
ciple of bivalence’: every sentence is true or false.
The following quotation seems to represent some fairly general attitude.

The typical Hilbert-style calculus is inefficient and barbarously unin-
tuitive. But they do have two merits. The first is that their mechanics are
usually very simple to describe — many Hilbert-style calculi for propo-
sitional logic have only one derivation rule, namely modus ponens. This
makes them suitable for encoding into arithmetic. The second merit is that
we can strengthen or weaken them quite straightforwardly by tampering
with the axioms, and this commends them to researchers in non-classical
logics.
Comment. With an eye to what I am going to do in the groundworks, I wish to
emphasize that I strongly disagree with Hodges’ view regarding the “second mer-
it” of the Hilbert-style calculi, viz., “that we can strengthen or weaken them quite
straightforwardly by tampering with the axioms”. To my mind, H-style calculi are
actually dangerous for research in non-classical logic because they invite tinkering
with ‘principles’ of logic and/or tamper with axioms on a superficial basis. Given
the usual lack of an underlying systematic structure of a set of rules and axioms,
one never really knows what is actually excluded when a certain axiom is sacrificed.
Hodges’ recommendation of Hilbert-style calculi for non-classical logic — “tamper-
ing with the axioms” — strikes me as either just silly or extremely cynical. Still, it
does, indeed, represent a certain tendency which accounts for the patchwork char-
acter of much of the research in non-classical logic.
I continue with one of the principal custodians of orthodoxy, Quine, and his
view on deviant logics.

Revision even of the logical law of the excluded middle has been proposed
as a means of simplifying quantum mechanics; and what difference is there
in principle between such a shift and the shift whereby Kepler superseded
Ptolemy, or Einstein Newton, or Darwin Aristotle?

My view of this dialogue is that neither party knows what he is talking
about. They think they are talking about negation, ‘∼’, ‘not’; but surely the
notation ceased to be recognizable as negation when they took to regarding
some conjunction of the form ‘p . ∼ p’ as true, and stopped regarding such
sentences as implying all others. Here, evidently, is the deviant logician’s
predicament: when he tries to deny the doctrine he only changes the subject.
Comment. The problem lies with the formulation that negation “ceases to be rec-
ognizable”.9
(2) Quine [1970], p. 85.
The classical logic of truth functions and quantification is free of paradox,
and incidentally it is a paragon of clarity, elegance, and efficiency. The
paradoxes emerge only with set theory and semantics. Let us try to resolve
them within set theory and semantics, and not lay fairer fields waste.
Comment. This conveys an attitude which is commonly ascribed to oriental po-
tentates: to blame the messenger who brings the bad news. What Quine avoids is
going beyond the obvious phenomenon, viz., where the problem shows up. It will
be clear that I am not going to follow his invitation to resolve the paradoxes where
they show up but investigate their environment more broadly, i.e., in particular the
various forms in which the fixed point property makes its appearance.
(3) Quine [1981], p. 91.
We stalwarts of two-valued logic buy its sweet simplicity at no small price
in respect of the harboring of undecidables. We declare that it is either true
or false that there was an odd number of blades of grass in Harvard Yard at
the dawn of Commencement Day, 1903. The matter is undecidable, but we
maintain that there is a fact of the matter. Similarly for countless similar
trivialities. Similarly for more extravagant undecidables, such whether there
was a hydrogen atom within a meter of some remote point that we may
specify by space-time coordinates. And similarly, on the mathematical side,
for the continuum hypothesis or the question of the existence of inaccessible
cardinals. Bivalence is, as Michael Dummett says, the hallmark of realism.
Comments. (α) The first point is intriguing if read as saying that undecidables are
the price for two-valued logic; an almost dialectical insight. I doubt, however, that
this is what Quine wanted to say, and I doubt even more that he is prepared to
actually pay the price as long as there is a chance to escape into floury10 language.
9 As regards this: cf. Read [1988], pp. 153 f; pp. 153 ff: rigid connectives.
10 Read: flowery; misprint intended.
(β) The last point is interesting in view of Priest’s position (dialetheism) which
claims to be both, realist and paraconsistent.
Quotation 96.21. R. M. Martin [1965], p. 280.

Every logic has a syntax and a semantics — in fact, [[. . .]], is a syntax and
a semantics in the sense that it is formulated within a syntactical or se-
mantical meta-language. These meta-languages in turn may be formulated
as formal systems and hence contain a logic as a part. This logic in turn is
the classical logic, irrespective of whether that of the object-language is or
is not.
Comment. I see two dogmas here: a) the distinction of syntax and semantics, b) the
meta-logic is classical. As regards b, I want to reiterate that the point of the present
study, like that of Hilbert’s proof theory, is not to have a classical meta-logic, i.e.,
for instance, not to have full tertium non datur available.
(2) R. I. G. Hughes [1981], p. 157.
Although the non-classical logics may have specialized applications, the
logic employed for abstract reasoning, including reasoning about logic, will
probably continue to be classical logic.
(3) Hacking [1979], p. 319.
If a nonstandard logic is possible, in a way that is not parasitic upon classical
logic, then a nonclassical notion of truth and consequence is possible. But if
a nonstandard logic must ultimately be explained using classical logic, then
indeed we would have found something that “our thought can overflow, but
never displace.”
Comment. Contrast this with the formulation in quotation 96.21 above. The phrase
that I find problematic is: “ultimately be explained using classical logic”. Is this
the case if a non-classical logic cannot be given an interpretation (realization and
translation) in a theory based of classical logic?
§ 97. Paraconsistent logic and dialetheism
The idea of admitting contradictions in a formalized theory is commonly seen as

being in close proximity to dialectics. In the particular form this idea has been
given under the label “dialetheism”, I regard it as mistaken in two respects: firstly
as regards the nature of the paradoxes in (higher order) logic and semantics, and
secondly, as regards the nature of dialectics (in the transcendental-speculative tra-
dition). Admittedly an expression such as ‘nature of the paradoxes’ is vague and
the topic will be discussed in chapter XXVIII in the groundworks. The present
paragraph is meant to provide some background.
97a. The ‘intuitive’ background. Not surprisingly with anything as unin-
tuitive as paraconsistency, there is no one ‘intuitive’ background. Worse, there is
hardly more than just a variety of attempts at picking up some ready-made historical
background without paying much attention to underlying philosophical problems.
§ 97. PARACONSISTENT LOGIC AND DIALETHEISM 1353
It is, in this respect, on a par with Tarski’s conception of truth or the notion of
possible worlds in modal logic.11 Plenty of missionary zeal makes up for that.
Quotations 97.1. (1) Routley [1979], p. 301.

As with the thesis that God exists, so with the Consistency Hypothesis that
the world is consistent, there are three main positions that can be taken;
namely, a theistic or classical position which accepts the hypothesis, an
agnostic or relevant position which suspends judgement, and an atheistic
or dialectical position which rejects the hypothesis.
(2) Routley [1979], p. 304.
A necessary condition that a sentential logic be dialectical is that it is
closed under modus ponens, i.e. when A and A → B are theses so is B, and
that it is simply inconsistent, i.e. contains contradictory theses of the form
A and ∼ A, but non-trivial, i.e. not every wff is a thesis.
Comment. It may be worthwhile mentioning that I do not regard the second one
as a “necessary condition”.
(3) da Costa [1974], p. 508.
Dialectic logic is intimately connected with the theory of inconsistent sys-
tems. There are several conflicting conceptions of dialectic logic and for most
specialists it is neither formal, nor even in principle formalizable. Nonethe-
less, employing techniques used in the theory of inconsistent systems, it is
apparently possible to formalize some of the proposed dialectic logics.
(4) Asenjo [1965], p. 321.
Other authors hold that dialectic does not involve rejection of the law of
contradiction. Hegel belongs to that group. His argument is that violation of
the law of contradiction makes it impossible to disprove any proposition at
all: it is impossible to assert anything because statements become indifferent
to proof, so to speak[[.]]
Comment. The last sentence refers to McTaggart, essentially as presented in quo-
(5) Arruda [1980], p. 4.
Hegel’s thesis (or Heraclitus-Hegel’s thesis) is the statement that there
are true contradictions [[. . .]]. Sometimes, Hegel’s thesis is also formulated as
to imply that consistency is a sufficient but not necessary condition for the
existence of abstract objects; concerning the existence of concrete objects,
consistency is neither necessary nor sufficient. Clearly, Hegel’s thesis can
only be supported with the help of a paraconsistent logic.
Comment. I regard this as a pretty lousy, but quite common form of Hegel “inter-
pretation”.12
11 Cf. quotation 90.8 as regards Krikpe’s introduction of the notion of possible worlds.
12 Cf. also Bunge in quotation 67.28 (1) in these materials.

Instead of trying to dissolve [[the logical paradoxes]], or explain what has
gone wrong, we should accept them and learn to come to live with them.
Comment. Whatever it is that speaks out of these lines, be that a pragmatic attitude
towards the paradoxes, or a conciliatory, or just resignation, it is certainly not an
attitude that regards them as sources of new thought forms, and it is in this respect
that I include it here as an antidote to what I regard as a sine qua non of any
Hegelian approach to the paradoxes.
(2) Priest [1979], p. 220.
Suppose we stop banging our heads against a brick wall trying to find a
solution, and accept the paradoxes as brute facts. That is, some sentences
are true (and true only), some false (and false only), and some both true
and false!
(3) Priest [1995], pp. 3 f.
Limits of [[a certain]] kind provide boundaries beyond which certain
conceptual processes (describing, knowing, iterating, etc.) cannot go; a sort
of conceptual ne plus ultra. [[. . .]] [[S]]uch limits are dialetheic; that is, that
they are the subject, or locus, of true contradictions. The contradiction, in
each case, is simply to the effect that the conceptual processes in question
do cross the boundaries. Thus, the limits of thought are boundaries which
cannot be crossed, but yet which are crossed.
(4) Priest [1995], p. 186.
If [[. . .]] → is a non-material conditional (for example, a strict conditional),
then α → ⊥ and ¬α are quite different notions. [[. . .]] In this case, the curried
versions of the paradoxes belong to a quite different family. Such paradoxes
do not involve negation and, a fortiori, contradiction. They therefore have
nothing to do with contradictions at the limits of thought.
Comment. What Priest seems to be saying here is that all paradoxes at the lim-
its of thought must involve negation. This strikes me as a severe presumptive
limitation which, however, makes good sense for dialetheism: dialetheism is mod-
eled on ‘kontradiktorische Widersprüche’ and essentially fails to provide any non-
classical strategy towards the limitative results of the kind presented, for instance,
in lemma 48.25iii–vi in the tools, i.e., limitations which arise from self-referential
paradoxes which do not involve negation.
Quotations 97.3. (1) Rescher and Brandom [1979], pp. 60.

Throughout history, a sort of horror contradictionis has been endemic
among rigorous thinkers, and it has been the view of the logical guild that
once a contradiction has been encountered nothing more remains to be done
but to leave the scene with proper expressions of disapproval. The indica-
tions are that a new spirit is abroad nowadays. Logicians and philosophers
are coming to take a new and more tolerant view of inconsistency.
(2) Rescher and Brandom [1979], p. 4.

A world-description can be viewed as making an assertion with claims
certain things to be the case. In admitting inconsistent worlds we ourselves
are accordingly safeguarded from self -contradiction because (as it were)
another assertor—the world-description at issue—is effectively introduced
as intermediary between us and the contradiction. It is the world at issue
that is inconsistent, but not necessarily our own discourse about it. (We
ourselves do not declare both P and ∼ P , but only Tw (P ) and Tw (∼ P ).)
In mooting the prospect of inconsistent worlds, one thus takes a position
that is—or can be—perfectly cogent and consistent within itself.
The avoidance of inconsistency unquestionably presents an important
principle for the claims of our own discussions. It is, no doubt, desirable
to avoid the inconsistency in our own thought and our own assertions. But
it is by no means equally imperative to shun the recognition of possible
inconsistencies in the objects of this thought and assertion. (Thought—
to reiterate the point—need not share the features of its objects.) The
consistent theoretical scrutiny of inconsistent worlds is not only attainable
but perhaps even a useful goal.
Comment. “Thought need not share the features of its objects” — whatever it is,
it sounds quite different from Hegel’s phrasing of “thought as the heart and the
soul of the world”,13 perhaps closer to Kant, in the sense that our representations of
objects may have properties, such as consistency, which the objects in themselves
don’t have.
97b. Claims. Paraconsistent logic is proposed as a way out of the trouble

caused by paradoxes and related phenomena. In what follows I present a few quo-
tations to give an idea of the kind of claims actually put forward.

Sooner or later we must come to claims we know to be true without a proof,
where the question of proof does not, as it were, arise. These are axioms
in the old-fashioned sense: self evident truths. (Why they are self evident,
I need not go into. It suffices that there must be some such things.) In our
present case these are presumably facts about numbers, such as that every
integer has a successor different from anything gone before or the basic
facts about addition. These are the sort of things that children become
familiar with when they learn to count and to do arithmetic. (We can, of
course, look for ‘proofs’ of these axioms in some foundational system such
as Principia Mathematica. However, these ‘proofs’ are not proofs in the
sense we are concerned with — means of coming to know that the things
proved are true.)
13 Hegel [1830], p. 83; see quotation 65.12 (1) in these materials.

(2) Priest [1979], p. 224.

We might well ask where exactly Gödel’s proof goes wrong. The place
is not difficult to locate. For, of course, his proof works only if the theory
under consideration is consistent. It is well-known that semantically closed
theories are inconsistent. [[. . .]] It is in this way that Gödel’s result is avoided.
Comment. In view of the results listed in lemma 48.25 in the tools this strikes me
as a terribly myopic view. True, Gödel’s original proof required (ω-) consistency; but
where would there be a justification for Priest’s claim that “his proof only works”
under that assumption? The notion of deduction, on which Gödel’s proof depends, is
represented on the formal level as a primitive recursive function the values of which
are either 1 or 0. From here it is a short step to formulate the part of Gödel’s result
that claims the unprovability of Gödel’s wff in such a way that the only assumption
is non-triviality.14
The next part of the claim concerns the proof relation of the naive theory.
Quotations 97.5. (1) Priest [1984], pp. 165.
[[T]]he argument concerns our naive proof procedures. These are informal
methods of proof which are used by working mathematicians (and logicians)
to settle the truth of some matter. Undoubtedly these change over time.
The informal methods of the 17th century are not those of the late 19th
century. Hence, to be precise, by “naive proof methods” I mean those in
operation now. If something can be established by our naive methods of
proof, I will call it a naive theorem, and the naive proofs and theorems, I
will call the naive theory.
We are now in a position to spell out the argument. My claim is that
treating the logical paradoxes as true contradictions resolves the problem
posed by the following two claims:
Claim 1. Let T be any consistent theory which can represent all re-
cursive functions, whose proof relation is recursive, and whose axioms and
rules are naively correct. Then T is incomplete in the sense that there is a
naively provable sentence that is not provable in the theory.
Claim 2. The naive theory can represent all recursive functions, and its
proof relation is recursive.
The problem in nuce is that the naive theory is, by definition, such
that anything which is naively provable is provable in it. Assuming its
consistency, it would, therefore, seem to be both complete and incomplete
in the relevant sense.
[[. . .]]
The content of [[. . .]] claim [[1 ]] is essentially Gödel’s theorem, and the
sentence in question is the Gödel “undecidable” sentence. [[. . .]] Any doubt
about claim one is likely to attach to the part which says that this sentence
is naively provable.
14 Cf. remark 114.4 (1) in the groundworks. Readers interested in further details are referred
to remark 49.22 (2) in the tools, as well as my considerations in chapter XXIX in the ground-
works, in particular the considerations following question 120.4 on pp. 1634 ff.
(2) Priest [1984], p. 173.

[[A]]ny consistent theory cannot be semantically closed. Hence semantic rea-
soning about the system cannot be represented in the system. But it is
essentially semantic reasoning which allows us to prove the Gödel sentence
to be true.
(3) Priest [1987b], p. 160.
What the second Goedel incompleteness theorem shows is that, clas-
sically consistency can be maintained only by giving the semantics of a
theory in a different theory. Thus (any) consistent theory must fail to be
capable of giving its own semantics either by the requisite notions failing
to be expressible in the language of the theory, or by requisite principles
about them failing to be provable in the theory. The theory must therefore
be either expressively incomplete or proof-theoretically incomplete.
To summarize: incompleteness is the price paid for consistency.
Comment. The problem of expressive completeness is quite central to my enterprise
and I take the situation to be considerably more complex than presented in the
above quotations. First of all, it is not just consistency which figures prominently in
Gödel’s second theorem, but non-triviality.15 Secondly, the incompleteness caused
by self-reference is where I see a chance to locate something like Hegel’s speculative
step, i.e., positive dialectic. After all, Hegel’s Logic is a dynamic theory in the
sense that it aims at developing thought forms as a result of some sort of logical
experience, as it were.16 In any case, giving up consistency does not save one from
incompleteness; as theorem 51.35 in the tools shows, the incompleteness result
extends to paraconsistent arithmetic as well; however, it is also complete.17 At
best what the dialetheist can achieve is to turn the incompleteness of classical
arithmetic into a dialetheia for paraconsistent arithmetic: paraconsistent arithmetic
is then complete and incomplete. But it should be clear that even that comes at
the ridiculously high price of having inconsistency spread to primitive recursive
predicates.
(4) Routley [1979], p. 324.
[[D]]ialectically, the coding into a language of a semantical antinomy forces
no [[. . .]] rejection of a truth definition. Only if further crucial assumptions
are added, namely that the class of all provable sentences of the metalogic
is consistent [[. . .]] and that the metalogic is classical, does Tarski’s negative
solution result, ‘that it is impossible to construct an adequate definition of
truth in the sense of convention T on the basis of the metatheory’. The
crucial assumptions are however dialectically unacceptable, so the negative
conclusion never ensues.
[[. . .]]
15 Cf. the deduction on p. 1634 in §120 in the groundworks.

16 Interested readers are referred to §120 in the groundworks for an elaboration on the
concept of logical experience.
17 Cf. remark 51.36 in the tools.
If Tarski’s limitative theorems concerning the definability of seman-

tical notions such as truth fail to establish results of the generality that
has commonly been accorded to them, and do not stand up without es-
sential but dialectically unacceptable assumptions, how, it is likely to be
asked, do Gödel’s limitative theorems, and all the results that bear on arith-
metic and its extensions, fare? Since all these results are permissed [[sic!]]
on consistency assumptions, it is evident that, correct or not, they can
be dialectically escaped. This elementary observation is virtually enough
to dispose of a great many exaggerated claims made on the strength of
limitative theorems[[.]]
Comment. In view of the comments made to the previous quotation and quotation
97.4 (2) above, there is not much to add, except perhaps that this kind of swift
reasoning has put me off paraconsistent logic.
Quotations 97.6. (1) Chihara [1984], p. 119.

[[C]]onsider the following step in Priest’s “paradox”: it is claimed that we
could prove, from self-evident truths, that the Gödelian sentence undecid-
able in P is true. But to do this, we would need to prove from self-evident
truths that P is consistent; and on the assumptions that the axioms of P
need not be genuine truths — only propositions we take to be self-evident
— it becomes unclear just how we would prove the consistency of P . For
in general, proving the consistency of complicated formal systems is no-
toriously difficult when we are allowed to use only self-evident truths. (It
is doubtful, for example, that even first-order Peano Arithmetic has been
proved to be consistent using only self-evident principles).
(2) Priest [1984], p. 166.
Contrary to what Chihara says, soundness proofs are not difficult: they
are very easy, almost trivial. Finitary proofs are difficult, but non-finitary
proofs are simple and can normally be given with the corresponding sec-
ond order machinery — which falls comfortable within the bounds of our
informal proof procedure. For example, to prove the soundness of Peano
Arithmetic we merely define the standard model of arithmetic and verify
that the appropriate axioms are true in it. Of course, the verifications use
the very principles (at the “meta-level”) whose truth we are supposed to be
verifying. The proof might not, therefore, convince a sceptic. It is, none the
less, a quite legitimate proof.
Comment. This is meant as a response to the foregoing quotation from Chihara
[1984]. Note that Chihara is not speaking of soundness proofs, but consistency
proofs from “self-evident truths”. In a footnote Priest points out that Chihara is
actually talking about consistency proofs, but he is quick to add that “classically
soundness implies consistency and Chihara is thinking classically” (ibid., p. 175,
n. 44), still neglecting the restriction to “self-evident truths”. Perhaps my comment
to the next quotation can give a clue why this restriction may not carry much weight
in Priest’s considerations.
(3) Chihara [1984], p. 119.

I find it hard to see how Priest could claim to be able to prove from self-
evident truths that P is consistent, given that he believes that P is incon-
sistent.
Comment. The point is, of course, not at all hard too see, at least for a dialetheist:
the consistency of P is a dialetheia: P is consistent and P is not consistent (incon-
sistent). So everything is actually quite self-evident: the truth of the axioms, the
consistency of P , the truth of Gödel’s sentence, and the inconsistency of P .
Quotations 97.7. (1) Priest [1987a], p. 20.

[[T]]he denial of the law of excluded middle would still not avoid dialetheism.
This is for the very simple reason that there are proofs of contradictions
which do not use it. Take Berry’s paradox, for example[[.]]
Comment. In his paper [1983a] Priest presents a proof of Berry’s paradox which he
claims makes no use of “the law of excluded middle”.18 As regards the issue of the
“law of excluded middle”, compare section 113c in the groundworks; as regards
Priest’s claim, in particular, I refer readers to remarks 113.16 (3) and (4) in the
groundworks. Cf. also Hazen [1987a], p. 420, regarding Priest’s formulation of
Berry’s paradox.
(2) Priest [1987a], pp. 36 f.
[[S]]et theoretic paradoxes can be produced which do not use the law of ex-
cluded middle or reductio. In Burali-Forti’s paradox, a direct argument is
given that the set of all (von Neumann) ordinals is not an ordinal, and a
different argument that it is. An example with fewer technical presupposi-
tions is Mirimanoff’s paradox concerning the collection of all well founded
sets.
Comment. This question is of a certain relevance in the context of how to deal with
the logical paradoxes. Again, I refer readers to remark 113.16 (3) in the ground-
works; compare also the following quotation.
(3) Priest [1983], p. 161.
[[A]]lthough no set theoretic paradox may be provable without the law of ex-
cluded middle, the case is different with the semantic paradoxes. Although
some of these, such as the heterological paradox, go via an assertion of the
form ϕ ↔ ¬ϕ, and hence use the law of excluded middle, the definability
paradoxes, such as Berry’s, Richard’s and König’s do not.
Comment. As for the two preceding quotations.
I close this section with a quotation concerning the possibility of providing truth
conditions for the ∈ -relation.
18 Section 95b of my [1992a] (version 0.5 of the present study) contained an L-style version
of that proof bringing to light the contractions involved. In view of Hazen’s point that the use of
the ǫ-operator invalidates the claim (because it hides features of excluded middle), I currently see
no need for a presentation of the proof, until, at least, the role of the ǫ-operator has been clarified.
Quotation 97.8. Priest [1983b], p. 55.

The truth conditions of arithmetic are well-founded whilst those of set the-
ory are not: the truth conditions of ∈ are given in terms of the truth of
compound sentences, which is itself given ultimately in terms of, among
other things, ∈. Similar loops do not arise with the truth conditions of
arithmetic. It might be thought that this vitiates my whole procedure. But
it does not; for well-foundedness is not required for my enterprise. It is for
some. For example, it was Tarski (1936) who first showed explicitly how to
give truth conditions. For his purpose well-foundedness was necessary: he
was giving a definition of ‘is true in L’, and an adequate recursive definition
requires a base clause which ensures a ground. However, I am not trying
to give a definition of ‘is true’. I am stating the conditions under which
certain sentences are true. [[. . .]] There is no reason why well-foundedness
should be necessary for this.
§ 98. Quantum logic
98a. Intuitive background. The refreshing thing about quantum logic is

that its intuitive background is not just another exercise in private opinions, but
algebraic models derived from the mathematical treatment of quantum mechanics.
The depressing thing, at least for me, is that I am not familiar with the latter.
Still, as a non-classical logic there is a point of contact which, hopefully, the present
collection of quotations will illustrate.

I regard the analogy between the epistemological situation in logic and the
epistemological situation in geometry as a perfect one.
(2) Putnam [1980], p. 78.
[[J]]ust as the almost unimaginable fact that Euclidean geometry is false —
false of paths in space, not just false of ‘light rays’ — has an epistemological
significance that philosophy must some day come to terms with, however
long it continues to postpone the reckoning, so the fact that Boolean logic
is false — false of the logical relations between states of affairs — has a
significance that philosophy and physics and mathematics must come to
terms with.
(3) Putnam [1980], p. 78.
[[I]]f quantum logic is right, then not only the propositional calculus used in
physics is affected, but also set theory itself.
Quotations 98.2. (1) Finkelstein [1969], p. 208.

To some the fact that there are vectors neither in A nor in ∼ A suggests that
there is not a two-valued logic in quantum mechanics, that the tertium non
datur breaks down. But we have already seen that even in quantum me-
chanics A ∪ ∼ A = I, A ∩ ∼ A = 0, expressing the fact that the tertium non
datur was not the weak point of classical logic. [[. . .]] [[I]]n fact the fracture
§ 98. QUANTUM LOGIC 1361
is in the distributive law. All the anomalies of quantum mechanics, all the
things that make it so hard to understand, complementarity, interference,
etc., are instances of non-distributivity.
(2) Finkelstein [1969], p. 212.
When we work problems in quantum mechanics, we use classical logic in
carrying out our computations, and this can obscure the fact that some
of the expressions are themselves statements in a non-classical logic. This
confusion is understandable, but avoidable. We must merely remember that
a state-function is a statement about an electron, say, and is not an electron
itself. The distinction is not a particularly subtle one: electrons are emitted
by cathodes, state functions are emitted by physicists. Therefore a property
of a state-function is not the same thing as a property of an electron, and
the two obey quite different logics.
(3) R. I. G. Hughes [1981], p. 152.
[[A]]ny operation on an elementary particle that determines the value of some
quantum-mechanical variable must simultaneously randomize the value of
at least one other variable; two variables that are linked in this way are said
to be incompatible.
(4) Finkelstein [1969], p. 214.
Today we speak of a curved geometry; the best term I can think of for such
a new logic is warped. New fundamental fields would enter to describe the
way propositions at one point are combined logically with propositions at
another point.
Comment. The conception of propositions at one point being logically combined
with propositions at another point is very close to my approach to dialectical logic.
The main problem is to specify how this logical combination is to be taken into
account in the logical laws.
(5) Dalla Chiara [1986], p. 429.
There are three possible states of the truth: true, false and indetermined. In
other words, we have a violation of the meta-theoretical tertium non datur
(according to which any proposition is either true or false). In spite of this,
the proposition ‘either X or not-X’ is always quantum logically true for
any X (in other words, the theoretical tertium non datur holds)!
This apparently curious logical situation can be explained by the fact
that, in this logical framework, the truth of a disjunction does not generally
imply the truth of at least one member (we may have ‘X or Y ’ is true for ψ
even if X, Y are both not true). This causes an asymmetrical behaviour of
conjunction and disjunction which determines the failure of the distributive
laws.
(6) Dalla Chiara [1986], p. 429.
[[I]]n the famous ‘two-slit experiment’, one deals with a physical situation
where, for a certain particle ψ, it is true that ‘either ψ has gone through
slit A or it has gone through slit B’; nevertheless one can neither maintain
that it is true that ‘the particle ψ has gone through A’ nor that it is true
that ‘it has gone through B’. Such a situation, which has been described for
a long time as intuitively paradoxical, represents nothing but a particular
example of the fact that the quantum logical truth of a disjunction does
not imply the truth of at least one of its members.
(7) Dalla Chiara [1986], p. 464, citing Jauch [1968].
[[Quantum logic]] is the formalization of a set of empirical relations which
are obtained by making measurements on a physical system. It expresses
an objectivity [[sic]] given property of the physical world. It is thus the
formalization of empirical facts, inductively arrived at and subject to the
uncertainty of any such fact. The calculus of formal logic, on the other hand,
is obtained by making an analysis of the meaning of propositions. It is true
under all circumstances and even tautologically so. Thus, ordinary logic is
used even in quantum mechanics of systems with a propositional calculus
vastly different from that of formal logic. The two need have nothing in
common.
98b. Aspects of the logic of quantum logic. There are various ways of
formulating quantum logic, not all of which seem very perspicuous. This makes the
problem of the meaning of the logical connectives the more urgent.
Quotation 98.3. Putnam [1969], pp. 232 f.

[[I]]s the adoption of quantum logic a ‘change of meaning’ ? The following
principles:
(1) p implies p ∨ q.
(2) q implies p ∨ q.
(3) if p implies r and q implies r, then p ∨ q implies r.
all hold in quantum logic, and these seem to be the basic properties of ‘or’.
Similarly
(4) p, q together imply p · q.
(Moreover, p · q is the unique proposition that is implied by every position
that implies both p and q.)
(5) p · q implies p.
(6) p · q implies q.
all hold in quantum logic. And for negation we have
(7) p and – p never both hold. (p · – p is a contradiction)
(8) (p ∨ – p) holds.
(9) – – p is equivalent to p.
Thus a strong case could be made out for the view that adopting quan-
tum logic is not changing the meaning of the logical connectives, but merely
changing our minds about the law
§ 99. SUBSTRUCTURAL LOGICS 1363
(10) p · (q ∨ r) is equivalent to p · q ∨ p · r (which fails in quantum

logic).
Only if it can be made out that (10) is ‘part of the meaning’ of ’‘or’
and/or ‘and’ (which? and how does one decide?) can it be maintained that
quantum mechanics involves a ‘change in the meaning’ of one or both of
these connectives.
Comment. In view of the position I take here: the operational rules are not changed;
hence no change of meaning.
As regards the notion of a change of meaning, Putnam’s view is as follows.
Quotation 98.4. Putnam [1969], p. 233.

[[W]]e simply do not possess a notion of ‘change of meaning’ refined enough
to handle this issue.
§ 99. Substructural logics
Dialectical logic as presented in §27 in the tools and further in the ground-
worksis a substructural logic.
Literature 99.1. Ono [1990] provides a survey. Restall [2000].
99a. Intuitive background.
Quotation 99.2. Ono [1990], p. 95.

Lambek (1958) introduced a Gentzen-type system for the implicational frag-
ment of the intuitionistic logic without any structural rule, now called the
Lambek calculus, which serves as a first major paper on logics without
structural rules, though until recently it was not well-known.
Quotation 99.3. Girard [1989], p. 80.

Relevance logic accepts contraction on both side, and removes weaken-
ing: the result of this cocktail of structural rules is very awkward since it
seems that the good combinations are: C+W+E (classical), W+E (affine),
E (linear), nothing (linear non commutative). The awkwardness of the logic
is made even worse by the adjunction of ad hoc distributivity rules, which
come from an attempt to stick —as much as possible in the absence of
weakening— to classical logic. In terms of resources, relevantists correctly
stated that the premise must be used in a causality, but their acceptation
of contraction now says that resources may be used ad libitum: from two
pieces of bread you will never go to one without eating, but from one, you
can get 1000, like in Jesus’s miracles . . .
99b. Contraction free logics. Of particular interest for the present study
are those substructural logics which have no contraction. The reason for this is to
be seen in their capacity to allow unrestricted abstraction.
Remarks 99.4. (1) Regarding the label “contraction”. In the present study I am us-
ing the label “contraction” for the sequential rules of inference introduced in Gentzen
[1934] and called “Zusammenziehung”. In Kleene [1952], p. 443, they are listed un-
der the name “contraction”, just as in the translation of Gentzen’s original work
in Szabo [1969], p. 84, and Takeuti [1975], p. 10. This use of the label “contrac-
tion” is not equivalent with its use for the formula schema which I listed as axiom
schema HA4 in section 16b, such as in Brady [1991] or a corresponding inference
rule which used to be called “absorption” in Moh Shaw-Kwei and others, still in
Priest [1987a], for instance, but seems to have been replaced by the label “contrac-
tion” since. Łukasiewicz’ infinite valued logic, for instance, is not a contraction free
logic in the sense here, because it allows a form of contraction, viz., E-contraction
in the sense of Slaney [1989], p. 105, which manifests itself in the provability of
distributivity. Or, in other words, I am not dealing with absorption free logic here;
there are a number of absorption free logics which are not contraction free in the
sense used here.19
(2) Shoenfield [1967], p. 21 uses the label contraction for the rule A ∨ A ⊢ A which
is valid in contraction free logic as presented here.
(3) Propositional calculi can be formulated without an explicit contraction rule,
as is done in Dyckhoff, “Contraction-Free Sequent Calculi For Intuitionistic Logic”,
The Journal of Symbolic Logic 57 (1992), 795–807. But this is not a contraction-free
logic in the sense I am using the term.
There exists a variety of logics without contraction, quite a number of which

take intuitionistic logic as a starting point. Interested readers are referred to Ono
[1990], in particular p. 97.
Quotations 99.5. (1) Pałasiński and Wroński [1986], p. 87.

As a logical system [[contraction free implicational logic]] was isolated as
early as in 1934 by A. Tarski [[1934/5]] who proved that all implicational cal-
culi containing BCK have the implicational fragment of the classical propo-
sitional logic as their Post-complete extension (see [[Tarski [1956],]] p. 397,
Theorem 2). Another interesting early result about the BCK-calculus was
found by S. Jaśkowski [[1963]] who proved that an implicational formula is a
theorem of the BCK-calculus if and only if it can be obtained by applying
the substitution rule to some classical tautology in which no variable oc-
curs more than twice. It is easy to notice that the above result of Jaśkowski
implies that the calculus BCK is decidable.
Literature 99.6. Grišin [1982]; Ono and Komori [1985].

19 The notion of robustly contraction free logics in Restall [1993] is based on a formulation of
contraction in terms of some implication like theoretical constant. In view of the foregoing remarks
it will be clear that I do not adopt this notion.

The logical laws extracted from mathematics are only adapted to eternal
truths; the same principles applied in real life, easily lead to absurdity,
because of the interactive (causal) nature of real implication.
(2) Girard [1989], p. 72 ([1995], p. 1).
Classical and intuitionistic logics deal with stable truths:
If A and A ⇒ B , then B, but A still holds.
This is perfect in mathematics, but wrong in real life, since real implication
is causal. A causal implication cannot be iterated since the conditions are
modified after its use[[.]]
Comment. Although this sounds very close to what I will be doing later on in this
study, it is not the same; the point being the idea of what makes a condition being
modified after its use. I shall run a line based on a systemic ambiguity of expressions
due to the possibility of arithmetization.
99c. Linear logic.20 I must admit at the outset that I have never made a
detailed study of linear logic nor am I familiar with the programming background
which motivated Girard’s development of linear logic; I only recognized familiar
thoughts and what I quote mirrors what I recognized.

There is a philosophical tradition of ‘strict implication’ amounting to
Lewis. In some sense, linear implication agrees with this tradition: in a
linear implication, the premise is used ‘once’, in the sense that weakening
and contraction are forbidden on the premise [[. . .]]. Even if this philosophical
tradition has not been very successful, the existence of linear logic gives a
retrospective justification to these attempts.

The program is essentially about the development of a logic of actions, i.e.
of non-reusable facts (versus situations)[[.]]
(2) Troelstra [1992], p. 1.
Linear logic may be viewed as an example of a “resource-conscious” logic,
where the formulas represent types of resource, and resources cannot be
used ad libitum. That is to say, asserting a sequent A, A ⇒ B means some-
thing like: we use two data (resources) of type A to obtain one datum of type
B. A truth on the other hand is something which can be used freely, as often
as we like. In Gentzen-style sequential formalisms “resource-consciousness”
shows itself by the absence of (some of) the so-called structural rules.
20 I wish to thank Greg Restall for providing me with heaps of material regarding linear logic
at a time when I had not even heard of it.

(3) Girard [1989], p. 73.

In linear logic, two conjunctions ⊕ (times) and & (with) coexist. They
correspond to two radically different uses of the word “and”. Both conjunc-
tions express the availability of two actions; but in the case of ⊕, both will
be done, whereas in the case of & , only one of them will be performed (but
we shall decide which one). To understand the distinction consider A, B,
C:
A: to spend $1
B: to get a pack of Camels
C: To get a pack of Marlboro
An action of type A will be a way of taking $1 out of one’s pocket (there
may be several actions of this type since we own several notes). Similarly,
there are several packs of Camels at the dealer’s, hence there are several
actions of type B. An action of type A → B is a way of replacing any specific
$1 by a specific pack of Camels.
Now, given an action of type A → B and an action of type A → C, there
will be no way of forming an action of type A → B ⊕ C, since for $1 you will
never get what costs $2 (there will be an action of type A ⊕ A → B ⊕ C,
namely getting two packs for $2). However, there will be an action of type
A → B & C, namely the superposition of both actions. In order to perform
this action, we have first to choose which among the two possible actions we
want to perform, and then to do the one selected. This is an exact analogue
of the computer instruction if . . . then. . . else. . . : in this familiar case,
the parts then. . . and else. . . are available, but only one of them will be
done.
(4) Girard [1989], p. 79.
In linear logic, both contraction and weakening will be forbidden as struc-
tural rules; but it would be nonsense not to recover them in some way:
we have introduced a new interpretation for the basic notions of logic (ac-
tions), but we do not want to abolish the old one (situations), and this is
why special connectives (exponentials ! and ?) will be introduced, with the
two missing structurals as their main rules.
Comment. The task of recovering contraction, at least to a certain extent, is tackled
in chapter XXXIII in the groundworks (for the situation of dialectical logic).
(5) Girard [1995b], p. 6.
The basic point is that linear logic connectives can express features that
classical logic could only handle through comples and ad hoc translations.
Typically the update of the position m of a pawn inside a chess game
with current board M into m′ (yielding a new current board M ′ ) can be
classically handled by means of an implication involving M and M ′ (and
additional features like temporal markers), whereas the linear implication
m ⊸ m′ will do exactly the same job. The introduction of new connectives
is therefore the key to a more manageable way of formalizing; also the
restriction to various fragments opens the area of languages with specific

expressive power, e.g. with a given computational complexity.
(6) Seely [1989], p. 376.
Γ, A → ∆
(der) : (“dereliction”)
Γ, !A → ∆
Γ→∆
(thin) : (“thinning” or “weakening”)
Γ, !A → ∆
Γ, !A, !A → ∆
(contr) : (“contraction”)
Γ, !A → ∆
Γ→A
(!) :
!Γ →!A
In (!), !Γ means !A1 , !A2 , . . . , !An . Girard actually gives the rules for
the de Morgan dual ?[[.]]
Comment. These rules are quoted from Seely [1989] rather than Girard [1987], p. 26,
because their formulation here fits better with my terminology: only → would have
to be replaced by ⇒. Similar rules can also be found in Ono [1993], p. 21. I shall not
place much emphasis on them because they are not compatible with unrestricted
abstraction.21
To illustrate the presence of non “resource conscious” reasoning in every day

life, I quote as a “real live example” an advertisement from a German computer
journal.
Quotation 99.10. 64’er, Das Magazin für Computer-Fans, issue 12, 1985, last page.
Udo macht aus 14,– DM satte 5019,– DM. Bei Schwäbisch Hall. Durch
vermögenswirksame Leistungen beim Bausparen im Tarif B. Udo, 16, Kfz.-
Schlosserlehrling. Er zahlt monatlich effektiv 14,– DM. Vom Staat erhält
er 12,– DM. Sein Chef zahlt 26,– DM. Das macht im Jahr 624,– DM und
nach sieben Jahren 5019,– DM.22
Comment. The point is that Udo does not turn a single 14,– DM into 5019,– DM,
but 84 times 14,– DM, i.e., he turns 1176,– DM into 5019,– DM,23 in other words,
he turns 14,– DM into 52,– DM every month, 84 times. Someone must have thought
that this was not impressive enough and added a substantial pinch of what I called
“DD-reasoning” (“Donald Duck’s reasoning how to become a millionaire”, or the
21 Cf. proposition 143.19 in appendix A1.
22 The comma in German notation stands for what is the decimal point in English notation.
Roughly, what the ad says is: Udo turns 14 Deutschmark into full 5019 Deutschmark by saving
money according to a Government supported scheme. Monthly, he pays 14 Deutschmark, the state
pays 12 Deutschmark and his boss pays 26 Deutschmark. That amounts to 624 Deutschmark in
one year, and 5019 Deutschmark in seven years.
23 12 times 52 is 624, but 84 times 52 is not 5019 but 4368; there must be some interest
included. (I don’t have the ad anymore.)

miraculous multiplication of money) on p. 959 of my [1992a] (version 0.5 of the

present study).24
99d. Contraction-free logic and unrestricted abstraction. This section,

finally, is concerned with the approach taken in the present study and the reason
that it is so meagre is partly that little work seems to have been done regarding
unrestricted abstraction in the context of a contraction-free logic, but also because
I have been so busy pursuing my own ideas.
Quotations 99.11. (1) Grišin [1982], p. 42.

Logic without contractions reveals two facts connected with the para-
doxes of set theory. First, derivations of contradictions from comprehension
axioms must use contraction rules or their equivalent, since it was proved
in [[Grišin [1974]]] that the class of all comprehension axioms is consistent
in logic without contractions. Second, if one adds to the comprehension
axioms an axiom of extensionality, then one can obtain a contradiction
even using the logic without contractions. [[. . .]] Moreover, it turns out that,
in logic without contractions, from the extensionality axiom and certain
comprehension axioms one can derive all the tautologies of classical logic.
For this derivation we need axioms of comprehension which are part of the
well-known systems of set NF. Thus, the problem of their consistency is
equivalent to the problem of their consistency in logic without contractions.
(2) U. Petersen [1980], p. 97; my translation.
Having inferred B from A and A → B we cannot expect [[. . .]] that A and
A → B are still available as presuppositions. It is possible that they have
changed in the process of inferring, that they have been exhausted, so to
speak. This means we interpret the implication A → B as “A transfers into
B”. In this way we want to take account of the peculiarity of unrestricted
abstraction.
Comment. Sometimes in retrospect I impress myself with my own insights. The
point, however, that was not at all clear to me when I wrote this, was why and how
in higher order logic or mathematics assumptions could ever be “exhausted”. My
answer today is worked out in §111 in the groundworks.
(3) Girard [1995a], p. 28.
Naive Set Theory has been the starting point of LLL: I was looking for
a system in which the complexity could be expressed independently of the
cut-formulas. In particular, it would also word for naive set-theory, since
there is a well-known (non-terminating, for obvious reasons) cut-elimination
procedure for it; by the way it had been observed long ago by Grishin [[Grišin
[1974]]] that, in the absence of contraction, cut-elimination works.
24 I have since abandoned the interpretation of (dialectical) implication in terms of using
resources and this is the reason why I no longer present the DD-reasoning.

Diagonal Method and Dialectical Logic Bo

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Diagonal Method and Dialectical Logic Bo

Uploaded by

Copyright:

Available Formats

DIAGONAL METHOD

Tools, Materials, and Groundworks

BOOK TWO: HISTORICAL-PHILOSOPHICAL

Part C. DIALECTIC, IDEALISM, AND

Chapter XIV. PREPARATIONS 712

Chapter XV. KANT’S TRANSCENDENTAL PHILOSOPHY 760

§62. The Antinomy of Pure Reason 817

Chapter XVI. HEGEL’S SPECULATIVE PHILOSOPHY 830

Chapter XVII. AFTER THE GREAT MASTERS 894

69c. Forms of thought and the idealistic conception 962

Part D. THE PHILOSOPHY OF MATHEMATICS AND LOGIC 979

Chapter XVIII. THE EMERGENCE OF ANTINOMIES IN THE

Chapter XIX. EARLY REACTIONS TO THE PARADOXES 1042

76d. Brouwer’s neo-intuitionism 1079

Chapter XX. FORMALIZATION, METAMATHEMATICS, AND THE

Chapter XXI. PHILOSOPHY OF MATHEMATICS AND LOGIC TODAY 1117

85e. Cut elimination and the ω-rule 1193

Part E. OUTSIDE THE TRANSCENDENTAL TRADITION 1199

Chapter XXII. ANALYTIC EMPIRICISM 1202

Chapter XXIII. PARADIGMS OF METHOD AND CONFIRMATION 1270

93d. Uncertainty and complementarity 1294

It is conceivable that a well-organized collection

This second book of my study concerning Diagonal Method and Dialectical

DIALECTIC, IDEALISM, AND

DIALECTIC, IDEALISM, AND THE TRANSCENDENTAL APPROACH

The historical-philosophical background of my theory of dialectic does not come

§ 56. Methodological preliminaries

A central issue in the present approach to dialectic and speculative philosophy is

I shall conﬁne myself to dialectic in the tradition of transcendental philosophy,

Quotation 56.1. Wang [1987], pp. 254 f.2

Quotations 56.2. (1) R. Robinson [1941], p. 1.

Quotations 56.3. (1) Cooke [1938], p. vii.

5 I am not at all convinced of N. K. Smith’s competence as regards the understanding of the

(3) French [1967], p. 624.

Quotations 56.5. (1) Hegel [1830], p. 181.

(2) Wallace [1873], p. 109; ‘translating’ Hegel [1830], p. 181.

Quotations 56.6. (1) Tredennick [1933], p. xxiv.

Denoting”: Frege “distinguishes . . . meaning and the denotation” as “unterscheidet

After these quotations, allow me to introduce my maxim of English common

8 In: W. Stegmüller (ed.): Das Universalienproblem. Wissenschaftliche Buchgesellschaft,

Darmstadt 1978; p. 26.

in German with Frege.

are cool. I shall not use them.

The issue of translation is diﬀerent according to the language in question. Ger-

my comment to quotation 66.14 (1).

enough, it is translated as “decision” in N. K. Smith’s Commentary. I have come across other

replace “just” by “precisely”.

thinks or feels diﬀerently, it is unnatural, deviant, abnormal, actually insane.16 All

Quotation 56.8. Burnet [1914], pp. v f.

§ 57. Ancient roots of the notion of dialectic

might remember ‘Un-American activities’.

19 Cf. quotations 92.22 in these materials.

57b. Platon’s dialectic. Prima facie, Platon could be expected to be a prime

Quotations 57.8. Lee [1955], p. 342; translating Platon II.532A/B (Republic).

Ideas (forms), universals, contradictions, and dialectic: This is the framework

(2) Fowler [1926/39], p. 229; translating Platon III.135A/B (Parmenides).

(AR2) metaphysics (ﬁrst philosophy)

In view of what is to come it might be worthwhile pointing out here that it is

(3) Rolfes [1925], p. 44; translating Aristoteles 1 b 9–15 (Kategorien).

Quotation 57.16. Tredennick [1933], p. 35; translating Aristoteles 986 a 23–986 b

instances of contrariety, but the Pythagoreans pronounced how many and

Und mit der Kontradiktion hat es dieselbe Bewandtnis. Es ist notwen-

τὴν οὺσίαν λεγομένον, ἢ τούτων τινὸς ἀποφάσεις ἢ οὐσίας· διὸ καὶ τὸ μὴ ὂν

(3) Schwarz [1970], p. 165; translating Aristoteles 1028 b 2–8.