A Commentary On Kant's Critique of Practical Reason (UCP 1960) - L.W. Beck

Lewis White Beck
A Commentary
on Kant's
Critique of Practical
Reason
Midvvay Reprint '

A Commentary on Kant's
Critiqu.e of Practical Reason
LEWIS WHITE BECK
THE UNIVERSITY OF CHICAGO PRESS
CHICAGO" LONDON
The Cniversity of Chicago Press, Chicago 60637
The university of Chicago Press, Ltd., London
© 1960 by The Cniversity of Chicago

All rights reserved. Published 1960
Midway reprint 1984
Printed in the Cnited States of America
ISBN: 0-226-04076-3
LCl": 60-5464
Foreword
Odd though it may be, the Critique of Practical Reason is a neglected

work. There is no study of it in any language that can be compared
favorably with the commentaries on the Critique of Pure Reason by
Kemp Smith, Paton, or Vaihinger. Not even briefer commentaries like
those of Cohen, Ewing. or Weldon on the first Critique exist in Eng-
lish for the second; there seems to be only one (that of Stange) in
German. Nothing like the uncompleted commentary on the Critique
of Judgment by Baeumler has been published to give the full histori-
cal background for the development of Kant's ethical theory. H. W.
Cassirer, who has written commentaries on the first and third Critique's,
has not done so for the second. So far as I have been able to discover,
no book even with the title of commentary has been published on the
Critique of Practical Reason since the eighteenth century.l Every book
on Kant in general and on his ethics in particular, of course, deals with
some of its contents; and surveys and abstracts of it as a connected
whole appear occasionally as chapters in larger works. But no one, I
believe, has essayed on a large scale to study the Critique of Practical
Reason as a literary unity and integral philosophical treatise.
There are at least two reasons for this comparative neglect of one of
Kant's greatest works. First, a commentary is not so absolutely indis-
pensable to the study of this Critique as it is to that of the others. The
book has few of the stylistic difficulties and philosophical obscurities
of the other Critique's. In contrast to the bewildering complexity of
the first and the third, most of the second Critique is straightforward
and almost simple. Or at least it appears to be on first reading. which
is perhaps the only reading most people give it.
The second reason lies in the understandable preference that many
readers have for the Foundations of the Metaphysics of Morals. This
1 Unless one excepts August Messer's Kommentar zu Kantl ethischen und ,eli-
gionspbilosopbiscben Hauptschriften (1929). in which he attempted to cover all the
principal treatises and succeeded in giving little more than paraphrases. The title, of
course, is not the important thing. Barni's Ezamen . • . (1851) is a commentary in
most senses of the word, though it neglects to explore historical and literary con-
nections. Bami says that "it is not meant to be a sterile commentary but an advanced
and fruitful criticism," though its analytical chapter "will offer at least an exact and
useful commentary" (pp. vi, vii).
v
Vl Foreword
small work has had three English commentaries (those of Duncan,
Paton, and Ross) in the last ten years and is perhaps the most widely
read of all Kant's works in England and America, existing in no less
than five current English translations (by Abbott, Beck, Friedrich,
Manthey-Zorn, and Paton). The Foundations does not, at least in its
first two parts, presuppose knowledge of the rest of Kant's writings
and is directed to those not involved in, or committed to, the Kantian
system. As a contribution to moral philosophy it must stand on its own
internal merits, and this fact makes it pedagogically more useful than
the larger Critique. While there are topics discussed in the Critique
which are omitted from the Foundations, there is also much overlap-
ping. In these overlapping areas, preference has almost always been
given to the simpler exposition in the Foundations, and the more obvi-
ously systematic and in part divergent treatments in the Critique have
been regrettably overlooked. Only the teachings of the Critique which
have no counterpart in the Foundations find their way into most expo-
sitions and criticisms of Kant's work. It has been wittily remarked that
if a reader of books on Kant noticed the few isolated passages from the
Critique of Practical Reason which are repeatedly cited, he might con-
clude that the Critique was a lost work, only a few fragments of which
have survived.
Nevertheless, as a part of the system of the critical philosophy, the
Foundations does not have the full merit of the larger work to be ex-
amined here. For it is only in the Critique that all the various strands
of Kant's thought are woven together into the pattern of his practical
philosophy. This pattern, in turn, can be understood only in the entire
fabric of the critical philosophy, and that rich design can be clear only
to those who have understood each of its three principal parts, which
are the three Critique'S and not shorter and more popular works like
the Prolegomena and the Foundations.
The absence of a commentary is a sign, in part an effect but perhaps
in part also a cause, of neglect of the Critique. It is my hope here to
take a step toward removing this cause. The commentary has a two-
fold purpose. First, as a hermeneutic study, it will seek to place the
Critique of Practical Reason in the context of Kant's philosophy and
of the moral philosophy of the eighteenth century insofar as it was the
body of thought to which Kant responded and which he wished to
modify. The Critique was written as a technical, professional work by
a philosopher who saw, and who insisted that his readers should see, the
way in which every part involved the others. Kant's architectonic has
been much derided as a pedantic crotchet of the aging baroque philos-
Foreword Vll
opher. But it is a mistake to believe that what is of utmost importance

to an author can be justly neglected by the reader. Parallels and differ-
ences in content and structure and terminology between this Critique
and the others cannot be passed over in silence if we are to penetrate
below the surface of any of them. In attempting to show how this
treatise stands out against a background often discerned only vaguely
by the reader, I shall have to recall again and again the doctrines of
Kant's contemporaries and the revolutionary theses in other parts of his
own work, knowledge of which was tacitly assumed when he came to
write this Critique. I hope to provide this basic information with suffi-
cient brevity and accuracy not to drive away the Kant-scholar-a man
known in the legends of scholarship for the vastness of his learning and
the shortness of his temper, a trait which has given a notorious acerbity
to Kant-literature for nearly two centuries. But I hope also to be able
to provide it with sufficient fulness and simplicity to throw light on
the text needed by those to whom memory of the first Critique gives,
at best, only wavering illumination and uncertain guidance. I cannot
flatter myself into thinking that I may not at times offend either the
Kant-adept or the beginner; I hope only that if I tire or irritate the
former, I shall not, on the same page, puzzle or discourage the latter.
My second purpose is to examine the contents of this work on their,
philosophical merits. We seem to be going through one of the periods
-the third in English philosophy, the second in American-in which
the study of Kant provides stimulation and sustenance to new and crea-
tive work in philosophy. In the past few years there has been a note-
worthy increase in the amount and improvement in the quality of stud-
ies devoted to Kant in France, England, Italy, and America; there seems
to be a heightening of interest in Kant even in Germany, where the
number and quality of Kant-studies have always been high. It seems as
if a period of thought in which the creative and critical work and
spirit of David Hume are dominant (as in America and England) is to
be followed by one in which Kantian criticism and reconstruction-
perhaps not recognized as such-revive. In this book, therefore, exposi-
tion of Kant may well be combined with philosophic estimation inso-
far as the latter does not require a whole treatise to itself.
But the first task is to find out what Kant said, how he said it,_ and
why. Only then can evaluation have before it a firm object, not an
amorphous mass that varies in shape with the degree of sympathy or
hostility with which it is approached; too often in the past, debates on
whether Kant was right or wrong have been vitiated by lack of the
most rudimentary agreement as to what he actually said and meant.
Vlll Foreword
It is a genuine pleasure to record my thanks to those who made it
possible for me to write this book. The University of Rochester al-
lowed me to be free of all academic duties during 1957-58 so that the
book could be completed. That year was spent in Germany and Italy
under a fellowship from the John Simon Guggenheim Memorial Foun-
dation and with the generous assistance of the Committee on Grants in
Aid for Research of the College of Arts and Science of the University
of Rochester. During the time I was in Germany, I was guest of the
Philosophical Seminar of the Albertus Magnus University in Cologne
and used its extensive facilities. Members of my family and colleagues
gave encouragement, advice, and assistance; each of them will see, on
page after page, evidence of my indebtedness. I am particularly grate-
ful to Professor C. B. Earp, of Wake Forest College, for help with the
eighteenth-century Latin authors.
The Frontispiece is by Heinrich W 01££, and was first published in
the bicentenary memorial volume for Kant issued by the University of
Konigsberg.
L. W.B.
SAN DOMENICO, FIRENZE
Note on the Citations
References to the Critique of Practical Reason are given as follows.

The first number refers to the page in the edition of the Konigliche
Preussische Akademie der Wissenschaften, Berlin, Volume V (ld ed.,
1913), edited by Paul Natorp. The second number, in parentheses or
brackets, refers to the Abbott translation, sixth edition, reprint of 1954.
When no book title is given but only page numbers, they always refer
to these two editions of the Critique. If the reader be annoyed by what
may no doubt appear to be an excessive number of specific page ref-
erences to the Critique, let him be reminded that they have not been
heaped up in this quantity out of any desire to document and redocu-
ment the obvious, but to make Index I, to passages in the Critique,
full enough to be useful to anyone who wishes to survey all my com-
ments on any specific passage.
References to the Foundations of the Metaphysics of Morals refer to
the Akademie edition, Volume IV, with the number in parentheses <1:-
brackets referring to the Abbott translation in the single-volume edi-
tion containing the Critique, as mentioned above.
Since my own translations of the Critique and Foundations, from
which quotations are drawn (with occasional, but not indicated, modi-
fications), include the Akademie pagination, it is not necessary to cite
the pages in these editions for the benefit of those who may wish to
use them in preference to Abbott's.
References to the Critique of Pure Reason indicate pages in the first
edition by the letter "A" and in the second edition by the letter "B."
The translations are those of Norman Kemp Smith, occasionally
modified.
References to the Critique of Judgment, in many cases, cite only the
paragraph number. When it is necessary to indicate a specific page,
the first number given refers to the page in the Akademie edition,
Volume V, and the number in parentheses or brackets refers to the
translation by J. H. Bernard in the edition of the "Hafner Library of
Classics:'
Kant's other works, with very few exceptions, which are clearly
noted, are cited by giving the title or a short title and a paragraph num-
ber or the volume and page in the Akademie edition. If an English
IX
x Note on the Citations
translation is easily available, this is indicated by giving the title in
English and, after the Akademie reference, the name of the translator
and the page in the translation. Full bibliographical data on the trans-
lations thus cited will be found in the Bibliography, Section III.
Occasionally it has been necessary to cite passages which, because
there is not yet an index, I have been unable to find in the Akademie
edition; these references are made self-explanatory.
All references to Kant's Lectures on Ethics, prepared for publication
from students' notes, are exclusively to the translation by Louis Infield.
Kant's correspondence is cited by stating the addressee and date of
each letter and its location in the second Akademie edition, Volumes
X-XIII, and in Volume XXIII, first (only) edition.
Kant's fragments (Reflexionen) are identified only by their number
in the Akademie edition, except those taken from Volumes XX and
XXIII. In these volumes the fragments are not numbered, and they
are cited by volume and page number. For specific places in long,
numbered fragments, I have also added the Akademie pagination.
References to works not by Kant are generally given by short title
with a minimum of bibliographic information. The Bibliography, Sec-
tion IV, supplies the information needed on these titles. Other works
touching on only a specific point and having little or no bearing on
the whole of the present study are cited in the notes with sufficient
bibliographic information. Authors named in the footnotes are listed
in Index, II; authors listed only in the bibliography are not indexed.
Table of Contents
PART I
I THE WRITING OF THE "CRITIQUE OF PRACTICAL REASON" 3

I Introduction 3
1 The Deferred Plan To Write a Metaphysics of Morals 5
3 The Decision To Write the Critique of Practical Reason 13
II THE LIMITS OF THEORETICAL REASON 19
I Introduction 19
2 The Problem of the Critique of Pure Reason 10
3 The "Copernican Revolution" 11
4 Theoretical and Speculative Reason 13
5 The Third Antinomy 14
6 Transition from the Problematic to the Assenoric Judgment
of Freedom 26
III THOUGHT, ACTION, AND PRACTICAL REASON
I
1
Two Aspects of Conduct
Conative and Cognitive Elements in Action
•
3 Practical Reason and Will
4 Pure Practical Reason
IV NAME, PURPOSE, AND STRUCTURE OF THE "CRITIQUE"; CoMMEN-
TARY ON PREFACE AND INTRODUCTION 41
I Introduction 42
1 Title of the Book 41
3 The Purposes of the Book 45
4 The Unity of Theoretical and Practical Reason 47
5 How Is a Synthetic a Priori Practical Proposition Possible? 51
6 Relation to the Metaphysics of Morals 53
7 The Structure of the Book 55
8 The Polemics of the Preface 56
PART II
V SURVEY OF THE ANALYTIC OF PRACTICAL REASON 6S

I Meaning of the Tenn "Analytic" 65
2 The Structure of the Analytic 66
Xl
""
Xll Table of Contents
3 The Problems of the Analytic 68
4 Division of the Commentary on the Analytic 69
5 Summary of Chapter I, §§ 1-8 70
VI THE ANALYTIC OF EMPIRICAL PRACTICAL REASON. I. FORMAL CON-
SIDERATIONS; COMMENTARY ON § 1 75
1 Empirical Practical Reason 75
2. Practical Principles 76
3 Rules 79
4 Maxims 80
5 Are There Empirical Practical Laws? 82

6 Hypothetical Imperatives 84
7 Conclusion 88
VII THE ANALYTIC OF EMPIRICAL PRACTICAL REASON. II. MATERIAL
CONSIDERATIONS; COMMENTARY ON §§ 2,3, AND PART OF 8 90
I Introduction 90
2 Desire 90
3 Pleasure 92
4 The Lower Faculty of Desire 94
5 Maxims of the Lower Faculty of Desire 95
6 Happiness 97
7 Egoism 100
8 Heteronomy 102
VIII THE "METAPHYSICAL DEDUCTION" OF THE MORAL LAW; COMMEN-
TARY ON §§ 4. 5, 6, 7, AND PART OF 8 109
I The Idea of a "Metaphysical Deduction" 109
2. Principal Features of the Moral Experience 112
3 First Approximation to the Principle 116
4 The Three Maxims in Moral Decision 117
5 Definition of the Principle 121
6 Autonomy 122
IX PRACTICAL CONCEPTS AND JUDGMENT; CoMMENTARY ON ANA-
LYTIC, CHAPTER II 126
I Introduction 126
2 Principle, Concept, and Judgment 12 7
3 The Concepts of Good and Evil U9
4 The Categories of Practical Reason 13 6
5 The Table of the Categories of Freedom 144
6 The Categories of Quantity 145
7 The Categories of Quality 146
8 The Categories of Relation 147
9 The Categories of Modality 149
10 The Typic of Pure Practical Judgment 154
Table of Contents Xlll
X THE "TRANSCENDENTAL DEDUCTION" OF THE PlUNCIPLE OF PURE

PRACTICAL REASON; CoMMENTARY ON S7 AND Ax., 41-50 (AB-
BOTT,13 1-40) 164
1 Introduction 164
1 The "Fact of Pure Reason" 166
3 The Deduction 17°
XI FREEDOM 176
Introduction 176
1 Two Conceptions of Will and Freedom 17 6
3 Kant's Recapitulation of the Argument of the First Critique 181
4 Freedom as a Theoretical Idea 18 3
5 Resolution of the Third Antinomy 186
6 Practical Freedom 188
7 Criticism of This Doctrine 19 1
8 Freedom as an Actor-Concept; Spontaneity 194
9 Freedom as Supreme Legislation; Autonomy I¢
10 The Synthesis of the Two Conceptions 198
II Summary 101
12 Moral Evil 103
13 Freedom and Creation lOS
14 Freedom as a Postulate 107
XII THE "AESTHETIC" OF PORE PRACTICAL REASON; CoMMENTARY ON

ANALYTIC, CHAPTER III; PART OF DIALECTIC; AND METHODOLOGY 109
1 The Psychological and the Ethical Problem
2 A Different but Related Problem
, 209
lU
3 Some Earlier Attempts at an Answer 21 3

4 Incentives 216
5 Desire and Pleasure 117
6 The Genesis of Respect 119
7 An Alleged Inconsistency 221
8 The Moral Feeling 21 3
9 Duty and Personality 12S
10 Virtue 227
II Moral Satisfaction 21 9
11 Moral Education 233
PART III
XIII THE DIALECTIC OF PURE PRACTICAL REASON; CoMMENTARY ON

DIALECTIC, CHAPTERS I AND II (EXCEPT SEcs.IV AND V) 139
1 What Dialectic Is 139
z The Dialectical Illusions of Pure Practical Reason 14°
XIV Table of Contents
3 The Concept of the Highest Good 141
4 The Antinomy of Pure Practical Reason 145
5 The Primacy of Pure Practical over Theoretical Reason 149
6 The Postulates of Pure Practical Reason 151
7 Belief and Ignorance lSS
XIV THE POSTULATES OF PURE PRACTICAL REASON; CoMMENTARY ON
DIALECTIC, CHAPTER II, SECTIONS IV AND V; AND CoNCLUSION 159
I Introduction 159
1 The Structure of a Moral Argument :&60
3 Ideas and Postulates :&~
4 The Immortality of the Soul :&65
5 The Existence of God 27 1
6 Morality and Religion :&79
7 The Two Awesome Things 181
BIBLIOGRAPHY
I TEXTS OF "KRITIK DER PRAKTISCHEN V ERNUNFT" 187

II TRANSLATIONS OF THE "CRITIQUE OF PRACTICAL REASON" 187
III TRANSLATIONS OF OTHER WORKS BY KANT AS CITED IN CoMMEN-
TARY 188
IV STUDIES OF KANT 188
INDEXES
I INDEX OF PASSAGES CITED FROM "CRITIQUE OF PRACTICAL REASON" 197

II INDEX OF NAMES AND SUBJECTS 300
Part I
I
The Writing of the

~~CritUjue of Practical Reason"
§ I. INTRODUCTION
Few substantial philosophical treatises have been written in the haste in
which the Critique of Practical Reason was composed. The book, how-
ever, shows few of the signs of hurry that marred some of Kant's other
works. Indeed, the short time he devoted to thinking it out as a single
self-contained book and to its actual writing may be responsible for a
virtue singularly lacking in the Critique of Pure Reason, which he med-
itated on for twelve years and "brought to completion in some four or
five months,"l presumably by making more or less judicious use of
manuscripts composed at various times during the twelve-year period.
For, unlike the first Critique, the second adheres to one single straight
line of argument, developed without exploring blind alleys and with-
out getting almost lost on tangents. In it, we do not, as it were, listen to
Kant thinking aloud or watch him trying first this and then that way
of making himself understood. The book has the magisterial tone and
direct style that come to a man who has thought out all he wants to say
before he puts one word on paper.
By Kantian standards, at least, it is an exceedingly well-written book.
To many, this may seem damning with the faintest praise possible. Cer-
tainly Kant's style is not to everyone's liking, not even to his own. But
few men have had juster estimates of their own style than Kant had of
his, and his decision not to try to write in a popular way was justified
by the nature of the materials he was dealing with and by the demands
he rightly made of him who ventures to read philosophy.2 And it can
1 To Garve, August 7, 1783; to Mendelssohn, August 16, 1783 (X, 338,345). The
last pan of this sentence does not commit one to the "patchwork theory" (cf. H. J.
Paton, Kant's Metaphysics of Ezperience [London, 1936], I, 41).
2 For an explicit caveat, see Prolegomena, end of Introduction. In an undated
note <504"), he says: "If, like Hume, I had the power to embellish my work, I would
hesitate to use it. It is true that some readers will be frightened away by its dryness.
But is it not necessary to frighten some of them away, with whom the matter would
come into bad hands?"
3
4 Tbe lVriting of the "Critique of Practical Reason"
be excused, if an excuse is needed, by the urgency which marked his
literary activity in his old age. The fastidious reader should remember
that all the best-known works of Kant were written in excusable haste
by an old man who had first discovered his message to the world late
in life and who was never robust but carried (by modern university
standards) a crushing load of academic duties.
Given the circumstances, what is surprising is how well the book is
written. Style cannot be considered apart from content, and the
strongly masculine, sachlich, and encyclopedic style of Kant's thinking
permits and perhaps requires3 the often condemned Kantian style of
writing, which Schopenhauer called "brilliantly dry." It does not ex-
clude occasional eloquence and poignancy of expression, and it never
allows the tone to be less than elevated. Unfortunately, however, these
high virtues do not entirely exclude the vices of pedantic distinctions
followed by carelessness in their use, overcondensation 4 alternating
with repetitiousness, and lack of the art of perspicuous paragraphing.
These weaknesses cannot be denied, but they are frequently exag-
gerated by philosophers who write no better. It is often remarked by
Kant-scholars that when, after dogged effort, they come to the root of
some perplexing sentence or paragraph, they find that the complexities
of the writing cannot always be attributed to the ineptitude of the
writer, but more often to the demands of the thought expressed. They
often realize that it would be very difficult, if not impossible, to say
better exactly what Kant was saying and that any simplification of style
would almost inevitably involve oversimplification of the thought. At-
tempts at simplification have usually produced only faint and distorted
echoes of Kant's meaning, and most of them, if they succeed in making
him seem simple, do so only by making his views seem silly.
Except in a few passages, simplification of style is not needed in the
Critique of Practical Reason. There are few passages in which the at-
tentive reader will experience any real difficulty in discerning what
Kant means. Diagramming of sentences, often said in the legends of
3 Cf. Critique of Pure Reason, A xviii-xix.
4 Kant could appreciate a joke about his own style, at least this aspect of it. Zeiter
told the following anecdote to Goethe and said that Kant had been amused by the
event reported. Wlomer, a banker, told Kant that he had read some of his works
and would have read more if he had had more fingers. "How so?" inquired the
philosopher. "Well, my friend," replied the banker, "your way of writing is so rich
in parentheses and conditional clauses which I have to keep my eye on that I put
one finger on this word, a second on that, and so on for the third and fourth. Before
I get to the end of the page all my fingers are used up" (Vorlander, Immanuel
Kant: Der Mann und das Werk, II, 99).
The Writing of the "Critique of Practical Reason" 5
scholarship to be prerequisite to understanding the Critique of Pure
Reason, is not called for here.
We do not know much about the actual writing of the book. But in
the light of evidence about to be adduced, it is probable that the book
was begun in the spring of 1787, brought nearly to completion by June,
and finished by September. I speak only of the actual writing; most of
the contents of the book had been clear in Kant's mind at least as early
as 1785. But the plan of writing a Critique of Practical Reason as a
single, integral work did not slowly mature. The book was not long
anticipated, and Kant came suddenly, and presumably with reluctance,
to the decision to write it.
In order to understand this decision, it is necessary to go back to-
ward the beginning of Kant's literary career to see what the book was
that he did want and plan to write.
§ 2. THE DEFERRED PLAN TO WRITE A METAPHYSICS OF MORALS
Kant's writings in the 1750'S and 1760's were almost all in the field of
the natural sciences. But his interest in them was genuinely philosophi-
cal, and he was concerned most of all with questions of the method and
scope of the sciences. Even in these works, however, when occasion
offered, he commented freely upon ethical and religious questions
which had become involved in the cosmological speculations of the
day. This direction of his thinking is especially clear in his most impor-
tant scientific work, the Universal Natural History and Theory of the
Heavens (1755)'
The so-called Prize Essay: An Inquiry into the Evidence of the Prin-
ciples of Natural Theology and Morals, published by the Berlin Acad-
emy in 1764, is the first of his works which deals more than inciden-
tally with questions of moral philosophy. Even here, as is seen from its
title, it was the foundations and methods of ethics that were in the cen-
ter of his interest. This essay, often interpreted as manifesting a com-
mitment to the moral-sense theory of Shaftesbury and Hutcheson, and
hence to empiricism in ethics, ends with a query the answer to which
gives direction to all his subsequent work in moral philosophy. He says
there that "it is still a question to be settled whether it is simply the
cognitive faculty or whether it is feeling ... which decides the basic
principles of practical philosophy."1S
Metaphysics is regarded in the Prize Essay as a "philosophy of the
ultimate grounds of our knowledge," as philosophy "applied to the
more universal rational insights." While rejecting the alleged identity
15 Prize Essay, 11,300 (Beck, 28 s).
6 The Writing of the "Critique of Practical Reason"
of the methods of mathematics and those of metaphysics-an identity
almost definitive of the dominant metaphysics of the time-Kant be-
lieved that metaphysics, following its own method of analyzing experi-
ence, could find indemonstrable propositions as certain as those of
mathematics. 6 The basic indemonstrable propositions of metaphysics
would be self-evident rational principles, though not definitions or in-
tuitively certain axioms as in mathematics. He thought metaphysics to
be possible as a rigorous science, and already to exist in natural theol-
ogy. The metaphysical foundations of morals, however, were yet to be
found: "The primary grounds of morals are not yet, in their present
state, capable of all requisite evidence."7
Kant must soon have felt capable of supplying this lack. In a letter to
Lambert the next year, he announced a fateful plan that, in one form
or another, was to attract, even almost to haunt, him throughout his
mature life. It was the plan to write a work "on the metaphysical foun-
dations of practical philosophy.us The title of this work and a report on
his progress with it were given in a letter to Herder in 1768, in which
he said that he was then at work on a "Metaphysics of Morals" and that
he hoped to complete it within a year.'
We can only speculate on what would have been the contents and
structure of this work, had it been completed at that time. But we do
not need to make any conjectures about it for our present purposes.
61t was in this spirit that he favorably compared Rousseau to Newton, for both
discovered "the hidden law the observation of which justifies providence" (Bemerk-
ungen zu den Beobllchtungen, XX, 58-59). Similarly, he esteemed the work of
Shaftesbury, Hutcheson, and Hlime as containing the method by which "the abiding
nature of man" can be discerned in the variety of his empirical appearances. This,
he said, is "an excellent discovery of our days," and accordingly he promised to
evaluate historically and philosophically what happens before showing what ougbt
to happen (Nllcbricbt von der Einricbtung seiner Vorlesungen in dem Winterbalb-
jabr von 176$-1766, II, 311). Each of these passages indicated the analytical method
he was to use, beginning with observations and rising by abstraction from empirical
to metaphysical concepts. The passage in the announcement of his lectures, however,
does not indicate that Kant at this time thought a descriptive, merely empirical,
ethics to be possible, though it has been interpreted as evidence for this by some
writers.
7 Prize Essay, n, z9B (Beck, z8z).
8 To Lamben, December 31, 1765 (X, 56).

'To Herder, May 9, 1768 (X, 74). The book itself was announced as fonhcoming
by the publisher Kanter, under the tide "Critique of Moral Taste." The term "meta-
physics of morals" was little used before Kant. It seems to have originated with
Canz, DisciplilUle morales omnes ([739), according to Max Wundt, Die deutscbe
Schulpbilosopbie im Zeitalter der Aufklarung (Tiibingen: Mohr, (945), p. zZ3. Cf.
also ibid., p. Z5I, on A. F. Hoffmann's conception (Vernunftlehre [1737]) of meta-
physics and moral philosophy, which was closer to Kant than to Wolff.
For though Kant's thoughts on ethics were undoubtedly undergoing
changes during this period, the most important change at this time oc-
curred in his view of the nature of metaphysics itself. His views on
this, unlike his ethical views at this time, can be documented and de-
scribed with confidence.
These changes were clearly shown in his Inaugural Dissertation: The
Forms and Principles of the Sensible and Intelligible World, published
in 1770 on the occasion of his accession to the professorship in Konigs-
berg. This is a brief systematic treatise on metaphysics. It not only dis-
tinguished but also sharply separated the sensible and the intellectual
elements in knowledge, and it discussed the "real use" of the intellect in
metaphysics whereby truths are discovered, contrasting it with the
"logical use" in merely drawing inferen~es from given judgments. It
contained a clear, though later surrendered, claim on the proper method
and realm of metaphysics as knowledge of an intelligible world, with
nothing borrowed from sensible knowledge of phenomena. Metaphys-
ics, as pure knowledge dealing with concepts not derived from experi-
ence but "given by the pure intellect itself," was held to be possible
and valid only when pursued according to a rule for preventing the
principles of sensibility from "passing their boundaries and meddling
with the intellectual."lo Such a rule, of course, depended upon a clear
and systematic discernment of the roots of our concepts-a discernment
that he believed he had achieved and which he never seriously revised,
even when he later concluded that metaphysics, as envisioned in this
work, was impossible.
Metaphysics of morals thereafter took on a very different look.
Whereas "metaphysics" had previously meant hardly more than the
most general conceptual knowledge issuing from an analysis of experi-
ence, it now came to mean systematic philosophy containing nothing
empirical but referring to a world beyond experience. l l Metaphysics is
knowledge of things as they are, and concepts of the appearances of
things, i.e., space and time, do not belong within it. A metaphysics of
morals, therefore, could not be a continuation of the empirical-anthro-
pological investigations of Shaftesbury; thenceforth Kant invariably
made clear the independence of metaphysics of morals from all anthro-
pology, even from "pragmatic anthropology," which deals with how
men should conduct themselves in the ordinary affairs of life. Rather,
metaphysics of morals had to have its basis in a Platonism of a non-
phenomenal world. "Moral philosophy, so far as it supplies the first
principles of moral judgment, is known only through the pure under-
10 Inaugural Dissertation, ~ %4 (Handyside, 73).
11 Cf. Critique of Pure Reason, A 843 = B 871.
8 The J-Vriting of the "Critique of Practical Reason"
standing, and itself belongs to pure philosophy," i.e., metaphysics.1 2 It
contains no empirical concepts of human nature.
That he was satisfied with this new path to metaphysics and with its
goal, at least for a while, is shown by a letter of the same year to Lam-
bert, in which he says he planned "this winter to complete my investi-
gations concerning pure moral philosophy and ... the metaphysics of
morals, "13 in which there would be no empirical principles. Contrary
to this fond hope, however, the Inaugural Dissertation marked the be-
ginning of the "silent decade" in which he published almost nothing
but during which the herculean labor of writing the Critique of Pure
Reason was performed, while the "Metaphysics of Morals," still re-
peatedly promised, was delayed again and again.
Less than a year later he wrote Marcus Herz that he was at work on
a book to be titled "The Boundaries of Sensibility and Reason"-the
book the world now knows as the Critique of Pure Reason. The book
as then planned was to contain not only a theory of appearances ("Phe-
nomenology") but also the essentials of a theory of morals, taste, and
metaphysics. 14 On February 21, 1772, he described, again to Herz, his
plan for this book. It was to contain a section on metaphysics, one part
of which was to give "the ultimate grounds of morality," and it was to
be published within three months. 15 In 1773, in still another letter to
Herz, he announced his plan to complete "my transcendental philoso-
phy, which is really a critical examination of pure reason,"16 and then
"to get to metaphysics, which has only two parts, the metaphysics of
nature and the metaphysics of morals, the latter of which I shall finish
first, and towards which I am looking forward with pleasure."17
The letter of I 772 shows Kant's dissatisfaction with the method of the
Dissertation in establishing the possibility of a priori knowledge of a
12lnauguraJ Dissertation, § 9 (Handyside, 49). Moral concepts, though confused,
are not sensible or empirical but are known by the pure intellect itself. Earlier
rationalistic philosophers had thought of "sensible" and "confused" as corollaries in
descriptions of modes of cognition, and the obscurity of ethical principles had
therefore led them to combine, in a most astonishing manner, empiricism in ethics
with rationalism in their philosophy as a whole. Kant shows in the Dissertation
that this confusion no longer troubles him, that he has now outgrown it, if he
ever had fallen into it.
18 To Lambert, September 1,1770 (X, 97).
U To Herz, June 7, 1771 (X, (13).
15 To Hen, February 11, 1771 (X, 131),
16 Literally, "Critique of Pure Reason," though probably these words were not
meant then as a title of the book.
17 To Herz, end of 1773 (X, 144).
purely intelligible world; but there was not yet any sign of his having
to give up the hope and expectation of being able to establish specula-
tive metaphysics on a solid ground. After the letter of 1772, however,
Kant's "recollection of Hume" awakened him from his "dogmatic
slumber," and he gave up forever the dream of a speculative metaphys-
ics of the intelligible world. Still, the transcendental philosophy or
critical examination of pure reason, as elaborated in the Critique of
pure Reason, did provide the prerequisites, in Kant's opinion, for a
"Metaphysics of Morals" in two senses: a systematic presentation of the
a priori laws of morality, and a practical rather than a speculative
answer to traditional metaphysical questions.
So in the Critique of Pure Reason, as it was at last published in 1781,
Kant still spoke of the metaphysics of morals. He distinguished two
parts of philosophy: critique, or propaedeutic philosophy, which "in-
vestigates the faculty of reason in all its pure a priori knowledge," and
metaphysics, or the "system of pure reason" which "exhibits in sys-
tematic connection the whole body . . . of philosophical knowledge
arising out of pure reason." "M:etaphysics," he continued, "is divided
into that of the speculative and that of the practical employment of
pure reason, and is therefore either metaphysics of nature or metaphys-
ics of morals."l8 The entire faculty of reason, including the practical,
was subjected to critique; that is, the Critique of Pure Reason was
meant to be propaedeutic to metaphysics in both its divisions. 19
In spite of mentioning the metaphysics of morals as a division of pure
philosophy, in the first edition of the Critique of Pure Reason Kant did
not promise to write such a work; he projected only a "Metaphysics
of Nature."20 Transcendental philosophy, which is the system of all
principles of reason so far as it concerns knowledge of objects, con-
tains only pure a priori concepts and principles, and thus it excludes
ethics. For though the highest principles of ethics were said even then
to be a priori and intellectual, "in the construction of a system of pure
morality empirical concepts must necessarily be brought into the con-
cept of duty."21
18 Critique of Pure Reason, A 841 = B 869.
19 Ibid., A xii, note.
20 I bid., A xxi.
21 Ibid., B 29. The Critique of Pure Reason excludes moral philosophy from
transcendental philosophy, not because the former is not pure, but because the
latter is concerned solely with the cognitive (cEo A 801 = B 829). Kant tacitly
widened the concept of transcendental philosophy (as he narrowed that of meta-
physics) until it is hardly distinguishable from critique itself; and though he never
explicitly included moral philosophy in transcendental philosophy, we shall see the
Instead of a "metaphysics of morals," therefore, we have in the Cri-
tique of Pure Reason a "Canon of Pure Reason," i.e., a statement of the
a priori principles of the correct employment of pure reason. This em-
ployment is entirely and exclusively practical, not speculative. The
Canon does not answer the practical problem as such, viz., "What ought
I to do?" but the problem he calls both theoretical and practical, to wit,
"What may I hope if I do my duty?" Thus was introduced the discus-
sion of the immortality of the soul and the existence of God, which are
Ideas of pure theoretical reason, dialectical and empty for theoretical
cognition. It is only in relation to man's will that these Ideas have any
necessary use, as guides or regulative principles for the pursuit of hap-
piness through becoming worthy of it. 22
Again, Kant did not long remain satisfied with a mere Canon of Pure
Reason. He soon returned to the long-deferred plan, not mentioned in
the Critique itself, to write a systematic work on ethics, presumably to
include a metaphysics of morals, for he referred in 1783 to being at
full apparatus of transcendental philosophy in discussions included in the Critique
of Practical Reason.
The tacit inclusion is due not merely to a widening of the scope of transcendental
philosophy, however; it is also to be attributed to a deepening of the level of
moral analysis itself. Thus the passage JUSt quoted says that in a system of moral
philosophy empirical concepts will have to be brought into the concept of duty.
whereas in A 15, the corresponding passage in the first edition, he had said, with
less caution, that "the concepts of pleasure and pain, and of desires and inclinations,
etc. will have to be presupposed." In both editions he distinguished between the
doctrine of virtue, or applied ethics, and pure ethics, the fonner of which alone is
in part dependent on empirical or psychological principles (A 54-55 = B 79). The
Critique of Practical Reason and the Foundations of the Metaphysics of Morals are
meant, of course, to be pure ethics in the sense of this passage, even though in the
Critique he makes use of definitions of psychological concepts needed in the concept
of duty and in depicting the relation of man to the moral law, which requires no
psychological concepts for its fonnulation. But in actual composition of the second
Critique and even of the Metaphysics of Morals, as we shall see (below, p. 53),
Kant made no very consistent effort to separate discussion of pure from applied
ethics, or metaphysics from critique, or either from system.
We shall see that the ascribing of something very like, if not identical with,
transcendental status to practical principles was one of the factors leading to the
decision to publish the second Critique on the same level with the first and not
subordinate to it, as a metaphysics of morals would have been subordinate to it.
Later Kant came to regard the first Critique as not a sufficient propaedeutic to all
metaphysics, and he indicates this implicitly in constantly referring to the first
Critique not by its correct title but by such expressions as "the critique of specu-
lative reason," as in the first paragraph of the second CTitique (3[87]).
22 Critique of Pure Reason, A 806 = B 834. It is not quite accurate to say that this
is their only use; for they are also regulative of inquiry, but they contribute nothing
substantively to knowledge.
The Writing of the "Critique of Practical Reason" I I
work on "the first part of my ethics."23 We do not know whether he

was referring to the Foundations of the Metaphysics of Morals or not.
But when the Foundations was published in April, 1785, he again men-
tioned his plan, now twenty years old, for a "Metaphysics of Morals"
to which the Foundations was only preliminary.24
The Foundations differs sufficiently from the Canon of Pure Reason
to make it easy to explain Kant's decision to write the Foundations as
still another propaedeutic to the ever receding "Metaphysics of Mor-
als." The Canon presented a view of moral problems without the con-
cept of autonomy and independent of the solution of the problem of
freedom that had already been worked out in the first Critique-the
twO foundation stones on which all his later work in practical philoso-
phy was to be based.
Fragment 6 of the Lose Bliitter,2fS which I think must have been writ-
ten between ]78] and I78~ shows the transition to the new position in
asking of practical judgments the question that the Critique of Pure
Reason raised with respect to theoretical judgments: How are synthetic
judgments a priori possible? For he realized at this time that morality
requires synthetic a priori judgments, that these judgments cannot be
justified in exactly the same way that their theoretical counterparts had
been justified, and that to justify them requires a more positive concept
of freedom than that used in the Canon. Justification of synthetic a pri-
ori judgments everywhere in Kant requires what he invariably calls a
"Deduction," not a canon of use. And the proper place for a deduction
is a critique.
But, instead of a critique, in ] 785 Kant presented the Foundations of
the Metaphysics of Morals. And yet:
There is. to be sure. no other foundation for such a metaphysics of morals
than a Kritik of pure practical reason, just as there is no other foundation
for a metaphysjcs than the already published Kr;t;k of pure speculative
reason. But. in the first place. a Kritik of pure practical reason is not of such
extreme importance as that of speculative reason. because the human rea-
son, even in the commonest mind. can easily be brought to a high degree
of correctness and completeness in moral matters, while, on the other hand.
in its theoretical but pure use it is entirely dialectical. In the second place,
I require of a Kritik of pure practical reason, if it is to be complete, that
the unity of the practical reason and the speculative be subject to presenta-
tion under a common principle. because in the final analysis there can be
but one and the same reason, which must be differentiated only in applica-
23 To MendelsSohn, August 16,1783 (X, 346-47).
,
24 Foundations of the Metaphysics of Morals, 391 (6).
2fS Reflexion 7202 ,

12. Tbe Writing of tbe "Critique of Practical Reason"
tion. But I could not bring this to such completeness without bringing in
observations of an altogether different kind and without thereby confusing
the reader. For these reasons I have employed the tide, Foundations of the
Metaphysics of Morals, instead of Critique of Pure Practical Reason.26
Notwithstanding this, the third section of the Foundations is entitled

"Final Step from the Metaphysics of Morals to the Critical Examina-
tion [Kritik] of Pure Practical Reason." Presumably all that he re-
garded as essential in a critical examination of pure practical reason was
given here, for in a letter to Schlitz written five months later he said
that he would immediately begin writing his "Metaphysics of Morals."27
Yet it appears a bit odd that a metaphysics of morals has been devel-
oped in Section II of the Foundations and that a transition is made from
it to a "Critical Examination of Pure Practical Reason." The explana-
tion of it is this. In the first and second sections of the Foundations,
Kant's method had been analytical, i.e., he began with the "moral facts"
and analyzed them to find what must be true if they were not illusory.
The moral imperative and the esteem for the good will are chimerical
unless the thought of duty can be a motive to action; and it can be so
only if the will is free. But is the will free? This is a question which can
be answered only by a synthetical method, i.e., a deduction of its pos-
sibility. Only this can justify any categorical statements about morality;
without it, all our statements must be hypothetical and problematic.
Hence the Foundations is, as a whole, propaedeutic to a "Metaphysics
of Morals," even though the internal structure which a metaphysics of
morals must have is achieved long before the end of the book; the last
section is aimed at justifying the assertion of what was only entertained
in the first two sections.
Some scholars28 have offered the hypothesis that at this time Kant
planned next to write his "Metaphysics of Morals" and then a "Critique
of Pure Practical Reason," which would complete his system by prov-
ing what neither the Foundations had shown nor what a yet-to-be-
written "Metaphysics of Morals" would show, to wit, the unity of the-
oretical and practical reason. But the evidence for this hypothesis seems
to me to be utterly unconvincing. It seems much more likely that the
Foundations constituted, in Kant's mind, precisely what its title indi-
cates and that no other "Critique" was planned. Even as late as April,
1787, when the Critique of Pure Reason had been revised, Kant spoke
26 Foundations of the Metaphysics of Morals, 39 1 (~7).
21 To Schiitz, September 13, 1785 (X, 406).

28 E.g., Paul Natorp, in the Akademie edition notes to the Critique of Practical
Reason (V, 496), and A. R. C. Duncan, Practical Reason and Morality, pp. :lJ. 35, I3:l.
of saving time for his "proposed scheme of providing a metaphysics of
nature and a metaphysics of morals"29 and makes no mention of the
work that was actually to occupy him that summer-the writing of the
Critique of Practical Reason.
§ 3. THE DECISION TO WRITE THE "CRITIQUE OF PRACTICAL REASON"
In 1786 a preliminary work on the metaphysics of nature (Meta-
physische Anfangsgrunde der Naturwissenschaft) was published, and
Kant turned to the heavy labor of preparing the second edition of the
Critique of Pure Reason. It was begun by April, 1786, and completed
by April, 1787. With his usual optimism in such matters, he had esti-
mated that the revision would take six months. He told Bering that his
"System of Metaphysics" would be delayed by two years, because he
was also planning a "System of Practical Philosophy" to be published
before the book on metaphysics proper.so
What was this "System of Practical Philosophy"? We do not know,
but we may presume that it was the "Metaphysics of Morals," so often
promised, so long postponed. Certainly it was not the Critique of Prac-
tical Reason; Kant almost always distinguished between the words
"critique" and "system" and "metaphysics," even though their scopes
in fact overlap to a large extent. So, as late as April, 1786, we know that
the Critique of Practical Reason was not planned as such.
We first hear of a "Critique of Pure Practical Reason" as a specific
literary project on November 8, 1786. Born, in a reply to a letter from
Kant not now extant, spoke of the new work as an addition to the Cri-
tique of Pure Reason.3t The Allgemeine Literaturzeitung (Jena) on
November 2 I, in an announcement of the future publication of the sec-
ond edition, said: "To the Critique of pure speculative reason contained
in the first edition, in the second there will be appended a critique of
pure practical reason."32
29 Critique of Pure Reason, B xliii. In the Preface to A, only a "metaphysics of
nature" was promised.
so To Bering, April 7, 1786 (X, 441). Bering. in his reply of May 10 (X, 445)
regretted the delay and expressed the wish that he could talk with Kant. "Perhaps
soon," he wrote, "our aeronauts will make their trips less expensive and dangerous.
and then a trip of 140 miles [sic-from Marburg to Konigsberg] will be a trivial
thing."
81 Born to Kant, November 8, 1786 (X, 471).
32 The notice is reprinted in Ak. III, 556. That this infonnation came from
Kant himself, or at least that it did not originate with the editor (Schutz). is made
clear in Schlitz's letter to Kant of November ], 1786 (X, 469). Kant had written
to Born and Schlitz on the same day (May :6) and may have mentioned his plan
to them at that time, for Born clearly had the infonnation before it was published
by Schutz. Unfortunately, the letters of May :6 are not extant.
During 1786 and 1787, therefore, Kant must have entertained at vari-
ous times the following plans: (a) to write a "Metaphysics of Morals"
based on the Critique of Pure Reason and the Foundations; (b) to write
a "System of Practical Philosophy" as soon as he had completed the re-
vision of the Critique of Pure Reason; (c) to append a "Critique of
Pure Practical Reason" to the new edition of the Critique of Pure Rea-
son; (d) to write the Critique of Practical Reason, as we know it today.
The first two projects were no doubt substantially identicaJ.33 What in-
terests us is the shift from a and b, considered together, to c, and the
final decision to go forward with d. As we shall see, Kant did not go
directly from c to d.
The step from b to c represents the magnitude of development in
Kant's views from the Canon of Pure Reason to the Foundations. The
Critique of Pure Reason, when it was written, was regarded as a pro-
paedeutic to both divisions of metaphysics; but by 1785 the proper
basis of a metaphysics of morals was located in the concept of auton-
omy, a concept not so much as mentioned in the first Critique. But plan
c was dropped, not at first ill favor of d but because of a return to a or
b, as shown in the Preface to the second edition of the Critique of Pure
Reason.
Some of the reasons for dropping it must have been external. The
success of the Foundatio1ls had created a demand for a new edition of
the first Critique, and there was considerable urgency in meeting it. By
November, 1786, the "six months" that Kant had estimated for the re-
vision had already elapsed, and the work was still six months from com-
pletion; and, even so, all the revisions were in parts which have little or
no direct bearing on questions of practical philosophy. He left un-
touched the rest of the Dialectic (with the exception of the Paralo-
gisms) with the surprising explanation that he had not found any serious
misundersanding of the other parts34-misunderstandings which he
found in plenty when he came, in 1787, to deal with the critics who
had charged that the Critique and the Foundations were incompatible.
The differences between the teachings of the Canon and of the Foun-
33 The twO divisions of work, when it was no longer a question of two different
literary projects, were later distinguished from each other in the statement that a
"system of practical philosophy" would contain anthropological data and would
presuppose a metaphysics of morals, which would take nothing from anthropology
(Metaphysik der Sitten, Einleitung, § ii [VI, 2[6-17]). In actual execution, however,
the Metaphysik der Sitten is more like the projected "system." The manifold
changes in Kant's conception of what "metaphysics of morals" should contain are
traced by Georg Anderson, "Kant'S Metaphysik der Sitten-ihre Idee und ihr Ver-
halmis zur Ethik der Woltfschen Schule," Kant-Studien, XXVIII (1913), 41-61.
3. Critique of Pure Reason, B xli.
datio1lS were too great to add an architectonic paraphrase of the latter
to the former; there would have had to be extensive rewriting as well"
and the architectonic structure of the Methodology of the first Cri-
tique would have been completely destroyed. His solution of this prob-
lem was to provide a new and extensive Preface to the second edition,
which showed his thinking on the ethical problem, while leaving the
Canon unchanged.
Furthermore, the original Critique was too long to add another major
work to it, even if it could have been fitted into its shape; the revisions
he did make added another thirty to its original 850 pages. Again, the
winter of 1786-87 was a time when Kant was rector of the University,
and on this occasion, at least, this was not simply a sinecure or position
of honor. There were the ceremonies attendant upon the death of Fred-
erick the Great in August, 1786, and the accession of Friedrich Wil-
helm II later in the year, in all of which Kant had a major role to play.35
All these facts help explain Kant's desire to restrict the revisions of the
Critique of Pure Reason to the absolutely essential, and we know his
impatience to get on with plan a or b, now deferred nearly twenty
years.
We may then suppose that the plan to write a separate Critique of
Practical Reason as a separate work was formed later than April, 1787,
the date of the Preface to the second edition of the first Critique, re-
placing the plan mentioned there for going ahead with a "Metaphysics
of Morals." What reasons led to this final, fateful, decision?
I shall suggest several. The first was Kant's acknowledgment, even in
1785, that a complete critical examination of practical reason was not
given in the Foundations. Two fundamental topics were not dealt with
in that work, and at least one of them was a topic both intrinsically im-
portant and fascinating to Kant's turn of mind: the proof of the ulti-
mate unity of theoretical and practical reason. The second was the con-
nection between the moral law, applicable to rational beings in genera~
and man, a connection not to be based upon anthropology but on an
a priori connection that Kant discerned between will and feeling. This
transition to specifically human reason was adumbrated in the Founda-
tions, but it was essential to plans a and b that it be fully elaborated.
Thus arose the necessity for what we now have as chapter iii of the
Analytic of the second Critique.
Another must have been the desire, natural to Kant and welcome to
his audience, to develop more systematically his concept of "the key-
III It should perhaps be remembered also that ] oseph Green died in 1786, and
Kant's earliest biographers, e.g., ]achmann, tell us how much this event affected the
philosopher and the daily course of his life.
16 Tbe Writing of tbe "Critique of Practical Reason"
stone of the whole architecture of pure and even speculative reason"
(3 [88]), i.e., the concept of freedom. This concept had been estab-
lished only as a possibility, i.e., as not self-contradictory, in the first
Critiqz.e; discussion of it in its full depth was ruled out as being unnec-
essary in practical philosophy as such; and it had been explored, but not
systematically, in the third section of the Foundations; but it needed
full encyclopedic and critical treatment.
Fourth, there was the objection that in the Foundations Kant had
gone against the injunctions made against speculative knowledge in the
first Critique. This was a charge made by a man he respected (Pisto-
rius), which, if valid, threatened the entire critical philosophy, in both
its theoretical and its practical parts. It was a charge that had to be met
at all costs. 36 The method of postulates, used in the Canon, easily led to
such criticisms; but the development of an antinomy and its successful
resolution was always Kant's way of exposing illicit pretensions, and he
could not renounce the opportunity to develop an antinomy in the
concept of the highest good as a way of showing that he was not, in
fact, advancing speculative claims.
Very closely related to all these reasons was Kant's desire to answer
the critics who had raised still other objections, before building on
foundations that had been widely challenged.
Finally, I think Kant saw an opportunity to strengthen the convic-
tion that he had sought to establish in the first Critique by writing an-
other book, from another point of view, which would lead to some of
the same conclusions by a different route. The second Critique is not a
continuation of the first, though the Preface to it may make the reader
forget that it is not. The second Critique made an entirely fresh begin-
ning in another realm of experience; and Kant warned repeatedly
against self-consciously trying to avoid discussions already completed
in the first and against keeping the first so much in mind that the natu-
ral path of the second would be affected by extraneous considerations
(7, 106 [92, 201]). The first point of fruitful contact between the two
books was reached only in the Deduction. At that place it was essen-
tial to Kant's argument that the independence of the two works be
granted; the argument required that there be a common focus from
two quite different angles, and, at the end of the Analytic, Kant insisted,
perhaps more than was justified, upon this independence of the two
works-as if the common focus were a gratifying surprise to him be-
cause "their agreement was by no means sought after" (106 [20d).
Such a "confirmation," however specious it may seem to a reader who
36 Critique of Practical Reason, 6-7 <9H)1): "Only a detailed Critique of Practical
Reason can set aside all these misconceptions."
is not convinced that there can be such single-minded ness in any man's
work on a single part of a larger whole,37 was one that could not even
have been proposed in a work that was either a part of the Critique of
pure Reason (plan c) or in a metaphysics of morals based on the first
Critique (plans a and b).
These and perhaps still other reasons determined Kant, after April,
17 87 (the date of the Preface to the second edition of the Critique of
Pure Reason), to undertake the new and unannounced book. On June
z5 of that year he wrote Schlitz that the Critique of Practical Reason
was nearing completion and would be sent to the publisher within a
week. We do not know when it was actually finished, though in Sep-
tember he wrote that it was in the hands of the publisher. ss There was
a delay in getting a font of new and sharper type for it, but it was de-
livered to Kant in December, with the publication date given as 1788.
Even this book, however, he called only preliminary to a "Metaphysics
of Morals" (161 [260]).
Taking all the evidence in hand, the book cannot have taken more
than fifteen months to write; and this maximal estimate is based on the
supposition that he actually worked on plans a, b, and c and that he
worked on them at the very same time that he was making the very
heavy revisions, in quite other topics, of the first Critique. The second
supposition conflicts with all we know of Kant's working habits; and
if the first is correct, there is no evidence of it in any "stratification" or
"patchwork" structure of the completed book. Moreover, this maximal
estimate leaves quite unexplained Kant's silence on plans for such a book
in April, 1787. Apart from the initial incredibility of supposing that a
man could do such a feat, it is in best accord with all our definite evi-
dence to conclude that the book was begun not earlier than April, 1787,
and finished before September. The initial incredibility weighs little,
however, when we remember that the Critique of Pure Reason was
"brought to completion" in an equally short interval.
Such was the long series of deferred plans, evolving through more
than thirty years from the Prize Essay to the final Metaphysics of Mor-
37 And it will appear especially unconvincing to one who remembers that Kant
said that it was practical concerns that led to the distinction between phenomena
and noumena, and his remark that the foundation of the critical philosophy lay in
the concept of freedom, in considering the imputability of actions (c!. Lose Bliitter
zur Preisschrift iiber die Fortschritte der Metaphysik, XX, 335; Refiexion 6339).
In the Fortschritte der Metaphysik itself (XX, 311), however, he does modify this
and state that there are "tWO angles" around which metaphysics or critique (it is
not possible to tell which is the antecedent of sie) turnS: the theory of the ideality
of space and time and the reality of the concept of freedom.
3sTo Schiitz, June 25, 1787 (X, 490); to Jakob. September II. 1787 (X. 494).
als, and having almost as by-products the Critique of Pure Reason,39
the Foundations of the Metaphysics of MOTals, and the Critique of
Practical Reason. Even if the Metaphysics of Morals were not an in-
trinsically important and interesting work, it is one for which we
should have a feeling of gratitude. As a goal, it inspired Kant to labors
which produced other and greater masterpieces. Had the promise to
Herder been kept earlier, the ]oss to moral philosophy would have
been very great indeed.
89 Cf. Lose Blatter zur Preisschrift uber die Fortschritte de, Metapbysik, XX, 335.
That this was indeed the lay of things is the thesis of Richard Kroner's Kant's
Weltanschauung (Chicago: University of Chicago Press, 1956).
II
The Limits of Theoretical Reason
§ I. INTRODUCfION
More than half the Preface of the Critique of Practical Reason is con-
cerned with issues that had been raised and, in Kant's view, settled in
the Critique of Pure Reason or deals with objections that had been
made by readers who detected what they regarded as inconsistencies
between the first Critique and the Foundations of the Metaphysics of
Morals. Since the Critique of Practical Reason was to repeat and elab-
orate some of the doctrines of the Foundations, Kant regarded it as es-
sential to his purposes to show that his doctrines here and in the F oun-
dations were consistent with those of the first Critique and, indeed, that
they gave additional support to his findings there. While the full proof
of consistency and mutual support could be given only in the body of
the Critique, there are many assertions in the Preface that can hardly
be appreciated unless the main conclusions of the earlier Critique are
kept in mind. In writing this book, Kant certainly assumed familiarity
with the former work. Yet he explains the fact that "concepts and
principles of the pure speculative reason are now and again reexamined
in this work"; only in this way, he says, can the "old and new," i.e., the
theoretical and the practical, uses of the "concepts of reason" be com-
pared and connected, so that the "new path" of the Critique of Practi-
cal Reason can be "clearly distinguished from the previous one."1
In this chapter I shall attempt to give a summary statement of the
argument and conclusions of the first Critique, sufficient only to make
the Preface easily comprehensible. From time to time in later parts of
this commentary it will be necessary to study specific parts of the ear-
lier work in order to provide the proper background for understand-
ing specific parts of this one. The reader who is generally familiar with
the Critique of Pure Reason will not need to go through this entire
chapter but is advised to go directly to § 6, which is more directly con-
cerned with the text of the Preface.
1 Critique of Practical Reason, 7 (91); d. Reflexionen 5019, 5036.
20 The Limits of Theoretical Reason
§ 2. THE PROBLEM OF THE "CRITIQUE OF PURE REASON"
The Critique of Pure Reason is a sustained effort to determine sys-
tematically the answer to the question "What can I know?" The an-
swer is "I can know the truths of mathematics and of the sciences of
nature, but 1 cannot know the objects of traditional speculative meta-
physics."
This answer is reached by examining the presuppositions of knowl-
edge. In this examination, it is found that the ground which makes it
possible for us to have knowledge in mathematics and science does not
extend to the alleged knowledge of metaphysical objects. This ground
is the givenness of the objects of knowledge to our sensibility, or to
what Kant calls "intuition."
All our knowledge and alleged knowledge is expressed in one of
three kinds of judgments.
i) A judgment may be analytic, such as "Every red object is col-
ored" or "Every body is extended in space." Such a judgment is called
"analytic" because its predicate is found by analysis of the concept of
the subject. Such judgments are certain, and they are important in or-
ganizing and articulating our knowledge. But an analytic judgment
does not extend our knowledge; such a judgment does not tell us more
than we already know, at least implicitly, in entertaining the concept
of the subject. Moreover, it does not tell us anything whatsoever about
the existence of the subject it mentions. "A triangle has three angles"
is absolutely true and would be just as true if no triangle had ever ex-
isted in the world; "God is a perfect being" is such a judgment, and we
do not need to know whether God exists in order to know that the
concept of God contains the concept of perfection as one of its predi-
cates.
ii) A judgment is synthetic if the predicate is not logically included
in the concept of the subject. Thus "This girl is young" and "All men
are less than two hundred years old" are synthetic judgments. Whether
they are true or false must be found out by experience; they are based
upon experience and refer to experience if they are true. Even if true,
however, such a judgment is not necessarily true; it is quite conceivable
that the girl is not as young as she looks and that human life might
extend beyond two centuries. Such synthetic judgments, whose truth
can be judged only in the light of experienced facts~ are called "a pos-
teriori," in distinction to the analytic judgments, which are a priori.
iii) Kant believed, however, that there are some synthetic judgments
which are necessarily true and therefore can be based neither upon ob-
servation and induction from experience nor upon a merely logical
analysis of the concepts they contain. These are synthetic a priori judg-
The Limits of Theoretical Reason 21
ruents. He held that mathematical judgments, such as the theorems of

mathematics, and the principles of a "pure science of nature," like "Ev-
ery event has a cause," are a priori synthetic. Hume, as Kant tells us
(13, 52 [99, 14 2 ]), regarded mathematical judgments as analytic be-
cause he recognized that they were a priori, and he believed that scien-
tific principles like that of causation were a posteriori because they
were not analytic. The consequence of regarding the causal principle
as merely a product of custom or habit, built upon induction and there-
fore a posteriori, is skepticism in science as well as in metaphysics; and
Hume, Kant says, was saved from universal skepticism only by retain-
ing the apriority of mathematical knowledge, which he saved from
skepticism only because he mistakenly believed that it was analytic.
It is easy enough to explain the possibility of judgments of the first
two kinds. Previous philosophers had not even noticed that the third
kind of judgment existed; for Kant, however, they are essential even
for synthetic a posteriori judgments, since any judgment based on ex-
perience, such as "The sun warms the stone," presupposes an a priori
synthetic judgment of the connection of one event to another as cause
to effect. The problem of the Critique of Pure Reason is, therefore,
How are synthetic a priori judgments possible?
§ 3. THE "COPERNICAN REVOLUTION"
The answer to the question is vividly described in what has been
called Kant's "Copernican Revolution" in philosophy. In a justly fa-
mous passage in the Preface to the second edition of the Critique of
Pure Reason, Kant compares his new theory of knowledge to the Co-
pernican system in astronomy. The predecessors of Copernicus had had
difficulty in explaining the apparent motions of the planets on the as-
sumption that they all revolved around the earth. Before Kant, it was
similarly impossible in philosophy to explain how there could be a pri-
ori knowledge of things on the assumption that knowledge is a passive
conformity to an object. "Failing of satisfactory progress in explaining
the movements of the heavenly bodies on the assumption that they all
revolved around the spectator," Kant says, Copernicus "tried whether
he might not have better success if he made the spectator to revolve
and the stars to remain at rest."2 By analogy, Kant did the same thing.
If the phenomenal characteristics of objects, the way they appear to
us, are explained in terms of the conditions of our knowing them, it is
possible to see how knowledge of these characteristics can be a priori,
because they are dependent in part at least upon the functions of the
spectator. Then it becomes necessary, of course, to make a clear dis-
2 Critique of Pure Reason, B xvi.
tinction between the phenomenal and the real characteristics of objects,
just as Copernicus, in rejecting the Ptolemaic epicycles, had to make a
clear distinction between the real and the apparent planetary motions.
Let the things in themselves be as they may, the objects of knowledge,
which are their appearances to us, must conform to the structure and
synthetic activity of the knowing mind.
The faculties of the knowing mind which make knowledge of phe-
nomena possible are sensibility, or receptivity to data, which presents
the sensations for our conceptions and through which our conceptions
are related to actual objects, and understanding, which connects the
conceptions into synthetic judgments about objects. The a priori forms
of the data, to which they must all conform, are space and time. There-
fore, all objects that we can know must be spatiotemporal. The a priori
rules for the synthesis of concepts into judgments about objects are
twelve categories of the understanding, which are derived from the
forms of judgments in formal logic.
Both the forms of intuition and the categories may be called "sub-
jective" in the sense that they are forms of our experience, not of meta-
physical realities or things in themselves. But they are "objective" in
the sense that they are not personal, psychological features. of this or
that mind, but are rules for the conduct of experience from the recep-
tion of data to the establishment of knowledge of public objects in one
space and time, the same for all observers. They are thus the founda-
tion for the kind of objectivity that characterizes knowledge and dis-
tinguishes it from mere fancy and error, to wit, objectivity as universal-
ity and necessity, producing a standard for all knowing minds and
underlying agreement among various observers about their common
objects.
Both sensibility or intuition and understanding or concepts are nec-
essary to knowledge. Intuitions without concepts are blind, a blooming,
buzzing confusion. Concepts without intuitions are empty, an unearthly
ballet of bloodless categories.
From this it follows that we can have knowledge only of a phenom-
enal world, for we have no intuition of things as they are. Since intui-
tion is necessary to knowledge and touches only on things in space and
time, what is metaphysical, in the literal sense of the word as that
which lies beyond physics, is not attainable by human knowledge.
Thus we can say that that which makes objective knowledge of na-
ture possible, namely, the a priori forms of intuition, is lacking in al-
l~ged metaphy~ical knowledge, and this lack makes metaphysics impos-
Slble .. Hume rejected. me~aphysics, rightly, but on the wrong grounds,
for hIS grounds of reJectmg metaphysics forced him also to reject nec-
The Limits of Tbeoretical Reason 23
essary judgments in science too. Kant, through the doctrine of the
a priori form of intuition, saved science from Humean skepticism and,
by the same device, destroyed traditional metaphysics.
The knowledge we have of nature is real knowledge, not a subjec-
tive substitute for it, even though the objects of our knowledge are not
things in themselves. The phenomena we know through intuition and
the conceptions we have of their connection through the synthesis
guided by the categories are not just thoughts in our minds; they are a
system of phenomena under law, identical with what is meant by "na-
ture." This systematic organization of experience is what the scientist is
talking about when he talks about nature. It is not something in my
head, different from what is in your head; it is sticks and stones, sealing
wax, cabbages and kings, stars and atoms, and nebulae not to be seen
until we have a larger telescope. The "metaphysics of nature," of
which Kant often spoke, is the systematic development and exposition
of all the a priori principles involved in our knowledge of nature, in-
dependent of what specific discoveries are made about what is in nature.
§ 4. THEORETICAL AND SPECULATIVE REASON
Besides sensibility and understanding, knowledge also requires reason.
Reason is the faculty of systematic thought, of providing a wherefore
for every therefore. In reasoning, we press on from partial knowledge
toward complete knowledge. Nature comprises series of phenomena in
space and time, and these series can and must be infinitely extendable,
for every phenomenon has other phenomena before it and after it and
beside it, all of which determine its spatiotemporal position and empiri-
cal character. Every cause is likewise effect, and every law seems to be
a special case of a more general law. It is this aspect of our experience
that it is the office of reason to explore.
Theoretical reason, as we see it at work in science, attempts to or-
ganize our knowledge into a logically rigorous system which will give
parsimonious explanations for phenomena. The most successful scien-
tific theory is one which explains the most facts with the fewest as-
sumptions. But in our sciences of nature there are always assumptions;
the fundamental propositions are fundamental only because we do not
yet know how to explain them. No final answers can be found in na-
ture to the questions we ask of her; science only defers the time when
we have to say, "This is the way things are; they might have been
otherwise, and I cannot say why they are not."
If we are to satisfy our demand for final answers, philosophers have
always thought that we must go outside or beyond our knowledge of
nature. The attempt to do so is the work of the speculative metaphysi-
ciano To complete the work of theory, theoretical reason must become
speculative reason and must leave physics in order to try to get an-
swers from metaphysics. Metaphysics is the product of speculative
reason, and it consists of alleged knowledge of things as they are which
will explain why things appear as they do.
Speculative reason, in going beyond phenomena, must cut off the
line tethering knowledge to the world of experience; but, since it is a
way of thinking, it still must use the categories of thought. The result
is the emptiness of thought without perceptual contact with things. It
is not knowledge, for a category can be applied in a definite way to
an object only through a sensuous representation. The categories,
therefore, while permitting us to think of supersensible objects, do not
permit us to know them.
Our thinking of them is not, however, idle fantasy. The categories
themselves and the demand for systematic unity determine what con-
cepts of objects must be used in order to complete, or to attempt to
complete, the search for ultimate principles which will explain every-
thing. We are not just permitted to think of the soul, God, and the
world as a whole as being explanatory of things in our experience; we
are compelled to do so if we follow out into the supersensible realm the
principles which function within experience. The categories, freed of
their anchorage in experience, become Ideas of reason. Ideas are con-
cepts to which no object in the senses can ever be adequate; but they
are not useless. They regulate the orderly pursuit of the whole. But if
it is erroneously thought that the Ideas refer to objects as they really
are, as the categories refer to objects of the senses, there arise various
kinds of illusions which it is the business of critique to expose.
The Critique of Pure Reason is, therefore, a negative critique and re-
jection of the pretensions of traditional metaphysics, in which, it had
been believed, we had knowledge of objects beyond the scope of sci-
ence. Speculative metaphysics is not a legitimate part of knowledge;
the only metaphysics which is possible is "immanent metaphysics," i.e.,
the systematic exposition of the a priori principles within experience
and of the regulative Ideas.
§ 5. THE THIRD ANTINOMY'

Kant was not content to "prove" that knowledge of the supersensi-
ble WQrld is impossible by showing that it lacked one of the conditions,
viz., sensibility, that he held to be necessary for there to be knowledge.
Such an argument would be a petitio principii which would convince
3 This is fully discussed in chap. xi.
no rationalist, who held that sensibility is only the lowest form of
knowledge and not a necessary condition of all knowledge. Kant tried
to show directly that such knowledge was in fact impossible, by expos-
ing errors involved in all arguments designed to prove the existence of
God or the existence and immortality of the soul. More dramatically,
he attempted to uncover an "antinomy" in speculative reason by show-
ing that for every synthetic a priori judgment it produced, an equally
good and necessary argument could prove its contradictory. For rea-
sons we are about to see, Kant called the antinomy "the most fortunate
perplexity" into which pure reason could ever fall (107 [203]).
An antinomy is a pair of contradictory statements, each of which is
validly proved and each of which expresses an inescapable interest of
reason. There are four in the first Critique . We shall here be concerned
only with the third, that between freedom and natural causation.
The antinomies strictly limit theoretical reason to the world of space
and time, nullifying all speculative flights from the results of science
and all attempts to use scientific hypotheses in speculations beyond the
limits of sense. But their resolution permits an altogether different use
of reason; their occurrence and resolution indicate reason's broader
competence as a faculty not exclusively devoted to cognition. 4
This is very clear in the third antinomy. This arises from the conflict
in the idea of causality-if every single thing must have a cause, then
all causality is in time under the law of nature; but if all things have a
cause, there must be a cause which is not an event in time under the
law of nature. Each of these is essential if we are to give absolute valid-
ity to the causal principle; yet both of them, it seems, cannot be true.
The resolution is this: The thesis, which asserts the reality of causes
not subsumed under the law of nature, and the antithesis, which asserts
that all causation is under the laws of nature now known or yet to be
discovered, may each be true if their respective scopes are distin-
guished. 5
The field of application of each is defined by the nature of the argu-
ment supporting it, and neither can be validly employed beyond the
area to which the respective proofs extend. The proof of the thesis pre-
4 Acrually, Kant had discovered the antinomy before he had fixed the final lines
of his theory of knowledge, and it was probably the discovery of the later antino-
mies which led him to retreat from the position taken in the Inaugural Dissertation
that metaphysical knowledge was possible. He told Garve that the discovery of the
antinomy was the beginning of his critical philosophy (September 2 I, 1798 [XII,
257 ]).
15 This is strictly true only of the third and fourth antinomies, as pointed out jn
Critique of Practical Reason, 104 (199).
sents the interest of reason, which requires a sufficient cause for each
and every phenomenon. The sufficient cause cannot be found within
phenomena, because every phenomenal cause is itself the product of
prior causes and hence not, by itself, a sufficient explanation of su~se
quent phenomena. The proof of the antithesis, on the other hand, pre-
sents the claim of the understanding in applying the law of natural
causation to all members of a series of events in space and time. The ar-
gument shows that the assumption of a free cause (i.e., of a cause that is
not it~elf an effect) within phenomena would disrupt the reign of law
required by our conception of nature. The counterargument, however,
shows that if we do not assume a free cause, we cannot assume a first
cause, and therefore that we cannot give a complete causal explanation
of anything, regardless of how much progress we may make in knowl-
edge.
The antinomy is resolved by showing that the thesis can be applied
to the relationship between noumena (things in themselves) and phe-
nomena, and the antithesis is restricted to relations among phenomena.
These separate and distinct but compatible applications are all that is
legitimized by the two proofs. The solution is attained by a distinction
between the world of appearance and a noumenal world. This dualism
is a necessary presupposition of Kant's ethical theory and is the princi-
pal conclusion of his criticism of speculative metaphysics.
By this dualism, science is limited in two respects: a boundary is fixed
beyond which scientific knowledge cannot aspire, and the possibility is
established that natural law is not the only formula of causality. But
beyond the scope of science, there may be another use of reason. "I
have therefore found it necessary," says Kant, "to deny knowledge in
order to make room for faith."6 If this denial of knowledge had not
been effected-and effected on solid epistemological grounds and not
by human wish and obscurantism-it would be morality and not science
that we should have to surrender.
§ 6. TRANSITION FROM THE PROBLEMATIC TO THE ASSERTORIC

JUDGMENT OF FREEDOM 7
Note that only the possibility of another kind of causality with its
own law is established by the resolution of the third antinomy. Noth-
ing is said (except incidentally) in the Critique of Pure Reason to show
that freedom as a mode of causality is actual or that there is an a priori
law for such causality. If there is not such a causality, however, the at-
8 Critique of Pure Reason, B xxx.
7 This transition is fully discussed in chap. x.
tempt of theoretical reason to establish in principle a complete system
of causes is condemned to failure; even theoretical reason needs such a
concept for its own completion, but cannot establish it. Without such
a conception, however, the very being of theoretical reason is endan-
gered, and its lack may "plunge it into an abyss of skepticism" (3
[88]).
Such a concept can be established only by showing that it alone can
do for some realm of experience what the principle of natural causation
does for the sciences of nature. That is, there must be some realm of
experience which, upon analysis, shows the necessity of some a priori
synthetic judgment which is possible only if free causes are asserted
actually to exist. The Foundations of the Metaphysics of Morals and
the Critique of Practical Reason show that there is an unconditional
necessity in the moral law. The moral law is an a priori synthetic prac-
tical proposition, and these two works show that it is possible. It is pos-
sible if and only if the will is a free cause. "There really is freedom,
for this Idea is revealed by the moral law." The moral law is the ratio
cognoscendi of freedom, and freedom is the ratio essendi of the moral
law (4 n. [88 n.]).
Nevertheless, we thereby have no knowledge of freedom. A cate-
gory, that of causality, is applied to a supersensible object, viz., our-
selves as noumena. We think ourselves free, though in another context
(nature) we know ourselves as phenomena under the law of nature.
The contradiction between what we must think and what we know is
resolved just as the third antinomy was resolved: we distinguish our
reality from our phenomenal appearance. We thereby gain no knowl-
edge which has been interdicted by the Critique of Pure Reason, but
likewise, if we properly understand the meaning of practical reason, 8
we involve ourselves in no contradiction.
In the Dialectic of the first Critique, Kant considers two other Ideas,
viz., that of the soul as substance characterized by immortality and that
of God as a perfect being. He refutes arguments for each, but not to
prove that the soul is not immortal or that God does not exist. He
proves only that theoretical proof of each is impossible. Each is a neces-
sary object of thought, playing a regulative role in the guidance of our
search for completeness in theory; but neither is an object of knowl-
edge.
The Critique of Practical Reason converts the problematic judg-
ments of the Critique of Pure Reason (such as "The soul may, for all
8 Critique of Practical Reason, 6 (90): "This must have seemed an inconsistency so
long as the practical use of reason was known only by name," i.e., so long as it was
thought that practical reason was only a special kind of cognitive faculty.
we know, be immortal and God may exist") into assertoric judgments.
They were "mere Ideas, . . . unsupported by anything in speculative
reason," but they now "attach themselves to the concept of freedom
and gain, with it and through it, stability and objective reality" (4
[88]). This they do by being shown to be conditions of a necessary
object that must be possible if the law is not to be vain and delusive,
namely, the summum bonum (4 [89]).
But neither do these assertoric judgments, employing the categories
beyond the limits of experience, express or add to our knowledge.
They are postulates of a practical but rational faith, necessary because,
without them. moral experience could not be made fully intelligible.
There is nothing in them of a subjective will to believe or an emotional
need.9 They are rational, but not cognitive.
9 In the Preface, Kant is mainly concerned to deny the accusation that he has
treated these judgments as if they gave knowledge. Later (143 n. [242 n.]) he deals
with the opposite criticism that, since they are not cognitive, they must be arbitrary
and subjective.
III
Thought, Action, and Practical Reason
§ I. TWO ASPECfS OF CONDUCT

That thought and action are intimately related in Kant's philosophy is
evident from the very title of the book. "Practical reason" is not now
a widely used term, and it has connotations which do not recommend
it to many recent writers in moral philosophy. Yet that which is named
by this term in the philosophy of Kant is an important and obvious
element of human action. I shall try to describe it, eschewing at first
Kant's somewhat perplexing terminology. We shall then see that many
features of Kant's view of thought and action are not markedly differ-
ent from those of the legendary man in the street, always assuming that
this exemplary figure is not wholly un-introspective and that he knows
a little psychology.
Human behavior presents two quite different faces to human beings:
it is Janus-like and never seems to the person who is acting exactly like
what it seems to the person who is watching the action. We may look
upon a man's action in the way a psychologist does, as a series of events
whose connection illustrates highly complex descriptive laws. Given
the whole set of conditions, constitutional and environmental, it is sup-
posed by the psychologist that a man's behavior can be predicted. The
viewpoint of the psychologist who makes this assumption and who at-
tempts to make specific predictions under it is the viewpoint of a spec-
tator or observer. He seeks to understand and to predict behavior, but
he does not actively participate in it. Granted that the requisite laws are
not yet known in full; granted, indeed, that they may never be known
exhaustively; granted that the laws so far discovered have only low
statistical validity compared to those of the sciences of inanimate na-
ture; granted, finally, that for even the best established of these laws
the facrual data necessary for their application to an individual case are
usually lacking at the time they are most needed-granted all this, nev-
ertheless and in principle, most psychologists would agree with Kantl
1 " • • • If we could exhaustively investigate all the appearances of men's wills,
there would not be found a single human action which we could not predict with
certainty, and recognize as proceeding from antecedent conditions..•. It is only in
30 Thought, Action, and Practical Reason
that human behavior might be predicted with the same certainty as an
eclipse of the sun or moon (99 [193])'
What the content of the necessary laws would be-whether like that
of the laws of physics or physiology or economics-must be determined
by the success with which each kind of attempt at formulating them
will be rewarded. But that there are such laws is a necessary presuppo-
sition of a psychology or anthropology that claims to be a science.
The other aspect of human behavior is one seen by the person in-
volved in the action. All the facts seen from one point of view may be
needed in and present to the other, but the entire mise-en-scene is dif-
ferent. In the second mode of regarding and understanding conduct, one
deals with his own behavior as the actor, but not as an actor who rou-
tinely runs through a fixed role in the human drama, a role well known
to the outward observer. No; human beings are actors who do not quite
know what is expected of them by the audience. The actor's consider-
ing his act is itself one of the determining causes of the specific action
he ventures upon, and, until his consideration is carried through, he
does not know what his action will be. Though it may be true that the
spectator, as a professional psychologist or wise Menschenkenner,
might foretell the precise effect that the actor's concern with his own
action will have upon his conduct, this specific cause-effect relation is
not known to the actor himself. The taking thought of his action is
one of the causes of his action; but, until he himself has taken thought
and thereby reached his own decision, the specific direction of the ac-
tion that will issue from his taking thought cannot be known to him.
To him, the act of taking thought is not so much a cause of a particular
action, related to it by causal laws, as it is a search after a ground or
good reason which will lead him to choose a specific act. The action is
a product of choice or decision which is reached just in this process of
taking thought. That much of the taking thought is a rationalization
may be insisted upon by the observer and suspected even by the actor.
After the choice is made, laws may be adduced to show that the actor
could have been expected to do as he in fact did; they may be adduced
by a behaviorist who might refuse to consider seriously what the ac-
tor's thoughts and inner struggles were; indeed, they may be adduced
even by the actor himself, who, with growing sophistication about him-
self and his own quirks and turns of mind, may discover that his action
falls into a pattern that applies to men in general.
the light of this [empirical] character that man can be studied-that is to say, we are
simply observing, and in the manner of anthropology seeking to institute a physio-
logical investigation into the motive causes of his actions" (Critique of Pure Reason,
=
A 550 B 578).
Thought, Action, and Practical Reason 31
Even before the event, the psychologist may be able to say that such
and such a percentage of men brought up as his subject was and placed
in the situation in which his subject is found could be expected to de-
lay their action for a day, saying that they wanted to think it over be-
fore deciding, and then that a specific percentage of them would do
precisely so and so. A certain number of men will try to decide wheth-
er they should continue smoking when their physician advises them not
to, and they will try to decide in long deliberations whose outcome
they do not know. But the psychologist, armed only with statistics on
past cases, might be able to say, "Whatever they think, x per cent will
go right on smoking and cite y as a good ground for doing so."
But none of this perhaps frightening knowledge, even supposing that
the psychologist had it and gave it to the actor, in the least serves as a
sufficient condition for the actor's making up his mind in a particular
way. He does not know whether he will belong to the fraction of men
who will do this predicted so and so or whether he will belong to the
other group; and the only way for him to find out is to do the consid-
ering and decide the issue, which should, in principle, have been pre-
dicted by the observer. The actor may know the statistics and be wise
to the little hypocrisies and rationalizations he practices; but he must
decide, and not merely know, whether he will be guided by this knowl-
edge to do what rationalization suggests he will do or whether he will
do the other, precisely because he recognizes rationalization for what it
is. If it were merely a matter of knowing, so that he could predict his
behavior with the same certainty that the perfect psychologist could
have, the experience of deliberation, taking thought, and deciding
would not be just illusory, as the observer may believe it to be; it would
not even occur.
In a word, from the point of view of the actor making a decision,
there is the experience that deliberation is effective, that thinking makes
a difference, that one is free and not wholly determined by causes be-
yond his control. From the spectator's point of view, this may be an
illusion: "Du glaubst zu schieben und du wirst geschoben," as Mephis-
topheles said. 2
Kant's theory that man's actions are both free and predictable is,
apart from its metaphysical explanation, a report on the distinction be-
tween the two points of view and the assumptions which define each
of them. When wishing to avoid having to "prove freedom in its theo-
retical aspect," Kant avails himself directly of the different assump-
tion necessary in the attitude of the actor, and he says: "The laws
2 Faust, Pan I, Walpurgisnacht.
which would obligate a being who was really free would hold for a
being who cannot act except under the idea of his own freedom"3-
and all of us are such beings whenever we face a choice. Put more gen-
erally and without restricting it to the Kantian problem, we may say
that those considerations without which a person cannot act deliber-
ately must be included in a full account of his behavior. In acting de-
liberately, a person does not need to show that these considerations are
really psychological causes effective in his behavior; he does not even
need to know, in advance, whether he will in fact be able to carry out
his decision when once it is made (15 [IOrJ). All this concerns only
the observer; and if the observer's knowledge of this is made available
to the actor, this new knowledge is just one more factor to be taken
into account in the process of deciding. For what the actor is con-
cerned with is this: Are the reasons I have for making this choice sound
reasons? It is not this: What makes me decide to do x or y, and will the
causes which make me decide to do x suffice to make me do x in fact?
Decision on the soundness of doing x is an entirely different matter
from the cognitive or theoretical decision that the causes exist which
justify the prediction that x will be done or that the actor will decide
to do x. The cognitive prediction belongs to the observer, the practical
decision to the actor.
Every time we act deliberately, we evaluate the soundness of reasons.
The psychological causes why these reasons occur to us, and why we
assign to each the specific weight we do assign to it are ,matters which
it is very important for us to know; self-knowledge consists in this
knowledge, and such knowledge may moderate our dogmatism and
prevent our fanaticism about them. But to make use of the reasons we
have for doing something, in mathematics as in morals, does not re-
quire that we know the psychological facts underlying our awareness
of them and the effectiveness that they have in the guidance of our
behavior. The principles themselves, not the• psychological causes of
our acceptance of them, are what deliberating human beings do almost
invariably consider. The attitude of the spectator is, in comparison with
it, artificial, difficult to maintain, and coldly indifferent to the issues to
be adjudicated in periods of painful indecision and more painful
decision.
§ 2. CONATIVE AND COGNITIVE ELEMENTS IN ACTION
The actor in a situation demanding decision can discern in himself
two very different but interrelated factors. The first is impulse. There
is, as a dynamic moving factor, some want, need, desire, or wish. Pre-
3 Foundations of the Metaphysics of Morais,+f8 n. (67 n.).
Thought, ActiO'Tl, and Practical Reason 33
sumably animals are conscious of this; certainly men feel pushed or
pulled in this way by their wants, needs, desires, passions.
But it is equally certain that a man in full possession of his powers
does not observe in himself that these wants always appropriate his en-
tire resources and lead automatically to the action which the need seems
to demand. Action is deferred; the need may be denied. Sometimes, of
course, a reflex or an uncontrollable force from some importunate need
may, as it were, seize him and force upon him or wring from him some
action against which thoughts of wisdom and prudence are impotent.
When this occurs, the person is no longer an actor; he no longer even
seems to himself to control his actions, and he disclaims responsibility
for them. He is like the spectator of a person in a drama (who hap-
pens, tragically, to be himself) swept along without effective delibera-
tion, choice, conscious purpose, decision, or responsibility by forces
either unknown to him or uncontrolled by him.
Happily, however, the more usual situation is one in which behavior
is not automatically triggered by impulse, but one in which impulse is
moderated, directed, redirected, and sometimes thwarted by our tak-
ing thought of the meaning of the impulse or spur to action. Perhaps
"meaning" is not an entirely suitable word for this phenomenon, and
whence the phenomenon itself comes-from the ego or superego, from
some overarching sentiment, from the prefrontal lobes, from some plan
of life projected by reason-may well be disputed. Yet the phenomenon
is indisputable, even if the taking thought which goes into the phenom-
enon is itself just a necessary part of our psychological makeup.' It may
be decried by the cynic who doubts that man deserves the honorific
name of Homo sapiens; but the soi-disant Homo sapiens feels it keenly,
poignantly, and often painfully.
In saying that the meaning of an impulse is a factor in conduct, I re-
fer to two closely related facts. First, a person regards the dynamic
factor which pushes or pulls him as being a certain kind of push or pull.
It is not just a vague and amorphous restlessness, which could call forth
nothing worth the name of action. And, second, a person has a con-
ception of the kind of response which is appropriate to it. For instance,
I have a feeling which began as a vague discomfort but which I easily
recognize and classify as thirst; and I have a conception, perhaps vague
around the edges but with a nucleus of clarity, of what sorts of action
would quiet this need and end this discomfort, e.g., drinking water or
beer certainly would, wine might, milk or coffee would not. In the
course of our experience, not only do the internal drive and the re-
sponses to it become associated together through trial and error in
4. 0. Critique of Pure Reason, A 803 = B 831.
matching this response to that need, but they each become generalized.
Such generalization is the essential step in learning. Such association and
generalization occur in animals, and they occur below the level of con-
sciousness in us men. But when a situation is complex and highly novel
and problematic, a conscious effort must be made toward identifying
the stimulus and the possible responses, toward determining the bound-
aries of the two generalizations, and toward accumulating the factual
knowledge needed in order to associate the most effective class of re-
sponses with each class of needs. In this process the impulsive element
must be held back while the cognitive exploration goes ahead and even
issues forth in tentative experiments. All this is what I mean to suggest
when I say that an intelligent being is able to discriminate, conceptual-
ize, formulate, and act in the light of the meanings of his needs. These
meanings appear as habits in his behavior and as rules of his conduct.
The spectator has to deal with all this; in fact, the previous para-
graph is a typical "spectator paragraph." He knows the habit or dis-
position of a particular person who discovers particular meanings and
who adopts specific lines of action in the light of these meanings. Psy-
chological laws can be formulated which will help predict behavior
even this complex. Such laws will hold whether they are known to the
actor or not; indeed, they can best be discovered by observing cases in
which the actor does not know them. The actor is seen as illustrating
these laws, not as really obeying them; I no more need to know the
laws of habit formation in order to act in accordance with them than
the planets need to know Kepler's laws.
What the actor has and what the planets lack, however, is a concep-
tion of the connection of things and events; and for our present ana-
lytical purposes, it does not matter at all whether this conception is
accurate or not. This conception may be hardly more than an associa-
tion of ideas such as the barely conscious "Things that look like that
usually taste good." It may be a conception of an objective law of na-
ture, which will lead the actor to act in accordance with his belief that
if he jumps from a certain height, he will be injured. It may even be a
conception of what is right, such as the moral precept that, when
tempted to lie. one should not do so. In any of these cases, the judg-
ment may be correct or incorrect, and its truth or falsity is irrelevant
to the initiation of the actual behavior predicted by the spectator as
that which could be confidently expected of a man with this or that
conception or misconception.
But that it should be a correct conception is of utmost importance
to the actor, and he acts on the assumption that his conception is a cor-
rect one. If the action is not an automatic response to a stimulus and
there is time permitted for deliberation, there is always some concep-
tion of the kind of thing or situation which he must handle, of the
kinds of promises or threats of future experience which will eventuate
from the alternative actions, and the proper rule for choice among
them.
\\7hat the thing promises, what it means in future experience, may
be so different from what it seems to be in its present impact upon us
that the action we undertake in the light of its meaning can be pain-
fully antagonistic to the impulses salient at the moment. When I sub-
mit myself to present pain in the interest of avoiding future pain, I
have this experience of allowing, or rather of making, the meaning of
the present situation dominate over its hic et nunc character in the de-
termination of my behavior. In the light of what I think is knowledge
of myself and of the way of the world, as a being whose thought is to
be effective in conduct, I act on a policy and not on an impulse, though
the impulse is always there.
This is possible only because impulse, through the medium of mean-
ing, can be integrated into and controlled by interest. Though interest
has a dynamic character which comes from impulse,~ it also has a dis-
positional governing character which is sustained by putative knowledge
of the meaning of situations and of the consequences of alternative ac-
tions. Interest is impulse conceptually weighed and in part conceptu-
ally directed. Intelligent action is action whose motive is an interest
guided by appropriate conception and not a blind and naked impulse.
Impulse leads to trigger-happy action or outbursts, to the jerks and
twitches that come and go with passing moments; interest leads to ac-
tions directed according to policies and plans. Only a being having the
power to reason can act from an interest. "Interest is that by which
reason becomes practical, i.e., a cause determining the will. We there-
fore say only of a rational being that he takes an interest in something;
irrational creatures feel only sensuous impulses."6
I may put the policy in favor of an interest into words and tty to
live according to it. Policies may be formulated in very specific rules
which may express a specific habit, as I may say, "I make it a practice
to write my name on the flyleaf of all my books." They may be ex-
pressed in general maxims that cover a wide variety of different be-
haviors, such as Carpe diem. They may be highly artificial and only
IS Up to this point Kant would, I believe, agree with our general account of action.
But he would assert that this definition of interest is too narrow, for there is an in-
terest which is not derived from impulse, though always associated with it in some
way (cf. below, § 4).
fj Foundations of the Metaphysics of Morals, 459 n. (80 n.); cf. Critique of Prac-
tical Reason, 79 (17:), and Critique of Judgment, S :.
slightly effective good New Year's resolutions. They may represent a
settled policy of life which can perhaps be clearly formulated only by
one's biographer or psychiatrist. Finally, and most important for our
purposes in moral philosophy, they may represent a consciously chosen
or invented aspiration for a life integrated by respect or reverence for
an ideal value, an aspiration that remains even when actual life repeat-
edly falls below it and mocks it with failures.
A person who tries to act on a policy of any of these kinds, which
Kant summarily groups under the name "conception of a law,"7 is said
by him to have a will. A voluntary action differs from an impulsive
action in that there is in it some control through a policy supposedly
based upon knowledge of ourselves, our circumstances, and the conse-
quences of our acts. "The will is never determined directly by the ob-
ject or by our conception of it," he says (60 [r 5 I]). "Rather, the will
is the faculty which makes a rule of reason [i.e., the conception of a
law of the connection of these meanings] the efficient cause of an ac-
tion which can make an object real." Policy, expressed in these rules or
conceptions of laws, gives direction and stability to a complex life that
could not long subsist upon the vagaries of passion. Only a being who
claims knowledge of the connections of acts with one another and with
consequences and sees them all in the context of a pattern of life can be
said to have a will. Such a conception of laws and patterns, which are
not mere tracings of past experience, is the product of our ability to
reason. It is this simple thought that Kant had in mind when he iden-
tified will with practical reason.
One hears little of "practical reason" in present-day psychology, be-
cause behavior can be described by the spectator without reference to
will, which seems to many psychologists to be a "ghost in the machine"
or a relic of an obsolete faculty psychology. This should not disturb
us, however, or make the moral philosopher apologetic to his scientific
colleagues when he is concerned in action or in evaluating the actions
of others. It should certainly not have the effect, which it may in fact
have had, of occasioning a neglect of the peculiarities of voluntary ac-
tion because of the greater psychological respectability and acceptabil-
ity of concepts of impulse and emotion. Give it any name we wish in
order to placate the behaviorist in his own field, the phenomenon of
responding to meanings according to conceptions of rules and laws and
of mobilizing our resources to withstand the importunities of momen-
tary impulse is the essential, though often agonizing, kernel in con-
sciousness of the kind of action which makes us men.
7 Foundationr of the Metaphysics of Morals, 411 (29); Critique Of Practical
Rearon,60 (151); Critique of Judgment, §§ 4t 10.
Thought, Action, Il1ld Practical Reason 37
While psychology has vastly deepened our knowledge of the impul-
sive in its many and mysterious ramifications in normal life, neurosis,
art, religion, politics, philosophy, and even science, Kant was primarily
concerned with the cognitive or conceptual factor in willing. What he
has to say about this has been, I think, little affected except in termi-
nology by recent science. For science is developed from the spectator's
point of view, from which the cognitive condition of action is likely to
be sometimes tacitly assumed and at other times overlooked or even
denied.
§ 3. PRACTICAL REASON AND WILL
Though Plato had distinguished willing from mere desiring, Aris-
totle was the originator of the distinction between practical reason
(nous praktikos) and theoretical reason (nous theoretikos). 8 The
Schoolmen translated the former as intellectus practicus,9 and they also
used the terms intellectus activuslO and ratio practica. l1 The Wolffians
did not use these terms in their Latin works or give literal translations
in their German works, but nevertheless they maintained the distinc-
tion in their terminology of cognitio movens and cognitio iners12 and
recognized the cognitive as well as the conative elements in volition in
such terms as appetitus rationalis. 13 Kant originated the term praktiscbe
Vernunft in 1765.14
8 De anima 433 a IS ff.; d. PoUtics 1333 a 18 ff.
9 Thomas Aquinas, In decem libros ethicorum Aristotelis ad Nicomachum ex-
positio § IIp.
~o Thomas Aquinas, Summa theologica, Prima, Q. 79, art. II; Secunda secundae,
Q. 179, art. 2. A fourteenth-century translation (Middle High German Translation
- of the Summa theologica, ed. B. Q. Morgan and F. W. Strothmann ["Stanford Uni-
versity Publications in Language and Literature," Vol. VIII, No.1 (1950)], p. 371)
renders intellectus practicus as da5 wUrkliche verstan. Cf. also M. Grabmann, Mittel-
alterliches Geistesleben (Munich, 1926), p. 434.
11 Summa theologica, Sec. sec., Q. 83, art. I, ad 3.
12 Baumgarten, Metaphysica, §§ 669,690.
13 Wolff, Psychologia empirica, §§ 880 ff.; cf. also Vernunftige Gedanckell von
den Kriifften des menschlichen Verstandes, § IS, where the term lebendige Erkiinnt-
n;s is used.
14 Nachricbt von der Einrichtung seiner Vorlesungen im Winterhalbjahr von
1765-66 (11,312); at least, no earlier Beleg is given by Grimm. Mellin (Kunstsprache
der Kantischen Philosophie [1798], p. 283) says: "The express10n praktische Vernunft
was not usual before Kant; one spoke only of Verstand and Wille." (I am indebted
to Professor Paul Schrecker for calling Mellin's remark to my attention.) I can find
no English use of the words "practical reason" before Richard Burthogge's Organum
'Vetus et novum, or Discourse on Reason and Truth, sec. 61 (1678). The term does
not, I think, occur in any of the British moralists Kant knew, with the exception
of Reid, Essay on the Active Powers of the Mind, Book III, Part iii, sec. 2; but this
Was published in 1788, after the Critique.
38 Thought, Acti01l, and Practical Reason
Even the words used give some indication of the progress that Kant
had made beyond the Wolffians in the conception of the will. In mak-
ing his distinction between reason and understanding,111 Kant ascribed
to reason the task of going beyond the order of things, as given, to an
ideal order of systematic connection of experience, a systematic con-
nection that is never passively found in knowledge but must be striven
for according to regulative Ideas. Reason is spontaneous in formulat-
ing Ideas that can never be adequately represented in our sense experi-
ence of the actual, even though this is categorized by the understand-
ing. The understanding is, of course, spontaneous, but its spontaneity is
restricted to a re-working of what is or can be given in perception.
Though Kant acknowledged a practical function of the understanding
( 2 3, 55 [109, 1451), in making reason the prime practical facul ty he
did three things. First, he called attention to the manner in which it is
theoretical knowledge in its systematic and ideal integrity that is rele-
vant to the act of voluntary choice, and not some isolated bit of experi-
ence or rule of thumb. Second, he called attention to the fact that in
practice we sometimes demand an unconditional certainty comparable
to that which reason alone is supposed to afford us in our theoretical
occupations. Third, and most important, through the connection as-
serted between reason and will, he prepared the way for a new defini-
tion of will itself, with all the moral consequences to be drawn from
this conception.
Will is the faculty of acting according to a conception of Jaw, which
is not a product or discovery of understanding but of reason. In con-
trast, his predecessors had thought of will as only rational desire, i.e.,
the faculty of acting according to a clear (rational) representation of
the object of desire. 16 They could discern a difference only between
the lower and higher faculties of desire and were never able, according
1~ Wolff translates ratio as Vernunft and defines it as "insight into the connection
(Zusammenhang] of truth," and hence as the art of inference; ;Tltellectus is trans-
lated as Verstand and is the faculty of clearly representing the possible. Pure under-
standing (intellectw purus) is understanding separated (abgesondert) from senses
and imagination, but human understanding is never completely pure (d. Verniinf-
tige Gedancken von Gott, der Welt und der SeeJe des Memcben [1736], §§ 368,381,
277, 181, 185). Logically, this corresponds very well to Kant's distinction between
reason as the faculty of inference and understanding as the faculty of concepts; but
the important Kantian distinction between the real and the merely logical use of
reason and understanding and the equally important theory of the generic difference
between sense and the discursive faculties are not anticipated.
16 Cf. Wolff, VeT'lliinftige Gedllncken von Gott, der Welt, und de.,. Seele des
Menschen, ~ 492, and Psychologia empirica {(737 )', § 880; Crusius, ETltwurf der notb-
'We1Uiigen VeT7IUnft'Wahrbeiten «(753), § 445. The gradualisric conception of the dis-
tinction between ratio and intellectUS is fundamental to this.
to Kant, to single out the unique feature of willing and, a fortiori, of
moral willing. For these reasons, Kant rejects, as inadequate to ethics,
their conception of a universal practical philosophy and, as inconsistent
with their own views, their attempt to distinguish io any ethically sig-
nificant way between the lower and the higher faculties of desire (12 f.
[109 f.]).
Yet there are two puzzles which arise from Kant'S way of speaking
of the relationship among will, practical reason, and theoretical reason,
and these must be cleared up before we come to the most important of
his doctrines, to wit, that pure reason can be practical.
First, Kant identifies will with practical reason, but he often con-
fuses the reader by speaking of reason as the determiner of the will. The-
oretical reason, which demands an order in the totality of the data of
possible experience, is practical when, through the order it projects as
possible if such and such an action is executed, it becomes a determi-
nant in behavior whose dynamic component is provided by impulse or
desire. That is, theoretical reason provides the knowledge of the law
which can be applied in the satisfaction of desire, and, insofar as it does
so, it is practical reason. Thus far, at least, there are not twO reasons, a
theoretical and a practical, but one reason-the faculty of formulating
laws and principles-which has two applications. One gives knowledge
of things as they are (or appear); the other gives direction to the
changes we introduce into this natural order by means of voluntary
action. The following two sentences therefore mean the same thing:
(a) Will is impulse guided by reason; and (b) Will is practical reason.
From a it is easy to move to another sentence, (c) Reason can deter-
mine the will, which seems to be incompatible with b, for b identifies
them. But, properly understood, band c are not incompatible. The last
sentence means simply: (c') Reason determines the action by which
impulse is to be satisfied; when it does so, it is called "practical reason,"
and the action chosen is called an "act of will."
Second, another puzzle is presented by Kant's oftell speaking of prac-
tical reason as a cognitive facult y 17 and as a faculty of desire. I8 He
mentions the danger of taking the words "practical reason" as if the
"object" of practical reason were comparable to an object of theoreti-
cal reason, i.e., as an epistemological object and not as an object of de-
11 Critique of Pure Reason, A 633 = B 661: "The practical use of reason is that
through which it is known a priori what ought to take place"; also draws distinction
between theoretical and practical knowledge. Critique of Practical Reason, 66 (157):
pure practical concepts are directly cognitions, not having to wait upon, or be
applied to, intuition.
18 Critique of Practical Reason, 14 (III); Erste Einleitung in die Kritik der Urteils-
kraft, XX, 145-46; Critique of Judgment, Introduction, III (V. 177 [13]).
sire or volition. 19 We should be warned by this against taking "practi-
cal reason" to denote merely the faculty by which we gain knowledge
of right and wrong, though we should not forget that practical reason
does have this cognitive function. It provides the cognitive factor in
the guidance of action whose dyna111is is impulse. The sentences of the-
oretical reason or understanding, such as "A is the cause of B," become
the practical proposition or cognition of practical reason, "If you de-
sire B, do A" (26 n. [II 3 n.]). The latter proposition might better be
called a cognition of technical or practical understanding when B is
some specific, well-defined object or situation. A prudent policy of life,
on the other hand, as an allegedly unconditional ground for choice not
only of means to happiness but also and more importantly of the genu-
ine composition of such an ideal, could properly be called the object
of a cognition of practical reason, as its pursuit is undertaken because
of a maxim or practical reason. We should then retain the name "cog-
nition of pure practical reason" for knowledge of the moral law and of
the highest moral aims.
§ 4. PURE PRACfICAL REASON

If practical reason can hold before us a law valid for practice but not
derived from our experience of the way things go in the world when
we attempt to satisfy some specific desire, this would be a law of a kind
entirely different from those of interest in theory. The relevance of the
latter kind of law, though essential for intelligent practice, is always
contingent upon there being some experienced situation by reference to
which we choose, from all the actual laws of nature, those which are
concerned with the causes of the object of the specific desire. Such laws
in their theoretical formulation may be necessary; but, when formu-
lated as practical rules, they are always contingent upon there being in
us the desires which can be satisfied through successful application of
our knowledge of them. If there is an unconditional practical law, it
could be discovered only by a reason that is intrinsically practical, and
not by a theoretical reason which is only extrinsically and contingently
practical, Le., one issuing laws which mayor may not be applicable in
practice, depending upon the desires and the situation. Such an intrin·
sically practical reason is called pure practical reason. 20
19 Critique of Practical Reason,s (90). We shall see later, in chap. ix, Kant's justi.
fication for referring to both as objects.
20 Kant uses the word "pure" to refer to both cognitions and faculties. In the for-
mer usage it refers to (a) cognitions which are independent of experience and (b)
cognitions in which there is no empirical content. Meaning II is equivalent to
a priori, and Kant said later (not quite accurately) that the Critique of Pure Reason
Thought, Action, Il1ld Practical Reason 41
That pure reason can be practical is the chief thesis of the Kantian
moral philosophy; it is equivalent to the assertion that there are uncon-
ditional practical laws. Kant rejected the Aristotelian thesis 21 that rea-
son alone cannot move us and the Humean thesis 22 that reason is and
ought to be the slave of the passions. Reason is concerned not with the
choice among ways to some end projected by desire; this is its merely
logical use. It establishes the goals of action through the formulation of
an intrinsically practical and unconditional law. This is its real use. 23
Pure reason, in its real use, is always concerned with unconditioned
conditions. Pure practical reason is the faculty of providing an uncon-
ditioned condition for voluntary action, which is a law demanding di-
rect obedience without a quid pro quo. As Kant is to show in the early
parts of the Critique of Practical Reason, such unconditioned condi-
tions cannot be found in or by an empirically practical reason, which is
indeed the slave of the passions. If pure reason is practical, however,
there is some intrinsically practical law and some motive independent
of the contingent and empirically discovered human desires. This mo-
tive must be our knowledge of the law itself through the respect that
it creates in us. An action having this motive is moral, and a being
which acts from this motive has a good will. The Foundations of
the Metaphysics of Morals stated, in this way, the requirement that an
action must meet if it is to be counted moral, and it concluded that pure
reason must be practical if morality is not a mere chimera.
But can pure reason, in its real use, be practical? Or is this analysis
of morality an analysis of a vain and empty delusion? To show that
pure reason can be practical, we are told in the first paragraph, is the
prime task of the Critique of Practical Reason.
was concerned only with "pure" in this sense ('iJber den Gebrauth teleologischer
Prinzipien in der Philosophie, VIII, 183-84). When used in reference to a faculty,
"pure" indicates that the faculty is a priori legislative (Critique of Judgment, V, 179
[.s)). It is imponant not to confuse these two meanings, though they are closely
related. The moral law is pure in senses a and b; the concept of duty is pure only
in sense a; practical reason is pure, or may be pure, in the sense that it is an a priori
legislating faculty, giving the moral law.
21 De anima 433 a 23.
22 Treatise of Human Nature, Book II, Part III, sec. iii (Selby-Bigge ed., p. 415).
Kant was not the first in his time to state that reason is both the necessary and the
sufficient condition of action; in this he was anticipated by a critic of Hutcheson's,
namely, John Balguy (The Foundation of Moral Goodness [1718] [Selby-Bigge,
British Moralists, II, 91-<}3]). and by Price (Review of the Principal Questions of
Morals [1758], chap. viii). There is no evidence that Kant knew of Balguy or Price.
23 For this distinction d. above, p. ,.
IV
Name, Purpose, and Structure of the ("("Critique";

Commentary on Preface and Introduction
§ I. INTRODUCTION
The Preface of the Critique of Practical Reason discusses the follow-

ing topics: the title of the book, its purpose, and its relation to other
works; and it gives a preliminary exposition of Kant's theory of free-
dom and a defense of his views against his crltics. A major part of the
Preface deals with criticisms that had been made of the first Critique
and of the Foundations of the Metaphysics of Morals; by answering his
chief critics, Kant hoped to show better the consistency among the
various parts of his philosophy. In addition, there are brief replies, or
anticipations of replies to be given later, to isolated criticisms of specific
points in his ethical theory.
The Introduction again discusses the title of the book and its pur-
pose, and then proceeds to a description of the structure and organiza-
tion of the new Critique.
In this chapter, I shall deal with these matters as follows: In § 2, I
shall discuss the title of the book and Kant's reason for choosing it; § 3
will discuss the purposes of the book as stated here and elsewhere; and
§§ 4 and 5 will deal in more detail with two of these purposes. Then I
shall take up, in § 6, the relation of this work to the M etaphy sics of
Morals; § 7 will be concerned with the organization of the Critique;
and, finally, in § 8, I shall review the polemics of the Preface, laying
primary emphasis upon their historical occasion, since most of their
philosophical import can be better discussed in the more systematic and
less historical studies to follow.
§ 2. TITLE OF THE BOOK
Even before he planned to write this book, Kant referred to the
"critical examination [Kritik] of pure practical reason," as we have
seen in chapter i. Yet when the book was written, he called it simply
Critique Of Practical Reason, though, as he says in the first paragraph,
Ntrme, Purpose, Il7ld Structure of the "Critique" 43
"the parallelism between it and the critical examination of speculative
reason seems to demand" the longer title (3 [87]).
In interpreting this first sentence, we meet with a difficulty that
exasperates all translators of Kant: in what sense is the word Kritik
being used? It may be part of the title of a book, in which case it
should be rendered as Critique. It may mean merely a critical examina-
tion of something. It may mean that formalized division of philosophy
that he called the propaedeutic to the development of a system of
reason; in this case, it should be translated as critique. The typographi-
cal practices of the eighteenth century and the fact that all nouns are
capitalized in German make it impossible to base a decision on the
letter of the text. One must try to understand what Kant meant at
each particular place where he used the word Kritik. It is obvious that
in the first paragraph the words Kritik der praktiseben Vernunft refer
to the book in hand and that the words Kritik der spekulativen
Vernunft refer to the Critique of Pure Reason, though they do not
properly state its title, nor do they give an entirely correct intimation
of its contents.
Kant here suggests that there is a parallelism between two Critique's.
But when he added the second Critique to the corpus of the critical
philosophy, he surrendered a belief that he had had when he wrote the
first; for the first was to give a critical examination of the entire faculty
of reason, both theoretical and practical. When he wrote the F ounda-
tions, he referred to the "critical examination of pure practical reason"
as doing for practical reason what the "critical examination of pure
speculative reason" (se. Critique of Pure Reason) did for metaphysics,
in the sense of transcendental philosophy and the metaphysics of na-
ture. 1 In 1785, therefore, Kant recognized a parallelism of function of
the two "critical examinations," one of which had already been writ-
ten and the other one of which was not even at that time a separate
literary project. In 1787, however, though mentioning the parallelism
and suggesting that it might seem to justify the use of the title men-
tioned in 1785, he was more intent upon denying the parallelism of
the function of the two works. For the first' Critique was concerned
with denying the pretensions of pure theoretical reason, while the
second is concerned with denying those of empirical practical reason.
In both the first paragraph of the Preface and in the second and
third paragraphs of the Introduction he says that pure practical reason
requires no critical examination. All that is required is to show that
pure reason can be practical, and, in order to show this, it suffices to
44 N trme, Purpose, and Structure of the "Critique"
examine practical reason in general. It is discovered, in such an ex-
amination, that only the empirically conditioned practical reason, not
the pure practical reason, "presumptuously overreaches itself," as pure
theoretical (se. speculative) reason does (3, 15-16 [87, IOI-Z]).
This justification for the title, however, is not coercive. For there
is a duality of meaning in the word Kritik as a name for a division of
philosophy, and this equivocation is bound to make any such decision
somewhat arbitrary. That the word Kritik even in the title of the first
book has various meanings is shown by the fact that its object, what
it is a critique of, is variously given by the following descriptive but
inaccurate titles: "Critique of Reason," "Critique of Theoretical Rea-
son/' "Critique of Speculative Reason," and even "Critique of Pure
Understanding. "2
Kant formally defines Kritik as "a science of the mere examination
of reason, its sources and limits,"3 and it is propaedeutic to a system
of pure reason. There are two functions of Kritik. Negatively, Kritik
fixes the boundaries of the competence of reason; this is its "police"
function 4 in preventing or exposing the dialectical illusions of specula-
tive metaphysics. Positively, Kritik is to secure to reason the "sure
path of science" against the import of skepticism from regions where
it is justified (speculative metaphysics) into those where it is not
ju~tified (science and morals). Kritik in the negative sense is Kant's
answer to rationalistic metaphysicians; .Kritik in the positive sense is
his answer to skepticism based on empiricism.
Since Kant favors the title Critique of Practical Reason, he seems
to be using the word here primarily in its negative sense and to have
as his chief aim the limiting of the claims of practical reason based on
empirical motives. But in the negative sense of Kritik, there is not, as
we have seen, a parallelism between the two works, since it is pure
speculative reason and empirical practical reason that stand in need of
negative criticism.
Yet negative critique of practical reason as such is not, in fact, the
entire purpose of the work before us. Even pure practical reason has
its dialectic. This dialectic is not a conflict between sensuously de-
termined and pure practical reason but a conflict among the Ideas of
pure practical reason itself. Hence pure practical reason also stands in
need of negative critique.
On the other hand, if we take the word "critique" in its positive
2 Critique of Judgment, Introduction, III (V, 179 [IS]).

8 Critique of Pure Reason, A II = B 15 .
• Ibid., B xxv.
Name, Purpose, and Structure of the "Critique" 45
sense, an equally good justification for calling this book "Critique of
Pure Practical Reason" can be given. For only as pure can practical
reason be legislative, and though the yet-to-be-written "Metaphysics of
Morals" would have the task of spelling out that legislation, the funda-
mental constitutional law of moral experience must be given in cri-
tique, just as the fundamental principles of theoretical reason, which
"gives the law to nature," are uncovered in the Critique of Pure Rea-
son. Accordingly, he says that the Critique of Practical Reason "gives
an account of the principles of the possibility of duty, its extent and
limits . . ." (8 [94]), and this is critique of pure practical reason in
the positive sense.
In view of these arguments and counterarguments, it seems to me
that equally good grounds can be found for either decision. For the
actual choice the decisive ground seems to have been the importance
he attached to critique in the negative sense, which he made his first
task in the book. Once the title was chosen, a historical accident was
responsible for the first paragraph of the Preface. Kant means that
the parallelism seems to demand the following pair of titles: "Critique
of Pure Practical Reason" and Critique of Pure Reason. But it does not.
The correct parallels would be either "Critique of Practical Reason"
and "Critique of Speculative [or Theoretical] Reason" or "Critique of
Pure Practical Reason" and "Critique of Pure Theoretical [or Specula-
tive] Reason." There is no book with the title of "Critique of Specula-
tive Reason" (where "speculative" means "pure theoretical"); yet
Kant suggests that his readers may expect the present book to have a
different title from the one he gave it, all because of a fancied parallel
with a title he had not used for another work!
But the historical accident was this. There was a parallelism in titles
used in the announcement of the new edition of the Critique of Pure
Reason in 1786. There it was said that "to the critique of pU,r e specula-
tive reason" there would be added a "critique of pure practical rea-
son." When, finally, for reasons that I have suggested were not par-
ticularly convincing, he chose the shorter title, the first paragraph was
necessary to tell his readers that this was indeed the work they had
been expecting since 1786. Either Kant had a very different conception
of the work in 1786, or at the time of the announcement he had not
formulated a very clear conception of what its contents would be. I
suspect that the latter was the case.
§ 3. THE PURPOSES OF THE BOOK
In the preceding section, when evaluating Kant's reasons for having

chosen the title, we have had to mention some of its purposes as he
46 Name, Purpose, and Structure of the "Critique"
states them in the Preface and Introduction. The purposes may be
stated more systematically as follows:
I. To examine reason's entire practical faculty so as to show
a) That pure reason can be practical,5
b) That empirically affected practical reason makes presumptuous
claims and must therefore be restricted to its proper bounds
(16 [102]),
c) And that pure practical reason makes no claims that are in-
terdicted by the Critique of Pure Reason (5, 6 [<)0]). But
2. Without collusion of the two Critique's, to establish the reality of
certain Ideas needed even by theoretical reason, which theoretical
reason could show only to be logically possible but not to be
really possible or to have actual objects. 6
3. To establish the principles of the possibility of duty, as laws of
pure practical reason, so that they can be applied to man as a
particular kind of empirically discovered being, in the "Meta-
physics of Morals" to which the Critique is a propaedeutic (8
[93-94])·
4. To resolve an inevitable dialectic in the judgments of pure prac-
tical reason (16 [102]).
In addition, twO other purposes of at least major parts of the book
are not explicitly mentioned in the Preface. They are:
5. To investigate and justify the fact that moral philosophy must
have precisely the systematic form it does have and no other (89
[182]).
6. To determine the way in which we can secure to the laws of pure
practical reason access to the human mind and an influence on
its maxims (151 [249]).
II Critique of Practical Reason, 3, 15.45 (88,101,134); Reflexion 7101. The answer
to the question as to how reason can determine conduct is erroneously ascribed to
metaphysics. not critique. in F oundQtions of the Metaphysics of Morals, 416 (44).
8 Critique of Practical Reason, 3, 5 (87, 89). This is shown with respect to free~
dom in the "credential of the moral law" at ibid .• 4B (137), from which the reality
of the other Ideas is inferred. Thus the Critique of Practical Reason "accomplishes
something I denied to speculative reason" (to Schutz, June 15, 1787 [X. 490)). viz ..
the step to a "practical-dogmatic metaphysics" as a pan of a metaphysics of nature.
Such a practical-dogmatic metaphysics must be distinguished from the metaphysics
of morals as a "moral-practical science of reason" (Fortschritte der Metaphysik,
XX. 193). Purposes 1 and 3 may accordingly be restated as follows: (1') to provide a
propaedeutic for a practical-dogmatic metaphysics of God. freedom. and immor-
tality and <),) to provide a propaedeutic for an immanent metaphysics of morals
as a system of aU a priori principles of morality.
Ntrme, Purpose, and Structure of the "Critique" 47
To these reasons must be added two others that are not mentioned
in the Preface or Introduction, but which have been stated in the
Foundations to require a critical examination of pure practical reason.
Since these purposes are among the most important that Kant had in
writing this book, it is astonishing that they are not given a prominent
and, indeed, dominant position here at the beginning of the book.
These purposes are:
7. To show the unity of the speculative and practical reason under
a common principle, because "in the final analysis there can be
but one and the same reason which must be differentiated only
in application."1
8. To show that a synthetic use of pure practical reason is possible,
i.e., to show how a synthetic a priori practical proposition is
possible.s
We shall consider purposes 7 and 8 before proceeding further into the
text of the Preface.
§ 4. THE UNITY OF THEORETICAL AND PRACI'ICAL REASON

The unity of theoretical and practical reason is asserted in the Cri-
tique (12 I [2 17] ), and almost the entire book can be considered an
elaboration of this. Yet one can only regret that at no point in the
book does Kant, as it were, take the reader by the hand and say, "Now
I shall show you precisely why I think theoretical and practical reason
differ only in being two applications of the same faculty." And in one
place he writes as if he does not have a proof that satisfies him; for he
says that comparison of the structures of the two Critique's "correctly
occasion the expectation of being able some day to bring into one view
the unity of the entire pure rational faculty (both theoretical and
practical) and of being able to derive everything from one principle"
(9 1 [ 184]).
Still it is fortunately not difficult to state what may have been this
"one principle," even though Kant did not and perhaps at this time
could not spell it out. Kant is concerned, in his mature ethical treatises,
to show that there are no moral concepts or practical principles that
have any basis other than the "legislation of pure reason." Reason is
the faculty of principles, and it brings all that is thought by the under-
standing under the highest unity of thought. Now if there are valid
practical principles whose necessity is not derivable from universal
1 Foundations of the Metaphysics of Morals, 39 1 (7).
a Ibid .. 445 (64). This is, of course, closely related to task III.
and necessary principles conceived only by reason, then the internal
unity of practice would itself be nonexistent or, at best, contingent.
Only reason can supply universal and necessary principles, whether to
knowle3ge or to conduct.
Kant's effort, through the first parts of the Foundations and much of
the Critique, is to show that non-rational grounds for decisions to act
are neither internally consistent, necessarily binding, nor universal in
application. Practically, they are divisive instead of unifying; no "moral
order" can be constructed on their foundation (35 [I 24]). But reason
serves the same function in the practical that it does in the theoretical
realm, that of systematizing, integrating, universalizing, and rendering
necessary what appears prima facie to be contingent.
While, in the theoretical realm, the Ideas used for these purposes are
only regulative and provide mere maxims for the conduct of inquiry
without determining its outcome or being necessary (for inquiry itself
is not necessary) (5 [89]); in the practical, the Idea of freedom is
constitutive of the experience to which it applies, for the experience
is of what ought to be (as defined by the Idea) and not of what hap-
pens to exist independently of it. We may and do mistake the function
of theoretical reason and think that its Ideas are constitutive of an in-
telligible world-of the world of things as they really are and not as
they appear. When we make this mistake, transcendental Ideas become
transcendent, and philosophical thinking falls into antinomies, paralo-
gisms, and other fallacies exhibited and eradicated in the Dialectic of
the first Critique. The same reason, following our demands for un-
conditional conditions for every motive and for the unity of motives
in a pattern of life, is, on the contrary, an immanent reason, actually
producing the objects to correspond to its Ideas. These objects, pro-
duced by us in acting in accordance with the demands of these Ideas,
are not things in the outer world, which we may have or lack the
power to effect; they are motives or states of mind or decisions of will
which directly express in actual experience the Idea of freedom, of
which the moral law is a necessary consequence. Edward Caird has
excellently said of the transition from theory to practice: "Just be-
cause reason cannot find its ideal [of necessary and universal systematic
unity] realised in the world, it seeks to realise that ideal for itself."9
Reason becomes practical, generating an Idea of a world that, through
our actions and attitudes, may be established with immanent com-
pleteness, order, and systematic unity, whether it can be actualized in
9 The Critical Philosophy of bmmmuel Kant, II, 164.

the products of human skill or not. 10 Theoretical reason, unless these
Ideas are established, vainly pursues an ever receding goal. l1
Still, if one remembers the discussion in the F oundatio1lS on "The
Extreme Limit of All Practical Philosophy," one may doubt that the
question as to why reason is practical can be answered. To show how
reason can be practical, he there tells us, is as impossible as (and es-
sentially the same as) to explain how freedom is possible. Like the
question as to why we intuit objects only in space and time, these
questions cannot be answered with an answer derived from any higher
principle. We show that they must be as they are because they are
necessary presuppositions for something that is actual; we do not
comprehend them, but we can at least comprehend their incompre-
hensibility .12
Nevertheless, an explicit answer to this question is given, not in the
Critique of Practical Reason, where, if anywhere, one might expect it,
but in the Critique of Judgment. 1s He tells us that reason always makes
a demand to find the unconditioned condition for all conditions. In
experience, this demand cannot be met, and it leads to theoretical spec-
ulation about an unconditioned thing in itself, an unconditioned sub-
stance, and an unconditioned cause. But in speculation these Ideas are
only assumptions, not cognitions, and no theoretical conclusions can
be drawn from the fact that we assume them for the guidance of in-
quiry. In the practical realm, the unconditioned necessity is reason's
own causality, under the name of freedom, and this is revealed to us,
3;S a fact, in our awareness of the claims of duty. What is commanded
by reason's imperative, however, may not take place in the phenomenal
world. We therefore draw a distinction between what does and what
ought to happen. Practical reason establishes the latter concept, theo-
retical reason the former. There seems to be an unbridgeable concep-
tual chasm between them, even if, in fact, what ought to be can be
made actual by an act of will and the ensuing behavior.
This conceptual distinction, however, is drawn by reason only be-
10 Cf. Critique of Pure Reason, A 548 = B 576; Critique of Practical Reason, 15

(JOI) •
11 Critique of Pure Reason, A 7¢ = B 814-
12 Foundatiom of the Metaphysics of MOTals, 463 (84); d. Critique of Pure
Reason, A 393.
13 ~76• The idea is put forward in Critique of Pure Reason, A 816 = B 8440 that
the unity of the two legislations is to be based on the regulative Idea of God as the
author of both types of law, the theoretical and the practical.
50 Nrrme, Purpose, tmd Structure of tbe "Critique"
cause of a peculiarity of the human mind. If our reason were an au-
tonomous sufficient cause of its objects, the distinction between what
is and what ought to be would never arise, just as in a world in which
reason had the power to produce objects by merely thinking of them,
the distinction between the necessary, the actual, and the merely pos-
sible would have no place. It is only owing to the subjective constitu-
tion of our minds-that reason as a cognitive faculty is applicable to
objects only under the condition that we passively receive sensible
data, while as a pure practical faculty it is independent of what is
given by the senses, since it is not in the least a faculty of knowing
objects14-that we do not immediately see that there is a single reason
in two different relations. For a differently constituted intellect, an
intuitive understanding, the distinction would not arise. It is a con-
tingent fact that our reason is discursive, i.e., that it must begin with
data and move to universals, and is not capable of producing data
simply by thinking of them. This fact is as contingent and inexplicable
as the fact that we experience objects only in space and time. We can
imagine another kind of mind, for which pure reason would directly
give knowledge of objects, and of objects as they are and not as they
appear to us. We have no grounds for supposing that there is such an
intellect, but we do not need to assume its existence; its mere possibility
is sufficient to render intelligible to us the way in which a distinction
which seems at first to be absolute can, in fact, be thought of as only
a difference in relation. From the possibility of an intuitive intellect
and from the fact that our intellect is not intuitive, we can see that it
is only for us (and, so far as we know, all actual minds) that there is
a distinction between what is and what ought to be, between the
necessary and the possible, between the mechanically and the teleolog-
ically caused.
So far from there being an opposition between theoretical and prac-
tical reason, therefore, Kant believes that there is one reason that carries
out the same function in two different applications-this function
being to supply unconditioned conditions for all that is empirically
14 The practical counterpart of the intuitive intellect as a possible cognitive faculty
is the holy will (Critique of Practical Reason, 81 [175]). Each is possible (conceiv-
able), but we have no reason to think that either of them exists, except insofar as we
have reason to believe that God exists. But each is an important conception, for, by
comparison with it, Kant brings out more clearly the peCuliarities of the human
cognitive or practical faculty. For a holy will, the concept of duty would have no
application, since duty presupposes that there are other grounds of choice, different
from and in conflict with the legislation of reason. For it, the difference between
what is willed and what ought to be willed would not arise. Similarly, for an intui-
tive intellect the distinction between the actual and the necessary would not arise.
Name, Purpose, tmd Structure of the "Critique" 51
conditioned. But the Critique of Practical Reason is very much con-
cerned with the apparent opposition between practical and theoretical
reason, because this felt opposition is the setting for the concept of
duty, respect for law and virtue, and the distinction between knowl-
edge and faith.
S5. HOW IS A SYNTHETIC A PRIORI PRACTICAL PROPOSITION POSSIBLF.?
This question is different from that of task I (a), which is to show

that pure reason can be a directly determining ground of will and have
a direct influence upon conduct. It also differs slightly from the
formulation in the Foundations of the Metaphysics of Morals, where
Kant asks, How is a synthetic use of pure practical reason possi~le? lIS
The question as stated here does not occur in the Critique but first
appears in print in the Metaphysics of Morals. i6 But that Kant knew
that the question needed an answer is shown by its formulation in
Fragment 6. 17 In view of the prominence given to the analogous theo-
retical question in the Critique of Pure Reason (especially in the second
edition) and in the Prolegomena, it is a bit strange that it is not even
mentioned in the second Critique, where one would expect it to have
a programmatic position.
The first two sections of the Foundations examine the phenomena
of morality in order to discover what would be its formula for action;
and this formula is presupposed by the Critique (8 [93]). This inquiry
issues, in the first part of the Foundations, in the statement that moral-
ity is solely a matter of will or motive, not of circumstance or con-
sequence, and that a good will is a will that acts from the motive of
doing an action because it is the right action, though any will may in
fact do the right action from other motives. A good will is one which
does the action from respect for law, not from the desire for this or
that consequence, though there is always some desire to effect some
change in the world by any action voluntarily undertaken. In Part II,
this examination leads to the formula of the categorical imperative as
the presupposition of and the criterion for a motive which could be
considered morally good.
But the first two sections, though containing many statements that
can be most readily interpreted as asserting real facts and expressing
Irs Foundations of the Metaphysics of Morals, 44S (64). My translation. in the
Chicago edition (p. 101) is in error. The penultimate sentence of the last paragraph
of Section II should read: "But we must not venture on this [synthetical) use with·
out first making a critical examination of this faculty of reason."
16 Metaphysik der Sitten, Rechtslehre. § 6 (VI. 149).
17 Refiexion 7101 (XIX, z8z).

52 Name, Purpose, I111d Structure of the "Critique"
valid moral commands, also contain statements that remind the reader
that the purpose of these sections is analytical and that their conclu-
sions are put forth problematically; they give the structure of morality,
whether or not such a thing is real. The principles are enunciated as
those which must be true "if duty is not a vain delusion and a chimeri-
cal concept,"l8 and their conceptual correctness is independent of an
answer to the question as to whether "true virtue can be found any-
where in the world."l9
In bringing these principles to light, Kant has followed an analytical
method in beginning with "common rational knowledge of morals"
and "popular moral philosophy" assumed as the prius, just as in the
Prolegomena he assumed the validity of mathematics and natural sci-
ence and asked what must be true if they are valid knowledge. But
just as the Prolegomena is a grand petitio principii if taken as an answer
to skepticism in isolation from the Critique of Pure Reason, the first two
parts of the Foundations are a petitio principii if taken in isolation from
the third part or the Critique Of Practical Reason. (Or it would be
better to say that they would have been a petitio principii, had not
Kant kept clearly in mind throughout that they constitute an exposi-
tion of a concept, not an argument.)
It does not follow that the analytical method will lead only to ana-
lytical judgments, and, indeed, it there leads to the formulation of the
categorical imperative, which Kant calls an "a priori synthetic practical
judgment." Synthetic propositions cannot be justifi~d by analysis,
though they may be discovered in this way; and if they are a priori,
they cannot be justified by any appeal to the facts of experience. They
must be justified, if at all, by a critical inquiry like that of the Critique
of Pure Reason. We must ask, How are such judgments possible? What
right do we have to assert such propositions? Only when we have
answered this question satisfactorily can a synthetical use of the prin-
ciples be made; only then can we know that the formula of a possible
moral command is the fonnula of a real and legitimately issued com-
mand.
When Kant thus states the problem of Section III of the Foundations
and reverts to it in the Critique (89 [I82]) in a passage that he espe-
cially urges in the Preface (8 [93]) upon the reader's attention, he is
saying that the Critique of Practical Reason has the task of showing (a)
that pure reason does have a real use in practice, not a merely logical
use in the hypothetical analysis of a possible phenomenon of morality
18 Foundtztiolls of the Metaphysics of Morals, 445 {64>.

19 Ibid., 407 (24).
and not a merely logical (technical) use in organizing our experience
for the sake of pragmatic success,20 and (b) that the synthetic a priori
judgments found in the analysis of morality are actually justified. Pur-
pose a is accomplished in the Analytic as a whole, and purpose b is
assigned to that part of the Analytic called the Deduction.
§ 6. RELATION TO THE "METAPHYSICS OF MORALS"

I have already discussed some of the historical relations between
this book and the Metaphysics of Morals, and in § 8 I shall have to
return to some of them in explaining the polemics of the Preface. In
this section, I shall consider one paragraph and one footnote in which
Kant speaks of the relation of the Critique to the "system of science" (8
[94]) and to the "metaphysics of morals" to which the Critique is only
a preliminary (16 I [260]).
We are told that this Critique is not to give us a classification of the
practical sciences, as the first Critique did of the theoretical sciences,
for the division of theoretical sciences there (into mathematics, natural
science, and metaphysics) is said to be entirely a priori, while the di-
vision of practical sciences (into the Doctrine of Right and the Doc-
trine of Virtue) requires empirical knowledge of man for the specific
definition of duties as human duties. Such knowledge is not a part of
the critical examination of practical reason but presupposes both cri-
tique, the elaboration of its a priori principles, and empirical knowledge
concerning human nature.
Yet the footnote, in answering a possible criticism of his general
position that it was based on the alleged facts of psychology, gives
definitions of life, faculty of desire, and pleasure, as the Critique of
Pure Reason said would be necessary,21 and declares that these are all
the definitions or facts that he needs that are empirical in character.
Even these facts are not introduced as direct evidence but, as it were,
obliquely, in order to show that psychology does not need to define
these concepts in such a way that pure practical reason would be
impossible.
But even this much empiricism seems to be incompatible with the
purposes of a critical examination of pure practical reason, and it
raises the following question: Just how "pure" can moral philosophy
be? Kant is not very explicit or consistent in his answer to this ques-
tion; he seems always to have been striving for a degree of purity
which could be obtained only in the emptiness of Wolff's "universal
20 This is equivalent to purpose lao
21 Cf. above, p. 9, n. Zl.

practical philosophy." But we may discern at least five levels of purity
involved in the articulation of Kant's total plan:
I. Moral philosophy independent of the peculiar nature of human
reason?2 (e.g., that we do not have an intuitive understanding) and
dependent only upon the fact of pure reason, which is not an empirical
fact (31 [I2oD-metaphysics of morals as envisaged in the Foundations.
2. Moral philosophy dependent upon level I and upon the three
definitions drawn from psychology, which give the ground for the
concepts of imperative, respect, and duty-Critique of Practical Reason.
3. Moral philosophy as the systematic development of principles in
2, independent of, but applicable to, human nature in the variety of its
forms as empirically known-metaphysics of morals in the book with
that title. 23
4. System of practical philosophy ("system of the science") as sys-
tematic exposition of 3, along with the empirical facts-repeatedly men-
tioned, but never written.
5. Moral and pragmatic anthropology-episodic elaboration of prac-
tical rules-Lectures on Ethics and Anthropology.
So the sharp line separating critique and system is blurred, if not
breached. And the book Metaphysics of Morals is much more like 4
than like 3. The only metaphysics of morals that Kant ever wrote, in
fact, was the Critique of Practical Reason, for only it is "a system of
knowledge a priori from mere concepts"-if, indeed, it is that.
Kant never wrote, or at least never completed, the "metaphysics of
nature" to which the Critique of Pure Reason was propaedeutic; all he
published in this direction was still another propaedeutic, the Meta-
physische Anfangsgrunde der Naturwissenschaft. Metaphysics that
could "come forth as science" had already come forth in the Critique's
themselves. The two metaphysics and systems were e,ver receding
ideals, and any further metaphysics, strictly conceived, after the Cri-
tique's would have been supererogatory,24 though they might have
22 Foundations of the Metaphysics of Morals, 41 %(%8).

23 Metaphysik der Sitten, VI, %16-17.
24 As critique grew and system receded, Kant once denied that the Critique of
Pure Reason was propaedeutic and stated that it was itself system (Erklarung in
Beziehung auf Fichtes Wissenschaftslehre, XII, 37()-71). This statement, I believe
unique in Kant, can best be explained by his pique at Fichte, who claimed to have
completed the work which his teacher had only begun, and probably it does not
represent any fundamental surrender of his own architectonic ideal. It nevertheless
expresses an opinion to which one is brought by examination of the actual contents
of the "critical" and the "systematic" works.
Name, Purpose, and Structure of tbe "Critique" 55
given Kant occasion to do what he had repeatedly said would be a
pleasure he had to refrain from-the systematic, analytic, or encyclo-
pedic unrolling of consequences from his a priori principles and
concepts.
~ 7. THE STRUCTURE OF THE BOOK
In this section I shall deal with the over-all structure of the Critique.
Details of the various subdivisions will be discussed as they are reached
in the course of the commentary.
Kant's architectonic has been treated sometimes with amusement,
often with contempt, and almost always with impatience. It has been
argued that, had he not held to the rigors of his architectonic schemes,
his works could have been organized in a less forbiddingly scholastic
pattern and could have been made shorter, since his ideal of complete-
ness of articulation led him to write whole chapters (e.g., at the end
of the first Critique) in which he had little interest and little to
contribute.
The Critique of Practical Reason is almost wholly free from exces-
sive architectonic zeal; only two parts may seem to owe their origin
to a forced parallelism with the first Critique, and in the course of
this commentary we shall find reason to believe that they are inte-
grally necessary parts of the organic structure of this work.
The organization is easily grasped but presents some interesting
features that call for discussion. There are two statements of the plan
of the work (16, 89-91 [102, 183-84]). In the latter statement there
is a comparison of the structure of the two Critique's; but, in making
this comparison, Kant incorrectly recalled the division of the first
Critique. This error is itself instructive. The accompanying tabulation
presents the structure of the relevant parts of the first Critique in fact,
their structure as Kant recalled it, and the structure of the correspond-
ing parts of the second Critique.
Actual Structure of Remembered Structure of Structure of Second

First CriliqlU First Criliqtu Critique
Transcendental Aesthetic Transcendental Analytic Logic

Transcendental Logic Transcendental Aesthetic Analytic
Analytic Transcendental Logic Of Principles
Of Concepts Analytic of Concepts Of Concepts
Of Principles Analytic of Principles
" Aesthetic"
Dialectic Dialectic Dialectic
The first column, showing the actual scheme of the Critique of Pure
Reason, gives a division according to the cognitive faculties involved,
viz., sensibility, understanding, and reason. The second, showing the
scheme as he remembered it in 1787, is a division according to the
objects of knowledge, with the Analytic including the Aesthetic be-
cause both are involved in the knowledge of phenomena and both are
set apart from the Dialectic, which deals with alleged knowledge of
noumena. In one respect, the division in the second column represents
the more important division of the material as a critical ontol ogy 25
of two realms, one theoretically valid and one theoretically invalid, the
second to be given practical validity in the second Critique. 26
§ 8. THE POLEMICS OF THE PREFACE

I have suggested that the reviews of Kant's earlier works contributed
to his decision to write the Critique of Practical Reason. Certainly,
they are among the first things he takes up, and large parts of the book
can be interpreted as responses to the criticisms mentioned in the
Preface. Kant's views were not discernibly modified by any criticisms;
the parts of the Critique of Pure Reason which were surrendered when
he came to the second were not surrendered because of any external
criticism of those parts but because of his own development and self-
criticism. There was little dialectical play between Kant and his critics.
Yet their statements were vivid in his mind when he set about to write
this book, and it will be worthwhile to consider them, if not as
philosophically significant, yet as stimul~ to some of the particular
turns of his mental operations in 1787.
Immanuel Kant was not a contentious man,27 and in his long career
of over forty years few of his works can properly be called "polemi-
cal." Yet he was bent upon making himself understood and was re-
peatedly disappointed to discover how little he had succeeded in his
25 Critique of Pure Reason, A 247 =
B 303: the proud name of "Ontology" must
give way to that of mere "Analytic." But Analytic is called "Ontology" in Fort-
schritte der Metapbysik, XX, 260.
26 For the meaning of "Analytic" and the reasons for the parts of the Analytic
being taken in an order the reverse of that in the first Critique, see chap. v, ~§ I and 2.
On the concept of an "aesthetic" of practical reason see chap. xii, §§ I and 2. On the
tenn "dialectic" and the structure of the Dialectic of Pure Practical Reason, see
chap. xiii, H I and 2.
27 At least, this seems to have been his estimate of himself (d. letter to Reinhold,
December 28 and 31, 1787 [X, P4]). His biographers frequently remarked on his
dislike of contradiction and argument, though he valued stimulating interchange of
different opinions. But as he grew older, they said, he became less and less able to
enter sympathetically into any views not his own.
Name, Purpose, tmd Structure of the "Critique" 57
efforts. We today ought to have a certain sympathy for those poor
men who were given the task of writing the first reviews of his difficult
books, which were far more puzzling then than now, when we have
the benefit of the work of generations of expositors, critics, and com-
mentators. Kant himself, however, usually responded to the reviewers,
not with the sympathy that can be expected from few authors but
with impatience and occasionally with justified contempt. He was held
back from more frequent explicit replies only by the importunities of
his friends, who saw better uses for his time and who took most of
the polemical burdens on their own shoulders.28
But inasmuch as one could not go forward in the critical philosophy
without understanding what had already been accomplished, Kant in
the Critique of Practical Reason (7 [92]) and in other works felt con-
strained, from time to time, to "re-examine" the "concepts and prin-
ciples of pure speculative reason" which, he learned from t.he reviews,
had been misinterpreted. Granting the need for answers to them, he
spoke of the Critique as being "better than all controversies with [his
reviewers] Feder and AbeI."29 Yet the reviews of the Foundations and
the rankling he continued to feel at the Garve-Feder review of the
Critique of Pure Reason certainly affected the work he did in 1787;
and it is possible to show in detail what these effects were and to
discover their causes.
The Critique, though born in controversy, is not, however, ex-
plicitly polemical after the Preface, with the exception of one passage;
it is dialectical (in a non-technical sense), in that it presents a struggle
among opposing ideas, but not forensic, as a struggle between opposing
men. The Preface, however, is in large part polemical, for he pays
personal attention to his reviewers; and the objections to which he
replies and the way in which he replies to them indicate the situation
which underlay the decision to write the Critique and which are re-
sponsible for some of the major parts of the book.
The polemical issues referred to in the Preface may be divided into
two kinds. The first concerns the criticisms repeatedly made which
permitted no specific answer because of their blanket character. Such
were the criticisms of those who boasted of knowledge that Kant had
28 Cf. letters from Biester. June II, [786; from Schiitz, March 23, [787; and from
Bering, May 28, [787 (X, 457, 480, 488); also Critique of Pure Reason, B xliv.
29 To Schiitz, June 25,1787 (X. 490). Schutz had written on March 23: "It is fine
that you hold yourself back from refutations, and plan to follow your own path
undisturbed" (X, 480). Cf. also the letter to Reinhold accompanying the gift of the
book, December 28 and 31, 1787 (X, 5(4).
declared and, he believed, proved to be impossible. so To this class be-
long also the criticisms by those who could not believe that ethics re-
quired so subtle a foundation as Kant had supplied,31 who charged
unspecified inconsistencies in the critical philosophy, who blamed Kant
for inventing "a new language" with "far too frequent use of abstract
terminology" in which to enunciate "in incomprehensible language
long-known things as if they were new."32
A second group of criticisms dealt with in the Preface provide a
more specific target and often show more philosophical acumen. They
concern issues which Kant mentions in the Preface, but the full answer
to all of them except the last is found only in the body of the Critique
itself. There are four such criticisms, and we shall deal briefly with each.
a) The alleged inconsistency involved in "the denial of the reality of
the supersensible use of the categories in speculation" and the assertion
of their validity "in respect to the objects of pure practical reason" (5
[90]). This "most weighty of the criticisms of the Critique [of Pure
Reason]" (6 [9 I]), involving also the questioning of the distinction be-
tween the noumenal and phenomenal selves, was made by H. A. Pis-
torius, in his review 33 of the Foundations and of Schultz's Erliiuterun-
gen uber des Herrn Prof. Kants Kritik der reinen Vernunft. 34 It is like-
30 "They wish to prove; very well, let them prove, and the critical philosophy
will lay down its weapons before them as victors" (5 [89]). The allusion is to Horace
Satires I. I. 19. This is an echo of the challenge to Garve and Feder in Prolegomena,
IV, 278, 368 (Beck, 25, 117). Certain turns of phrase here and in the review of J. A. H.
Ulrich's Eleuthereologie (VIII, 453-60) published by Kraus on the basis of exten-
sive notes prepared by Kant lead me to conjecture that Ulrich may have been one
of the targets of this paragraph, though the Eleuthereologie was apparently (but
not certainly) published too late for Kant to have read it when writing the Preface.
We must remember, however, that Kant saw many books before publication and
that he undoubtedly knew Ulrich's somewhat similar Institutiones logicae et meta-
physicae of 1785 (cf. Ulrich's letter to Kant, April 21, 1785 [X, 402-3]). Vaihinger,
the discoverer of the Kantian provenance of the Kraus review of Ulrich's book,
studied this controversy but did not suggest that Ulrich may already have been in
Kant's mind as an opponent to be answered in the Preface (d. Hans Vaihinger,
"Ein unbekannter Aufsatz Kants liber die Freiheit," Philosophische Monatshefte,
XVI [1880], 192-209).
31 These criticisms were reported to Kant, but not indorsed, by Jenisch in his
letter of May 14> 1787 (X, 4B5-87); they seem to be alluded to in the paragraph be-
ginning on Ak. 8 (93) of the Critique.
32 Critique of Practical Reason, 10 (1]6), and the important footnote on "formula"
at ibid., 8 n. (93 n.), directed against G. A. Tittel, tiber Herrn Kants Mora/reform
(Frankfort and Leipzig, 1786). Cf. also Critique of Pure Reason, A 831 B 859, =
and Critique of Practical Reason, 105 (100).
33 Allgemeine deutsche Bibliothek, LXVI (1786).447-63.
84/bid., pp. 92 if.
Ntm1e, Purpose, and Structure of the "Critique" 59
wise present, in less pointed form, in the criticisms by J. F. Flatt.35
Kant refers to Pistorius as "truth-loving and acute, and therefore
worthy of respect,"36 and it is to his criticism that it is not permissible
to apply the categories to the noumenal that Kant makes the statement
quoted earlier: "Only a detailed Critique of Practical Reason can set
aside all these misconceptions" (6-7 [91-<)2]). Few other topics occupy
a larger part of the Critique proper. It is the central theme of Section
II of chapter i of the Analytic, "Of the Right of Pure Reason to an
Extension in Its Practical Use Which Was Not Possible to It in Its
Theoretical Use."37
b) The alleged circularity in the relationship of freedom and the
moral law in the Foundations. It had been said by Flatt38 that each was
used to prove the other and that this was a vicious circle. Kant men-
tions and replies to this in a very concise and effective manner in the
Preface (4 n. [88 n.]), and the full refutation of it occupies the main
35 Tiibinger gelehrte Anzeigen, May 13, 1786. Kant returned to deal with Flatt,
one of the critics of whom he was most contemptuous, in Metaphysik der Sitten, VI,
1°7·
36 Critique of Practical Reason, 8 (94). This "truth-loving critic" is even more
fulsomely praised in Opus posttmtum, XXI, 416. Contrast with the contempt for
Flatt in Critique, 14 (100).
lIT Pistorius' review must have been known to Kant when he wrote the Preface to
the second edition of the Critique Of Pure Reason. But he said there (B, xli) that he
made no changes after the tirst chapter of the Dialectic "because I have not found
among competent and impartial critics any misapprehension of the remaining sec-
tions." Yet Pistorius' questions did concern the status of the Ideas and the noumena-
phenomena distinction. I do not know what may have been Kant's intentions with
respect to the Pistorius criticism in April, 1787. Perhaps the review by Pistorius was
in part responsible for his rewriting the section on the Paralogisms, though credit
for this is generally given to the Garve-Feder review, with its accusations of ideal-
ism. Whatever the cause, only a few weeks later he found that the understanding of
the Dialectic as a whole was so inadequate that it had to be explained again in the
second Critique. It may be worth noting that Kant praises Pistorius both times (ef.
n. 36 above) in the context of criticism (c) instead of (a).
88 Kant himself had mentioned the apparent circle in the argument (Foundations
of the Metaphysics of Morals 450 [6c))), and he explained it again to Kiesewetter
(letter of April 10, 1790 [XI, ISS]). But in the Foundations he pointed out a some-
What different route of escape. There he emphasized that there is in the experience of
thought itself an awareness of free spontaneity, and this is independent of the ques-
tion of the validity of the moral law. Even if one denied the reality of freedom in
words, he would be expressing his freedom as the precondition for his argument's
having any legitimate claim to validity. This argument is repeated in the review of
Schulz's Sittenlehre (VIII, 13) but is absent from the second Critique. The thought
was an enduring one, however, and the term "autonomy" is extended tinally to cover
theoretical as well as practical reason in Opus postu'I7rUm, XXI, 93, [00.
portion of Section I of chapter i of the Analytic, "Of the Deduction
of the Fundamental Principles of Pure Practical Reason."
c) The error, alleged again by Pistorius, that "the concept of the
good was not established before the moral principle."39 This is admit-
ted, and it is defended in chapter ii of the Analytic.
d) The alleged discovery that "there can be no a priori knowledge
at all" (I2 [97 D. This was the view of Feder,40 who had revised the
review submitted by Garve of the Critique of Pure Reason. By this
time, of course, Kant knew to what extent Feder was guilty and Garve
was innocent of the worst features of that review, to the stimulus of
which we owe so much of the Prolegomena. So Kant allowed himself
to indulge in bitterness and irony in dealing with this alleged dis-
covery.41
In reply to this, Kant takes the following course. He first shows that
it is inconsistent, for what we know by reason is a priori, and to prove
that we have no a priori knowledge would be to prove by reason that
reason is not valid. Next he criticizes the Humean conception':'! of cau-
sation, on the ground that its substitution of common human consent
for the a priori principle is inexplicable if the concept is "fundamen-
tally false and a mere delusion of thought," which he believes is the
correct inference to be drawn from Hume's premises. But, third, he
exculpates Hume from the absurdity of the doctrine of attempting to
prove by reason that there is no reason, by pointing out that Hume,
unlike the unnamed critic Feder, did not make his empiricism so uni-
versal as to include mathematics in his strictures. Mathematics escaped,
however, only because of Hume's error in regarding mathematical
39 Critique of Ptactical Reason, 8--<) (94--<)5) i cf. 62-64 (154-55).
'0 Uber Raum und Caussalitiit [sic] zur Prufung der Kantischen Philo sophie
(Gottingen, 1787). Kant identifies Feder as his target in the letter to Schutz, June 25,
1787 (X, 4(0).
41 With his characteristic and unvarying admiration for Hume, Kant first dis-
tinguished between the genius who made the original discovery and the epigone
who reduced it, in his opinion, to an absurdity, and he resen'ed his scorn for the
latter. The note at the end of the passage (13 n. [100 n.]) is a defense against mis-
labeling Hume a "genuine empiricist"; but in this defense Kant could not resist
making one more indirect rejoinder to Feder's having labeled him an idealist (cf.
the comparable footnote in Prolegomena, IV, 375n. [Beck, 124n.]). The allusion to
William Cheselden (J688-J752) refers to the physician who cured a blind man and
whose report on his adjustment to seeing was repeated by Herder (Vierter Wald
[J769], Werke [Suphan ed.]. IV. 49) and thus, probably, passed on to Kant.
42 I see no reason to think that Tetens was the opponent whom Kant had in mind
here, as asserted by Cay von Brockdorff, Die deutsche Aufklarungsphilosopbie
(Munich, 1916), p. 105.
N tr111e, Purpose, and Structure of tbe "Critique" 61
judgments as analytic. But though Hume was wrong in holding them
to be analytic,43 what is important here is that he correctly held them
to be a priori and thus maintained an internal standard of criticism for
other forms of knowledge. Genuine empiricism is not just skeptical of
philosophical speculation but would destroy both mathematics and
morality, and on these two counts it is not, according to Kant, to be
taken seriously, since here a philosophical theory, itself speculative,
directly collides with the "highest evidence."
A lesser man might have spent his limited time in extended debate
with his critics or (at his own expense) have ignored their objections
and misunderstandings. But it is one of the marks of Kant's greatness
that he was able to deal systematically, if not sympathetically, with
other men's errors and misunderstandings and out of the bits and pieces
of scholarly journalism to take materials to be built into the permanent
edifice of philosophy. It is to this aspect of his talent that we owe much
of the Critique of Practical Reason. The signs of it are scattered
throughout his work, but they are concentrated here in the 'Preface.
43 In Enquiry concerning Human Understanding, Sec. VII. Elsewhere (Critique of
Practical Reason [53 (143) J) Kant says Hume's views lead "inevitably to skepticism"
even with respect to mathematics. This might refer to Hume's actual, though
passing. skepticism of geometry in the Treatise, while the remark in the Preface
refers to the Enquiry. But it is more likely that the judgment does n()t pretend to
state a historical fact, but means only that empiricism, once on the skeptical path,
stops neither at the "chief branches of knowledge" nor at "ordinary reason," which
was the refuge of the common-sense philosophers, who wished to be skeptical of
rationalistic metaphysics while remaining credulous of other kinds of knowledge
that would be destroyed in a genuine and thoroughgoing skepticism.
Part II
v
Survey of the Analytic of Practical Reason
§ I. MEANING OF THE TERM "ANALYTIC"
Throughout the Critique, Kant uses the word "analytic" as though it
were entirely familiar to his readers. But we must refer back to the first
Critique if we are to understand precisely what he means by this word
and why he chose this name.
General logic, he there tells us, contains an analysis or resolution of
the fonnal procedure of understanding and reason into their elements
and exhibits them as principles of a logical criticism of knowledge; as
general logic, it entirely abstracts from all content of our judgments
and concepts and deals exclusively with their fonnal relations.! Tran-
scendental logic, on the other hand, does not disregard the important
distinction between a priori and a posteriori knowledge and is con-
cerned with the grounding and structure of concepts and judgments
known a priori. Transcendental logic, therefore, contains only rules of
the a priori thought of objects, and hence it excludes other concepts
and judgments which are fonnally identical with them as respects their
logical interrelations but which, as regards their cognitive content, are
drawn from experience.2 The analytical part of transcendental logic
resolves the a priori or "transcendental" operations of understanding
and reason into their elements, and thus it serves as the standard of
criticism for all our knowledge of objects, including even empiTical
knowledge. The transcendental analytic, therefore, is a "logic of truth"
and not, like the analytical part of general logic, a logic of mere formal
consistency. Transcendental analytic consists in the "dissection of our
a priori knowledge that pure understanding of itself yields"; the con-
cepts and principles which it exhibits must be pure and not empirical,
intellectual and not intuitive, simple and not composite, fundamental
and not derivative, systematically complete and not rhapsodically ag-
gregated.a
1 Critique of Pure Reason, A 60 = B 84-
2 Ibid., A 56 =B 60. a Ibid., A ~ =B 8<}.
65
66 Survey of the Analytic of Practical Reason
Except that it is practical reason and not understanding or theoretical
reason that is here under examniation, the term "analytic" has the same
meaning in the present Critique. The analytic is the "rule of truth" in
practical philosophy, and its most important tasks are the differentia-
tion between the (empirical) doctrine of happiness and the (pure) doc-
trine of morality (92 [186]) and the systematic exposition of the a
priori principles and concepts of the latter.
§ 2. THE STRUCTURE OF THE ANALYTIC

Kant says: "We begin with principles and proceed to concepts, and
only then, if possible, go on to the senses, while in the study of specu-
lative reason we had to start from the senses and end with principles"
(16 [102]; 90 [184]). There are two points to be examined here: the
causes of the reversal of the order of the parts of the Analytic, and the
presence or absence of an Aesthetic.
a) The order of the parts.-Two slightly different reasons are given
for the reversal of the order: (i) Since we are here dealing with a will
and not with knowledge of objects, the determination of the will by
principles and not by concepts of objects must first be established (16
[ 102]). This follows from the definition of will. (ii) It is easy to estab-
lish the principles, since there is in every natural human consciousness
a recognition of the principle as a law completely a priori and inde-
pendent of the senses (91 [185]), and therefore independent of con-
cepts of objects and of the givenness of objects to the senses.
The concepts of the objects of practical reason (either empirical, as
the concept of the object of some specific desire, or the pure concepts
of the objects of a pure practical reason, which, as we shall see, are the
moral good and evil) cannot be given prior to, or independently of,
this principle. If it were, either we should have no a priori knowledge
of good and evil (as we have no a priori knowledge of empirical ob-
jects of desire or of the empirical characteristics of objects of empirical
knowledge), or else we should have to have a moral sense, as a kind of
a priori intuition of moral value. But Kant asserts that the principle is
the prius of the concept, for the concept can be drawn only from a
principle or from an intuition like that of the alleged moral sense; and
he denies the cognitive validity of the so-called moral sense" The rela-
tion of the principle to the concept is accordingly compared with that
of the major to the minor premise in a syllogism and not to that of
generalization to fact of observation (90 [184]).
4 Critique of Practical Reason, 38-39 (129): Such an intuition would have to be a
"feeling for law as such" and would "regard as an object of sensation what can
only be thought by reason."
Survey of the Analytic of Practical Reason 67
b) An aesthetic of pure practical reason?- Transcendental Aesthetic, in
the Critique of Pure Reason, has nothing to do with the theory or judg-
ment of beauty but is "a science of all principles of a priori sensibil-
iry."15 It is concerned with intuition, the faculty by which our concepts
are brought into relation with their objects. It is a peculiarity of the
human intuition that it is always sensible, sensibility being the mode in
which we are affected by the objects. 6 Hence the forms of our sensible
intuition, which are space and time, are necessary in our knowledge of
objects as appearances. A critique of knowledge must contain an anal-
ysis of the a priori principles or forms of sensibility, and, since sensi-
bility furnishes objects for our concepts and principles, such an analy-
sis appropriately precedes the Analytic as a part of Transcendental
Logic.1
Practical reason, however, is not concerned with the relation of
knowledge to objects which may be given to us, but with the bringing
into existence of objects which are not passively given to us. Hence it is
not surprising that there should be nothing in practical reason exactly
corresponding to the Transcendental Aesthetic and that anything analo-
gous to it should come after the discussion of the principles and con-
cepts, not before it. So Kant says that, after establishing the principle
and the concepts, he will, "if possible, go on to the senses." "Only then
could the last chapter [of the Analytic 1, dealing with the relation of
pure practical reason to sensibility and with its necessary influence on
it, i.e., the moral feeling which is known a priori, close [the Analytic]"
(90 [ 18 3]).
There are two things to be noted here: (i) We are concerned not
with the effect of objects on the sensibility, since we are not concerned
with relating our concepts to given objects as objects of knowledge,
but with the effect of reason itself and its principles and concepts upon
our sensibility. The problem is that of the subjective determination of
the will, the subjective factor being provided by sensibility or feeling
and the determining factor by the principles of practical reason. The
analogous problem in the Critique of Pure Reason is perhaps that of the
pure imagination. which represents in the sensibility the synthetic op-
erations of the understanding. (ii) The mode of sensibility which is
II Critique of Pure Reason, A %1 = B H.
'Ibid., A 19 = B 33.
T We have noted above that in the Critique of Pure Reason the Aesthetic precedes
and is not a part of the Logic, of which the Analytic is the first pan, whereas in the
Critique of Practical Reason the part analogous to Aesthetic is made a part of
Analytic but is still separated from the Logic, which is only a subdivision of the
Analytic (el. chap. iv, § 1).
relevant here is not sensation, referred to outer objects, but feeling,
which is without cognitive function.
The concepts of pure theoretical reason do not apply directly to in-
dividual sensuous representations. That which mediates between them
and makes the application of the categories to experience possible is the
schema of the concepts. There is something analogous to this mediating
function in pure practical reason in what Kant calls the "type of pure
practical judgment." The analogy between the typic and the proper
function of an aesthetic is perhaps closer than that between the chapter
on incentives and the Aesthetic of the first Critique. But the closest
analogy is between the discussion of typic and the discussion of the
schematism; and, since Kant himself uses the word "aesthetic" to refer
to the material of chapter iii instead of that at the end of chapter ii of
the Analytic, we shall follow his suggestion.
§ 3. THE PROBLEMS OF THE ANALYTIC
If morality is not to be declared "a vain and chimerical notion," Kant
must solve three problems. They are as follows:
I. To formulate the law the knowledge of which could, independ-
ently of impulse, be a motive for action.
II. To prove that pure reason can be practical, i.e., the conception of
the law formulated in answer to Problem I can indeed be a motive for
action.
III. To bring to light, in the nature of a rational being in general and
in man as a rational being sensuously affected, those factors which
make it possible that knowledge of this law can in fact be a motive for
action.
Kant's solutions to these problems constitute the main body of the
Analytic. Since the problems are not isolated from one another, he at-
tacks first one and then another, using partial solutions to the one in
order to gain a foothold in approaching the others, in such apparent
confusion that the entire Analytic has sometimes seemed to be a lifting
by transcendental boot-straps. But this apparent confusion is a conse-
quence of the actual relations existing among the problems; each one is
so intimately related to the others that a full answer to anyone im-
plicitly contains answers to all of them. Roughly speaking, however,
chapter i gives his solution to Problems I and II, though chapter ii con-
tinues the solution of Problem I, reformulated as a question concerning
what concepts can legitimately enter into a law the knowledge of
which is to be a motive for action. Chapter iii constitutes the main
answer to Problem III.
While it is hazardous to press analogies between the first and second
Critique's further than Kant himself, the great analogizer, carried them. 8
it is interesting to observe a parallelism that Kant does not point out
between these three problems and those of the Analytic of the first
Critique. \Ve shall find that everyone of the three principal questions
has its counterpart in clearly separated problems of the first Critique.
Problem I is a problem formally like that of the Metaphysical De-
duction of the Categories, i.e., that part of the Analytic which shows
what the pure concepts of the understanding are and hence indirectly
determines the a priori principles of understanding and the Ideas of
reason. Here there is likewise a "metaphysical deduction" of the prin-
ciple of pure practical reason, i.e., a specific a priori definition of it and
of its concepts.
Problem II is a problem formally like that of the objective Transcen-
dental Deduction of the Categories, namely, the problem of showing
that the categories which are discovered in answering Problem I do
have objective validity. In the second Critique, there is a "transcenden-
tal deduction" which shows that pure reason can be practical, not
merely what its law would be if it were practical.
Problem III has some resemblances to that of the so-called Subjective
Deduction of the Categories in the first edition of the Critique of Pure
Reaso17, i.e., to show what mental operations are necessary if pure con-
cepts are to be applied in experience. I have already pointed out that
this problem might also be compared with that of the schematism. But,
inasmuch as Kant compares it with the problem of the Transcendental
Aesthetic, it is perhaps well to maintain this analogy.
§ 4. DIVISION OF THE COMMENTARY ON THE ANALYTIC
In this commentary we wish to separate out the various strands of
Kant's arguments and to present his conclusions as specific answers to
these specific problems. The Commentary on the Analytic is therefore
divided as follows:
Problem I.-Chapters vi and vii present an analytic of practical rea-
son in general and, more specifically, an analytic of an empirically af-
fected practical reason, in order to show the characteristics of its max-
ims and their relation to moral laws. These chapters are primarily com-
mentary on chapter i, §§ 1-6, of the Analytic .
. Chapter viii gives the derivation of the one principle of pure prac-
tIcal reason and is, accordingly, entitled "The 'Metaphysical Deduc-
tion' of the Moral Law." This chapter continues the commentary on
§§ 1-6, and moves forward into §§ 7 and 8.
Chapter ix, "Pure Practical Concepts and Judgment," is, in sub-
8 Cf. the remark at Critique of Practical Reason, 91 ( 184).
stance, a continuation of the "metaphysical deduction," though it in-
terrupts the serial movement of the analysis of the argument as Kant
gives it, by being devoted to an interpretation of chapter ii of the Ana-
lytic. The reasons for this interruption will be made clear in chapter
ix.
Problem 1I.-Chapter x discusses Kant's solution to this problem and
is accordingly entitled "The 'Transcendental Deduction' of the Prin-
ciple of Pure Practical Reason." It is commentary on §§ 8 and 9 of
chapter i and on the section entitled "On the Deduction of the Princi-
ples of Pure Practical Reason."
Chapter xi, "Freedom," examines Kant's theories on this concept, as
expounded here and in other places. Because this concept is used so ex-
tensively here and in most of his other books, it is not feasible to give
exclusive attention to the treatment in any specific section of the Cri-
tique of Practical Reason; but chapter xi deals fully with two parts of
the Analytic, namely, "Of the Right of Pure Reason to an Extension in
Its Practical Use ..." and "Critical Examination of the Analytic of
Pure Practical Reason."
Problem lII.-Chapter xii, following the suggestion made by Kant, is
entitled "The I Aesthetic' of Pure Practical Reason" and is a commen-
tary on chapter iii of the Analytic and on the Methodology.
§ 5. SUMMARY OF CHAPTER I, §§ [-8
We have set out three problems of the Analytic but pointed out that
these problems are so intimately entwined that Kant does not take up
and finish one and then move, in stately fashion, to the next. In almost
every section he is concentrating on one but keeping the others in
mind, and he constantly directs his readers' attention to the ramifica-
tion of each answer into the spheres of the others. Any attempt to iso-
late the problems, therefore, involves some artificiality, an artificiality
that must jeopardize the exegetical work before us unless the reader is
prepared to see the context from which the artificial abstractions have
to be drawn for the sake of microscopic examination. To aid in this, I
give now, in conclusion, a brief running account of Kant's exposition
in its printed form.
"Practical propositions" are those propositions the knowledge of
which plays a part in determining the will to make a specific choice
among possible actions. They are called "principles" if they are general,
i.e., if they express a general determination of the will and if other
practical propositions, called "rules," are subsumable under them or
derived from them in their application to specific circumstances. A
principle is called a "maxim" if the motive which is involved in obedi-
ence to it is a motive only for the person who actually embraces this
maxim as expressing his own policy in life. A principle is an "objective
law," however, if the motive which it formulates and to which it gives
expression is recognized as proper to the will of every rational being.
Every principle to some extent constrains the person who acknowl-
edges it. Even if my principle is a mere maxim that holds only for my-
self, such as the maxim of not allowing any wrong done me to go un-
avenged, it constrains me, at least sometimes, to bring my momentary
impulses (e.g., fear) into line with this general purpose or determina-
tion of the will. Even such a principle, therefore, can give rise to rules
which determine what I, with this motive, ought to do and would do
if I (a) had this policy and (b) were completely rational in the choice
of actions with respect to this policy. Such rules are called "impera-
tives" for a being who, like man, does not always willingly and spon-
taneously do what is prescribed by reason as necessary for the carry-
ing-out of the purpose. It is only by reasoning that we know what we
ought to do in order to carry out the policy expressed in the maxim,
but no one is so rational that he does what he ought without more or
less frequent conflict with his inclinations.
If a principle is really a maxim, so that the motive for action in ac-
cordance with it is some subjective condition, the corresponding im-
perative, which tells us what a reasonable man would do in order to
satisfy this desire if he had it, is a "hypothetical imperative." It com-
mands, or rather counsels, a man only if he has the desire in question.
The conative, dynamic factor in obedience to such an imperative is de-
sire or impulse; the cognitive factor is a theoretical cognition or some
belief about the relation of means to end.
A law, on the other hand, such as "lying is wrong," is not, at least
prima facie, addressed just to a man who wishes for honor or some
other specific goal. The imperative which expresses this law to a man
who does 'not obey it by nature is a categorical imperative. It does not
tell us to avoid lying if we would obtain a good reputation; it tells us
not to lie, period. It seems to be addressed to rational beings generally,
not just to those men having specific desires that can be satisfied through
obedience to it.
All principles based on any object of desire apply only to those who
actually have the desire. All such principles are mere maxims, not laws.
They cannot be laws even for those beings who do have the desire in
question, such as the desire felt by all men for their own happiness. A
law must have objective necessity, recognized by reason; but the pres-
ence or absence of a specific desire can be known only empirically.
Funhermore, a law must give rise to imperatives which are definite and
7'1. Survey of the Analytic of Practical Reason
specific, yet universal in application; but the diversity of desires is so
great that even if they are all subsumed under the general desire for
happiness, they do not issue forth in anything more than general coun-
sels, proverbs, and good advice which is sensitive to the variety of men
and circumstance.
Every principle or rule which presupposes, for its application, some
specific desire falls under the general principle of self-love or the desire
for one's own happiness; for a state of happiness is one in which there
is continuous satisfaction of all desires. Those philosophers who make
the desire for happiness the proper motive for morality cannot derive
from it any universal precepts, for each man's conception of happiness
differs from that of others, and anyone man's conception varies from
time to time according to the state of his specific desires. No rule de-
rived from the desire for happiness is more than a hypothetical impera-
tive, and it therefore lacks the a priori necessity characteristic of law.
It follows from this that if a rational being regards his maxims as
universal laws, as he does when he says that some action that he does is
the kind of action that all men (or other rational beings) should do, it
cannot be by vinue of the material of the maxim, which refers to the
object or the purpose of his will. This is true even if the maxim should
fortunately, in a person of benevolent or sympathetic disposition, be a
desire for the general welfare or the happiness of others. If the mate-
rial or goal of desire is presupposed in a principle, whatever it may be,
there is no universality in the principle, and the corresponding impera-
tive is not categorical.
Besides the material of the maxim, however, there is only its form.
The form of the maxim as expressed in an imperative is "ought," just
as the form of every theoretical proposition is some mode of "is." As
form, it is independent of any specific desire, which constitutes the
content of specific maxims. If we abstract from an imperative all con-
tent by virtue of which it is addressed to a person motivated by a spe-
cific subjective desire, we are left with only the form, the skeletal
"ought." What is derivable from this, unlike what is derivable from
any specific content, is addressed to all rational beings who act, and the
rules derived from it are fitted to be universal in application. That is,
the form of a maxim and not its content determines whether it is a law
or a mere maxim.
If a principle is a law, its form must be such that it applies to all ra-
tional beings, and the corresponding imperative must be directed to all
rational beings who do not, by nature, observe the law automatically.
Thus only a law can generate a categorical imperative. The categorical
imperative tells a partially rational being to act on a maxim that a
wholly rational being would act upon without being commanded to do
so. If a rational being can decide upon his actions under a maxim sim-
ply because the maxim is a law valid for all rational beings, this being
can obey a categorical imperative, and pure reason can be practical.
Assuming that a person can obey a categorical imperative, Kant
sho\Ys that the will of this person must be free in the strict, transcen-
dental sense. That is, it cannot be entirely determined by the person's
conception of his sensuous impulses, for this would make his actions
only an effect of natural phenomena. Only reason can present the con-
ception of a universal law as a motive, and of a being that acts upon
this motive we say that his will is "free."
Conversely, if we assume that the will is free from the mechanism of
nature, the will must be determined by the form and not by the con-
tent of the maxim or law. It must be determined by some conception
of some law, for otherwise it would not be will but mere caprice. And
if it were determined by the content, i.e., what the law held before the
person as a way of satisfying one of his desires, the will would not be
free from the mechanism of empirical nature.
Hence the concept of freedom and that of a universal practical law
reciprocally imply each other. We are not directly aware of freedom,
but we are directly aware of the binding quality of a universal law, for
we have it presented to us in our consciousness of the moral law. The
moral law thus leads us inevitably to-assert the existence of freedom; it
is the ratio cognoscendi of freedom, while freedom is the ratio essendi
of the moral law.
The moral law is this: A purely rational being acts only on maxims
that he would will to be maxims for all rational beings, i.e., only on
maxims that could be willed to be principles universally binding on all
such beings. For men, who are not completely rational, this is expressed
in the categorical imperative: "So act that the maxim of your will could
always hold at the same time as the principle for giving universal law."
Only pure reason could be the source of such a law and imperative.
Reason discovering sensuous motives and the laws of nature by which
they might be managed or satisfied would not be able to formulate any
laws having the universality and necessity that we find in the moral
obligation we experience. This law is not derived from any observation
of empirical fact; it is not a theoretical law of what is. It is a practical
law that pure reason itself prescribes as the ground of its own actions.
Thus pure practical reason, as the source of its own law, is autonomous
or self-legislating in a way in which an empirically conditioned practi-
cal reason could not be.
The principles of the empirically affected will are based upon the
contingent fact that certain desires are felt and upon our knowledge of
the way in which they may be satisfied in the course of nature. They
are not, therefore, products of the autonomous lawgiving of reason,
and they are not, consequently, absolutely binding or obligatory. All
moral systems except the one based upon pure reason as providing the
motive are heteronomous and are unable to account for the absolute,
unconditional, universal, and necessary constraint that we experience
in moral obligation. Either we must explain away these characteristics
of moral obligation by showing them to be illusory products of a psy-
chological mechanism, or we must accept the thesis that pure reason
can be practical, i.e., can give a law the knowledge of which can and
should be a sufficient motive for action.
VI
The Analytic of Empirical Practical Reason

I. Formal Considerations; Commentary
on Section 1
§ I. EMPIRICAL PRACTICAL REASON

"Practical reasont' and "will" generally, for Kant, mean the same thing.
We have already discussed in chapter iii this identification of the two
concepts. It suffices, therefore, to remind ourselves here that reason in
its logical use is the faculty of drawing inferences and of systematizing
knowledge, of finding a "wherefore" for every "therefore," and that
in its real use it posits certain a priori synthetic propositions or princi-
ples which are supposed to state the unconditioned conditions for all
that is found in experience. If the then-clause in an inference is a
statement that a certain action should he done or an imperative or a de-
cision to do it, the reasoning that leads to this conclusion is an inference
of practical reason. Reasoning as thus determinative of an action does
what the common man believes his will to do, the will being thought
of as an almost palpable organ for decision and execution, often war-
ring with a recalcitrant or weak flesh. 1
Yet the terms "practical reason" and "will" are not in every respect
synonymous; they are so little synonymous that one is, I believe. sur-
prised when he learns that Kant identifies them. There often seems to
he something especially unreasonable about will, so much so that Scho-
penhauer could easily see them in ineluctable conflict. If one does not
follow Schopenhauer in thinking that there is unceasing opposition be-
l Especially in the early parts of the Critique but also quite generally elsewhere.
Kant uses two words which are ordinarily translated as "will": Wille and Willkiir.
Though they are later distinguished and the distinction becomes important, I can-
not see much consistency in Kant's choice between them. Roughly speaking, we
may surmise that practical reason is Wille, and what I have called "the almost palpa-
ble organ for decision and execution" is WiIlkii,. We shall attempt later (pp.
J76 If.) to clarify the distinction as Kant subsequently drew it; but nothing, I think.
is gained by insisting upon it here. The words seem to be meant as synonyms in S 1.
7S
76 The Analytic of Empirical Practical Reason
tween them, one can still discern a great difference in the connotations
of the two words. "Will" suggests more directly the dynamic impulse
involved in action; "practical reason" suggests something cold and de-
liberate and without any "push." Such difference in connotation must
have been present also in Kant's mind, for we have seen how he often
speaks of reason as determining the will, as if there were two factors
here in opposition to each other.
However little he is justified by his own definitions in speaking in
this way, it is easy to understand why he did so. In acts of will, there
are two distinguishable factors: a want, which I have called the "dy-
namic" or "conative" factor, and what we recognize as that which we
ought to do, the recognition of which I have called the "cognitive"
factor. What I want appears as an incentive, and it speaks to my in-
clination, impulse, drive, and propensity. What I ought to do in order
to achieve what I want is discerned by reason. Reason may discover a
rule for what ought to be done if we are to obtain what we want; in
this respect, Hume was right in calling reason the slave of the passions;
but it is an intelligent slave, which, while serving its master, neverthe-
less guides, disciplines, and in part controls the master for the master's
own best interests. When the conative factor is the condition which
makes a specific rule of action relevant to choice, the reasoning in-
volved in formulating or picking out and applying the rule to the spe-
cific situation is based on empirically given data of two kinds: the
present data (feeling), which indicate the existence of a particular wish
or want, and the facts of past experience, which indicate how this wish
or want can be satisfied.
Reasoning functioning in this way is called "empirical practical rea-
son." The analytic of empirical practical reason is the study of the fun-
damental principles and rules of the conduct of reasoning in this con-
currence with desire, whereby blind pushing desire is converted into
interest. Such a study to be undertaken here has two parts: a study of
the formal aspects of such reasoning (in the present chapter) and a
study of the dynamic factors and the specific kinds of volitions result-
ing from them (in the following chapter).
§ 2. PRACTICAL PRINCIPLES
Practical propositions "assert [aussagen] an action, through which,
as a necessary condition, an object becomes possible."2 The term
"practical proposition" therefore refers to all sorts of rules of action,
2 Vorlesungen uber Logik, S 31.
The Analytic of Empirical Practical Reason 77
such as directions for a construction in geometry, a recipe, a counsel,
or any statement of a means to an end. Any proposition which is
effective through being entertained in deliberation on action is a
practical proposition, even though its content may be the same as that
of a theoretical proposition expressing mere knowledge without direc-
tion for use. 3 The familar theoretical proposition, "Alcohol causes
drunkenness," becomes, when related to a want, a practical proposi-
tion, "Drink alcohol" or "Don't drink alcohol," depending upon the
direction of the specific want that makes theoretical knowledge of
the effect of alcohol practically relevant. Not every theoretical prop-
osition, of course, has a practical counterpart; a practical proposition
concerns only what is possible through will, and it states what, in some
sense, ought to be done instead of what is done.
The first definition in § I is that of a species of practical propositions,
viz., principles (GTU17dsatze). "Practical principles," we are told, "are
propositions which contain a general determination of the will, having
under it several practical rules." Practical propositions are thus of
different levels of generality, from a specific command like "Shut the
door now" or a general precept like "Do not take unnecessary chances
when driving an automobile" to a general policy of life such as "Do
not permit any offense to go unavenged." Kant wishes to preserve the
name "principle" for those on the third level of generality. This is in
keeping with the requirement, implicit in the word GTU17dsatz, that
principles must be basic, not derived from more fundamental proposi-
tions. In the Lectures on Logic/, he restricts the word "principle" even
further to propositions which are a priori synthetic. At this stage of
his exposition in the Critique of Practical Reason, however, he is not
prepared to assert that there are any practical principles in this strict
sense;5 here he means by "principle" only a proposition expressing a
3 Practical propositiOns differ from theoretical propositions only in their V orstel-
lungsRrt, not in their Inhalt, with the exception of practical propositions which
concern freedom under laws (Erne Einleitung in die Kritik der Urteilskraft, XX,
1915). The statement above, therefore, is true of "technically practical propositions"
but not, according to Kant, of "morally practical propositions." This is not quite
correct, however. If we maintain, more successfully than Kant did, the distinction
between imperative and law, a morally practical imperative (i.e., the categorical
imperative) will correspond to the moral law, which is a theoretical statement about
the freedom of rational beings in general.
f Vorlesungen iiber Logik, § 34 (IX, IlO); cf. also Critique of Pure Reason,
A J49 = B J88.
• They would be law,s, and he is not yet in a position to assert that there are any
practical laws.
determination of the will which is not regarded as being taken for the
sake of some prior commitment or policy. 6
Fundamental practical propositions, then, are propositions which
contain a general determination (Bestimmung) of the will, having
under it several practical rules. What is a general determination of the
will? The word Bestimmung is one of Kant's favorite words, and he
overuses it. Among the English words needed to translate it are "defini-
tion," "determination," "predicate," "decision," and "motive." Here it
seems to have the following meanings: a determining cause of an
action, in which sense it means "motive" (Bestimmungsgrund), and a
decision that is taken, 1 by virtue of whkh the will is not indefinite
volition but a specific or determinate volition with a specific direction
and goal. The essential point in "general determination of the will,"
therefore, is that the practical fundamental principle must contain or
express a lasting policy or settled disposition of will, not a capricious
resolve or a variable rule of thumb.
Also in the first sentence the word "contain" requires brief comment.
Enthalten is a curious word here; one might say that a proposition
"contains" terms, but hardly that it "contains" a determination of the
will. One would be tempted to translate enthalten as "asserts," but this
would disjoin the proposition too much from the determination of the
will and make it merely theoretical. I think Kant means this: the
practical proposition formulates what the determination of the will is,
gives expression to it, and does not merely refer to it. It is a proposition
the knowledge or entertainment of which is itself a factor in the ~ill's
decision.
The question for an analytic of empirical practical reason is this:
What principle or principles exist which do indeed contain general
determinations of the will, so far as these determinations can be found
in experience? The question for an analytic of practical reason in
general is this: Are there any fundamental principles in the strict sense
of a priori synthetic propositions which contain a determination of the
will; and, if so, what are they?
6 Having formally introduced the word Grundsatz, Kant in characteristically
regrettable fashion frequently ignores it. (It has been said of Kant that he succeeded
in being technical without being precise.) Almost invariably in its stead he uses the
word Prinzip, which. though declared synonymous with Grundsatz (VorJesungen
fiber Logik, § 34). is much looser in meaning. Thus a rule or precept (V orscbrift)
is often caned a Prinzip, while a genuine Grundsatz is also called Prinzip (e.g., in
§ 4. p. 17 [ I1 4}). Where necessary for the sake of clarity, I shall refer to Grundsatz
in the strict sense as a "fundamental principle," even though this is strictly pleonastic.
T For example. Critique of Practical Reason, 90. four lines from end of page (184.
1.18).
§ 3. RULES
Kant gives no formal definition of "rule," and his usage is highly
variable. Sometimes it means "law,"s sometimes "imperative,"9 and
sometimes merely a precept or common maxim which guides us in
dubious cases-the kind of rule that serves as a "go-cart for the stupid"
who do not trust themselves to act wisely in the complexity of par-
ticular cases and who therefore fall back upon some simple standard
commonplace. 1o
In the Remark on § I, however, he shows us that he means a rule to
be concerned with the empirical contingent character of action in spe-
cific situations, given the general determination expressed in the prin-
ciples. Rules are thus distinguished from fundamental principles which
express an actual policy of life (maxims) and from fundamental princi-
ples (if there are any) which are universally vaIid.l1 Rules express what
is, on the average, right under a general principle, but they do not hold
necessarily and without exception. They require healthy common sense
and good judgment for their wise application, and they cannot be ap-
plied automatically by appeal to a general policy.12
Kant does not say that the rules fall directly under the principles, as
if they were logically contained in them. They are contained, he says,
in the general determination of the will, not in the principle. Thus rules
may be identical when principles are different, and different when prin-
ciples are identical; two men can aim at the same end and achieve it by
following different rules. It is important to note that Kant must insist
upon this loose relationship between rule and principle; otherwise he
could not distinguish, within a single type of action, those cases of it
that are moral from those that are merely legal. Precisely this impor-
tant loose relationship is what is overlooked in some of the most com-
mon criticisms of Kant's ethics, those which insist that uniformity in
moral actions is overemphasized by Kant.
Kant is not asserting the universal applicability of any rules. The uni-
8/bid., 31 (119). They are confused also in the first Critique; cf. Paton, Ktmt's
Metaphysic of Experience, I, 493.
9 Critique of Practical Reason, lO, second paragraph (106).
10 Reflexioncn 5235-8; cf. Critique of Pure Reason, A 134 = B 173.
11 Foundations of the Metaphysics of Morals, 389 (4), distinguishes rule from law
by reference to the empirical conditions involved in the former. But the proper
dichotomy is between rule and principle, and under principles it is necessary to
distinguish laws from mere maxims; then rules fall under both maxims and laws.
12 The same is true of rules of theoretical judgment (d. Critique of Pure Reason,
A 133 = B (72).
versality he insist!l upon is not that of a rule but that of a principle,
which permits latitude in the choice of rule. But his failure to abide by
his terminological distinction between "rule" and "principle" and his
tendency to regulate his own life by "maxims" (really rules, as de-
scribed by his acquaintances who wrote biographies of him) perhaps
led him into the error of sometimes thinking of rules as valid without
regard to circumstances, as in On an Alleged Right To Lie from Altru-
istic Motives written in reply to Benjamin Constant. I3
A rule is always a "product of reason," for only reason (as the fac-
ulty of thought in general, not as one of the three cognitive faculties)
can give us the knowledge of the relation of the means to an end which
is expressed in a rule. Because a rule has a cognitive component that is
invariant with respect to the purposes which may be served by using
the knowledge it expresses, Kant calls a rule for the use of a means to
an arbitrary end a "technical principle" (Prinzip; should, of course, be
Satz) and not a practical principle. 14 A technical rule is objectively
valid as a statement of what ought to be done by a rational being who
has the general determination under which the rule falls. But it is sub-
jectively practical, i.e., actually relevant in the determination of be-
havior only for a being who has the general determination in question.
We may therefore properly call rules which prescribe means to ends
"conditional rules," even though they embody knowledge independent
of the subjective condition of its use.
We cannot answer the general question, "What are the conditional
rules of an empirical practical reason?" The answer would be too com-
pendious to state: it would include all practical art, all applied science,
and all knowledge of the ways of the world. But we must ask, in an
analytic of practical reason generally, "Are there any unconditional
practical rules which follow directly from unconditional principles?"
And the answer to this question would have to be given in an analytic
of pure practical reason.
§ 4. MAXIMS
Practical principles are then classified as maxims or as laws, as sub-
jective or objective. Kant's fonnulation of this division is peculiar, how-
13 Recent studies have convincingly shown the relevance of circumstances to
rules in Kant's ethics; cf. especially H. J. Paton, "An Alleged Right To Lie: A
Problem in Kannan Ethics," Ktmt-Studien, XLV (1954), IC)O-203, and "Kant on
Friendship," Proceedings of the British Academy, XLII (1956), 45--66; Marcus G.
Singer, "The Categorical Imperative," Philosophical Review, LXIII (1954), 577-91;
and W.I. Matson, "Kant as Casuist," Journal of Philosophy, LI (1954),855-<)0.
14 Critique of Practical Reason, 26 n. (113 n.). In Critique of Judgment, Introduc-
tion, I. they are called "technically practical" as distinct from "morally practical."
The Analytic of Empirical Practical Reason 8.
ever, because the principium divisionis is not whether the principle
itself is regarded as objective or subjective but whether the condition
is valid for the individual only, or for every rational being. And, fur-
thermore, it is a logically faulty division, because "maxim" is broader
than "law" and, in fact, includes "law" as one of its species. We must
examine each of these peculiarities.
What is meant by "condition" (Bedingung)? Most simply, we can
say that it means the same as "determination" (Besti11lmung) in the
sense of a general setting of the wilf.1!S But why, then, in this paragraph
setting forth the formal definitions, where precision and elegance are
called for, did not Kant repeat the word "determination"? The reason
lies in his logical terminology. It was his practice to call the major
premise in a syllogism of the first figure a prinfiple (sometimes a rule),
and its middle term (the subject of the principle) he called the condi-
tion. 16 In the light of this usage, we can say that "condition" means
that which is involved in the maxim as the general determination of the
will, and thus it becomes the middle term in a practical syllogism:
To avenge a wrong is always my purpose-Maxim or principle.
To tell this lie would avenge a wrong-Rule.
Therefore, I purpose to tell this lie-Decision.17
HTo avenge a wrong" is the condition, and it may be held only by my-
self or ascribed to all rational beings.
I turn now to the relation between maxim and law. The term
"maxim," like "condition," is also taken from logic, deriving from senten-
tia maxima, the name of the first major premise in a polysyllogism. Logi-
cally, then, "maxim" and "principle" have the same meaning, and law
is only a species of them. In fact, however, Kant seems to distinguish
sharply between them; in the Foundations,18 he says that a maxim states
how we do behave and a law prescribes how we ought to behave. In-
stead of this dichotomy, however, he is interested in establishing the fol-
lowing possibilities: (a) a rational man acting according to some maxim
while holding it to be valid only for him, and for him only because its
condition is the actual state of his own motives; (b) a rational man rec-
ognizing a condition valid for, though not necessarily effective in, all
15 Indeed, Kaubler (Der Begriff der Triebfeder in Kants Ethik, p. 41) suggests
that Bedingung is a typographical error for Bestimmung; and BertiTnmung is used
in this sense in Metaphysik der Sitten, Einleitung (VI, 111).
16 Vorlesungen uber Logik, ~ 57 (IX, 110); Critique of Pure Reason, A 311 =
B 378, A 300 = B 357; cf. Critique of Practical Reason, 90 ( 184).
17 Critique of Practical Reason, 90 (184); Metapbysik der Sitten, VI, 437-38.
18 Foundations of the Metaphysics of Morals, 410D. (38 n.).
rational beings; (c) a rational man recognizing a condition as present
and effective in all rational beings as such and therefore as valid for and
applicable to himself. There is thus a trichotomy, not a dichotomy, to
wit: (a) mere maxim, (b) law, (c) law which is also a maxim. Ill
The questions requiring an answer are these: (1) What is the su-
preme maxim or maxims of an empirically practical reason? (z) Are
they laws, or mere maxims? (3) Can a law itself be a maxim? The first
of these questions must be answered in the light of empirical facts abOllt
man and will concern us in the next chapter. The third question belongs
in the analytic of pure practical reason. The formal question in an ana-
lytic of empirical practical reason is the second.
§ 5. ARE THERE EMPIRICAL PRACTICAL LAWS?
Kant answers that they are mere maxims, and he supports this answer
with many arguments. In this section, however, we shall be concerned
only with the answer insofar as it is based upon purely formal or epis-
temological considerations; the other arguments properly fall within the
scope of the next chapter.
A mere maxim must depend upon conditions which reflect individ-
ual differences among rational beings. These differences are differences
in their desires, in the conative and not in the cognitive components of
their volitions. A law, on the other hand, might seem to depend upon
what all men do, in fact, have in common, and this need not necessarily
be something common to their cognitive faculties but might be some-
thing lying in their conative makeup. In fact, Kant believes that all men
do have a common desire, namely, the desire for happiness. Why, then,
are we not, in Kant's opinion, able to state a practical law based upon
this putative empirical fact-or, indeed, upon any other alleged fact,
if we are able to find some other common conative component in all
men?
A law cannot be based upon any actual features inductively discov-
ered in the generality of mankind. For let us suppose that one desire,
upon examination of a fair sample of mankind, turns out to be present
in them. We might then make the statement, "All men have desire D,"
in which D would be a general determination of their wills. But, even
so, D would not issue in a law that all men should act so as to satisfy D.
For let us see what would happen if we found a man in whom D was
absent, as we might always possibly find one by continuing our induc-
tion. The fact that everyone else had D would not constitute the slight-
est reason why he should do actions called for by D, or should feel any
shame for lacking D; it would, in fact, provide a ground for rejecting
19 Cf. Critique of Pure Reason, A 812 =B 840.
the generalization. The generalization is a posteriori, but a law, says
Kant, must be universal and necessary, that is, a priori. This, he says, is
true of laws in natural science as well. 20
Few would agree with Kant in this sweeping statement. There is an
element of brute fact in what seem to us to be the supreme "laws of
nature," and they might, for all we know, have been quite different. It
is just a brute fact about our universe that bodies attract each other; yet
we call the description of this attraction a "law."
It is nevertheless clear that Kant is here pointing out an important
distinction in morals which depends upon our different responses to an
exception to an alleged law. There is no possible logical inference from
how men do, in fact, behave to how they ought to behave. No empiri-
cal generalization can do anything but surrender in the face of a well-
authenticated exception. An a priori principle or law, on the other
hand, can "condemn" an apparent exception, and it does so even in
natural science. A body that does not fall according to Galileo's law-
and none does-is simply not a freely falling body. We do not revise
the law when we find an "exception" to it; we reclassify the object. The
law which can be held in spite of apparent exceptions is accordingly
not a mere summary of experience; it is prescriptive, not merely de-
scriptive, even though experience may have suggested its form and may.
in fact, corroborate it. There are such laws in science; there seem to be
such laws for men's practical conduct.
A law is prescriptive, whether it is a theoretical or a practical law.
(Perhaps it would be better to say that a law is prescriptive but that it
may prescribe for theory-construction, in which case we call it "theo-
retical" even though it represents a practical decision, or it may pre-
scribe for conduct, in which case we call it "practical.") Kant, therefore,
in warning against allowing any empirical principle to be called a law
is calling attention to "the most important distinction which can be
considered in practical investigations" (26 [114]), to wit, that between
factual generalizations and practical!!1 prescriptions. No logical ir.fer-
ence between them is possible. There are, therefore. no empirical laws
of practical reaJon, but only empirical maxims (21 [108]). If there are
any laws, they are not empirical.
20 Critique of Prat;tical Reason, 26 (114); Critique of Pure Reason, A 159::::: B 198.

21 "Practicai" is here taken in a sense broad enough to include the practical (pre-
scriptive) aspect of even theoretical systems. In the Critique of Pure Reason Kant
often speaks of maxims for the conduct of speculative inquiry and inquiry in natural
scien("e; a maxim is such a rule which is subjectively necessary for the conduct of
inquiry but has no direct objective validity (d. Critique of Pure Reason, A 666 :::::
B 694).
Kant's point can perhaps be made clearer if we avoid the use of the
term "a priori," which seems to belong only to theoretical knowledge,
or at least to be univocal in meaning only when applied there. His point
is the essentially simple one that practical laws must be prescriptive, and
empirical laws, which are a posteriori, are not prescriptive. By exclu-
sion, then, he can call practical laws, if there are any, a priori.
§ 6. HYPOTHETICAL IMPERATIVES
Rules are formulated in the imperative mood. In an almost casual
manner, down in the Remark to § I, Kant reintroduces one of his most
significant distinctions, that between an imperative and a law,22 and
then the no less important, but more celebrated, distinction between
two types of imperatives.
As an excuse for this apparent casualness, let it be said that the sec-
ond of these distinctions, so elaborately developed in the Foundations,
really plays only a comparatively minor role in the remainder of the
Critique. The reason for this is that the analytical method of the Foun-
dations begins with the phenomenon of moral constraint, while the
synthetic method of the Critique commences with the principles, and
the imperative is only the appearance of such principles to a less than
perfectly rational being, to one not ipso facto determined by the ra-
tionallaw. The doctrine of imperatives is therefore at most a coronary
to "the investigations of the Critique. It is the discussion of imperatives
and the formula of the categorical imperative that is presupposed in the
Critique. For this reason, it is appropriate to refer to the fuller discus-
sion of these matters in the Foundations.
Any practical proposition is a product of reason. For a being like
man, who is not wholly guided by his reason, a practical proposition is
presented as a command of reason and is expressed in the imperative
mood. It is objectively valid and thus differs from a mere maxim; but it
need not express a law in the sense of being unconditional in its prac-
tical function. It may be objectively valid as an imperative only for a
being with a specific condition of volition. We must, therefore, distin-
guish between a law which is necessarily binding on a rational being as
such, which gives an unconditional imperative to all partially rational
beings, and a law which is merely the objectively valid cognitive com-
ponent in the determination of choice by a rational being affected in a
particular empirical way. The latter kind of law is a theoretical propo-
sition functioning as a practical proposition only under the condition
22 A distinction he repeatedly ignores; d. Critique of Practical Reason, :U, 30
(107. 119); Foundations of the Metaphysics of Morals, 410 and n. (37,38 n.); Critique
of Pure Reason, A 802 =B 830; Metaphysik der Sitterz, Einleitung, VI, ZU-Z3.
of the subject's volition, which makes some among the myriad of theo-
retical propositions practically relevant in the detennination of his con-
duct. An imperative to take an action, so far as it is determined by the
actual condition of the subject's will and by a theoretical component
based upon knowledge of fact, is one that commands hypothetically, or
is a hypothetical imperative.
In the Foundations two types of hypothetical imperatives were dis-
tinguished: the problematic or technica}23 imperative and the assertoric
or pragmatic. The difference between them is that in the former the
condition is not asserted but mentioned only hypothetically (e.g., "If
you want to make bread ..."), while in the latter it is asserted (e.g.,
"Since you want to ..."). The former are also called "rules of skill";
the latter, "counsels of prudence." The distinction between the two
types of hypothetical imperatives is ignored in the Critique, but a cen-
tral problem of the Critique, to which the next chapter will be devoted,
is to determine what condition it is that is asserted in all hypothetical
imperatives of the second kind.
An analysis 24 of a hypothetical imperative would make the follow-
ing components yxplicit:
a) In the protasis:
(I) A conative element, directed to something, B, which is the purpose
of the commanded action, B being an object or state of affairs the
representation of which is one of the causes of its existence as an
effect of action undertaken by the subject; this is expressed in the
imperative by "If I want B" or "Since I want B."
(II) A cognitive element: knowledge of the causal relation between the
action commanded, A, and the purpose, B; equivalent, under condi-
tion I, to the statement "A is a means to B."
(III) A tacit premise or rule of practical inference: "If I fully will the
effect [B], I also will the action [AJ required for it."25
b) The apodosis: "Do A."
In the technical imperative, B may be the object of any desire; and in
many cases there are well-established laws of nature (II) which tell us
how we can achieve B. In the pragmatic imperative, B is one specific
object which Kant thinks we can assert and not merely entertain as pos-
sibly desired; we have the desire for happiness. But in this case, the con-
23 Erste Einleitung in die Kritik der Urteilskraft, XX. 200 n., criticizes the term
"problematic imperative" and substitutes "technical imperative" for it. Metaphysik
der Sitten, VI, 222, distinguishes only between technical and categorical.
24 The full analysis is given in my contribution to the Paton-Festschrift, "Apo-
dictic Imperatives," Kant-Studien, XLIX (1957-58), 7-24. and the remainder of this
section is taken largely from that essay.
25 Cf. Foundations of the Metaphysics of Morals, 4 1 7 (34).
tent of B is so vague and the knowledge of the 'world and its ways is so
inadequate that we cannot state universal rules that 'will permit us to
count on achieving B. Therefore, the range of the conative component
and the certainty of the cognitive component seem to vary im'crseIy:
where the end in view is highly specific, real laws of nature may enter
at II, and where it is a vague end, like well-being or happiness, the cog-
nitive component is likely to be empirical knov,:ledge not worthy of
the name of law but mere belief and opinion.
Kant tells us that the possibility of these imperatives is easy to e~tab
lish, since they are, "in what concerns the will," analyticaPG This is
correct, but it requires a little analysis to see what it means and why it
is true.
It is not entirely clear in what sense an imperative is the kind of judg-
ment that can be either analytic or !>ynthetic. Imperatives are not judg-
ments with a subject and predicate, and thus they do not fall under
Kant's explicit division of the types of judgment. And though for any
imperative a set of indicative statements (not necessarily factual; they
may be value-judgments) may be formulated from which the impera-
tive could be reached merely by a syntactical change of mood, in the
case of hypothetical imperatives some of the corre~ponding indicative
judgments are themselves hypothetical, and hence they do not fall un-
der Kant's official rubric for distinguishing analytic from s~·nthctlC
judgments. It is not entirely pedantic to insist upon these niceties; hut
the minor infelicities of expression should not excuse us from finding
the important and correct point that Kant is insisting upon.
If we take the statements of the conditions under which the apOd()~l~
is regarded as being binding and phrase them in declarative statemcnt<;,
e.g., "Under conditions C a rational being will do A," then the propo-
sition "A rational being will do A" \vill follow analytically from this
and another factual proposition, "A rational being is under condition
c." By "follow analytically" I mean that a denial of the conclusion will
contradict the joint assertion of the premises; hence to deny the propo-
sition corresponding to the apodosis will contradict the propositions
which are its premises. Since the apodosis is connected analytically with
the protasis, Kant says the imperative is analytic.
28/bid .• 417 (34).1 overlooked the following point in my paper presented to Paton.
Actually. Kant says that only the technical imperative is analytic "in what concerns
the will," while the pragmatic imperative would be analytic "if it only were easy to
give a definite concept of happiness" (ibid., 417 [35]). But in this distinction be-
tween the twO imperatives, Kant is incorrect. "In what concerns the will," both are
analytical in the sense to be described, in what concerns the understanding, i.e., the
cognitive content in the choice of particular means to the end in view, both are
synthetic.
But it is analytic in this broad sense only "in what concerns the will."
That A is necessary to B is not known analytically and that the rational
being is under condition C is not known analytically. All that is ana-
lytical is component III, and this concerns the form of the volition and
not the specific contents willed. Component III merely specifies the re-
lation between A and B as variables. Kant should say, therefore, not
that the hypothetical imperative is analytic, but that its formal princi-
ple, which is III, is analytic. It alone concerns the will of a rational be-
ing independently of the cognitive content and the contingencies of
human desire. Not to accept III is to fail to be a rational being con-
cerned with desires; but we fail, in varying degrees, to follow through
in our obedience even to this formal principle. We will health, but we
will acts that jeopardize it, and we fail to be prudent men.
The conclusion of the hypothetical imperative is reached only
through component III. Because of its formal character, even hypo-
thetical imperatives can be objectively valid; they are not persuasive or
emotive but rational, even though they are relevant to action only un-
der specific conditions which need not be true of rational beings as
such. The conditions that they are concerned with are conditions of
him to whom the imperative is directed, not of him who issues the
command. .
Since beings like us do not, in fact, always will the means necessary
to their ends, even when they know the means, the hypothetical im-
perative expresses a constraint of reason upon impulse. If we were com-
pletely rational beings, the maxim of doing whatever is necessary to the
end in view would be easy to follow. But since we are not, even our
desires and wishes can create constraints and not mere lures and entice-
ments. Pleasure itself, it has been ruefully discovered, can be a hard
mistress.
Kant's treatment of imperatives, under the supposition that the hypo-
thetical-categorical distinction taken over from formal logic is defini-
tive of two sharply distinguished kinds of constraint (the prudential
and the moral), has led to some unfortunate interpretations of his ethi-
cal doctrine. These misinterpretations would not have arisen, perhaps,
if he had followed one of his own suggestions that the distinctions of
modality of judgments rather than the distinctions of their form should
?e used in distinguishing among imperatives. A hypothetical imperative
IS defined by its form: "If you wish so and so, then do this or that," and
a categorical imperative is defined by its form: "Do this or that." It
would appear from this that all imperatives having moral import must
~e categorical imperatives, and Kant regularly states the moral impera-
tIve in categorical form. This in turn has led to the view that for Kant
in moral decision we should ignore all circumstances which would be
expressed in the antecedent of a hypothetical judgment or imperative,
and this has given rise to one of the interpretations of Kant that can be
considered only silly.
The important point is not that a moral imperative must have a spe-
cific grammatical or logical form different from that of a prudential or
technical imperative. There are morally valid imperatives that are hy-
pothetical in form: "If you have promised to return the book, do so";
and categorical imperatives which do not have any moral import or
claim: "Shut the door." What we must do jf we are to remain faithful
to Kant's purpose in drawing the distinction is to see that no impera-
tive is morally valid if it is directed merely to the man who has a spe-
cific desire, whether it is in the form of "If you wish to be warm, shut
the door" or in the categorical form, "Shut the door." We must be able
to discern that condition of will which must be present in a man if it is
intelligent to direct an imperative to him at all; this is the condition that
he be a person with a practical reason. Assuming only this and the fact
that he does not by nature do what the law would require either for the
achievement of some specific purpose or for intelligent action in the
pursuit of any purpose whatsoever, the imperative is directed to him as
a practically rational being regardless of his specific desires; under this
condition, the imperative is apodictic whether it is formally hypotheti-
calor categorical. In contrast to it there are assertoric imperatives-
imperatives relevant to a man of whom it can be empirically asserted
that he does, as a matter of fact, have such and such a maxim which can
be carried out in the way stated in the imperative. In an apodictic im-
perative, hypothetical conditions may enter into the determination of
the specific action required, but not into the determination of the one
condition that renders the imperative valid for everyone instead of
valid only for an individual who just happens to have the desire that
could be satisfied through obedience to this imperative. For example,
making a promise creates an obligation, and one can say (categorically),
"Keep your promise" or (hypothetically) "If you promised to return
the book, do so." But the hypothetical imperative, "If you want to be
able to borrow another book, keep your promise," cannot be put into
a categorical form that is apodictically necessary; such an imperative,
whether categorical or hypothetical, is only assertoric in modality.
§ 7. CONCLUSION
In this chapter we have studied the formal structure of practical rea-

son as employed in satisfying any conative or dynamic urge whose ex-
istence in us is empirically discoverable. In so doing, we have raised
The Analy tic of Empirical Practical Reason 89
questions that cannot be answered by purely formal considerations but
require knowledge of human nature and of the moral phenomenon. We
want to know:
,. What principle or principles are there which express a general de-
termination of the will so far as this can be discovered empirically?
z. Are these principles mere maxims, or laws, or both maxims and
laws?
3. Are there any principles which express a general determination of
the will independently of empirically discoverable dynamic or conative
factors?
The next chapter will answer questions I and z by tracing Kant's
theory (a) that all principles which express an empirically discovered
condition of will are "principles of self-love or private happiness" and
(b) that they are mere maxims and do not have the force of law. Chap-
ters viii and ix will then study Kant's affirmative answer to the third
question and show how such principles, which are the only laws, are
discovered and what they are.
VII
The Ana ly tic of Empirical Practical Reason

II. Material Considerations; Commentary
on Sections 2, 3, and Part of 8
§ I. INTRODUCTlON
Empirical practical reason is reasoning used in the guidance of behav-
ior for the sake of satisfying some desire. In chapter vi we examined
the formal aspects of this guidance as exercised through principles,
maxims, rules, and imperatives. In this chapter we shall be concerned
with the moving forces in such conduct, with its efficient and final
causes instead of its formal causes. This will require us to try to under-
stand Kant's views of and terminology in psychology or, as he would
say, anthropology. We must understand what he means by desire, pleas-
ure, pain, interest, and happiness in their empirical, psychological set-
ting. We shall then follow Kant's argument that these concepts and the
principles employing them are not competent to fill the office ascribed
to moral concepts and principles.
§ 2. DESIRE
Desire is the faculty of a being, through its ideas (V orstellzmgen,
representations or "ideas" in the Lockean sense), to cause the reality of
the objects of these ideas. 1 This succinctly indicates the two factors in
desire that we have already distinguished: the cognitive factor, or idea,
and the dynamic or conative factor. The latter is sometimes called "de-
sire proper" (Begierde) j when it is of a settled dispositional character,
Kant calls it "inclination" (Neigung)j often the name "impulse"
(Antrieb) or "incentive" (Triebfeder) is applied to it. 2
1 Critique of Practical Reason, 9 n. (94 n.) j Metaphysik der Sitten, VI, ZJ I. Fol-
lowing Wolff and Baumgarten, the faculty of desire (Begehrungsvermogen) js
distinguished from that of aversion (Abscbeuungrvermogen) at Critique of Prac-
tical Reason, 58 (148), but no further use is made of this formal distinction, and
nothing is lost in Abbott's not speaking of faculties at all in this passage.
2 The usage of Neigung to refer to settled disposition is not always observed. It
generally means any inclination and is approximately equivalent to Triebfeder ex-
Interest is that by which reason determines the will. 3 But since prac-
tical reason is identical with will, it is more accurate to say that interest
"indicates an incentive of the will as it is presented by reason" (79
[172]). In a dependent will, i.e., one not wholly rational, there is always
an interest which is expressed in a maxim; unless it is so expressed, and
thus subject to intelligent inspection, we do not have a will but only
blind impulse. Will differs from mere desire in that in the latter there is
an image of an object which is the target of behavior; while in willing,
on the other hand, there is also guidance by knowledge of a law or
principle which relates the action objectively to what it is that is de-
sired. Thus animals have desires, but only rational beings can have a
wiII.
The objects of the ideas-the objects being the final causes of action,
the ideas being among the efficient causes-are also called "objects of
interest or of inclination."4 The objects are the purposes of the action,
since purpose is defined as the concept of an object considered as the
cause of the reality of the object:~ The idea is one of the efficient causes
of the object by virtue of being one of the factors which determine the
cept when Triebfeder refers to the moral incentive or motive. Abbott translates
Triebfeder as "motive" or "spring." "Spring" follows a usage going back to the
early seventeenth century, but not common now. There is good etymological justi-
fication for it, since Feder refers, e.g., to the mainspring of a watch. "Motive" is a
less fortunate choice, because Kant carefully distinguished between Triebfeder and
BewegungsgTUntJ (= "motive") in the Foundations of the Metaphysics of Morals,
427 (45). Yet the fault is Kant's, not Abbott'S, for in this chapter of the Critique
he is using Triebfeder in the sense in which the Foundations defined Bewegungs-
grund. Picavet translates it as mobile, Capra as movente, Born (following a Kantian
parenthesis) as elater. I have followed the translation suggested by Greene and
Hudson in their rendition of Religion within the Limits of Reason Alone. But I
have done so with growing dissatisfaction, for the meaning of Triebfeder is obvious
to a Gennan, while incentive must be explained to a reader of English. It does
not seem possible to find an entirely suitable English equivalent, and I suspect that
the reason for this is that Kant himself did not use the word univocaJIy. On the
tenninology and its variations see Kaubler, Der Begriff der Triebfeder in Kants
Ethik (1917).
3 Foundations of the Metaphysics of Morals, 459 (80 n.); Metaphysik der Sitten,
VI, 212-13; Critique of Judgment,
§§ 2,3'
4i "Object of inclination" and "object of interest" must be distinguished from
"object of practical reason." It is only the latter that is defined as the good or evil.
"Object of choice [WiUkurJ" is used by Kant in the legal sense, referring to that
which I may dispose of as I choose, i.e., property (Metaphyrik der Sitten, VI, 246)
as well as in the sense of "that which I choose" (e.g., Critique of Practical Reason,
36 [125]).
GCritique of Judgmtnt, V, 180 (17).
person's action which, in turn, will produce the object. An object is a
purpose if we ascribe its realization to a chain of causes one member of
which is an idea or conception of the object which determines the per-
son's behavior in actually producing the object.
The word "object" denotes two quite different things. It may mean
an actual state of affairs, a physical thing and its psychological effects
that can be brought into existence by action. The production of such
an object requires empirical knowledge of its causes and skill in apply-
ing this knowledge. It is in this sense only that the word "object" is to
be used in analyzing an empirical practical reason. But the word has an-
other, quite unusual, meaning. It may refer to an internal setting of the
will, to an act of decision itself without regard to the causality of will
in bringing its object (in the first sense) into existence. This is the
meaning that the word will have in the analysis of pure practical
reason. 6
§ 3. PLEASURE
Kant's theory of desire is hedonism. "To wish for something and to
have satisfaction in its existence, i.e., to take an interest in it, are iden-
tical."1 Desire is always directed to something-its object-the existence
of which is expected to give pleasure.
The converse, however, does not hold. While his theory of desire is
hedonistic, his theory of pleasure is not exclusively conative. For it is
possible to experience pleasure without antecedent desire. Pleasure can
arise from the mere contemplation of an object or from the experience
of an idea in imagination. Such pleasure is called "contemplative pleas-
ure" and is best seen in the enjoyment of beauty, where the pleasure is
characterized as being disinterested. Pleasure which is the goal and re-
8 Critique of Practical Reason, 15 (101). Similarly, Foundations of the Meta-
physics of Morals, 431, 1. u. (Abbott [50, four lines from bottom] paraphrases and
misses the word "object.") It is essential for Kant to use the word "object" in this
unusual way, for otherwise moral action as such would have to be declared without
object, i.e., without interest or purpose. 'Vhen he says that there is an interest present
in all action but that interest need not be assumed as the determining cause of all ac-
tion, he is referring to interest in the existence of the object, which is the first sense.
But there can be an interest in making the right decision whether the decision can be
rendered effective in the world or not. The latter is the moral interest, and its object
is a certain setting of the will to act from a certain motive (d. Foundationr of the
Metaphysics of Morals, 413 n. £30 n.], and Critique of Practical Reason, H [Il2]).
This is the sense of "object" in chap. ii of the Analytic, when Kant is discussing the
object of pure practical reason, and it will be discussed in chapter ix of this com-
mentary.
1 Critique of Judgment, ~ 4 (V, 20<) [43]).
ward of action, however, is called "practical pleasure"-practical pleas-
ure is an "interest of inclination."8
Both kinds of pleasure may be defined by reference to the subjective
state of the person. Pleasure is the consciousness of the causality of an
idea to keep the subject in the state in which he is having this experi-
ence. This is true even of disinterested aesthetic pleasure. Pain is, by
corollary, our consciousness of an effect that an idea has in making us
try to, or wish to, change our subjective state.9 Pleasure is present when
there is harmony or facilitation of function; Kant calls it the idea of the
agreement of an object or action with the subjective condition of the
person (desire)1° or with the work of our perceptual and imaginative
capacities (as in the disinterested pleasure of art). It would perhaps be
clearer to say that pleasure is the feeling produced by such agreement.
Pain is the spur to activity to bring about such agreement and is thus
connected with desire. l l
The faculty or capacity for experiencing pleasure and pain is feeling.
Feeling is one species of the general affection of the sensibility (sensa-
tio ), the other species being sense proper (Sinn) .12 If the content of
our sensory experience is of such a kind that it can be related to an ob-
ject of knowledge and thus, notwithstanding its dependence upon the
subject's psychological constitution, can become a component in knowl-
edge of the object, the content is called "sensation." Receptivity to it
is sense proper. The color green, for instance, is a sensation and is sub-
jective, since it depends upon the person; yet it is related to the object
in such a way (by means of an a priori synthesis under the category of
subsistence and inherence) that one can correctly say that the object is
green. This objective attribution cannot take place with feelings, and
hence they are subjective in a double sense.
There are only two elementary feelings-pleasure and pain. All other
feelings, such as the feeling of the sublime, the beautiful, and respect
are defined by the accompaniments, contexts, causes, or "objects" of
the pleasure or pain we feel. Whether the origin of the pleasure lies in
some physical stimulation, the physical fulfilment of a desire, or some
8 Metaphysik der Shten, VI, 212.
II Anthropologie in pragmatiscber Hinsicht, § 60.
10 Critique of Practical Reason, 9 n. (94 n.), in the definitions drawn from psy-
Chology.
11 Antbrop%gie, § 60.
12 Metaphysik der Sit ten, Einleitung, Ii Critique of Practical Reason, 23 (109) j
Erne Einleitung in die Kritik der Urteilskraft, XX, 226. Pleasure is referred to
"inner sense." This is not the inner sense of the Critique of Pure Reason but the
inner sense or capacity of feeling as described in Antbropologie, § 15.
idea held in contemplation, the feeling is always an effect upon our
sensibility. In a hedonistic calculus, there is no place for qualitative
differences (23 [I 10]).
A feeling of pleasure or pain, we have seen, cannot be ascribed as a
property to an object. It cannot even be related to the existence of an
object by a necessary rule. For whether a feeling of pleasure or pain
will arise in the presence of an object can be learned only in the actual
experience. All our action aimed at the satisfaction of desire, i.e., at
pleasure in the existence of an object, is therefore based upon experi-
ence; all knowledge of the pleasurableness of any existing state of affairs
is, therefore, a posteriori. From this it follows that no law, i.e., no nec-
essary principle valid for all rational beings, can be based upon any
maxim for the realization of practical pleasure.
§ 4- THE LOWER FACULTY OF DESIRE
The distinction between the lower and the higher faculties of desire
( 22 [109]) is the scholastic distinction between appetitus sensitivus and
appetitus rationalis, which, in turn, derives from the classical distinc-
tion between passion and will.1 3 The paragraphs before us are directed
primarily against "the otherwise acute men," the Wolffians,14 who,
while drawing this distinction, did not sharply distinguish between the
sensitive and the rational faculties in general. In Wolff's doctrine, de-
sire has as its condition the knowledge of a perfection. 15 If this knowl-
edge is obscure or confused, corresponding to the lower faculty of
sense, the desire can mislead us into erroneous or bad conduct; if it is
clear and distinct, coming from the higher cognitive faculty of under~
13 Thus Hutcheson (Essay on the Nature and Conduct of the Passions and Affec-
tions [Selby-Bigge, British Moralists, I, 400 n.] which existed in a German transla-
tion of 1765) quotes the scholastic definition of appetitus rationalis: "A constant
natural Disposition of Soul to desire what the Understanding, or these sublimer
Sensations, represent as Good," and adds, "This many call the Will as distinct from
the Passions." On the concept of "passion" and Kant's views of its relation to feeling
and reason, cf. Karl Bernecker, Kritische DaTStellung des Affekthegriffes 'LIon
Descartes his zur Gegenwart (Diss., Greifswald, 19 1 5).
14 Wolff, Psychologia empirica, § 584; Baumgarten, Metaphysica, § 689; Baumeister,
Philosophia definitiva, §§ 849,852,891.
15 Op. cit., §§ 887-90' Pleasure necessarily accompanies the awareness of a perfec-
tion, according to Wolff. This is denied by Kant (Erste Einleitung in die Kritik der
Urteilskraft, XX, 226). But Wolff is not, at least by his own profession, a hedonist,
for the perfection is desired because it is good and not because its achievement
gives pleasure (d. Verniinftige Gedancken von der Menscben Thun und Lassen,
H 14, 139; Psychologia empirica, §§ 511, 558-59). Many of Wolff's followers did
not observe this nice distinction, which resembles Butler's, and were easy targets
for Kant's criticism here.
standing, the will is rightly guided to choose a real perfection. Hence
the work of reason or understanding in morals, as elsewhere in the ra-
tionalistic philosophy, is to bring our ideas to clearness and distinctness,
for the difference between a sensuous and a rational concept is a dif-
ference only in clarity and not in kind. Hence the Wolffians, Kant tells
us, should not have distinguished between two faculties of desire. For
them, consistency would require only one faculty, and only he is able
properly to distinguish two. The Wolffians should have admitted:
If the determination of the will rests on the feelings of agreeableness or
disagreeableness which [one] expects from any cause, it is all the same to him
through what kind of notion [whether intellectual or sensible] he is affected.
The only thing [one] considers in making a choice [if Wolff is correct] is
how great, how long lasting, how easily obtained, and how often repeated this
agreeableness is [23 (110)].
But the essential epistemological premise that Kant had held since the
Inaugural Dissertation was that there is a generic difference between
sensibility and understanding. The corresponding ethical thesis here is
that there is a generic difference between any sensuous desire, no mat-
ter how refined, with its maxim, no matter how clear and distinct, and
a higher faculty of desire and its principle, which is not determined by
desire for pleasure at all. In fact, this distinction is even sharper in
Kant's ethical writings than in his epistemology, since knowledge is al-
ways the joint product of sensibility and understanding, whereas mo-
rality depends solely upon the determination of conduct by the rational
faculty. Empirical practical reason is always concerned with the satis-
faction of the lower faculty of desire. If there is a pure practical rea-
son, it must be the higher faculty of desire which is not just a refined
Epicurean version of our empirical and animal nature. The higher fac-
ulty of desire must, of course, have an entirely different structure and
function from the lower faculty.
§ 5. MAXIMS OF THE LOWER FACULTY OF DESIRE
Our next task is to follow Kant's argument that the lower faculty of
desire cannot give rise to laws, but only to mere maxims. If he can show
this and if he can demonstrate that the moral law is a law in the strict
sense of the word, he will thereby show that an ethical system cannot
be based on an empirical foundation in the nature of the human lower
faculty of desire.
Theorem I is as follows: "All practical principles [Prinzipien] which
presuppose an object (material) of the faculty of desire as the deter-
mining ground [Bestimmungsgrund] of the will are, without excep-
tion, empirical and can furnish no practical laws" (2 I [107])' It will
follow from this that all imperatives which constrain us or require ac-
tion for the satisfaction of the lower faculty of desire are hypothetical
imperatives.
Before studying the proof of this theorem, let us examine for a mo-
ment the word "material." "Material" here seems to be equated with
"object of the lower faculty of desire," but the connotations of the two
words are distinct, and the distinction is of great importance. "Mate-
rial" is also contrasted with "form" and "formal," and Kant means
here: All practical principles which figure as the cognitive component
in volition because of their content, i.e., their reference to an object of
desire, and not because of their form, are empirical. It is of the utmost
importance not to fall into the common misapprehension of this theo-
rem and to conclude that Kant means that the presence of a desire and
hence of a material disqualifies a maxim from being a law. He says ex-
plicitly that there must be an object of desire if there is to be action at
all (34 [123]). The theorem disqualifies only those maxims which are
chosen to guide conduct because of their content, i.e., because of their
reference to an object of desire (material) as the determining factor.
All maxims have material; but only the latter are material maxims. Con-
tent (object of desire) without form is blind impulse; form without
object of desire is practically ineffective-this is as true of Kant's ethics
as the corresponding sentence in the first Critique is of his theory of
knowledge.
I now give Kant's proof of the first part of the theorem, i.e., down
to the last six words. If the desire for an object is the condition for
enunciating the proposition expressing the determination of the will to
realize this object, the practical proposition is empirical. This is because
desire for the object is the ground for determination, desire for the ob-
ject being desire for the pleasure expected from its existence; and
knowledge that the object will give pleasure is merely empirical and
hence, at be~t, only probable. Therefore, the maxim which calls for such
action is valid only under the empirical and therefore uncertain condi-
tion that the object will in fact give pleasure.
Kant's proof can be strengthened, so as to make it independent of the
debatable hedonistic theory of desire, by considering the rules in-
volved in such an action. A rule for the realization of any object can be
learned only empirically. Therefore, a determination of the will and its
train of specific rules guiding action is doubly uncertain: it is uncer-
tain whether the production Of the object will satisfy the desire (what-
ever it is), and it is uncertain whether the object will be produced by
the action initiated and guided by the acceptance of the maxim.
The second paragraph gives the proof of the last part of the theorem,
to wit, that such material principles can furnish no practical laws. It
would have been better to say. "cannot be practical laws," since his
own wording suggests that laws are related to principles in the way in
which a rule is related to them and thus creates the supposition that
there are a priori rules of morality.
A principle is a law when the condition of the general determina-
tion of the will is correctly regarded as a condition of the will of every
rational being as such. But whether the object will give pleasure is con-
tingent upon the condition of the subject, and this condition is known
only empirically and cannot be known to be necessarily and univer-
sally present in rational beings. To hold as a practical law, the condition
underlying the principle would have to be known to be necessarily
valid for rational beings as such, and this knowledge cannot be got by
any observation. When a theoretical proposition is made into a prac-
tical proposition because of our interest in its content, the practical
proposition is never a law; it is, at best, a mere maxim or a rule prac-
tically relevant on the assumption of the maxim.
With the proof of Theorem I, the stage is set for the analytic of
pure practical reason, which begins at Theorem III. Theorem II and
Remark II of § 8, however, are parts of an analytic of empirical prac-
tical reason, and to them we now turn.
§ 6. HAPPINESS
Theorem II is_"All material practical principles [Prinzipien] are, as
such, of one and the same kind, and belong under the general principle
[Prinzip; should be Grundsatz] of self-love or one's own happiness"
(22 [lOS]). In this section, we shall be concerned with the theorem
only in so far as it expresses the doctrine of hedonism and shall reserve
for later treatment those aspects of it concerned with the ethics of
egoism.
A rational being's consciousness of the agreeableness of life which
without interruption accompanies his whole existence is happiness. That
a being who desires pleasures desires happiness, and conversely, follows
from this. Therefore, any practical proposition which is based on the
desire for pleasure (a material principle) falls under the fundamental
principle of making desire for happiness the supreme motive in action
and choice. The desire for happiness, however, is not merely the ag-
gregate of our desires for pleasures. Happiness, unlike pleasure, is a
concept belonging to understanding, not to feeling; it is not itself a
direct object of anyone impulse. Its pursuit, therefore, is guided by a
maxim of higher order than the rules for the pursuit of a particular
pleasure and is therefore under the regulation of a general fundamental
principle, while maxims or rules for the pursuit of particular pleasures
can hardly be called "principles" at all. But as happiness is our concept
of a systematic whole of pleasures, the maxims for the pursuit of
pleasures are under the restrictive condition of a general principle
which regulates their relations to each other. In this sense they "belong
to" the general principle (Grundsatz) of happiness.
From this is derived the corollary that all material practical principles
place the motive in the lower faculty of desire and that, if all our
principles were of this kind, there would be nothing worthy of the
name of the higher faculty of desire (22 [109]).
The second Remark in § 3 discusses further the insufficiency of the
desire for happiness as supplying a practical law (25 [I 12]). The desire
for happiness is admitted to be necessary in any rational being endowed
with feeling and to be one~ though not always an effective, determinant
of the being~s interests. But because happiness is merely the "general
name for subjective grounds of determination" and since these grounds
vary from person to person and from time to time within one person,
the principle cannot be a law~ which must be the same for all rational
beings. The diversity of interests summed up under the abstract name
of happiness, even supposing that all men consistently follow the maxim
of trying to achieve happiness as they conceive it, for themselves or
even for others, would lead necessarily to conflicts in conduct. 1s
Kant~s argument at this point leaves much to be desired in cogency.
He is placing a demand upon hedonism that his own theory cannot
meet~ to wit, that identity of principle should lead to identity of rule
and action. While the Critique does not make it as clear as does the
Metaphysics of Morals, there can be a variety of rules of action even
under the moral law itself. To require that the desire for happiness
should issue in identical and universally applicable specific rules and
to conclude that, because it does not do this, it is an inadequate founda-
tion for law is to demand too much of any principle. Kant's own
formulation of the law would not pass this test, though he seems to
have thought, in his oversimplified examples, that it could. The moral
law requires that we should seek the happiness of others; but it does
not require that all benevolent men should do the same action, for the
happiness of others is at least as variable as my own. That my concept
of happiness can issue only in contingent rules, therefore, does not
show that this principle is not a law and does not provide a basis for
destructively criticizing it.
The more valid argument that an empirical principle cannot be or
16 Critique of Practical Reason, 18 (115-16); d. also Versuch den BegTiff nega-
tiver Grossen in die Weltweisbeit einzufiihTe7I., II, t81-81, versus Maupenuis.
The Analytic of Empirical Prl1ctical Reason 99
furnish a law is given in the second paragraph of Remark II under § 3.
This argument, which is formal in character, has already been discussed
in chapter vi, § 5.
Objectionable as the argument in the Remark to § 3 may be from
the standpoint of fairness to his opponents, however, we should not
overlook it in favor of the purely fonnal argument, for in it Kant is
enunciating a richer principle for the moral imperative. He is saying
that a maxim must be universalizable in the sense not only that it must
not be self-contradictory or contradictory to the purposes of the person
who accepts it in a particular instance but that the actions which it
dictates must be compatible with each other. It is not really essential
that the actions be identical, though he implies that it is. It is essential,
however, that the actions which one man undertakes under a maxim
not be incompatible with the actions which I or another man under-
takes under the same maxim; and if they are, then the maxims cannot
be instances of a single law. In a fragment he says this more clearly
than anywhere else: "The regulative principle of freedom: that [the
actions] do not conflict; the constitutive principle: that they reciprocal-
ly promote each other [for the] purpose of happiness."17
If we apply this richer principle to the case in point, I could without
logical contradiction will that each man should strive only for his own
happiness. If I do so, I am willing that each man should follow a self-
consistent maxim but, at the same time, that he should follow a maxim
that will produce inco,!l1patible and mutually frustrating actions. This
maxim cannot be a law, because a law must be a unifying and not a
divisive factor in the world; it must make a system of ends possible. 18
In § 3 Kant is thinking too narrowly of a logical criterion of law,
where logic admits of only identity and difference in terms. But his
theory requires him to think in terms also of real repugnance and
hannony, and not merely of difference and identity. His brief and
pointed statements and examples usually illustrate the latter concern,
even when some of the former concern is presupposed. In the Meta-
physics of Morals, however, the former principle comes into its full
rights, though it has been made explicit in the concept of a realm of
ends in the Foundations and that of a moral commonwealth in Religion
'Within the Limits of Reason Alone.19
1'1' Reflexion 72 51.

18 This is the principal component in the doctrine of the type of the moral law
(d. below, chap. ix, § 10).
19 Foundations of the Metaphysics of Morals, 430 (49); Religion within the Limits
of Reason Alone, VI, 98 (Greene and Hudson, <)0).
§ 7. EGOISM
Kant equates the principle of hedonism with that of egoism. To seek

happiness and to seek one's own happiness are tacitly identified in
Theorem II.
Self-love is "predominant benevolence towards oneself (philautia)"
and is also called "selfishness."2o Interest in one's own happiness is
perfectly natural and inescapable, and to seek one's own happiness is
at least indirectly a duty, since its lack is a rich source of temptation
against the perfection of one's self.21 There is a joy in the apprehension
of one's own perfection, or at least in recognizing one's progress to-
ward it. This may properly be called the moral feeling; but any theory
which places the worth or criterion of morality in the pleasure of this
feeling reduces to a eudaemonistic, egoistic theory that Kant thinks is
patently false, since the acknowledgment of what is right must precede
and be independent of the enjoyment of our possession of it (38 [I 28] ).
With apparently no apprehension that he is being inconsistent, Kant
admits that an interest in the happiness of others can be a part of our
nature and that such directly altruistic actions can occur without any
moral constraint to them. Such an interest is present in at least some
men as a sympathetic disposition. 22 We cannot suppose the interest or
disposition to be present in all men, and certainly not in all rational
beings as such. The reason we have a duty to consider the happiness
of others is entirely independent of the fact that some men do naturally
consider it.
20 Critique of Practical Reason, 73 (165). There is a difference between Kant's
concept of self-love in § 3 and that in later parts of the Critique and in the Lectures
on Ethics. In chap. iii of the Critique and in the Lectures, Kant is most directly
concerned with what I shall call an "egoism of the will," while here his target is
"egoism of the feelings." By "egoism of the will" I do not mean the selfish prefer-
ence of one's own happiness to that of others, which is what is ordinarily meant
by selfishness and which is the meaning it has in § 3. The principle of self-love ex-
pressive of the egoism of the will is "the propensity to make the subjective grounds
of one's choice into an objective determining ground of will in general" (Critique
of Practical Reason, 74 [J66J), and Kant calls it "moral self-love" and "an inclina-
tion to be satisfied with one's perfections" (Lectures on Ethics, 135). No doubt the
two vices go together, and if Kant is correct in Theorem I, they must go together;
but they are, nevertheless, different vices. The egoism of will approaches moral
conceit and arrogance and is a much more serious vice than mere preference for
one's own happiness to that of others.
21Critique of Practjc~l Reason, 93 (186); Foundations of the Metaphysics of
Morals, 399 (15).
22 Lectures on Ethics, 194: "We have an instinct to benevolence, but not an in-
stinct to righteousness."
The A1Ullytic of Empirical Practical Reason 101
The inconsistency between saying that all material maxims are max-
ims of self-love and that some men have a direct and even general
interest in the happiness of others, therefore, does not mean that there
is an inconsistency in Kant's ethical theory. Whether we have or have
not a duty to consider our own happiness or that of others is not af-
fected by a decision as to whether we do, in fact, treasure the happiness
of others only insofar as it contributes to our own. But a serious in-
consistency in his psychological theory and in the analytic of empirical
practical reason is present.
This inconsistency was often present in the writers whom Kant
most admired. Butler, whom Kant presumably did not know as a writer
on ethical subjects,23 had been able to point the way to making it
psychologically intelligible that men could have, and indeed sometimes
do have, a direct sympathetic or benevolent interest in the welfare of
,. others. For philosophers like Butler and his opponent Hobbes,24 who
based their (different) ethical theories upon their (different) concep-
tions of human nature, it was of the utmost importance to know
whether men could possibly be genuinely unselfish. For Kant, on the
other hand, to whom the facts about human nature (whatever they
might be) do not suffice to determine man's duty, the dispute between
the egoists and the altruists lies this side of the ethical question. The
ethical question is independent of questions of human nature. But it
is nonetheless regrettable that Kant did not clarify his own thinking
about human nature ip. this important respect.
Let us not, with some other writers on Kant, exaggerate the im-
portance of this error. It has been said that, in order to break the hold
of pleasure (especially of one's own) on the faculty of desire and to
find maxims which are not selfish and hedonistic, Kant had to leave
the empirical realm altogether and base his ethics upon what amounted
to an actual renunciation of human nature. The conclusion is accord-
ingly drawn that if Kant had had a more adequate psychology, like
that of Butler, he would not have had to reject so completely the con-
ception of human nature as a basis for morals.
23 But he might have found the same thing in Wolff, Verniinftige Gedancken
'/Jon der Menschen Thun und Lassen, § 139, and in Hume, Inquiry concerning the
Principles of Morals, Appendix II.
24 Because Kant distinguished, as we have seen, between practical pleasure and
contemplative pleasure, he is not guilty of the oversimplification of the hedonists
and egoists criticized by Butler. What Butler says of pleasure in general, Kant says
only of pleasure in moral action: there is a pleasure in the action, but the interest in
the moral action is not an interest in the pleasure itself (Critique of Practical Reason,
116,160 [z 12,258]; Metaphysik der Sitte'Tl, VI, 378).
This, however, is incorrect. Butler's analysis of desire may refute
Hobbes or Mandeville not only in psychology but in ethics putatively
based upon an analysis of human nature; but it does not touch Kant in
any essential point. For Kant's argument against the principle of private
pleasure is not that it is selfish in any morally repugnant sense, and his
argument that we should concern ourselves with the happiness of
others is not predicated upon the alleged fact that we do so concern
ourselves. Precisely the same infirmity in the moral argument for
selfishness is found in the moral argument for altruism, when each is
based upon real or alleged facts of human nature. 25 Neither desire-
that for my own happiness or that for the happiness of others-can
found a duty to seek either.
§ 8. HETERONOMY
All action directed to the satisfaction of an interest in an object of
desire is action according to rules which fall under the material of
some maxim. If constraint is necessary in order that these actions rather
than others take place, the action is commanded by a hypothetical
imperative. In actions under a hypothetical imperative, we have seen
that there are always two laws which are relevant. One of them comes
from reason itself and is the formal principle of hypothetical impera-
tives. The prudent man accepts this principle in the choice of actions
for the satisfaction of any desire. The other is a law of nature, or at
least some more or less trustworthy generalization about facts of ex-
perience, and states a connection between causes and effects. It supplies
the values, as it were, for the variables of the formal principle, so that
the prudent man not only decides to do that which will, in principle,
lead to his goal but decides what it is that will, in fact, lead to it. The
second of these Jaws does not come from reason itself but is borrowed
from experience. 26 In content, it is a theoretical proposition; only the
choice of a specific end by the acting person gives it practical relevance.
In following a hypothetical imperative, reason is thereby acting un-
der one law which it has not itself prescribed; it chooses a law, pre-
scribed as it were by nature to the person who has, as a matter of
empirical fact, the goal which can be achieved by use of this theoretical
knowledge. Kant borrows from politics his name for this mode of
action: he calls it "heteronomy" and distinguishes it from autonomy or
self-legislation. A reason which is the slave of the passions, a will which
follows the promptings of desire and chooses laws of nature as its
23 Foundations of the Metaphysics of Morals, 398 ([4); Critique of Practical
Reason, 34 (113)·
26 Foundations of the Metaphysics of Morais, 435-36 (54).
guide in satisfying them, a principle or a maxim whose content is the

condition of an act of choice, and the imperative 27 which directs this
choice of a specific action-all these can be called "heteronomous,"
even if the laws are laws of nature or even of God (38, 152 [127-28,
25 0 ]).
Remark II of § 8 contains (40 [129]) what Kant claimed was an ex-
haustive classification of all material practical principles which had
been proposed as a basis for morality, together with a summary criti-
cism and rejection of each. By such criticism and rejection, he claims
to have established the necessity for there being a purely formal law
under which and through which the will can be autonomous.
Kant was fascinated by such classifications, and there exist, in addi-
tion to the one he chose finally for the Critique, at least four others.28
In the Foundations, the classification is the following:
I. Empirical principles, drawn from the principle of happiness
I. Physical feeling
2. Moral feeling
II. Rational principles, drawn from the principle of perfection
3. Ontological concept: perfection as a possible result of action
4. Theological concept: an independent (antecedent) perfection, viz., the
will of God.
27 Kant does not apply the adjective "heteronomous" to imperatives, but its use
in this connection is suggested by Manfred Moritz, Studien ZU711 Pflichtbegriff in der
Kantischen Ethik. The important point here is not, however, terminological but
substantive; an imperative that is heteronomous in origin may nevertheless be cate-
gorical in form (cf. my "Apodictic Imperatives," Kant-Studien, XLIX [1957),7-14).
28 They appear in Reflexionen 663 [ and 6637, Lectures on Ethics, I2 ff., and Foun-
dllt;ons of the Metaphysics of Morals, 441 f. (60 f.). The Reflexions are undoubtedly
early, and the classifications given include Kant's own theory. They were written
before he had clearly formulated the distinction between heteronomy and auton-
omy, which does not make its appearance prior to the Foundations. The theory he
was later to espouse appears in them as "truth," as the source from which moral
principles are derived. By "truth" he refers to the character of those maxims which
are objective and necessary, as shown by the fact that they can be "openly ad-
mitted": "That the maxim of which can be openly admitted is good. Consequently
everything morally bad is against the truth, because it tacitly assumes a different
maxim from the one it professes" (Reflexion 6642). The language is that of Samuel
Oacke. Here is one of the earliest anticipations of the categorical imperative and
the "maxim of publicity" in Perpetual Peace.
The classification of the empirical principles in the Lectures is the same as in the
Critique, though Mandeville is cited as representing the doctrine of physical feeling
and Hobbes as the representative of "government." The intellectual principles
classified in the Lectures also include Kant's own theory that the "inner nature of
the action as comprehended by the understanding" is the valid principle. Because
his own theory is included, these early attempts at classification are not, strictly
speaking, classifications of heteronomous principles alone.
104 Tbe Analytic of Empirical Practical Reason
This classification, however, would no longer serve after Kant had
established, in Theorem II, that all material principles, including the
rational ones, were under the general principle of happiness or self-love.
Both principles I and II are derived from this fundamental principle,
as he explicitly states in the penultimate paragraph of the Remark. In
improving upon this classification, he goes back to the Reflexionen
and the Lectures and takes over the internal-external (i.e., natural-
conventional) distinction employed in them and thus creates a new
genus, "external subjective principles," and puts concepts I and 2 into
the same genus as being both subjective and internal. There then results
the following classification:
I. Subjective
A. External
I. Education (Monraigne)
1. Civil constitution (Mandeville)
B. Internal
3. Physical feeling (Epicurus)
4. Moral feeling (Hutcheson)
11. Objective
A. Internal
5. Perfection (Wolff and the Stoics)
B. External
6. Will of God (Crusius and other theological moralists)
The major criticism of these tables is the at least partial arbitrariness
of ascribing the views to specific thinkers; Kant, who has already com-
plained of the injustice of labels in philosophy (13 n. [lOon.]), here
applies labels to his opponents. The arbitrariness is made clear in the
difference in the positions assigned to the same thinkers in different
versions of the table. Thus in Reflexion 6637 and in the Lectures,
Mandeville and Helvetius are classified under what, in the Critique, is
given as physical feeling; and Hobbes in the Lectures is taken as rep-
resenting the political principle, since, in his view, the sovereign power
may permit or prohibit any action and thereby make it right or wrong.
The attribution of this view later to Mandeville may no doubt be
justified by citing his aphorism that "moral virtues are the political off-
spring which flattery begot upon pride."29 Montaigne is described, in
the Lectures, as "basing himself on examples and pointing out that in
29 "It is visible, then, that it was not any heathen religion or other idolatrous
superstition, that first put man upon crossing his appetites and subduing his dearest
inclinations, but the skillful management of wary politicians; and the nearer we
search into human nature, the more we shall be convinced, that the moral virtues
are the political offspring which flattery begot upon pride" (An Inquiry into the
Origin of Moral Virtue [1723], in Selby-Bigge, British Moralists, II, 353).
matters of morality men differ with their environment, and that the
morality of one locality is not that of another."30 There is, of course,
only a short step from "custom" to "education," especially when ed-
ucation is as fallible and variable (because not guided by an ideal of
universal validity) as Montaigne regarded it.31
The criticism of Epicurus has already been given in Kant's general
condemnation of the hedonistic foundation for ethics, though he ad-
mires the consistency of the Epicurean doctrine, which he holds to be
higher than that of the moral-sense theory (24 [J I I]). In the Inaugural
Dissertation, § 9, Kant had placed Epicurus and Shaftesbury together
in reducing "the criteria of morals to the feeling of pleasure or un-
pleasantness," and for this he was criticized by Mendelssohn, who
rightly saw that there was a difference in saying, with Epicurus, that
pleasure is what makes good things good and in saying, with Shaftes-
bury, that there is a particular pleasure in the contemplation of the
good which serves as a mark of its presence.32 In now separating them,
Kant criticizes the moral-sense theory because, to the degree to which
it tries to distinguish the pleasurableness of the moral feeling from that
of physical feeling,33 it must already presuppose that the concept of
the good is held in mind, conformity to which could give the special
feeling of moral pleasure. Feeling itself has no cognitive value; we have
"no feeling for law as such"; for consciousness of the law is a product
of reason and, without it, conformity to law would bring no pleasure
(3 8 [ 12 9]) .
.But Kant is clearly wrong in this argument. He erroneously thinks
that, by showing the pleasure of the moral sense to be like all other
pleasures, he has a right to say that the moral-sense theory "reduces
everything to the desire for one's own happiness" ( 38 [I 28] ). This
confuses two very different things that the British philosophers had
kept properly separate, viz., the disinterested pleasure we experience
in doing something righteous or in contemplating a righteous action,
whatever it may be that makes it righteous, and the interest we have in
the pleasure accruing to us if we do or contemplate a certain action,
30 Lectures on Ethics, 11.
31 Montaigne characteristically says: ''Whatsoever is told us, and whatever we
learn, ever remember: it is a mortal hand that presents it, and a mortal hand that
receives it" (Essays, Book II, chap. xii), Kant must have known that this did not
represent the whole point of Montaigne's thought on morals, however, for the same
essay gives an extensive criticism of the variability of customs,
32 Mendelssohn to Kant, December 15,1770 (X, 114).
33 But it cannot do so; cf. Critique of Prllctical Reason, 13 (109); Metapbysik der
Sitten, VI, 376--77.
which may be righteous or unrighteous. 34 If the former pleasure is
disinterested, as Kant might hold when he developed his views on moral
contentment (117 [2 14]), its presence could serve as a mark of ethical
value, comparable to the hedonic mark of beauty, without committing
him to the view that pleasure derived from the moral sense is the goal
of the action itself. If this feeling arises from the immediate apprehen-
sion of the fittingness of actions without the need of an antecedent con-
cept of this fittingness, which would be the concept of virtue, then the
sentence beginning "One must already value the importance of what
we call duty" (38 [128]) begs the question against the theory of the
moral sense. To the assumption that we pursue virtue for the sake of
the accompanying pleasure, both Hutcheson and Price anticipate Kant
in making precisely this objection. 35
In spite of his objections to it, however, Kant prefers the moral~
sense theory to an outright and consistent hedonism because "it pays
virtue the honor of ascribing the satisfaction and esteem for her direct-
ly to morality, and does not, as it were, say to her face that it is not
her beauty but only our advantage that attaches us to her."36
Perfection as an ontological c.oncept is rejected as empty and in-
definite.37 When made specifically moral perfection, it is circular to
argue that perfection is the source of the moral predicates, which must
have been added to it. 38 In the Critique he distinguishes transcendental
and metaphysical perfection from practical perfection, which is "fit-
tingness or sufficiency of any thing to any kinds of ends." This, how-
ever, is insufficient to define more than the highest form of talent and
skill, which are ethically neutral. Thus Wolff's concept of perfection
suffers from the same weakness as his universal practical philosophy,3D
to wit, it is so general that nothing specifically ethical can be derived
from it. Ends must be given, and, if ethical perfection is here in ques-
tion, ethical ends defined by some other principle than mere perfection
or a mere general will must be presupposed. The concept of perfection,
however, is esteemed by Kant more than any of the other heteronomous
84 Cf. Hutcheson, Inquiry concerning the Original of Our Ideas of Virtue or
Moral Good, Introduction (Selby-Bigge, British Moralists, I, 72).
35 Cf. Ibid., sec. II; and Review of the Principal Questions of Morals, chap. i, sec.
iii (Selby-Bigge, British Moralists, I, 92; II, 114).
37 As already recognized in the Lectures on Ethics (pp. 24, z6, 39) in his criticism
of Baumgarten, Wolff, and Cumberland that moral rules drawn from the concept
of perfection are "tautologous."
38 Foundations of tbe Metaphysics of Morals, 443 (62).
3D Ibid., 390 (5). Wolff's conception of perfection is given in Ontologia, ~ us.
The Analytic of Empirical Practical ReasO'Tl 107
principles, since it does preserve morality from all contamination with

empirical concepts.40
Finally, there is the external perfection represented by the will of
God as the putative source of ethical principles. t1 The same objections
apply to this as to internal perfection, but even more strongly, since
the consequences of this error are much more damaging. Either Crusius
surreptitiously introduces ethical predicates into the concept of divine
perfection,42 with the result that theological perfection no longer
grounds the moral principle but presupposes it; or a hedonistic motiva-
tion is postulated as the ground of obedience to God, whereby the
"concept of the divine will is made up [only] of the attributes of desire
for glory and dominion combined with awful conceptions of might
and vengeance,"43 to the detriment and, indeed, destruction of both
ethics and religion.
But everyone of the heteronomous principles, thus banished from
the foundations of moral volition, re-enters the moral scheme of things,
once the purity of the source of morals has been secured. The external
subjective grounds return not as grounds of duty but as the duty of
decorum 44 and obedience to authority. We have a duty to cultivate
moral feeling, a feeling of satisfaction or contentment in the per-
formance of duty.45 We have at least an indirect duty to achieve
happiness if possible, both in this life and in the next. 46 The will of
God is a symbol of the holiness of will, which is an ideal to be striven
for (32 [12 I ]); our own moral perfection is one of the ends which are
also a duty.H In rejecting heteronomy, therefore, Kant does not reject
the moral, political, social, religious, and physical goods which the
philosophers of heteronomy had rightly commended. They are all
affirmed, but under the condition that their pursuit be regulated by a
formal principle.
U Crusius, Anweirung vernii1lftig zu leben (1744), § 174-
42 On relation of conscience to obedience to God, ibid., is 132-33.
U Amhropologie in pragmatischer Hinsicht, § 14.
45 Metaphysik der Sitten, Tugendlehre, Einleitung (VI, 399-400); Critique of

Practical Reason, 38 (129).
46Fou1ldations of the Metaphysics of Morals, 399 (IS); Critique of Practical
Reason,93 (186).
47 Metaphysik der Sitten, VI, 385 f. Wolff, on the contrary, in making happiness
the mere consciousness of perfection, had included the perfection of others in the
end which is also a duty (d. Philosophia practica universalis, II, § z8).
Believing, then, that he has not only proved, in Theorem II, that,
in principle, no empirical motive can give a moral law, but that in
Remark II on § 8 he has shown that no proposed empirical principle
does, in fact, provide a law in spite of the pretensions of other moralists,
Kant concludes that "the formal practical principle of pure reason" is
the only one that can furnish categorical imperatives, i.e., "practical
laws which enjoin actions as dutiful" (41 [130]). To the positive
derivation and defense of this principle we now turn.
VIII
The t'~etaPhJsical Deduction" of the Moral

Law; Commentary on Sections
4, 5, 6, 7, and Part of 8
§ I. THE IDEA OF A "METAPHYSICAL DEDUCTION"
Chapter i of the Analytic of Concepts of the Critique of Pure Reason

is entitled "The Clue to the Discovery of All Pure Concepts of the
Understanding." In that chapter Kant argues that the table of judg-
ments, taken, with slight modification, from classical logic, provides a
table of categories for knowledge. For there to be knowledge of ob-
jects, which are not given to consciousness in propria persona, there
must be a synthesis of the representations we have of objects, since
only the connection between our representations makes it possible for
different ones of them to have a common focus in the object which
is the source of each of them. Synthesis of representations, of course,
may be mere psychological association of ideas; but this synthesis is
only subjective and holds only for the individual who happens to
associate this representation with that. For there to be knowledge, this
synthesis must be governed by rules which are valid for all knowers,
so that we can judge that an object has such and such properties and
not merely say that I happen to associate one representation with an-
other. Knowledge arises in the synthesis of representations and issues
in judgments of objects. The kinds of judgments that constitute the
corpus of knowledge, therefore, are a clue to the kinds of synthesis
which are necessary to knowledge. The categories, or pure concepts
of the understanding, are the concepts or rules of these syntheses, and
to each form of judgment there corresponds a category.
In a passage added in the second edition, Kant refers to this chapter
as the "metaphysical deduction." The adjective "metaphysical" refers
to the pure a priori conceptual knowledge which this chapter is sup-
posed to provide. The metaphysical deduction is distinguished from
the transcendental deduction. The latter has the responsibility for
shOWing that the categories do enable us to have a priori knowledge of
109
110 The "Metaphysical Deduction" of Moral Law
objects. The metaphysical deduction is Kant's effort to discover what
the categories are; the transcendental deduction is his effort to show
that they are valid. The two deductions are, in principle, independent
of each other, though they both have the same starting point in the
doctrine that knowledge requires synthesis and is always expressed in
judgments. Had the metaphysical deduction in some way discovered
other categories than those it did in fact discover, the transcendental
deduction need not have been affected in any significant way. The
specific categories derived in the metaphysical deduction are used only
illustratively in the transcendental deduction. The essentials of the
metaphysical deduction were clear in Kant's mind long before he even
realized the need for a transcendental deduction.
In the Critique of Practical Reason and even more clearly in the
sharply demarcated divisions of the Foundations of the Metaphysics
of Morals, there is an analogous separation of tasks. Kant first sets out
to discover what the moral law must be if the concept of duty and
therewith all morality are not spurious. Then he attempts to show that
the law thus discovered is valid and not just an empty conception.
Yet the separation is not so clean-cut in the second as in the first
Critique. The reader himself must draw the boundaries between them,
and the boundary is not a single straight line but a wandering line that
weaves its way in and out of the various sections. 1 There are several
reasons for this lack of formal neatness, which is elsewhere one of the
most characteristic marks of organization in Kant's larger works. The
moral law, he says, is in no need of "justifying grounds," for it speaks
with authority even to the commonest intelligence. Human beings, he
believes, acknowledge its validity even when they could not begin to
state it in formally exact terms. To formulate, in Socratic manner, and
not to teach and defend, as it were, from the outside is the method to
be followed in practical philosophy. Hence Kant often permits himself
to speak as if the moral law had been established (that is, transcenden-
tally deduced) when, in fact, at those points he has only succeeded in
formulating it (i.e., metaphysically deduced it).
Moreover, the fact that there is for Kant only one moral law makes
it more difficult, simply as a stylistic matter, to separate the task of
formulating it from that of defending it. In the first Critique, he had
twelve categories, any of which he might choose, from time to time,
for purposes of illustration, since the specific justification of anyone
of the twelve, in the metaphysical deduction, was easily distinguished
from the grounds of justification of any category whatsoever. Hence
1 He is more careful to preserve the distinction between the two tasks in Founda-
tions of the Metaphysics of Morals; cf. 416, 44S (44.64).
The "Metaphysical Deduction" of Moral Law I I I
illustration and justification were easily distinguished. But here It IS

only one principle that is under examination, and it is one that he
thinks is already given with full authority, needing only to be worked
out and brought to clear consciousness. Consequently, the formulation
and the defense of the principle can hardly be so clearly separated as
in the earlier work.
Nevertheless, he is aware of the dual nature of his task, and he points
it out explicitly in his distinction between the "exposition" and the
"deduction" (46 [135]). We could, in fact, as well speak of the "meta-
physical exposition" as of the "metaphysical deduction" in each of the
first two Critique's, for by "exposition" he means "the clear, though not
necessarily exhaustive representation of that which belongs to a con-
cept; the exposition is metaphysical when it contains that which ex-
hibits the concept as given a priori."2 But the established position of
"metaphysical deduction" in Kant's architectonic makes me prefer this
name for the material about to be examined.
Still, too much must not be expected from the proffered analogy to
chapter i of the Analytic of the first Critique, in spite of its obvious
existence. For there is no ready-made table of formal distinctions to
serve as a clue to the discovery of the principle of pure practical rea-
son. There is nothing in the Critique of Practical Reason comparable
to the synthetic procedure of the metaphysical deduction in the first.
Yet there is a parallelism between the larger movements of thought
that underlie the formulations in each work. For Kant himself says:
"We come to know pure practical laws in the same way we know pure
theoretical principles, [to wit] by attending to the necessity with which
reason prescribes them to us, and to the elimination from them of all
empirical conditions, which reason directs" (30 [uS]).
In spite of the differences in appearance, the argument of the Cri-
tique of Practical Reason in its discovery and formulation of the moral
law is comparable to that of the Foundations and the Prolegomena.
In all three of these works, Kant begins with a problem presented by
ordinary experience to philosophical analysis, and he reaches his con-
clusion by attempting to demonstrate that only one principle or theory
is capable of rendering the problematic experience intelligible. In the
Foundations and in the second Critique it is practical experience in
general and moral experience in particular. Our task in this chapter is,
with Kant, to "attend to the necessity" which is experienced in moral
concern, to show how this necessity requires the elimination, or at
least the transcending, of empirical conditions, and then to show what
remains as the formal principle of a pure practical reason.
2 Critique of Pure Reason, B 38. It is interesting to observe that the Analytic of
the Critique of Judgment is formally divided into "exposition" and "deduction,"
I 12 The "Metaphysical Deduction" of Moral Law
§ 2. PRINCIPAL FEATURES OF THE MORAL EXPERIENCE
In the analytical method employed here and in the first two parts
of the Foundations, Kant begins with an experience and regresses upon
those presuppositions without which it would have to be declared un-
intelligible or spurious. When the analytical method is employed at the
behoof of a metaphysical deduction or exposition, the presuppositions
are found as organizing concepts, principles, or forms of the experience
examined. The result of this regression upon the formal conditions is
the statement of the principle reached in the first two parts of the
Foundations and presupposed here (8 [93]).
There is no difficulty, he tells US,3 in discerning the possibility (i.e.,
in discovering the principles) of hypothetical imperatives. Granted
that men have desires that may be satisfied only by the application of
their knowledge of the world and granted that they are not sufficiently
rational in the pursuit of their aims always readily to do those actions
indicated by their knowledge as being necessary, it is obvious that they
will experience a constraint to do them, the constraint being expressed
in a hypothetical, conditional, heteronomous imperative.
The facts and principles which make heteronomous imperatives so
easily explicable, however, are wholly inadequate to account for some
of the central and characteristic features of moral experience. An
ethics which does not acknowledge the sharp and radical distinction
between prudence and morality and which thereby makes morals
heteronomous simply leaves unexplained, or explains away, certain ob-
vious characteristics of moral concern. Perhaps they should be ex-
plained away; it is conceivable4. that they may be spurious. Whether
they are spurious or not is the decision that must be made in the light
of the success or failure of the transcendental deduction. But, first, the
sensible thing to do is to discern what principles must be presupposed
if we take moral experience at its face value, though provisionally,
until we come to the deduction. The most important of these features
which require a different set of presuppositions from those involved in
heteronomy are the following.
a) The objective necessitation of the action.-In moral concern there
is a felt constraint to do a certain action, and this constraint has a quite
different tone from that which I experience when called upon to do
some action for the sake of some even long-range goal I have before
me. The latter is a constraint predicated upon a desire I have; the
fonner seems to be independent of and often in conflict with all my
3 Foundations of the Metaphysics of Morals, 4 1 7 (34).
"bid',407 (Z3-z4).
The "Metaphysical DeductionU of Moral Law 113
desires. This kind of constraint is called "duty," the objective necessita-

tion of an action that is not subjectively necessary but subject to my
free choice to commit or omit it.
The difference between being constrained to do something because
of my needs or wishes and being constrained to do it irrespective of
them is perhaps most easily discerned in the parallel distinction between
being "obliged" to do something and being "obligated" to do it. To
be, or to feel, obliged to do something is quite different from being,
or believing myself to be, obligated to do it. For instance, I am obliged
to put my name in my books, since I do not want them to be borrowed
and not returned; but I am obligated to return borrowed books, no
matter how much I desire to keep them as my own. It makes perfectly
good sense to say, "I had an obligation to tell the truth, but to get out
of that scrape I was obliged to lie." To be obliged to do something
means that, to accomplish a given purpose, I have to do something I
don't particularly want to do, or dislike doing. To be obligated to do
something means to be under the necessity of choosing to do something
without consulting my desires. In each case my cognitive intelligence
(for Kant, reason) must decide what it is that I ought to do; but, in
the former case, one of reason's premises is a desire, while, in the latter,
desire is not a premise (though, of course, if the phenomenon of
morality is in general spurious or in a particular case is only an instance
of hypocrisy and cant, it is a cause).
b) Motives as the object of moral judgment.-In passing a moral
judgment, I do not pass on the success or failure of an action in leading
to the object of some desire. I need not await the issue of an action to
decide whether it was good or bad; I need only know the motive which
led to it. In making a moral decision, I do not first need to know
whether the action I undertake will succeed or not; I do not even need
to know whether, in fact, I have the power to produce the effect. 1I
But in acts of skill and prudence, success is everything, and motives
are of importance only indirectly, if at all.
But, even more obviously, I do not morally judge actions themselves
so much as decisions to act. Though Kant often writes as if an action
is the object of moral judgment and is itself good or bad, it must be
remembered that action is not merely an item of external behavior
(e.g., the emitting of certain sounds) but includes, as its definitive
component, the motives, intentions, and decisions which lead to it. The
complex act of making a decision for certain reasons and not for others
is what we attend to in moral judgment; but in matters of skill and
prudence, it is the act and its outcome that count.
II Critique of Practical Reason, 15.45.68 (JOI, 135. 160).
c) The universality of 'moral judgment.-Moral judgment is no re-
specter of persons. Though Kant admits in his discussions of imperfect
duties that circumstances make legitimate differences in the specific
acts by which moral responsibilities are met, there is always in moral
judgment an appeal to a disinterested forum which makes no allowance
for the one thing pre-eminently important in questions of prudence,
viz., what one happens to desire. Not only is moral obligation inde-
pendent of my desires; it is independent of specific desires that any
men may have. I can and do make moral judgments of beings that are
not men (e.g., God); and though perhaps only men experience obliga-
tion, we hold a holy will to the same objective standard of right that
we apply to ourselves.
It must always be carefully noted that the universality of moral judg-
ment has nothing to do with the variability of custom and the variety
of moral judgments current in various societies and in different ages.
The fact that moral judgment varies from time to time and from place
to place has nothing to do with the kind of universality that Kant is
speaking of here. Regrettably, he never explained the precise sense in
which he meant, and the precise sense in which he did not mean, that
moral judgments are universal. Fortunately, he did so in his treatment
of aesthetic judgment,6 and we can easily apply his discussion of that
to moral judgment. Kant was not so blind to the fact that aesthetic
judgment is highly variable as he is sometimes believed to have been
to the variability of judgment on moral questions. When he insists
that, in passing an aesthetic judgment, we are making a judgment that
claims (though it does not receive) universal assent, he is distinguishing
between what we may call "normative" universality and "social" uni-
versality: an aesthetic judgment has, or claims to have, the former but
neither has, nor claims to have, the latter. Similarly, moral judgments
are normatively universal, but Kant, who knew as much about folkways
as perhaps anyone in the eighteenth century, was not concerned with
them in the analysis of moral judgment itself. Whether what I believe
to be universally valid is in fact universally valid is one question and
an important one; whether what I believe to be universally valid is in
fact universally believed to be valid is a quite different question and
one so easy to answer in the negative that it is of no importance at all.
But the obvious negative answer to the second question does not imply
that one must in principle give a negative answer to the first question.
It is not even a premise for an answer to the first question. Sometimes
one must answer the first question in one way, sometimes in another;
and Kant is attempting, in this part of the Foundations and the Critique,
6 Critique of Judgment, ~ 7.
Tbe "AJ'etapbysical Deduction" of .Moral La'll.' r 15
to find the criterion by which it must be answered. But even prior to
the erection of this criterion, it is essential to recognize that moral
judgment makes a universal claim of a kind not made by the expression
of a wish or a liking. 7
d) Direct interest in the action.-In all actions done for the sake of
some desired consequences, our direct interest is in the consequences,
the existence of the state of affairs desiderated; and we have only an
indirect interest in the action itself. Though all actions voluntarily
undertaken do have consequences and are intended to have conse-
quences, in which there is some kind of interest,8 Kant denies that the
actions necessarily occur because of this interest in the object.9 He
asserts that in moral action there is an interest in the action itself.
meaning by "action" the whole complex of motive and decision as well
as outward conduct. In modern terminology, Kant is saying that moral
value is an intrinsic value and that it interests us because we recognize
its intrinsic value. Success and failure in achievement are not moral
categories. but concepts of technique and prudence; worthiness to be
happy and culpability are the corresponding moral categories, and their
applicability is wholly independent of success or failure.
Other characteristics might be added to this list. For instance, Kant
speaks of the difference between mere regret and repentance as distin-
7 The reader who is interested in the bearing of the facts of comparative anthro-
pology on a universal moral criterion like that of Kant is advised to read A.
Macbeath's Experiments in Living (London: Macmillan & Co., Ltd., 1951). On the
basis of a wide collection of data, Mr. Macbeath concludes that folkways and
customs alone do not establish an acknowledged standard in the consciousness of
members of a group until a claim to universality and a bearing upon the general
welfare of the group (which is a step toward universalization) are present. He thus
finds something comparable to a categorical imperative present as a cultural in-
variant in all systems of folkways, defining for each the difference between the
customary and the moral. Also relevant to the question of universalization is the
level of abstractness of the rule or principle being tested. The "exceptions to moral
rules" which are paraded by ethical or cultural relativists are not relevant to the
evaluation of a universalistic ethics like Kant's unless they can be expressed, for in-
stance, in such a form as "It is sometimes right to act from bad motives" or "It is
sometimes right to treat others merely as means," and not merely "It is sometimes
right to take property that does not legally belong to you."
8 Critique of Practicd Reason, 34 (ll3). The interest need not be an empirical
interest in some sensuous satisfaction resulting from the satiation of a need. In
moral action the empirical interest cannot be denied to exist and need not necessarily
be frustrated in action; only "no regard must be given to the interest in the object"
When we are trying to decide whether an action is morally right. Cf. ibid., 93 (186);
Foundations of the Metaphysics of Morals, 413 n. (30 n.); Uber den Gemeinsprucb
... , VIII, 278.
9 Foundations of the Metaphysics of Morals, 413 n., 459 n. (30 n., 80 n.).
II6 The "Metaphysical Deduction" of Moral Law
guishing two kinds of failure 10 and of the specificity and clarity of
moral commands, which he contrasts with the vagueness that is an in-
escapable infirmity of counsels of prudence. l1 But this latter is certainly
an oversimplification of the facts, as Kant fully realized in the Meta-
physics of Morals. Since it not only is wrong but plays no part in the
derivation of the law, it need not be further discussed here. It should
be pointed out, however, that this alleged feature of moral principles
does playa very great role in the transcendental deduction. There it is
not any longer a question of the clarity of specific moral commands in
contrast to the obscurity of prudential counsels, but only of the clarity
of the moral principle itself, by virtue of which one readily recognizes
the difference between what is right and what is only advisable, with-
out having to feel easily assured that some specific act really is the right
one in given circumstances.
§ 3. FIRST APPROXIMATION TO THE PRINCIPLE
From these salient features of moral concern, judgment, and action,
Kant draws four conclusions concerning the principle involved in
moral action.
a) The principle must be a law and not a mere maxim. Only a law
can necessitate, and only a law can be universal in application. Such
universalizability is essential for normativity. A mere maxim, even if it
did express the condition of the will of every man, would not be neces-
sary for any of them, would not obligate any of them.
b) The principle must be a law by virtue of its form and not of its
content. The content is always some object of interest in that which is
to be achieved by following the law taken as a maxim; but such an in-
terest is neither universal, nor obligation-creating, nor directed to the
action as such.
c) But the law must be capable of being a maxim, i.e., of being an
expression of the actual condition of a will. Otherwise, the law or the
action's conformity to it could not interest us, and knowledge of it
would be practically ineffective except under the contingent condition
of an interest in the object of the maxim. Knowledge of the principle
itself must provide the motive to act in accordance with it, and, when
it does so, the action is not only done in accordance with it but because
of i~, i.e., out of respect for it. Otherwise reason is only the slave of the
paSSIons.
10 Critique of Practical Reason, 37 (r2 7). The remark on Priestley's view of re-
pentance (ihid., <)8 [191]) refers to The Doctrine of Philosophical Necessity (1777),
pp. 86 ff. (note taken from Natorp in Akademie ed.).
11 Critique of Practical Reason, 36 (n6); 0 her den Gemeinspruch .•. , VIII, 186.
The "Metaphysical Deduction" of Moral Law 117
d) The law must ground a categorical imperative, though it must be

distinguished from the imperative. It must be a law of how a rational
being as such would necessarily act. Only such a law can be a categori-
cal imperative to a partially rational being like man, who does not by
nature act in accordance with his conception of law. The "would" of
the moral law must be expressed as the "should" of the imperative, but
distinguished from it.12 The "should" of this imperative, unlike that of
the hypothetical imperative, is not contingent upon there being an in-
terest in the object of the action, or even upon there being an interest in
the moral law and its reign, as if those who felt no interest in it were
absolved of all blame for transgressing it. 13
Not every formally categorical imperatIve, however, is based upon
this law, for any imperative can be put into categorical form. Only
those imperatives which are "practically right" (2 I [107]) are here in
question, and these are imperatives which are unconditional or apodic-
tic, even though they may be hypothetical in form.14 That an impera-
tive can be apodictic in authority, yet not categorical in form, is a
nicety that Kant seems to have overlooked, with unfortunate conse-
quences for his concept of the universal application of the moral law.
It permitted him to write as if the categorical imperative allowed no
room for considerations of contingencies that would appear in the ante-
cedents of hypothetical, but still necessary, imperatives. 15
§ 4. THE THREE MAXIMS IN MORAL DECISION
In chapter vi, § 6, I analyzed action undertaken because it is believed
to lead to a desired end. In the hypothetical imperative, fully expanded,
we discovered a material maxim, a theoretical law or cognition of the
causal relation of an act to its consequences, and a formal principle that
a rational being who fully wills an end also necessarily wills the means
believed to produce it.
Analysis of moral action in this chapter shows that if morality is not
to be declared chimerical, a specific action must be decided upon in-
12 Foundations of the Metaphysics of Morals, 449 (68). Strictly speaking, of
course, there is no "should" in an imperative. A sentence "You should do X" is not
an imperative to do X. But both this sentence and the imperative proper are re-
lated to the law "A rational being as such does X" in the same way; each is directed
to a rational being who does not by nature do X because of the law.
13 Thus, while there is no duty to have a moral interest, a being who does not is
condemnable not as morally evil but as falling outside the scope of moral concepts
altogether (cf. Metaphysik der Sitten, VI, 4°0).
• 14 E.g., "If you have made a promise, keep it," is hypothetical in form but apodictic
10 modality; "Shut the door" is categorical in form but problematic in modality.
1~ 0. above, pp. 87 f.
dependently of whether there are actual desires that can be satisfied by
the action and independently of whether we know and can master the
means by which the object of desire can be realized. Kant must accord-
ingly show how a specific action can be decided upon and undertaken
without presupposing a specific desire to be satisfied by the action. If
this cannot be shown, all actions would have to be analyzed in the
manner given on page 85 above, and the universality, necessity, and
intrinsic value of moral action would have to be declared illusory. For
if the determining maxim is the material one, like I, A, in that analysis
(a specific setting of the will to produce a certain end, which can be
achieved only by the use of knowledge given in I, B), the imperative
cannot be categorical, and the corresponding principle is a mere maxim.
Yet we cannot simply eliminate I, A, and leave moral action unmoti-
vated. Kant rightly recognizes that something at least analogous to I, A,
is necessary to all action. "It is certainly undeniable," he says, in a sen-
tence which seems to have been overlooked by many of his critics,16
"that every volition must have an object and therefore a material." But,
he adds, "the material cannot be supposed for this reason to be the de-
termining ground and condition of the maxim" (34 [123]). Kant must
find a principle, not to replace I, A, but to control it.17
The control of one maxim by another is nothing strange. Material
maxims themselves stand in a hierarchy. My desire for money may be
expressed in one maxim; my desire to live slothfully in another; and one
of these can control the expression of actions done under the aegis of
16 The most notable of those who have fallen into this error is Max Scheler (Der
Formalimrus in der Etbik und die materiale Wertetbik, [Halle, 1916]; also in Jabr-
bucb fur Pbilosophie und phiinomenologische Forschung, Vols. I and II). Scheler's
Kant-criticism is that, for Kane, the a priori must be empty because it is purely
formal. But since ethical principles are also, for Scheler, a priori, there must be an
a priori which is material, i.e., given in an intuition of material value~essences. In
this way, Scheler believed he could avoid the "subjectivity" and "intellectualism"
of Kant's ethics, which are consequences, he held, of Kantian "empty fonnalism."
But this criticism misses the main point of Kant's distinction between the object of
the will, which is always present, and the object of the will as its detennining ground,
which is present only to empirical practical reason. Scheler, in his own construc-
tive work, is left with no rational criterion for material principles, which is what
the fonnal principle suppHes in Kant's theory. In the schema introduced below,
Scheler locates the a priori at a, Kant at h. But while b is purely fonnal, it is not
therefore empty; its content is a.
11 Cf. Reflexion 6633 (ca. 1770): "The supreme principles of moral decision are
rational, but they are only fonnal principles. They determine no purpose, but only
the moral fonn of every purpose. Therefore, according to this fonn in concreto
primary material principles occur [vorkommen]." On the control of one maxim by
another see Paton, The Categoricallmperat;ve, pp. 136-37.
The "Metaphysical Deduction" of Moral Law 119
the other. This is the sense in which the maxim of self-love is regarded
by Kant as the supreme material maxim. But, since even this maxim can
ground only hypothetical imperatives and the explicatum here is un-
conditional obligation, the maxim he seeks cannot be governing by vir-
tue of its material. The supreme maxim must select from among the
material maxims those which should be permitted in our practical syllo-
gisms,Is and it must do so by virtue of being a criterion of maxims with
respect to their form and independent of their content. That is what
Kant means when he says that the principle he is seeking must be
purely formal; that is his obscure way of saying that it must be a prin-
ciple regulating conduct through controlling the form of permissible
maxIms.
In the analysis of moral action, therefore, we must add one more
principle to the two practical principles uncovered in our analysis of
hypothetical imperatives. We must distinguish the following:
a) A material maxim setting forth some general condition of my
will.
b) A principle (law) which determines whether a is subject to mor-
ally permissible execution and which is itself a maxim.
c) The formal principle of the hypothetical imperative, which is the
condition of the choice of acts necessary to attain the end set forth in a,
the actual means being discovered by the cognition of causal connec-
tions between acts and consequences.
F or instance, my purpose (a) may be to promote the happiness of
others. The law (b) authorizes me to act upon this maxim. The prin-
ciple (c) selects from among my empirical cognitions those that make
specific actions incumbent upon me, granting the legitimacy of a. The
law (b) forbids certain types of maxim (a), allows some, and requires
still others.
The easiest case to discern in Kant's discussions is the first, so much
so that it is often taken as the only one. He is often interpreted lD as be-
lieving (i) that the moral law is only a criterion for the admissibility of
material maxims and (ii) that there is always a conflict between natu-
ral, spontaneous material maxims and the law. This interpretation is
only partially correct and must be supplemented in what follows. But
the reason why this half-correct interpretation is so easy lies in Kant's
realistic and unsentimental estimate of human character. Where a mate-
18 This is said in Foundations of the Metaphysics of Morals, 432, 444 (50, 63): a
will which made a material maxim its law would need still another law to give it the
proper universality.
19 If, that is, the objector has given up the still more common error of believing
that Kant thought that b had nothing to do with II at all.
rial maxim leads to action which does in fact accord with the law, any-
one with our philosopher's sense of hard fact and hatred of cant will
naturally regard the material as the motive and will distrust its embel-
lishment as duty (154 [252])' So that the role of each can be distin-
guished, Kant preferred to deal with the extreme cases in which the
two things, duty and desire, work at cross-purposes. 20 Hence his ethics
seems to be more an ethics of prohibition and of repression of the nat-
ural man than one of the celebration of spontaneous good nature. Un-
doubtedly, thIS semblance is based in part on Kant's own personal habitus
(at least by the time he reached the writing of this book); but even
more it is based upon many readers' failure to remember the polemical
situation in which Kant found himself, one in which he had to separate
and set apart and seem to set in opposition those things that had been
confused and even identified by others.
The most significant case, however, is that in which the moral law
requires me to act on certain maxims, whether they express an ante-
cedent condition of my will or not. If we are to justify the alleged di-
rect interest in moral action, the moral principle must become a maxim
with its own material. This material is not empirical, like happiness, but
is the morally good, which is defined by the principle itself and not de-
sired antecedently to it. Only if the formal principle can itself be taken
as a maxim, so that we strive to instill in ourselves a certain readiness to
act in a given way regardless of whether we have an antecedent dispo-
sition to these actions, can we say that pure reason, the exclusive source
of b, can be practical.
20 Cf. Critique of Practical Reason, 9Z--93 (186), where he compares his procedure
to that of a chemist in separating a compound into its elements. Kant's examples in
the Foundations are of this type. They concern, for instance, the man without sym-
pathy who acts from "practical love." If the purpose Kant had in devising these
examples is remembered, one can see the error of interpreting him as a proponent of
"sour duty." He is simply insisting that a and b be distinguished and that we not
attribute moral value to a alone but to a only when it is instilled in us or acted upon
because of b; what other values a may have is not being discussed here at all, but
the specific moral value is being made salient by contrast, i.e., by choosing cases of
a in which the only prima facie value is the moral one. These are not examples for
imitation but examples for illustration. We shaH return to this matter in chap. xii
when we discuss the relations between Kant and Schiller. It is a pity that the ques-
tion must be discussed so fully, but it is necessary because Kant did not make the
purpose of his examples absolutely clear with repeated statements that could not be
overlooked even by his opponents. Had he done so, the quantity of writing against
his ethics would have been halved and its quality raised immeasurably. I have dis-
cussed the misinterpretation of the examples which arises from neglect of the
polemical context in "Sir David Ross on Duty and Purpose in Kant," Philosophy
md Phenomenological Research, XVI (1955),98-107.
The "Metaphysical Deduction" of Moral Law I ZI
But we have not yet seen what this law is which requires that our
maxims have a certain form and which has an object that cannot be
derived from our natural desires.
§ 5. DEFINITION OF THE PRINCIPLE
The supreme formal principle states that a purely rational being
would act on no material maxim that could not be accepted as itself a
law for all rational beings. All maxims of a rational being as such are
independent of the differences between them, i.e., their desires; the
maxims of a rational being as such are, therefore, laws. The imperative,
which is the formula of this law's constraint on a partially rational be-
ing, is that a maxim should be followed only if it is a maxim that I
could will that all rational beings should follow. The imperative serves
as a criterion for maxims that may function in morally permissible
actions.
Since, however, to be moral and not merely legal, there must be a
direct interest in the action, the principle must not be merely a nega-
tive test for an antecedent maxim, as the previous paragraph makes it
seem. The principle must be something positive, something command-
ing us to establish maxims in our hearts that are recognized as valid for
us as rational beings because they are valid for rational beings generally
or willed to be the maxims that all rational beings should follow. In this
positive, constitutive sense, contrasted with the previous negative and
regulative sense of the principle, knowledge of the principle must itself
b~ a ground for action. We must not merely act on maxims that do pass
the test, but our supreme maxim must be to act upon them because they
do pass it.
ft is very easy to emphasize the negative, regulative aspect and to
neglect the positive, constitutive aspect of the principle and then to
suppose that Kant is refuted by showing that there are maxims which
could be universalized but which do not obligate me. 21 For instance, I
can easily will that everyone should write his name in his books, but
this has no effect on my duty. In the negative, regulative function,
there are, as it were, two volitions: a desire to do something, but only
if I can desire others to do it too. In the positive, constitutive function
there is a more intimate relation between them: J desire (or will) to do
something because I can will it for others too. Only if I will it for my-
self because I can will it for others, does the supreme moral maxim free
21 This error is attributed to Kaat by, among others, C. D. Broad, Five Types of
Ethical Theory (London, 1930), p. uS. For an analysis of the logic of this con-
troversy see Austin Duncan-Jones, "Kant on Universalization," Analysis, XVI
(1955),11-1+ It will be discussed morefully in chap. ix.
111 The "Metaphysical Deduction" of Moral LIl'W
itself from antecedent material maxims. This is, I think, intended in
Kant's statement of the imperative, in the words italicized: "So act that
the maxim of your will could always hold at the slD1Ie time as the
principle of a universallegislation."22
This is stated by Kant to be the sought-for moral law, or supreme
principle of pure practical reason. In fact, however, it is not a law, but
an imperative; and though he has distinguished between law and im-
perative, he nowhere in the Critique states the law in proper form. He
does so only in the Foundations, and there incidentally: "An abso-
lutely good will is one whose maxim can always include itself when re-
garded as a universal law," which is an a priori synthetic proposition.23
§ 6. AUTONOMY
A law is a principle that is universally valid for all rational beings. As
a principle, it must contain a general determination of the will. We
have been attempting to understand what this principle is, considered
as a practical proposition.
In the section immediately following upon his formulation of the
principle, however, Kant begins to use the word "principle" in another
sense. "Principle" has hitherto meant a basic, fundamental synthetic
a priori judgment; now it means the actual prius or condition that is
formulated by this judgment. The principle he now speaks of is not a
judgment but a condition of will or the "general determination" itself.
And he says: "The autonomy of the will is the sole principle of all
moral laws and of the duties conforming to them .... The moral law
expresses nothing else than the autonomy of pure practical reason, i.e.,
freedom" (33 [111]).
In §§ 5 and 6 of the Critique (18-29 [II 6- I 7]) Kant shows that a
formal (synthetic) maxim is sufficient to determine the will only if the
will is not wholly detennined by the material of its maxims. If a ra-
tional being follows a formal (synthetic) principle, he does so because
his will is free from empirical detennination. This is freedom in the
negative sense. And since a will requires some law for its determination,
such a being must follow a law not given to it by nature; this can only
be a law given completely by reason itself and not by reason working
22 Critique of Practical Reason, S 7 (30 [119]). The notion that there are two
volitions-one for my own action and another for the universalization of the muim
of the fanner volition-has not even any apparent verbal justification in later
fonnulas of the categorical imperative in the Foundations. Cf. especially the state-
ment of the "principle of autonomy"; "Never choose except in such a way that the
maxims of the choice are comprehended in the same volition as a universal law"
(Foundations of the Metaphysics of Morals, 440 [59]; italics added).
The "Metaphysical Deduction" of Moral Ltl'W 123
on the data of experience. Hence the will must be free also in a positive
sense, i.e., as self-determining.
Thus Kant concludes that a free will, a moral will, and a pure prac-
tical reason are identical. If reason were not capable of prescribing the
law to itself, it would have to borrow the law from nature, and such a
law can support only a heteronomous imperative. Only a law given by
myself as a member of the intelligible world can interest me directly as
a member of the empirical world; all other laws and actions under them
can interest me only indirectly.
Only a self-given law, therefore, can support the phenomena of
morality described above in § 3; and a self-given law of a rational being
must have the formula stated by Kant. With it, the metaphysical de-
duction of the principle is complete, though we shall continue, in the
next chapter, to follow Kant in matters corollary to it.
Before turning to these parts of the exposition of the concept, how-
ever, it is well to consider Kant's conclusions reached thus far, espe-
cially in the light of his philosophy of theoretical law. Some writers
have remarked on what they call the "paradox of Kantian ethics," the
paradox of his having begun with the most complete subjection of self
to law in the history of modern ethics and having ended with the com-
plete subordination of the law to the self. His central moral phenome-
non is the restricting of the self by law; its explanation is the lawgiving
of the self, or autonomy.
But there is no paradox here, and Kant's thought is a development of
an older concept in the light of his previous theoretical development of
the concept of the law in general. Wolff and other rationalists had
made the law of nature sovereign over man and had regarded reason as
man's moral mentor, since only reason could discover nature's law. 24
That obedience to the law of nature would bring men to perfer:tion and
happiness was obvious to them; but happiness was not, they thought,
the reason why men should obey the law of nature. Obligation fol-
lowed simply from an analysis of the concept of perfection. 25
Kant, however, was not, during the critical period, any more taken in
by this concept of natural obligation than by that of natural purposive-
24 Cf. Wolff, Verniinftige Gedancken von der Menschen Thun und Lassen, § 137:
The laws of nature require that we do that which contributes to Our perfection.
Our conscience requires that we act according to the laws of nature. The laws of
nature are the laws of morality and of reason, which discovers them. Similarly,
Crusius, Entwurf der nothwendigen Vernunftwahrheiten, H 133. 137: There is an
inborn idea of natural law, placed in us by God and recognized by conscience,
which is not a theoretical cognition but a Grundtrieb or has as its basis a Trieb of
the will.
25 In this they were followed by Kant in the Prize Essay, II, 199 (Beck, 183).
ness as a ground for morals. He saw that laws of nature are without
moral import and that they can be guides to action only if some pur-
pose is presupposed. The only purpose he could find which could be
served by our knowledge of natural law was happiness and all its pleas-
urable components. Hence he concluded that Wolff's practical philos-
ophy, though ostensibly an ethics of law, was really an ethics of desire,
unable to ground the authority of the law it celebrated.
Kant was not willing to surrender one jot of the authority of moral
law. His theoretical philosophy provided the clue to his concept of the
source of law. In his theoretical philosophy, reason is the source of the
law of nature; the laws of nature are not passively recorded but are
conditions we place upon experience as criteria of its objective signif-
icance. Only because of their origin can they have the necessity denied
by Hume or explained on the assumption that they are intrinsically and
self-evidently rational or revealed as the laws of God-explanations he
contemptuously rejected.
But if the source of the objective necessity of the theoretical laws of
nature lies in us as rational beings, it is easy to see how Kant could also
locate the source of moral law in autonomous reason and guarantee its
objective necessity at the same time. The only significant difference be-
tween them is that the former laws have a sensuous condition and are
therefore restricted to what is, where "is" means "is for us men, under
the sensuous condition of our experience and in accordance with the
second Postulate of Empirical Thought." The latter, on the other hand,
are not restricted to this sensuous condition and are regarded as laws of
a world we do not know actually to exist, but one which we know
ought, by our actions, to be brought into existence. Reason does not
find its demands for rationality fully satisfied in what is presented as
actual; its demands for complete rationality can be met only in a world
not actually given to us. The discrepancy between what reason de-
mands and what experience presents-this unsatisfied demand of theo-
retical reason-appears as the practical interest in the reign of a rational
law that does not actually reign in nature.
To a being like man, who legitimately uses existence only as a cate-
gory of what is presentable to the senses, such a law prescribes what
ought to be, not what is. Its claim on him lies in the fact that it is not
only a law for him but a law by him. If the law sprang from another
source, be it God himself, its claim on man would be only conditional.
It could be effective only through a system of reward and punishment
working upon his desires. The purity of morals would be destroyed.
Only autonomy can unite the two conceptions which are inherent in
Kant's ethics: that man is independent of everything outside him in the
The "Metaphysical Deduction" of Moral LIlW 125
determination of duty and yet that in the moral struggle we are com-
manded, and we are not "volunteers" (82, 85 [175, 178]).
Kant's ethics, historically, represents a transition between two great
conceptions of the relation of man to the world. Against the eight-
eenth-century position that man is a part of nature and ought to be
subservient to her laws, Kant reacted by inverting the order and mak-
ing nature what she is because of how she appears to us. Then he tran-
scended even this Copernican venture by daring to weigh nature in the
scales of reason and to declare that she is wanting and does not contain
the destiny of man. The practical-what man ought to be and how he
ought to transform his existence-in this conception takes precedence
over what nature is and what she demands of man as part of her order.
Nature produced man but brought him to the stage where he can final-
ly assert his independence of her.26
Continuance of this development of the theory of the creative self is
found in romanticism. For the romanticists, man stands above nature as
her author and judge; but the judge has lost his law. The universality
of law, which is Kant's heritage from the rationalists and naturalists of
the seventeenth and eighteenth centuries, was maintained and even mag-
nified by him, though he anchored it in the abstractly personal, in the
res cogitans. But with Herder and others, this was only a halfway point,
an unstable mixture of rationalism and subjectivism. The development
was to be completed, they thought, only by seeing the concretely per-
sonal, historical man as the source of law.
But, of course, under this dispensation law is no longer rational and
universal, and its necessity is only historical. It is only an expression of
feeling and history, of Zeitgeist, Volksgeist, and Sch'Wiirmerei. To
those who wished to take this last step, Kant addressed his essay on
"Orientation in Thinking": "Men of intellectual power and broad
minds! I honor your talents and love your feeling for humanity. But
have you considered what you do, and where you will end with your
attacks on reason? "27
With acute foresight, Kant told them what would be the end of those
attacks. But the attacks have continued, and we have seen their conse-
quences, and they are as Kant foretold.
26 Cf./dea for a Universal History, Thesis III (VIII, 19).
27 What Is Orientation in Thinking? VIII, 144 (Beck, 3°3).
IX
Practical Concepts and Judgment; Commentary

on Analytic, Chapter II
§ I. INTRODUCTION
Chapter ii of the Analytic, "The Concept of an Object of Pure Prac-

tical Reason," follows the Deduction. Logically, it should have pre-
ceded it and come immediately after the establishment of the formula
of the supreme principle in § 7. Because of the difficulty in some of irs
parts, however, one can well imagine that Kant wished the reader to
become thoroughly familiar with the basic principle and its ramifica-
tions before undertaking the work on concepts and moral judgment.
While its position after the Deduction permits Kant to speak in it
sometimes in the context of the law as having been deduced (estab-
lished as valid), and not merely as formulated, at other places he
recognizes that this chapter is a part of the Exposition, or what I have
called the "Metaphysical Deduction." In one place, indeed, he writes
explicitly as if the (transcendental) Deduction were yet to come. 1
While a few passages can best be interpreted in the light of this De-
duction as having been accomplished, as a whole this chapter seems
best understood as an elaboration and application of the formula, quite
independently of the warrant that the formula gains from the (tran-
scendental) Deduction.
Stylistically, the chapter is very uneven. In its early parts it is one of
the clearest and most straightforward sections of the Critique. Irs
middle part, on the other hand, is the most difficult and obscure in the
entire book. It is often regarded as merely a specimen of Kant's "vicious
1 Critique of Practical Reason, 67 (J 58): The sentence beginning "Finally. the
categories of modality •.." even suggests that at one time, in the course of writing,
the order of the parts may have been reversed. There are internal reasons which re-
quire that the Analytic of Principles should follow the Transcendental Deduction of
the Concepts in the first Critique. But there is no good reason why the study of con-
cepts and judgments should follow the Deduction in the second Critique, even
though it is necessary that the formulation of the principle should precede that of
the concepts.
116
Practical Concepts and Judgment 127
architectonic" and as having no integral place in the argument of the

book as a whole. Even the most charitable interpretation of it must
admit that it is cryptic and poorly constructed and, unlike the cor-
responding part of the first Critique, left largely unexploited in the
further development of Kant's doctrine. The very obscurity and inept-
ness of the construction, however, speak against this section as having
been an expression of a merely crotchety interest in architectonic;
for when Kant did follow this penchant, his tables had an elegance
and polish sorely missed here. It is at least arguable that he was here
exploring entirely new ground, ground which he did not succeed in
neatly cultivating. The last part of the chapter is again clear; though
it deals with a difficult and subtle topic, anticipations of it in the first
Critique and in the Foundations adequately prepare the reader for its
difficulties.
§ 2. PRINCIPLE, CONCEPT, AND JUDGMENT
Kant speaks twice (16, 90 [102, 184]) of the fact that the Analytic
of this Critique has the same parts as that of the first but in reverse
order. The Critique of Pure Reason first formulates and justifies certain
pure concepts (categories) that cannot have an empirical origin but
must apply to empirical objects if we are to have any knowledge of
them. Second, it shows how these concepts are to be applied in experi-
ence, i.e., how occasions for the application of the categories are found
and distinguished within experience. This is performed in the chapter
on "Schematism," which is concerned with the faculty of transcenden-
tal judgment, i.e., the faculty of subsuming intuitions under concepts
which are independent of intuition. Finally, it is shown how these
concepts can be synthesized a priori into principles, which are the
laws given to nature by the understanding. This is the subject matter
of the Analytic of Principles. If the concepts are not restricted to the
conditions of experience, they are called "Ideas of Reason,"2 and the
Transcendental Dialectic shows that no synthetic a priori judgments
or principles can be justified when the categories are treated as Ideas of
non-sensuous objects.
In the Critique of Practical Reason, however, the start is made from
principles and not from concepts. The reason for this is clearly stated
in chapter ii, 3 but it is implicit in the whole spirit of the Kantian ethics.
The moral fact-the phenomenon to be explained and rendered in-
2 Another, more positive, conception of the Ideas also appears in the Critique of
Pure Reason and passes over into the Critique of Practical Reason; it will be fully
discussed in chap. xiii.
a In answer to Pistorius, Critique of Practical Reason7 61-63 (154).
uS Practical Concepts tmd Judg;ment
telligible-is the consciousness of obligation to carry out the terms of
the moral law. \Ve have no independent faculty of intuition or moral
sense to give us the concept of the good as something to be achieved;
the mode in which moral concern first arises is obligation, expressed
by the law and its imperative, and not intuition or even a judgment
that something or other to be achieved by or realized in action is good. 4
Moral concepts-indeed, all practical concepts-are to be derived from
principles, and not conversely. If the reverse order is attempted, that
which is called "good" would have to be discerned by a receptivity
(feeling), and thus heteronomous principles would be derived. Ac-
cordingly, the original moral phenomenon of absolute constraint, which
requires an autonomous principle, would be left unexplained. Hence
Kant says that we should not beg the most important question in
moral philosophy by choosing a procedure that dictates its answer, as
beginning with concepts would entail heteronomy of principle. We
should at least leave open the possibility (and explore it) that principle
determines concept.
In his discussion of the difference between the order of topics, how-
ever, Kant omits discussion of the mediating role of the faculty of
judgment. Yet it has this mediating role, whether the principles and
concepts in question be those of theoretical or of practical reason. This
is easily seen jn the analysis of syllogisms that Kant gives in each Cri-
tique.r, In each, the subsumption of facts or acts in the world under the
principle or major premises is a task ascribed to judgment (Urteils-
kraft). Transcendental philosophy, in its doctrine of judgment, speci-
fies a priori the instances to which the principle or concept is to be
applied. "It must formulate by means of universal but sufficient marks
4 It is correct to insist with Paton (The Categorical Imperative, p. 45; cf. In De-
fence of Reason, pp. 1 57~77) that the concept of the good is basic to that of duty
in Kant's ethics. We have no obligation to do anything except under the idea that it
is good to do it. Duty is only the form in which the good appears to beings like man
who do not necessarily desire the good and must be constrained to seek it. But this
does not imply that the concept of good is basic to the moral principle. The con-
cept of duty can be derived from either, but the principle is basic to the definition
of the concept.
I) Critique of Pure Reason, A 304 == B 300-61; Critique of Practical Reason, C)O
(184). Elsewhere (Metaphysik der Sitten, VI,438) Kant ascribes to reason the re-
sponsibility for drawing the conclusion of the practical syllogism, but in this passage
he is obviously referring to the logical, rather than the real, Use of reason. In this
passage he calls practical reason in its real use (in the establishment of the principle
or major premise) the "practical understanding." In Foundations, 412 (29), he says
that, since reason is required for the derivation of actions from laws, will is nothing
else than practical reason. Here also we have to do with reason in its logical use,
and this follows Wolff's tenninology in Psychologia rationalis, §§ 494,528.
Practical Concepts tmd Judg;ment 129
the conditions under which objects can be given in harmony with these
concepts."6 Similarly, in the practic.al syllogism it must decide what
object or act falls under the principle which defines the good and state
the ground of this decision in the form of a rule of judgment, so that
moral judgment and possible facts and acts of experience can be
brought together in such a manner that the latter can be estimated.
But, though formally similar, the two syllogisms are quite different
in their epistemic or transcendental function. The concepts of pure
understanding or categories are applied to intuitions a priori, and the
intuitions are the "third thing" that mediates between concepts so that
an a priori synthesis of them is possible. But they do not apply directly
to intuitions; they apply only to certain formal patterns or structures
of them, which are isomorphous with the intellectual synthesis whose
rule is given in the corresponding category. These formal structures
are the schemata of the concepts. For instance, it is only because
"cause" and "effect" are both referred to intuitions in a specific tem-
poral relationship that we are able to make synthetic a priori judgments
of causality that apply to experiences. But the concept of the moral
good is an Idea that cannot be schematized; we cannot find a structure
of intuition that corresponds to it. The concept of lawgiving, which
is that of freedom in one of its senses, is the concept of a cause that
does not exist in time, and therefore the schema of causation does not
apply to moral decision and its expression in action with the same
epistemic consequences that it has in theoretical knowledge. Still, the
concept of the good, though defined by the principle, is to be applied
to particular acts or objects that can be created by our free choice in
the world of experience. The principle itself must be, as it were, sche-
matized (68 [160]), so that it does not remain empty and inapplicable
in the practical experience of the guidance of choice.
The chapter, accordingly, falls into three clearly distinguishable
parts. The first thirteen paragraphs define the concepts of good and
evil as the concepts of objects of practical reason. The next four para-
graphs develop the categories of practical reason in the light of the
principles. The last section prescribes the condition for the application
of these concepts and principles.
§ 3. THE CONCEPTS OF GOOD AND EVIL
An object of practical reason is an effect possible through freedom

(57 [148]). In this definition, two words require preliminary com-
ment. "Object" must be taken in a sufficiently broad sense to cover
two things: states of affairs produced by action and action itself. It
e Critique of Pure Reason, A 136 = B 175.
J 30 Practical Concepts tmd Judgment
must not be thought that "object" means only a thing in the world
created by action. "Freedom" in this definition is freedom of choice
(Willkur) and not necessarily transcendental freedom, since the object
need not be an object of a pure and autonomous will which is tran-
scendentally free. Since a will is never directly determined by an object
of desire, but only by a concept of an object and a rule which the
acting person takes as ground for a decision pro or con, freedom in
the sense of this sentence is characteristic of voluntary choice, regard-
less of whether we are concerned with the "free will," which is moral,
or only with "free choice," which is characteristic also of intelligent
choice and technique. 7 No question concerning transcendental freedom
need be raised in the interpretation of this chapter, though, because it
follows upon the Deduction, Kant permits himself from time to time
to use this deeper concept of freedom.
The concept of the object of practical reason is the object's repre-
sentation regarded as an efficient cause of the action that is to produce
the object. Object of practical reason and purpose of the will are thus
identica1. 8 Moreover, the material of a practical principle in contrast
to its form is the same as its object (27 [J 14]). It is essential, however,
to distinguish between the material of a principle and a material prin-
ciple. All principles have a material because all volition has an object
(34 [J2 3]); but a principle is called a "material principle" only if its
material is the condition of the action directed or elicited by the
principle. If there are laws of practical reason, Kant has shown that
there must be principles which are not material principles; but this
part of the chapter is devoted to finding out what the material of such
a non-material principle can be.
To say of an object that it is an object of practical reason is to
envision it in the causal nexus of free action, which, under a law or
rule, produces it. This is analogous to the condition under which we
say of any conception that it refers to an object. For our theoretical
understanding, to refer a datum to an object is to hold that there is
a rule according to which this datum is synthesized with others in one
spatiotemporal system and thus given the status of evidence of an
object as a "permanent possibility" of other data; a "wild datum" is
one not synthesizable with others under a rule and hence is not evi-
dence of an object. Object of experience is to theoretical understanding
what object of practical reason is to the faculty of rational choice;
7 Critique of Practical Reason, 60 (151); d. Critique of Pure Reason, A 801 =
=
B 830, A 548 B 576; Uber den Gemeinsprucb ... , VIII, 181; Metapbysik der
Sitten, VI, 381.
8 Critique of Practical Reason, 57 (148); Metapbysik der Sitten, VI, 381, 384~5.
Practical Concepts and Judg;ment 131
mere sensation is to theoretical understanding what mere object of

desire is to rational choice. Object of experience or object of practical
reason is that to which our conceptions apply when they stand in
necessary synthetic relationships to each other and not merely in casual
relationships of associations of ideas.
A rational being acts so as to produce an object only under the
thought that the object is or would be, in some sense, good; and it acts
to avoid a state of affairs only if it thinks that it is in some sense bad
or evi1. 9 Hence, Kant says, the only concepts of objects of practical
reason as such are the concepts of good and evil,1° There are, however,
three senses of good and evil. Something may be good for some arbi-
trary purpose (e.g., cyanide is good for suicide), for some actual pur-
pose (e.g., health is good for happiness), or for some obligatory and
necessary purpose to be held by a rational being as such. Kant ac-
cordingly distinguishes between bonitas problematica, bonitas pragma-
tica, and bonitas mOTl1lis, corresponding to the three kinds of impera-
tive,u In each of these kinds of good, and not merely in the third,
there is a necessary relation of the object to the will, necessary as
determined by some rule of reason.
Granted that suicide is the end proposed, then every man who is
rational and not merely he who is bent npon this act must necessarily
recognize the goodness of cyanide for this purpose. The necessary
relation of an object to a will which judges it to be good in this way
is only a hypothetical necessity, but, even so, it makes a necessary
claim on the assent of a rational being. However irrational it may be
9 I am not able to identify the "old formula of the schools" (Critique of Practical
Reason, 59 [Iso]) as Kant states it, but the sense (and some of the words) are in
Wol1£ (Psychologia rational;s, §§ 880, 881, 89Z) and in Baumganen (Metaphysica,
H 661,665). Baumgarten makes the mistake that Kant points out in the text, putting
the anticipation of pleasure in the object in the determining position. He writes:
"If I foresee something as pleasing, which is possible by my e1£on, I strive for it .•..
Hence I am able' to strive after many goods, and tum from many evils, sub ratione
{bon; et] mali" (ibid., § 665). Kant, of course, thinks Wolff is guilty of this error,
too, since perfection is a material principle to the extent that it has any moral im-
pon at all, and all material principles are ultimately hedonistic. But W 011£ himself
denies this, holding that perfection is the object of striving and that pleasure is only
an accompaniment of its attainment.
10 Critique of Practical Reason, 57 (148). Kant wrote: "By a concept of practical
reason I understand the idea [Vorstellung] of an object as an effect possible through
freedom." Vorlander inserted the words "of an object" after "concept," and this
emendation was adopted in my translations.
11 Lectures on Ethics, p. 15. They correspond to the three kinds of necessity in the
Yorlesungen fiber Metaphysik (Politz ed.), p. 186.
132 Practical Concepts tmd /udg;ment
to say "Suicide is good," "Cyanide is good for suicide" expresses an
objective decision of reason.
This is easy enough to see when we are concerned only with means
to ends. But Kant argues at length that even such a judgment as "Sui-
cide is good" must likewise present a claim to reason. If it were good,
it would not be because a despairing man desires it but because it would
have some necessary claim to the assent of reasonable beings. Otherwise
it would suffice merely to say, "I desire to commit suicide" or perhaps
"I desire to commit suicide; would that you did too." It is this neces-
sary relation to rational assent that Kant investigates here.
There are two possible relations between the principle or maxim of a
will and its object. The object can determine the concept through the
principle, or the principle can determine the concept of the object. The
question is Which is the prior notion, the good or the principle?
With respect to bonitas problematica sive pragmatica, the concept of
what is good is determined by the principle that we have called the
"formal principle of a hypothetical imperative." But what is good-the
denotatum of the concept-is not determined by this principle. That
which makes a good thing good in either of these senses is the pleasure
we have or expect to have in its existence or the causal relation between
something we do and the existence of this pleasurable state. Neither of
these is decided by principle or definition. Yet so intimate is the con-
nection in the bonitas prar;matictl between what is good according to
principle and the pleasure that the hedonist identifies the two concepts.
Kant challenged this identification; even common usage, he says (like a
twentieth-century philosopher) opposes the identification of the good
with the pleasant. 12 Good is a communicable, rational concept, and
pleasure is merely a private affection of the inner sense.t 3 The good is
judged by reason, not by private feeling, even though the judgment
may have a feeling as its premise. Reason judges what is desirable; feel-
ing decides what is desired. But the good is a normative concept, cor-
12 Critique of Practical Reason, 58 (149); Critique of Judgment, § 4 (V, 108 [41]).

The Stoic at whom Kant says we may laugh (Critique of Practical Reason, 60 [151])
for (tacitly) drawing the distinction between the two meanings of "evil" during the
sharpest paroxysm of the gout was Poseidonius (Cicero Tusculan Disputations ii.
1S)·
13 The merely pleasant represents the object simply in its relation to inner sense
or feeling; it is therefore contingent. The pleasant can be brought under principles
of reason by having applied to it the concept of purpose; then it is the good in one
of the senses of this term, and the good necessarily satisfies by vinue of the universal
concept involved in the effectuation of a purpose (d. Critique of /udlr"lent, § 4). Ii
something is judged to please without a concept, yet necessarily, it is not good
but beautiful.
Practical Concepts and Judg;ment 133
responding to the desirable and not to the merely desired. So the he-
donist, who avoids the erroneous identification of "good" with "pleas-
ure," yet who wishes to maintain a necessary relation between them,
comes to mean by "good" that which, according to the judgment of
reason, leads to pleasure. In this way, "good" denotes an object that
stands in a necessary relation to choice as judged by reason and is iden-
tical in meaning with "the usefuL" The sense in which pleasure is good
is then quite different from that in which the means to pleasure are
good.
"Good," then, has two quite distinct meanings, even if one denies the
main thesis of the Kantian ethics that there is a bonitas moralis wholly
distinct from bonitas pragmatica. It may mean das Gute as a character-
istic of actions, maxims, and character as judged by reason (even if
judged by reason to be good only because it leads to a desired end). Or
it may mean das Wohl ("well-being") which is decided by each per-
son according to his own feeling and without a necessary relation to
reason. To decide whether something is das W ohl or a part of it re-
quires an answer to the question Is the end to be achieved actually
pleasant in such a way that it is an ingredient in happiness? To decide
whether an action is good in the sense of bonitas pragmatica (das Gute
in one of its three subsenses) requires reason to give an answer to the
question Will this act lead to the achievement of the end, which is hap-
piness? The answer to each of these questions, though it can be given
only by reason (in a broad sense) working on the data of experience, is
empirical and only probable. Hence no principle derived from con-
cepts of the good as an object (das Wohl) can be a law, and no impera-
tive to seek a previously and independently defined good can be cate-
gorical.
On the other hand, it is possible that the principle might determine
what objects are good. This must be the actual case, in fact, if there is
any reason to choose a principle other than by reference to the empiri-
cal object of the maxim or rules subsumed under it. If there is a law
which expresses an absolute obligation, it must do so by virtue of its
form and not its object. Since every principle of volition, however, has
a material or an object, this is to say that the form of the principle,
which is one of categorical obligation, must determine the concept of
the good.
In making this determination, we do not have to inquire whether the
object, i.e., what is desired, be it an action of a certain kind or its con-
sequent end-state, will be pleasant, or whether it is physically possible
for us to achieve it in fact. The "moral possibility" of willing it takes
precedence over the psychological possibility of having a direct in-
134 Practical Concepts and Judgment
clination to it and the practical feasibility of reaching it. When decid-
ing on an object of pure practical reason, we consider only the possi-
bility of rationally willing it-whether, that is, if reason completely
determined our choice, we would choose it. "Good" in this sense is de-
fined as that which it is necessary to will by a law of reason whose
applicability is independent of actual physical capacity of achievement.
Hence "good," in the only sense in which anything can be absolutely
good--good independent of the state of my private desires and the ways
of the world-refers only to actions, the maxims that lead to them, the
will that produces them, and the character that supports them as their
subject. This is the only good that can be commanded universally; the
others can be recommended only "to whom it may concern." The nec-
essary relation of this willing to the good is categorical, not hypotheti-
cal; it is apodictic, not problematical.
Thus the connection between action and its object is much more in-
timate for pure practical reason than for practical reason in general.
One might say, though Kant does not say it here,14 that virtue is its
own reward (object). What he does say is that the good characterizes
actions directly and not derivatively (though it must be remembered
that action means not merely external behavior but the inner setting of
the will to behave in a certain way). The object that is absolutely good
can be obtained only in and by acting in a certain way-not as if it were
an end to be achieved by the use of certain means, the circumstances
being favorable. It can be obtained only in and by acting in a certain
way because acting in a certain way is the first thing-in fact, the only
thing-commanded by an imperative based on a law. The form and the
object, so far as the object is the moral good, of the maxim coincide.
The object of pure practical reason is not an effect of action but the
action itself; the good will has itself as object. HI The morally good ob-
ject is not a thing in space and time, but its concept is a supersensuous
Idea of a supersensuous nature. It is not a nature to be produced as an
effect by a certain kind of action but a nature that is at least partially
realized in the mode of action itself (43 [13 2]). The good is not tran-
scendent to moral action but is immanent within it.
Nonetheless, the good will is not by itself sufficient to determine a
specific action. Every volition has an object, though the object need
not itself be an unqualified good, something absolutely and in every re-
spect good. The categorical imperative, which commands the pursuit
H Though he does say it in Metaphysik der Sitten, VI, 377.
15 Cf. Foundations of the MetaphYSics of Morals, 437 (56): "Act according to
maxims which can at the same time have themselves as universal laws of nature as
their object" (italics added).
of the immanent moral good, always presupposes maxims which have a
material and which it controls, as we have seen.I6 Hence devotion to
the moral good does not require renunciation of other goods; it re-
quires only that, in determining what our duty is, we take no account
of our desires, which determine directly or indirectly the other kinds
of goods. I7 Some desires are compatible with devotion to the good or
can be made compatible with it. They should be cultivated and need
not be thwarted or eradicated by a single-minded devotion to moral
goOd. I8 The desire for our own happiness may be of this kind, and
"reason has an inescapable responsibility to attend to [this] interest and
to form maxims with a view to the happiness of this, and where possi-
ble, of '3. future life. HI9 To secure one's own happiness is at least indi-
rectly a duty (93 [186]). But my own happiness can be rationally
willed only under a law which likewise holds the happiness of others
as an object before me (34, 85 [ 12 3, 178]).
To be sure, a state of universal happiness might conceivably have
been produced by natural mechanisms not guided by reason or by the
grace of God working in wondrous ways. But something of inestimable
value would be lost in such an arrangement. The immanent object of
pure reason is something that only it could realize, at the cost of often
failing to realize it. In an Aristotelian vein, Kant here repeats an argu-
ment of the Foundations 20 that each of our faculties has its own proper
function and that of pure practical reason is "to consider . . . what is
good or evil in itself, which pure and sensuously disinterested reason
alone can judge, and furthermore to distinguish this estimate from a
sensuous estimation and to make the former the supreme condition of
the latter" (62 [153])' As the perfection of rational nature, this also is
an end or object completely determined by pure practical reason and is
an object which it is our unrestricted duty to pursue. 21
Object, in this sense, is not a terminus of action, not something to be
gained only through the use of means guided by a knowledge of causal
18 Cf. above, chap. viii, § 4.
17 Uber den Gememspruch .••• VIII, 178-79; Critique of Practical Reason, 93
(186) •
18 In agreement with Rousseau (cf. Mutmasslicher Anfang der Menschenge-
Ichichte, VIII, 117).
19 Critique of Practical Reason, 6[ ([51). Kant has been interpreted as denying
this, or at least of having no right to assert it. and is accused therefore of being the
proponent of "sour duty" (cf. Metaphysik der Sitten, VI, Z 14, and above, p. [20, n.
10).
20 Foundations of the Metaphysics of Morals, 395 (10-[ I).
21 Metaphysik der Sitten, VI, 385-87.
136 Practical Concepts tmd Judgment
law, not something to be hoped for only in the indefinite future when
just men shall have been made perfect, the crooked straight, and the
rough places smooth. No; the object is a will of a certain disposition, a
disposition to act in accordance with and out of respect for law. The
only purpose of moral action as such is to secure the reign of law, and
every moral action in part accomplishes this aim. It is in this sense that
Kant says that what the moral law requires of us can always be accom-
plished; its object is achievable, however unsatisfactory the state of the
world may be, because the object it holds before us and the action that
is to achieve it are at least partially identical.
§ 4. THE CATEGORIES OF PRACTICAL REASON
If the reader of the Critique of Practical Reason has been lulled into
a feeling of easy familiarity by the no doubt interesting, but certainly
not surprising, development of the relation between principle and con-
cept, between law and the good, in the beginning of chapter ii, he will
be suddenly shocked at the marked change in style and direction of the
argument at the beginning of the fourteenth paragraph. A fog seems
suddenly to settle over a road along which one expected to make
good time. This paragraph and the three succeeding ones are strikingly
unexpected, original, and obscure. They require, because of their dif-
ficulty, exegesis of each sentence. Without further ado, let us move
forward, slowly and cautiously, into the darkness:
Since the concepts of good and evil, as consequences of the a priori deter-
mination of the will, presuppose a pure practical principle and thus a causality
of pure reason, they do not (as determinations of the synchecic unity of the
manifold of given intuitions in one consciousness) refer originally to objects,
as do the concepts of the understanding or categories of the theoretically
employed reason [65 (156)].
Here Kant is speaking explicitly of the moral good and evil, not of
good and evil in the broad sense defined at the beginning of this chap-
ter of the Critique. Yet, as an introduction to the further development,
it is essential that "good and evil" be taken in the more general sense.
In the limited (moral) sense, they do not refer to perceptual objects at
all. In the broader sense, they do not refer to objects simpliciter but
only to objects in the context of the causality of their production by an
act of will. "Good and evil," therefore, do not even seem, like the the-
oretical categories, to apply to objects as they are given in their own
right, so that knowledge of them can be connected in one conscious act
of theoretical judgment. Nor do they, as the theoretical categories do,
actually make these objects possible by being rules for the synthesis of
the representations in one consciousness and one spatiotemporaI system;
Rather, they presuppose these objects as given, and they are without excep-
tion modes of a single category, that of causality, so far as their determining
ground consists of reason's idea of a law of causality which, as a law of free-
dom reason gives itself, showing itself to be a priori practical [65 (156)].
Again, this is explicitly restricted to moral good. But, in the light of

what follows later and especially in the light of the first clause of the
sentence, we must interpret Kant as still concerned with the more gen-
eral usage of the concepts. The concepts of good and evil, he says, are
not constitutive of objects of experience, as the theoretical concepts
are. Rather, they presuppose that the objects of experience to which
they are to be applied are already given under the theoretical catego-
ries. When we say of an object that it is good, in addition to all the
other categories that make it an object uberhaupt) we use the category
of causation to relate the object to its cause which lies in an act of the
will. When we say of a given object that it is good, we call attention
to the fact that, as good, it is given under the mode of causality by vir-
tue of which it is "an effect possible through freedom" (57 [148]),
i.e., that it stands in a causal relation to an act of choice. As an existing
or possible thing in nature, it is given by virtue of the synthesis of em-
pirical inruitions under the category of natural causation; but, as some-
thing good, it has a necessary relation to a specific cause, an act of will.
On the one side, the actions are under a law of freedom instead of a natural
law and thus belong to the conduct of intelligible beings, and on the other
side as events in the world of sense they belong to [the world of] appearances;
so that the rules of a practical reason are possible only with respect to the
latter and consequently in accordance with the categories of the understand-
ing [65 (156-57)].
Here it is clear that he is speaking of practical reason in general and

not just of pure practical reason. Kant tells us that any object is called
"good" only if it is made an object of practical reason by an act of free
choice. Yet free choice is possible only under the condition that the
object chosen is possible according to the rules (categories) of sensuous
experience. (Thus it is not open to me to make a square circle the ob-
ject of my will, however much I might desire to accomplish the feat of
constructing one.) But a pure practical reason can decide on its object
independently of whether I possess the capacity actually to produce
the possible object or not. The practical possibility presupposes the
theoretical possibility. No principle of practical reason, therefore, can
conflict with the conditions of the natural possibility of things, for the
objects of a practical reason in general must be possible objects of nat-
ural experience.
These rules, however, contribute nothing to the theoretical use of the un-
derstanding in bringing the manifold of the (sensuous) intuitions under one
consciousness a priori, but only to the a priori subjection of the manifold
of desires to the unity of consciousness of a practical reason commanding in
the moral law, i.e., [to the unity of] a pure will [65 (157)1.
"The rules of practical reason" must be taken very broadly to include

principles (maxims as well as laws). Now these "rules" determine what
objects in the world of sense are good and evil. But when we know that
an object, is good, we have not thereby added one bit to our theoretical
understanding of it. The object as a cognized object is presupposed in
the judgment that it is good, but "It is good" is not a theoretical state-
ment that adds another natural predicate to it.
Yet these "rules" have a role analogous to that of the theoretical cate-
gories. 22 The categories (and their schemata) are rules for the synthesis
of intuitions not only into representations of objects but into represen-
tations for one consciousness. The "I think" must be able to accom-
pany all my representations, and the theoretical categories are the
modes in which I can think these representations as related in one con-
sciousness of a single object or realm of objects and events in one space
and time. I do not say of any representation that it is of an object un-
less it stands in a determinate relation to other representations accord-
ing to rules which are the theoretical categories. A single object pre-
supposes a single integral consciousness, and conversely. Similarly, I do
not denominate any object "the good" except insofar as it stands in a
necessary relation to the faculty of desire. I do not identify the pleasant
with the good, because a pleasant thing does not have this necessary
relationship. My desires are manifold and conflicting; the pleasures of
the table conflict with those of sport, those of this life perhaps with the
hope of those of another. But if I can call anything "good," I mean
that it is an object which would necessarily be desired by a rational man
whose reason controlled his desires or at least controlled the choice he
makes among his desiderata. Hence "I desire, rationally" is like the "I
think": it must be able to accompany all my representations of an ob-
ject as good. Only if "I desire rationally" accompanies all the represen-
tations I have that I desire anything simpliciter, can I be said to have
achieved the integral state of a completely rational personality, to be
possessed of a holy will. But though I can desire many things that are
not good, i.e., are nor desirable but yet are desired, as a rational being
I can restrict my choices to those things which are rationally desired,
and the state of doing this is not holiness of will, which I cannot
achieve, but virtue, or "moral disposition in conflict" (84 [178]).
22 As pointed out in Fragment 6, i.e., Reflexionen 7101 and 7104 (XVIII, 178, 184).
The statement "I desire X" is a statement of fact, and at any moment
I may actually desire X and things I know to be incompatible with it.
But "I desire X rationally," i.e., "I will X, under a rule of reason that I
regard as valid not only for myself but for others" means "X is good."
This judgment can be a constraint on my wayward desires for those
things not compatible with X. The judgment "X is good" is a rational
estimate, believed to be objectively valid, and is distinguished from the
sensuous estimate of X, viz., "X would be pleasant" or "X would be
fun." Only a rational being can judge "X is good"-whether, in fact, X
does meet the requirements for being good or not-but an animal can
desire X. The manifold of desires is synthesized by reason according to
practical principles or rules, and their common focus is the object called
"good":
These categories of freedom-for we wish to call them this in contrast to
the theoretical concepts which are categories of nature-have a manifest ad-
vantage over the latter. The latter categories are only forms of thought.
which through universal concepts designate, in an indefinite manner. objects
in general for every intuition possible to us. The categories of freedom, on
the other hand, are elementary practical concepts which determine the free
faculty of choice [einer freien Willkiir]. Though no intuition exactly COrre-
sponding to this determination can be given to us, the free faculty of choice
has as its foundation [or: the determination of the free faculty of choice has
as its foundation] a pure practical law a priori, and this cannot be said for any
of the concepts of the theoretical use of our cognitive faculty [65 (I57)].
Before exploring and identifying this "manifest advantage," we must
pause a moment to consider what Kant has already said. "These cate-
gories of freedom"-earlier called "categories of practical reason" (I I
n. [96 n.])-what are they? The "these" indicates that they are the
"rules" previously mentioned; yet in the table itself it will appear that
Kant has not yet decided whether "rule" or "concept" is the better
name for them. Still, there is nothing wrong or even surprising in some-
times referring to a category as a rule and sometimes as a concept. 23 A
category, though a concept, contains (like any concept) a rule for the
synthesis of representations. (This is true not only of categories but of
every concept: the concept of "cat," for instance, though a posteriori,
is a rule for the synthesis of our perceptual and imaginative representa-
tions which are necessary for us to say, on the occasion of one or more
of them, "There is a car.") The elementary practical concepts are just
the concepts of good and evil in general; and the rules spell out their
Use in the synthesis of desires, observance of these rules being necessary
for the conversion of desire into rational desire, or will.
23 On the general problem of the relation of category to rule, d. Vaihinger, Com-
'lne1ZtilTzu Kantr Kritik der reinen Vermmft, 1,222.
I40 Practical Concepts tmd ]udg;ment
Next, Kant tells us, the categories of freedom are forms of determi-
nation of choice and therefore of the objects of rational choice, i.e., of
the good as object. They are not forms of objects simpliciter, which
would require that intuitions be synthesized; this would constitute
them objects of knowledge, not of practice. This independence from
intuition is the "manifest advantage" of the practical over the theoreti-
cal categories. Of any category that functions in the determination of
an object, it must be possible to exhibit an intuition or pattern of intui-
tions which corresponds to it; otherwise the concept is empty, and only
analytical judgments can be made upon it. The good and evil, how-
ever, have no corresponding intuitions, and no intuition can be given
corresponding to the determination of choice, while the determination
of an object of knowledge occurs through the schematization of the
concept.
What is actually given to keep the concepts of good and evil from
being. empty is not an intuition, the source of empirical factuality, but
a principle which Kant elsewhere 24 calls the "fact of pure reason." That
is to say, knowledge of the principle itself gives content to the practi-
cal concepts, whereas in the theoretical case the concepts must be es-
tablished with respect to intuition before synthetic principles can be
formulated with them. This constitutes the advantage of the practical
over the theoretical categories. Theoretical concepts can be synthesized
only by grace of intuition, which is wholly different from and inde-
pendent of the thought formulated in the categories. 25 Practical con-
24 Critique of Practical Reason, 31,41 (no, 131); cf. below, chap. x, § 3.
25 Thus practical concepts have a wider significance than theoretical concepts in
their cognitive function. See the remarkable statement that can hardly be under-
stood apart from the present context, though it occurs in the Foundations (411-11
[28]): speculative (theoretical) philosophy makes its principles depend upon "the
particular nature of human reason," where "human reason" refers to our entire
cognitive faculty for which intuition is always sensible, not intellectual. All princi-
ples of practical philosophy, on the other hand, depend solely upon pure reason
alone, in the strict sense of "reason." This, however, hardly seems compatible with
the earlier view in the Critique of Pure Reason that excluded practical philosophy
from transcendental philosophy, in spite of the fact that its principles were pure (d.
above, p. 9, n. 21) • But in the second edition, as we have seen, Kant pointed out that
empirical concepts were needed for defining the concept of duty. \Vhat Kant means
can best be stated as follows: The determination of the moral principle is inde-
pendent of the particular constitution of the human mind and is valid for all rational
beings, whether their reason be discursive, like ours, or intuitive and whether their
will be holy or only virtuous. But the concept of duty presupposes that there are
desires, and hence the concept of duty is not independent of our mental constitu-
tion. Concepts of objects used in our cognition are likewise dependent in their cog-
nitive use upon the fact that they are rules for the synthesis of intuitions which,
for us men, are given independently of the concepts.
cept5, on the other hand, are justified by the fact of pure reason, with-
out needing any appeal to empirical or pure intuitions or feelings of
pleasure. Accordingly (omitting one sentence just paraphrased), Kant
continues:
Since in all the precepts of the pure will it is only a question of the deter-
mination of the will and not of the natural conditions (of practical ability)
for achieving its purpose, it thereby happens that practical concepts a priori
in relation to the supreme principle of freedom thereby become cognitions,
not needing to wait upon intuitions in order to acquire a meaning. This
occurs for the noteworthy reason that they themselves produce the reality
of that to which they refer (the intention of the will)-an achievement which
is in no way the business of theoretical concepts [66 (157)].
These sentences explicitly concern only the pure will and are true only
of it. "Practical concepts a priorin are direct cognitions of what ought
to be, because they are consequences of a primordial fact, the law of
pure practical reason that a rational being as such necessarily wills to
act in such and such ways. Their object is directly given (as what
ought to be, not as what is): the concept itself is a cognition, not need-
ing to await the schematization of the concept or its exhibition or con-
struction in intuition.
Here, then, we have the one instance in the whole of knowledge
where thought can create its object directly, and not by being a cause
of a series of events which may terminate in an object. (This object is
respect, which is a necessary corollary of knOWledge of the law.)26 But
the knowledge thus gained is not knowledge that its object exists; that
would be theoretical knowledge requiring intuition. The knowledge
thus gained is knowledge that something ought to be, and the recogni-
tion of this is a necessary condition of its actual existence as an instance
of the good. Kant would go even further and say that the concept cre-
ates the existence of the object,27 since the object is a certain setting of
the will:
Freedom is regarded as a kind of causality (not subject to empirical grounds
of determination) with reference to actions possible through it, which are
appearances in the world of sense, and ... consequently [freedom] is referred
to the categories of the possibility [of actions] in nature, while, however,
each category is taken so universally that the determining ground of the
causality can be placed beyond the world of sense in freedom as the property
of an intelligible being [67 (158)].
26 Cf. below, chap. xii, §§ 6,8.

27 There are many expressions of this, e.g., Critique of Practical Reason, 66, 92,
153 (157,186,251); Foundations of tbeMetaphysics of Morals, 404 (20).
14Z Practical Concepts and Judg;ment
Freedom is a kind of causality in which the cause of an action is not
another phenomenal event. This kind of causality cannot be understood
by us; we can understand cause-effect relations only when both mem-
bers of the nexus are in one spatiotemporal series. Yet a category is a
pure concept of the understanding. It owes nothing to sensibility, and,
though we can use it to give knowledge only in its connection with
sensibility, we can and must think with it even when we are thinking
of that which can never be given to the senses. Since, however, one
member of the nexus is in nature and thus occurs under the schema-
tized theoretical categories, we can think of freedom of the cause un-
der the categories of the possibility of its effects in nature. We thereby
gain no knowledge of the cause, but we are enabled in this way to
think of freedom in definite and specific ways with respect to events
that we do know in nature. Thus, to take a simple illustration, every
causal change is the change in the accidents of some substance, which is
permanent. In practical action, every act must be thought of as the ex-
pression of a single unchanging subject to which the action is imputed,
and this subject is the person. Hence in thinking of free actions, we
must think them in relation to a person as an agent, though we can have
no theoretical knowledge of this substance.
Thus far, Kant may seem to have been repeating-on a repellently
technical level, of interest only to Kantian adepts-the conclusion of the
more straightforward exposition of the concept of the moral good. But
suddenly he bewilders the reader by then remarking that the categories
to which such remarkable prerogatives have been ascribed are catego-
ries of practical reason in general. (We have, however, taken this into
account in the preceding exegesis.) The categories
proceed in order from those which are as yet morally undetermined and
sensuously conditioned to those which, being sensuously unconditioned, are
determined only by the moral law.... The categories of mOdality initiate
the transition ... from practical principles in general to those of morality.28
The puzzle presented by the last sentence here can be resolved, if at all,
only after we have examined the Table of Categories. Before leaving
this section, however, I must again point out the source of most of its
difficulties, to wit, a confusing shift of attention back and forth from
the categories of pure practical reason to those of empirical practical
reason and practical reason in general. Had the section been written so
28 This citation is a con junction of two sentences, one on p. 66 (157) and one on
p. 67 (158); they are here combined by omitting the parts which are prima facie in-
compatible with each other. In their full fonn and relation to each other, they will
be discussed below, when the examination of the table has been completed and there
can be some expectation of understanding their relationship to each other.
that the categories of practical reason in general had been first elabo-
rated and then those of pure practical reason introduced as a subclass,
the section would have been much more lucid. I shall now attempt such
a rewriting of the section in its most important parts.
Since the concepts of good and evil presuppose a causality exercised
by reason through a principle, they do not refer, as originating and con-
stitutive concepts, to objects which are to be given; this is the task per-
formed by the theoretical categories and their schemata, which make
the objects first possible. Rather, the concepts of good and evil pre-
suppose the objects as given in possible experience. They refer only to
the specific mode of causation by grace of which they are judged to be
good or evil. This is their causal origin in an act of will, which takes
place under the guidance of reason's conception of some law. The con-
cepts of good and evil, therefore, implicitly refer to the cause of the
actions or objects called "good" or "evil," and a principle for decision
on the goodness or evil of an action or a thing is a principle which bears
upon the specific causality of the action or thing.
The actions of the will and their consequences are in nature, and
therefore are under the theoretical categories of their possibility. In or-
der to understand them, as natural events, we do not need the concepts
of good and evil or any principles of practical reason. Theoretical con-
cepts and principles suffice.
The principles of practical reason, however, have an analogy to the-
oretical principles and categories. They bring the manifold of desires to
the unity of practical reason and give rise to acts and decisions of will,
just as the theoretical categories bring the manifold of intuition to the
unity of self-consciousness and give rise to judgments of objects of ex-
perience. The principles of practical reason may therefore appropri-
ately be called" "categories of practical reason," or, as modes of the
causality of practical reason, "categories of freedom."
They have a manifest advantage over the theoretical categories. The
latter have a relation to objects of every intuition possible to us and
must await intuition before they have a definite constitutive use in
knowledge of any object. The former, however, are in themselves cog-
nitions, because they arise from the same reason as that which presents
the fact to be rendered intelligible, viz., the consciousness of the prin-
ciple of practical reason. They thus produce the reality of that to
which they refer, namely, an intention of the will, and do not need to
await intuitions to determine whether they refer to something real or
not. Theoretical categories can never do this in a mind like ours, for
which understanding is discursive instead of intuitive.
The categories of practical reason or of freedom are of two kinds.
144 Practical Concepts find Judgment
Practical reason may have for its material some actual desires which it
synthesizes under a principle. The principle is, of course, a priori, but
the intention of the will based upon it is contingent upon the presence
of the specific and empirically known desires. Such principles and the
corresponding concepts of good and evil may be called principles and
concepts of empirical practical reason. On the other hand, the principle
of the will may be wholly independent of any desire, and such a prin-
ciple will be one of pure practical reason. There should, therefore, be
morally undetermined, but sensuously conditioned, categories and also
sensuously unconditioned, but morally determined, categories.
§ 5. THE TARLE OF THE CATEGORIES OF FREEDOM
After giving his table, Kant says: "I add nothing here to illuminate
the table, for it is sufficiently understandable in itself." No one has ever
granted him this claim. Even Schlitz, one of his most faithful disciples,
felt called upon to suggest to the master a way in which it could be
improved. 29 Since, unlike the table of theoretical categories, almost no
further use was made of it30 in spite of Kant's statement that "such a
division based on principles is very useful in any science, for the sake
of both thoroughness and intelligibility" (67 [158]), we can read back
into it very little that can make it more intelligible. Most commenta-
tors have ignored it; only Mellin has wrestled with its many difficulties,
and, though some of his suggestions are shrewd and valuable, on some
points he was, I think, definitely in error. My own ingenuity is quite
insufficient to interpret all parts of it consistently with the ipsissima
'Verba. Nevertheless, the main features of it are not unintelligible.
Before embarking upon an interpretation of what Kant actually per-
formed here, let us consider what we might legitimately expect him to
do in formulating a table of the categories of freedom.
First, we should expect him to set out in a systematic order the ab-
stract and logically constant features of practical judgments. We should
expect some kind of systematization of various distinctions that he has
29 To Kant, June 23, 1788 (X, HI) ; cf. Ak., XIII, 219.
30 One might have expected the whole Metaphysics of Morals to be organized
with respect to it, as Metaphysical Foundations of tbe Natural Scie71ces 'was
organized according to the table in the first Critique. A few fragments suggest that
the Metaphysics of Morals may have been thought of in this way (d. Vorarhe;ten,
Ak., XXIII, 218, and the [unsent) letter to lung-Stilling written after March I, 1789
[XXIII, 495]). Once in the Foundations (prior to the construction of the table) he
made explicit use of the categories of quantity (436 [55]), and in the Metaphysics of
Morals he later made use of those of relation (Rechtslehre, § 4). Some (rather fancy)
use of the table is involved in Rell'k;on within the Limits of Reason Alone, VI, 101-2
(Greene and Hudson, 93).
Practical Concepts and Judgment [45
already made: laws and maxims, material and formal principles, cate-
gorical and hypothetical imperatives, imperatives of perfect and of im-
perfect obligation, legal and moral judgments, laws and imperatives,
etc.
Second, we should expect him to set up concepts (categories) of
good corresponding to each of the kinds of judgments. We should then
have a table of judgments or rules and a corresponding table of con-
cepts.
But he does not proceed in these ways. He assumes that the impor-
tant formal features of practical judgments are identical with those of
theoretical judgments, and thus he avails himself of the division of cate-
gories arising from the table of judgments in the first Critique. And he
does not clearly distinguish between the judgments and the concepts,
so that in the table we find sometimes one and sometimes the other.
Some of the difficulty is undoubtedly due to the fact that each of the
headings is, as it were, imbedded within the single category of causal-
ity. We are never told precisely what is meant by a "mode of a single
category," and we are left to imagine what is meant by the quantity,
quality, relation, and modality of the single category of causation.
Other difficulties arise from the fact that the categories of practical
reason are not purely formal like those of theoretical understanding,
but always have a content in the peculiar causality (freedom) which is
the ratio essendi of the moral good. The categories of good and evil do
not owe their origin solely to the forms of practical judgment, as the
logical organization of the table suggests, but to the classes of causes or
effects which are specified in various practical judgments. We can ex-
pect to find, therefore, purely logical distinctions mixed up with those
of the practical in general and with distinctions belonging to ethics
proper.
In order to interpret the table, I shall make one change in it. I shall
distinguish principle or rule from category or concept of the good;
where Kant states one, I shall try to supply the other.
§ 6. THE CATEGORIES OF QUANTITY
Practical judgments are either principles or rules. We are here con-
cerned only with principles. Principles are either mere maxims or laws.
Maxims are maxims either of an individual or of the generality of man-
kind. Laws are distinguished from mere maxims of either kind by be-
ing universally valid and by the requirement that wills under law
should coexist in the all-comprehensiveness of a total system of ends.31
We have, then, the following principles and categories:
[. Mere maxim of individual Subjective, good for the individual
1. Maxim of human beings generally Objective, good for the generality of
(precepts) men
3. Laws as maxims Good for rational beings as such, both
objective and subjective
Three comments are called for by Kant's own listings. First, he er-
roneously contrasts maxim and principle. A maxim is a kind of princi-
ple. Second, he explicitly makes the third a synthesis of the first two,
and thus interprets laws as maxims for individuals. He is able to do this
under the judgment of totality, by virtue of the concept of a realm of
ends as an organic unity of various rational beings. 32 Third, there is a
radical difference between the good in the first two and in the third
categories, a difference which is not adequately shown as a difference
in quantity.
What kind of good is Kant talking about? Perhaps the first two are
das Wohl, the third das Gute. But this is not a quantitative difference.
If we follow Kant in saying that das Gute in a heteronomous system
means that which leads to the end (das Wohl) , then all three are
equally objective as judged by reason, though in Kant's table only the
second and third are called objective.
Here and elsewhere 33 Kant makes it clear that, under the category of
quantity, the transition to a strictly moral concept is reached with the
third category of this group.
§ 7. THE CATEGORIES OF QUALITY
Theoretical judgments, with respect to quality, are either affirmative,
negative, or infinite; the corresponding categories are reality, negation,
and limitation.
Practical judgments are, with respect to quality, either rules (in a
loose sense) of commission or of omission (counsels or commands to do
something or to leave something undone) or permissions to make some
exception to a rule of the former kind. Hence the judgments and cate-
gories of quality are as follows:
I. Rules of commission The good as that which is com-
manded.3'
1. Rules of omission The good as that which is to be
achieved through the avoidance of
certain actions
32 Ibid., 436 (54-55); d. Critique of Practical Reason, %8 (115).
33 Critique of Practical Reason, 67 (159); Foundations Of the Metaphysics of
Morals,43 6 (55),
34 We are not here concerned with the source of the commands or counsels. They
may be arbitrary fiats, civil laws, moral commands, or recipes for making sauces.
3. Rules of exceptions The good as that which is appropri-
ate in obedience to a general pre-
cept
Rules of exception correspond to infinite judgments in theory, in that

they exclude certain actions from the scope of one of the earlier rules.
They are formally affirmative but have a negative predicate in their the-
oretical formulationj 35 in their practical form, we might refer to them
as "but-rules" which include both a positive and a negative element,
viz., "Do so and so, but ...." They are to be distinguished from leges
permissivae. 36 Such rules of exception are present in commands of im-
perfect obligation.
It should be noted that, while the transition from one kind of cate-
gory to the other is clearly marked in those of quantity, no one of the
subcategories of quality seems to be specifically moral. Even in pru-
dence and skill it is necessary to distinguish among these three kinds of
judgment.
§ 8. THE CATEGORIES OF RELATION
Matters become a bit more obscure when we reach the categories of

relation. All rules or principles of practical reason state, implicitly at
least, a relation of the act and locate the good in a relatum of the act.
According to the relations of judgments as categorical, hypothetical,
and disjunctive and the corresponding categories of subsistence-inher-
ence, cause-effect, and reciprocity, we have the following categories of
relation:
I. The rule that all acts be judged as The good as the character of the per-
acts of a subject son; moral good; dignity
2. The rule that all acts be judged (For these categories, see the text,
as having consequences for the below)
state of the person
3. The rule that in a moral commu- The good as justice or equality under
nity the acts of one person be de- a common law; the moral commu-
cided upon and judged as they nity
affect the state of another, and
conversely
The first of these categories corresponds to that of subsistence and

inherence in the first Critique and is sufficiently clear. Just as substance
is that which is the cause of and is preserved through change in its ac-
cidents, the person is an intelligible substance to be preserved in con-
35 Critique of Pure Reason, A 72 =B 97.
36 Cf. Perpetual Peace, VIII, 347 D. (Beck, 9 D.).
'48 Practical Concepts and Judg;ment
duct. Accordingly, the person is an end in itself as a setter of all ends,
and moral values are always values of a person.
The proper interpretation of the second category is a matter of con-
siderable difficulty. It is subject to at least three interpretations, each of
which can be supported by citing Kant's variable usage of the word
"state" or "condition" (Zusta12d). "State" may have reference to moral
condition, physical condition, or the state of well-being and its oppo-
site. The three interpretations are as follows:
a) The moral rule does not issue from personality as an abstract entity,
but from a person in a particular moral condition, a condition of "heart"
that requires self-knowledge for its estimation. 37
b) While our duties of perfect obligation are the same to all men, the
modes of their application differ according to the condition of men, such as
their age, their sex, their social position, their health. 38
c) "Condition" may mean merely the state of happiness or unhappiness 39
[60 (151)].
The following are permissible interpretations of the second category:

lQ) The rule that an action be Goods of action as derivative from
judged as issuing from a person specific goods of character, e.g.,
in a certain moral condition the courageous, the magnanimous
b) The rule of taking account of The good as the suitable to the pa-
the condition of those affected tient of the act
by an action
c) The rule that acts be judged by The good as the prudent
their consequences for the wel-
fare of the person
The third of the categories of relation, in which actions of one are to

be judged in relation to their effects on the state of the other, is likewise
subject to several interpretations according to the meaning of "state,"
but the third of the possible meanings is perhaps the most plausible. 40
37 Metaphysik der Sitten, VI, 441. This is relevant in the imputation of merit
(cf. ibid., 128).
38 Ibid., 469.
89 All three meanings seem to be combined in Reflexion 7111 (XIX, 186), where
Kant says that the "unity of will [WillkUr]" is "conditioned according to the degree
of capacity and happiness of one person with respect to another."
40 Professor Paton, in a personal communication, has proposed the following ele~
gant interpretation of the categories of relation. They are (a) relation to homo
noumen01l (personality) under the imperative of autonomy; (b) relation to homo
pbae7lomeno7l (person) under the imperative of the end in itself; (c) relation of
actions of different persons in a kingdom of ends.
§ 9. THE CATEGORIES OF MODALITY
Before considering the specific moments of the category of modality,
it is necessary to return for a moment to the last sentence before Kant
introduces his table. He says: " ... The categories proceed in order
from those which are as yet morally undetermined and sensuously con-
ditioned to those which, being sensuously unconditioned, are deter-
mined merely by the moral law" (66 [157])' This sentence is devel-
oped, in the last paragraph of this section, by specific reference to the
categories of quantity. The sentence just quoted seems to say that in
each group of categories there are two which apply to objects of an
empirical practical reason, while the third applies to the moral good.
But the sentence just quoted could not have been illustrated with the
categories of quality or relation. The question before us is Does it ap-
ply to the categories of modality, or do they have some peculiar rela-
tion to this transition?
Before answering this question, we must first remind ourselves of
some of the peculiarities of the modal categories. The categories of
modality in theor.;:tical knowledge "concern only the value of the cop-
ula in relation to thought in general" and "contribute nothing to the
content of the judgment."41 By analogy, we can say that the modality
of practical judgments tells us nothing of what is the good to which the
judgments are related, but only tells us the practical weight to be as-
signed to their copula, the "ought," and, per corollary, to the corre-
sponding good. The modality of the concept tells us whether what it is
about is possibly, really, or necessarily good, where "good" has been
previously defined under the other categories. The relevant judgments
would be problematic, assertoric, and apodictic imperatives.
Disappointingly, Kant does not seem to follow this apparently simple
clue to the modal concepts,42 though the first of the categories suggests
that he is abou{ to do so:
I. The rule that cettain actions are The permitted and the forbidden; the
to be judged as allowed and cer- possibly good and that which can-
tain ones as forbidden with respect not possibly be good on the as-
to the accomplishment of some sumption of an arbitrary purpose 43
arbitrary purpose
41 Critique of Pure Reason, A 74 = B 99-100.
42 Though at Critique of Practical Reason, 11 n.
(¢ n.), he suggests that he does so.
This would give rise to the three already recognized concepts of the good, bonitas
problematica, pragmatica, and moralis. There is much to recommend this division,
in spite of what I judge to be Kant's failure to carty it through; indeed, it is so
much to be recommended that others, as we shall see, have actually found it here.
43 The permitted and the forbidden (rmerlaubt) are defined ibid., II n. (¢ n.);
Foundations of the Metaphysics of Morals, 439 (58); and Metaphysik der Sit ten, VI,
222.
ISO Practical Concepts ttnd /udg;ment
The second pair of categories corresponds to the theoretical pair: ex-
istence and non-existence. It is as follows:
1. The rule that certain actions are Dutiful actions and actions contrary
required or forbidden by an ac- to duty; the actual moral good and
tuallaw lying in reason as such evil
The proper interpretation of the second pair of modality categories
is disputable. MellinH interprets "a rule lying in reason as such" to
mean a rule that prescribes an action by which an actual end, which is
always happiness, is to be achieved. This interpretation makes it possi-
ble to regard the three pairs of modal distinctions as corresponding to
the three kinds of imperative and to the three species of the good. But,
as he admits, it requires him to interpret the concept of duty in a non-
moral sense; and this seems to me to be a fatal objection to his view,
quite apart from the fact that "a rule lying in reason as such" could
hardly refer to the supreme principle of hypothetical imperatives, viz.,
the maxim of self-love.
Schtitz'lI interprets the pair to correspond to "that which is really
commanded" and "that which is not really commanded." This gives the
contradictories: "duty" and "not-duty" instead of the contraries listed
by Kant, "duty" and "contrary to duty." I accept this as the most rea-
sonable way of reading the table, though it conflicts with the ipsissima
verba of the author.
On either interpretation, a real surprise awaits us in the third pair.
Whereas, before, the third category has been an almost Hegelian syn-
thesis of the first two, '6 the third category here arises from a logical
division of the second.'7 Of anything that is actual, we can say that it
is either necessarily or contingently actual. Of an actual duty, Kant says
-by a very tenuous analogy-that it is a duty of perfect or of imper-
44 Encyklopiidisches W orterbuch zur kritiscben Philosophie, IV, 534; the same
interpretation had been given by Bendavid, Vorlesungen fiber die Critik der prak-
tischen Vernunft, p. 29.
'liTo Kant, June 23, 1788 (X, 541-42). Schutz in this letter also reports on and
criticizes the effort by A. W. Rehberg in Allgemeine Literatur Zeitung, III (1788),
353 ff., to improve the table of modalities by making the three moments: (a) the
permitted and forbidden, (b) the dutiful (pflicbtmiissige) or the virtuous and its
contrary, and (c) the holy and unholy. Rehberg's article is abstracted in Ak., XIII,
119·
'8 Critique of Pure Reason, B J 10.
'7 If Mellin is correct, the third appears to be a logical division of duty, but "duty"
must then have a moral or juridical meaning lacking in Mellin's interpretation of the
second category. Bendavid attempted to make the third a synthesis of the tirst two,
but failed completely.
Practical C01lcepts tmd ludlr'11e1lt 151
feet obligation. Thus, corresponding to the contrast between the apo-

dictic and contingent judgments, we have the following:
3. The rule that certain actions are Perfect duty; an action directly com-
made necessary by virtue of the manded; the necessarily good
actual rule lying in reason as such
and
The rule that, while a certain max- Imperfect duty; an action not fully
im is necessary for rational beings, determined by the actual rule taken
there is permissible latitude in the abstractly, but falling within the
choice of actions under it 48 scope of the rule as determined by
sound judgment using the rule of
exceptions; the contingently good
action under a necessarily good
maxIm
The distinction of duties of perfect and imperfect obligation follows

that adopted in the Foundations, though it diverged from the "usage of
the schools."49 A perfect duty is an act that can be directly commanded
or one whose maxim requires a certain act, since the contradictory
48 Mellin (op. cit., p. 537) interprets only perfect duty as moral, thus diverging
from Kant's meaning. Hence the order of the categories in his rendition is as follows;
1. Actually dutiful actions, as those conforming to a law actually lying in reason as
such; and actions opposed to this law
3. Duties conforming to a law necessarily lying in reason as such
a) Duties by a necessary law, required by an apodictic imperative, the law being
the moral law (e.g., the duty to tell the truth) = perfect duties
b) Duties under a contingent imperative which corresponds to "a contingent law
(rule) lying in reason as such" (e.g., the duty of a philosopher to teach the
truth) = imperfect duties
The scheme proposed here is:
1. Actions which are, in fact, dutiful or opposed to duty; the legally correct or
incorrect actions
3. Actions which are necessary because they are dutiful; the morally good action
a) Actions or duties of perfect obligation
b) Actions or duties of imperfect obligation
The best evidence for the latter interpretation is that it conforms to the doctrine of
perfect and imperfect obligation in the Metaphyrik der Sitten, where both are
treated as apodictically necessary and differ only in what is commanded (a certain
action, action under a cenain maxim).
49 FoundatiOn! of the Metaphysics of Morals, 411 n. (39 n.); cf. Retlexionen 7114.
71640 7170. The distinction of the schools was that perfect duties could be enforced
by external legislation, while imperfect duties could not. This is only a corollary of
Kant's distinction and appears again in the parallel distinction between Rechtrpflicbt-
en and Tugendpflichten. By internalizing the concept of legislation as autonomy,
the emphasis is shifted from the aspect of enforcibility.
152 Practical Concepts and ludg;ment
maxim is self-contradictory when made into a law. An imperfect duty
is one in which the rule of exceptions may apply, in order to prevent
obedience to one rule from conflicting with another. Moral duties are
always imperfect; legal duties are perfect:
If the law can command only the maxims of the action but not the actions
themselves, this is a sign that it leaves open to obedience a certain latitude for
free choice.... Under a broad [imperfect] duty there is no permit to make
exceptions from a maxim of actions, but only a permit to limit one maxim
of duty by another.... Such imperfect duties are only duties of virtue [and
not duties of law or jurisprudence].GO
Despite all these puzzles, the most puzzling thing in the account that
Kant gives of modality is not in the listing of the specific categories but
in the statement he makes of the relation of modality to the other cate-
gories. "Finally," he says,
the categories of modality initiate the transition, though only in a problemati-
cal way, from practical principles in general to those of morality; and only
later will it be possible to establish the principles of morality in a do wnatic
form through the morallaw [67 (158)].
How can this sentence possibly be reconciled with the one that says
that the transition occurs within each group of categories?51
Perhaps the principles of modality in the first Critique, instead of the
categories of modality in the first Critique, can supply an answer. For
we are here dealing with a question of real possibility, not merely logi-
cal possibility. And in these questions it is principles ot modality that
are important. The second Postulate of Empirical Thought in General
is "That which is bound up with the material conditions of experience,
that is, with sensation, is actual."52 If all our thought of duty and of the
good is not to be merely a logical exercise in the analysis of perhaps
empty thoughts which are only logically possible, then some actual
"material conditions of experience" must be given with which these
concepts can be "bound up." This material condition, of course, can-
not for practical reason be sensation, but it must be something that
could make the same contribution to the practical categories that em-
pirical intuition (sensation) makes to the theoretical. This material con-
dition, or rather this analogon of a material condition, is presented by
the "sole fact of pure reason," the direct consciousness of the moral law
to which Kant appeals in the Deduction. 53
Thus interpreted, the two sentences are not inconsistent. In each
50 Metaphysik der Sitten, VI, 39'>.
51 Cf. above, p. 14Z.
52 =
Critique of Pure Reason, A 218 B 265-66.
53 Critique of Practical Reason, 31 (no); d. below, chap. x, ~ 2.
Practical Concepts and Judgment r 53
family of categories he meant that the transition occurs from the con-
cepts of practical reason in general or empirical practical reason to those
of pure practical reason, though we have seen grounds to doubt that
this actually occurs except in respect to quantity. But there is not yet
given any warrant for saying that any experience requires to be judged
in the light of the categories of pure practical reason. The transition is
made (or attempted) in each of the first three groups; yet this transi-
tion as a logical development is given weight by the categories of mo-
dality which relate the copula in every practical judgment to "thought
in general" without adding any new internal form or content to the
judgment.
Even so, it is not quite correct to say that the categories of modality
initiate this transition problematically. They show what are the three
possible relations of the "ought" to the totality of practical experience,
and in this sense they prepare the way for it. The actual transition,
however, is not made by or within the table, but by our being directed,
by the categories of modality, to the "fact of pure reason" which will
tie all of them to the "material condition" of practical experience. (It
must be remembered that "material condition" here means the fact of
pure reason, not the material of a practical judgment in the sense in
which "material" refers to an actual empirical desire.) By reference to
the experience of choice, all the logical distinctions drawn in the table
become prescriptive of possible meanings of the concept of the good;
by reference to the fact of pure reason, some one or more of each
group comes to be the formula for a moral obligation or the concept
of the moral good.
This ends our tedious examination of the table of categories. It is pre-
sented with hesitation, and it contains more questions than answers,
more conjectures than decisions. We regret that Kant did not work it
out with his accustomed architectonic skill. Something comparable to
it-I make no extravagant claim for my own version-is surely needed
to bring the doctrine of principles and the doctrine of concepts to-
gether. It is perhaps symptomatic of the poor construction of the table,
or at least of its obscurity, that little use was made of it in the book that
it should have served to organize. The cause of the failure may well be
that Kant was here approaching the principles of a philosophy of action
that he never achieved, which we in the twentieth century do not yet
have in the formalized manner that Kant was accustomed to give to all
his deepest work. He failed to achieve it not merely because of the in-
herent difficulty of the task but because he carried over into practical
philosophy a theory of logical fonn that was not adequate to elucidate
all practical judgments. Even today we are still seeking a complete the-
ory of the fonns of decision and valuation; interesting work has been
done, but more important work remains to be done, on such notions as
the "quantity" of imperatives.~·
The "metaphysical deduction" of the principle of pure practical rea-
son lacked the elegance of the Metaphysical Deduction in the first
Critique because there was no sound "logical clue" in the table of
judgments suitable to the purposes of the second Critique. The attempt
subsequently made to use this table of judgments in elaborating a the-
ory of practical categories was therefore destined from the beginning
to be a hazardous operation, and we have sufficient evidence to justify
the verdict that in fact the results were somewhat artificial and arbi-
trary, without the completeness and elegance and necessity claimed for
the categories of theoretical reason. All the manifold objections made
to the listing and derivation of the categories of theoretical reason
seem to apply, a fortiori, to those of practical reason. This verdict must
stand, I think, until some Kant-scholar has been able to give a more
convincing exposition of them than I have achieved here.
§ 10. THE TYPIC OF PURE PRACTICAL JUDGMENTIIII
Judgment (Urteilskraft) is the art or faculty of applying a concept
or rule to a particular case.1I6 Kant shrewdly remarks that a weakness
uCf., for example. R. M. Hare, The Language of Morals (Oxford, 1951), pp.
187-<)1. If few effortS have been made by Kant~exegetes to make sense of the table
of categories, even fewer have been made to construct a better one in the spirit of
Kant's works. The only such attempt I have seen is by J. Stilling, "i)ber das Problem
der Freiheit auf Grund von Kants Kategorienlehre," Archiv fUr die Gcscbichtc der
Philosophie. XXI (1908), 518-34. XXII (1909), 1-17. Stilling says that his work
(which leads to sixteen "categories of freedom") is an elaboration of Albrecht
Krause's Die Gesetze des menschlichen Herzens dargestellt als formale Logik des
reinen Gefiihls (Lahr, 1876), a book I have not seen.
115 I take this occasion to call attention to the error in the Chicago edition of my
translation (p. 176), where the section title is incorrecdy given as "Of the Typic of
Pure Practical Reason." Barni made the Same mistake in his French translation.
116 Judgment is also the art of finding a rule for a case presented. This is called
"reflective judgment," in cOntrast to "determinative judgment" (Critique Of Judg~
ment, V, 179 [15]). In the Critique of Practical Reason Kant is concerned excusive1y
with detenninative judgment: the principle being given, find the case to which it
applies. That this is a fundamental insufficiency in his ethical theory is argued by
M. C. Nahm, " 'Sublimity' and the 'Moral Law' in Kant's Philosophy," Kant-Studien.
XLVIII (1957), 502-24. The creative venturesomeness of private moral judgment.
which may lead from new moral insights to new moral principles, is an important
aspect of moral life that does not receive proper attention in the Critique. For
Kant, moral concern lies in the application of known principles to given cases, not
in eliciting new principles from problems of conduct. Note Kant's remark (Critique
of Practical Reason, 8 n. [93 n.]) that no one would want to be the inventor of a
new moral principle.
in this talent, which he ascribes to mother wit, cannot be perfected by
the learning of still other rules, since this talent would again be re-
quired for the proper application of them. Thus a man learned in
medicine may possess sound theoretical principles and know the rules,
but only practice which perfects a native gift can make him a skilful
physician. tl7
When the judgment or rule to be applied is a priori, as in the case
of theoretical categories, the role of judgment is the same. But it is
more difficult to understand how it functions because the concept did
not arise from the kind of experience to which it is to apply. Kant
holds that it is easy to see how one can say, "This plate is circular,"
since the empirical concept of plate includes the geometrical concept
of circle, or at least has been associated with it in experience; and we
can both think and intuit a circle. It is more difficult to see how one
can say, "The sun causes the stone to become warm," since, as Hume
showed, there is no intuition or datum of cause from which the con-
cept has been abstracted and to which it ostensively applies.
Kant's solution is presented in his doctrine of schematism. A schema
is a representation that is homogeneous with both an intuition and a
concept, so that the concept can refer to it directly and thence in-
directly to the specific intuitions. The schema of an empirical concept
is a kind of generic image which contains as its core the defined prop-
erties of the definiendum but covers also a range of variation that
makes it resemble all the members of a class. tl8 There can be no image,
however, for an a priori concept. Its schema is not an image but the
representation of the procedure of the imagination in synthesizing
from possible data of intuition in such a way that this synthesis is
homogeneous with the conceptual synthesis whose rule is the category
itself. The schema is both a rule of this procedure and its product,
which is the required formal structure of intuition to which the
category is applied. Thus the category of causation is the concept of
the dependence of the existence of one thing upon that of another.
But Hume showed that we have no intuition of such dependence. Yet
we do apply the concept to intuition under one condition, to wit, the
regular succession of phenomena according to a (putative) rule. The
schema of cause, therefore, consists not in some (impossible) image
additional to the images of two events but in the "succession of the
manifold [of intuitions] in so far as that succession is subject to a
rule."119 Whether a spe~ific succession is subject to a rule, of course,
is a question for empirical research to decide. But the schema exhibits
117 Critique of Pure Reason, A 133 =B 171.
1i8/bid., A 141 =B 180. lUI/bid., A 144 = B 18J.
1;6 Practical Concepts rmd Judgment
the concept of cause in such a way that it can be applied to these
sequences which ~o fall under the rule and can be withheld from
those which do not. Without the schema, we could not move from
the concept of mere logical dependence, which is analytical, to real
dependence, which is synthetic. 60
An analogous problem arises in judgments which are to subsume
actions under practical rules or concepts, where subsumption may
mean the estimation of a particular action in the light of a general rule
or the decision that a particular action would satisfy the rule and
should, therefore, be performed. To decide whether a particular act
possible in the sensuous world falls under the principles which define
the good requires the power of jUdgment. The problem of practical
judgment, when it is a matter of skill and even of prudence, is not
essentially different from that met with in theoretical judgment, since
all the practical propositions of skill and prudence have theoretical
correlates and since the criteria of a successful technique and of a
satisfactory mode of life are to be found in experience.
With pure practical judgment the problem is more difficult, since
the law is a law of reason, not of understanding, and no intuition can
be adequate to it. We can never be sure, in any experience, whether the
full terms of the moral law have been observed. It therefore seems
absurd, Kant says, to wish to find a case in experience, which stands
under theoretical categories, to which the moral law applies so that
the ideal of moral good can be exhibited in concreto. 61 Similarly, the
60 I do not wish to appear to be falling into the error of thinking that the forms
of judgment which generate the categories are {oons of analytical judgment only.
Certainly I can say "Ghosts cause miracles," and this is a synthetic judgment. But
unless intuitions are supplied, corresponding to the concepts of ghosts and miracles,
and unless they follow the schema of causation, such a synthetic judgment cannot
be justified. The only kind of dependence that can be known purely conceptually
is analytical dependence; intuitions for the schema are necessary if the asserted de-
pendence is to be synthetic. Yet the hypothetical judgment and the category of
causation are meant to apply, respectively, to synthetic (causal) judgments and to
the synthetic connection of phenomena whereby the occurrence of one is dependent
upon the occurrence of the other.
81 Critique of Practical Reason, 68 (r59). In Critique of Pure Reason, A 425 =
B 453, however, Kant states that moral philosophy, like mathematics, "can present
its principles, together with their practical consequences, one and all in concreto, in
what are at least possible experiences." But the contradiction between the two
passages is more apparent than real, as can be seen from the context of the latter
passage. Kant is there trying to show that the "skeptical method," i.e., the method
of the antinomies, is especially serviceable in speculative philosophy but useless in
natural science, mathematics, and morals, for none of these "lay claim to insight
into what is beyond the field of all possible experiences." The moral philosophy
which presents its cognitions in concreto is a philosophy already armed with a type
Practical C01lcepts and /udg;ment 157
morally good is not a natural property of an act standing in causal
or other categorical relations to others. Yet pure practical concepts
and principles must be applied, and they are applied even by the "com-
monest mind" with often more skill than by the moral philosopher.
The questions are: How is it done? What is it that will bridge the
conceptual gap between what ought to be and what is, so that the
concepts of the former may be applied in a definite way to, or withheld
from, at least parts of the latter so as to show that they are, or are
not, as demanded by the moral imperative? Without an answer to
these questions, the normative-descriptive distinction, upon which
Kant lays such enormous weight (26 [114]), marks an uncrossable
chasm which is incompatible with the very notion of the ought and
which is not, in fact, present in our ordinary moral concern as some-
thing uncrossable; we cross it every time we make a moral decision
or pass a moral judgment.
Kant's answer to these questions is his theory of the type or typic
of practical judgment. 62 Discussion of this has a place in the second
Critique analogous to that of the schematism in the first.
A practical principle does not, as such, give knowledge of any
empirical fact. While a practical judgment presupposes that what is
commanded or counseled is possible in the empirical world, this pos-
sibility is established by pure cognitive (theoretical) procedures. But
that an action possible for me in the world of sense is an action that
is morally possible 63 requires a different kind of judgment. We are not
concerned, in practical decision as such, with an action to which a
specific practical principle applies descriptively by virtue of its cogni-
tive component or correlate; we are concerned only with the decision
of judgment (though, of course, the typic was formulated after the passage in ques-
tion). The su11Tl11Um bonum, on the other hand, cannot be presented in concreto as a
cognition of practical reason, and there the method of the antinomies again comes
into its own.
62 Type: "That by which something is symbolized or figured; a symbol, emblem"
(Shorter Oxford English Dictionary). For full discussion of the meaning of this
term see Paton, The Categor;callmperative, pp. 160-61. Neither the New English
Dictionary nor Grimm's W orterbuch recognizes the use of "typic" (Typik) as a
noun; but it is clear that Kant wishes to distinguish between the type of moral
judgment (Typus) and the theory of the type (Typik). On symbol and schema cE.
Fortschritte der Metaphys;k, XX, 17~o, and Critique of Judgment, ~ 59.
Kant uses the terms "type of the concept" and ''type of the principle" as equivalent
(Critique of Practical Reason, 70 [161]).
63 Critique of Practical Reason, 57 (148): "To decide whether or not something is
an object of pure practical reason is only to discern the possibility of willing the
action."
158 Practical Crmcepts and Judg;ment
as to what action ought to be done among all those that are theoreti-
cally possible. We have, therefore, an analogy: a schema of a case
occurring according to a law is necessary for knowledge of the case,
while a schema of the law itselfu is necessary to connect, in practice,
possible events in sense experience with a cause under a law which is
not a law of natural connection. Because the schema of a case has an
intuitive component, it is always cognitive in function. But no intui-
tion is available to a practical law of what ought to be, for two reasons:
(a) it is a law of what ought to be, not of what is, and (b) only one
member of the moral motive-act nexus is within nature. A schema of
the law itself must be provided by the understanding or the faculty
of thought and not by intuition or imagination.
Just as a schema was the "third thing" that could mediate between
pure concept and pure intuition, the type must be a third thing that
can mediate between the concept of nature, all that is, and the concept
of what ought to be. The third thing in the practical judgment is the
concept of law itself as definitive of a realm or kingdom." Nature is
phenomena under law, and natural law provides a type or model 66
by which we can think the practical law in concreto.
We ask ourselves, in face of temptation, "What sort of world would
this be if everybody acted the way I wish to act? Would I be willing
to create such a world, or to live in it if it existed?" But the desire to
live in a world in which, for example, everyone developed his talents
is not the reason why I should develop mine; it is not the motive 61
of my efforts at improving myself, for the obvious and simple reason
6t Kant points out the danger of the expression "schema of the law" (ibid., 68
[160]). He uses the word "schema" in a broader sense in Reflexion 5611; in Meta-
physik der Sitten, § 45 (VI,468) it means hardly more than example. In the Founda-
tions neither "schema" nor "type" is used, but there is a brief reference to the prob-
lem under the name "canon of moral estimation" (414 [4d). Similarly, Zum ewigen
Frieden in der Philosopbie, VIII, 410: "canon of morally practical reason." Re-
flexion 7160: "analogon."
66 Perpetual Peace, VlII, 371 (Beck, 37): Natural law is a "model" for constitu-
tional construction.
61 Critique of Practical Reason, 69-70 (161). Failure to recognize the difference
between the type of moral judgment and the motive of moral action has led some
critics to say that in the categorical imperative Kant has committed himself to a
utilitarian or even egoistic doctrine (cl., for example, Mill, Utilitarianism, chap. i).
But Kant did not fall into this confusion, which he warns against. He was quite well
aware of the fact that if I lie, it does not mean that all other men will lie, and there-
fore the fear of the consequences of my lying does not include the fear of the con-
sequences which would follow from the existence of a world in which everyone lied.
ef. especially Foundations of the Metaphysics of Morals, 438 (57).
that I am not the creator of a world, and in this world I know, all too
well, that not everyone is going to act the way I do, be it for good or
ill. But implicit in the notion of a moral order is that of an order of
interacting wills (the third category of relation), and the best model
we have for such a world is the order of nature under law. 68 That is,
what would not be possible in an order of nature under law is not
morally possible, though what is actual in nature (a matter which
empirical intuition must decide) has no judicative function in the
abstract determination of what is morally possible and necessary.6D
"The order of nature under law" means two things, one of which,
as Paton points out,1° is largely forgotten today. First, it means a
uniform sequence of phenomena under causal law, and its distinguish-
ing feature is the universal uniformity of nature. Nature, as a mechani-
cal system, was believed, in the eighteenth century, to be "governed"
by such laws having universal application. Before passing to the second
of the meanings of the "order of nature," we must inquire into the
practical significance of this concept of uniformity.
The first test of a maxim is the mere universalizability of the maxim,
i.e., the interpretation of it as a descriptive universal principle. Of some
maxims, Kant says that they destroy themselves if made universal. 71
68 From time to time Kant pushed the analogy very far. Thus he early compared
the good will in Rousseau to the force of gravity in Newton, seeing in each an organ-
izing principle in an orderly realm (Triiume eines Geistersehers, II, 330, 335; Re-
flexion 5419). He told Mendelssohn (April 8, 1766 [X, 72]) that this did not express
a serious opinion but was only an example of how far one could go in Erdichtung
where data are lacking for knowledge. But the thought evidently had a certain
fasc1nation for him, for we find it again in Metaphysik der Sitten, Tugendlehre,
S 14 (VI, 449) and Opus postumum, XXI, 35.
69 Though, of course, it has a contributory function, since every action is predi-
cated upon empirical knowledge, which provides the content for the abstract prin-
ciple of choosing means suitable to a given end.
70 The Categorical Imperative, pp. 150 ff., 161 ff. I hesitate to follow Paton in
only one point in his definitive and exemplary exposition: I do not think that he is
correct in asserting (p. 149) that only the second of these conceptions of the order
of nature plays a part in the typic. The concept of causal uniformity does have, I
think, at least a minor role, as I try to show in the text.
71 Kant says this repeatedly. In the Critique, however, he speaks also of actions as
being universalizable. (Note that in the first and second formulas in the Foundations
£42 r <38, 39)] he speaks of universalizable maxims, while in the typic it is the uni-
versalizability of the action itself which is mentioned.) This is an inaccuracy in the
Critique. If lying were universal, we would be able to get along far better than in
this world, where it is only frequent; we should simply interpret affirmative sen-
tences negatively and negative ones affirmatively. But if the ma:rim is to deceive
another person, the best way of doing it is by sometimes telling the truth and some-
times not.
160 Practical Concepts and Judgment
That is, the effectiveness of such a maxim is dependent upon the fact
that it does not correspond to a universal. or even general, description
of human behavior. It is not possible to will rationally that such a
maxim should be universal. even though the proposition "All men
should lie" is not logically self-contradictory. A maxim like "I should
lie" depends for its effectiveness upon the fact that it is not universal,
that its theoretical correlate "I lie" is not universalizable into a judg-
ment, "All men always lie"; for. if it were, there would be no such
thing as a lie at al1,12 One's lies show mendacity and cleverness only
because they are exceptions to a general rule. But general rules which
have exceptions are not laws of nature; the latter have no exceptions.
Attention should here be called to a matter already discussed above
(p. 12 I). The universalizability of a maxim is a negative test of its
validity as a law. But many maxims can in fact be universalized which
do not have the status of law. In this way, the type of the moral law
as a universal and uniform natural law is only a negative criterion for
moral judgment.
The second meaning of the order of nature is one according to
which all the laws and the phenomena under them are in such a rela-
tion that nature as a whole can be interpreted as an organic unity, 73
which suggested to the natural theologian that it had been designed
by a wise creator. For Kant, the inference to a wise creator is logically
invalid; but the thought of a teleological organization is required as a
regulative principle in our search for as yet unknown causal connec-
tions. Natural theology and natural teleology regard the world of
nature as such a realm. though for Kant only the Idea of it is a valid
methodological assumption. Moral teleology. on the other hand. sees
such a realm as an ideal to be achieved in action;74 it is a regulative
Idea for practice and not for knowledge only. This Idea is that of a
realm of ends, organized by the third category of practical relation.
that of community of persons under common law, the whole being
72 This is said independendy of the logical problem of the Epimenides paradox.

Let it be supposed that all men ex(.ept Epimenides did, in fact, invariably lie. The
logical paradox would not then ensue, but the practical effectiveness of lying would
disappear. We should simply say that Epirnenides said "Yes" when other men said
"No," and no one would be deceived, and if Epimenides willed that all other men
should lie, there would be no sense or purpose in his lying. Hence the maxim would
be "self-destructive" just as surely as his statement about all men was paradoxical.
73 The moral significance of this conception has already been touched upon lightly
(d. Critique of Practical Reason, 17-18 [115], and above, p. 99).
U Foundations of tbe Metaphysics of Morals, 436 n. (ss n.).
a corpus mysticum of rational beings. 75 Such a world is archetypal,

and the Idea of it "determines our will to impart to the sensuous world
the form of a system of rational beings" (43 [I 32 D.
This regulative conception of nature, believed by almost all eight-
eenth-century scientists and philosophers to describe the actual cosmos,
is the model for our thought of the moral realm. I do not merely ask
myself whether a realm of nature consisting of rational beings acting
uniformly in the way I propose to act would be possible, i.e., if the
maxim could be a universal law and accomplish the ends I have and
express in the maxim; I ask, further, whether I, as a creator of the
world in which every part should have its natural place and function,
would will that certain maxims should have the force of law. Would
it be a world in which the natural ends of things would be systemati-
cally thwarted? If so, though such a world is possible as a territory76
of uniform event-sequences, it would not be a realm which could be
rationally desired by a being who consistently traced out the implica-
tions of his desires. That is, when I will an immoral action, one that
would sow discord among rational beings, I will according to the
maxim of the act and also will (tacitly) that my maxim not be uni-
versal. And such a maxim is not then analogous to a law of nature.
Here it is necessary to point out an easy and frequent error in
Kant-interpretation. In the third example in the Foundations, Kant is
said 71 to have committed himself to the view that the motive of the
T5 Critique of Pure Reason, A 808 =B 836.
T8 On territory and realm d. Critique of Judgment, Introduction, II.
TT E.g., by Ross, Kant's Ethical Theory, p. 47. Can the statement (CritiqUe of
Practical Reason, 70 [162J) that the maxims of a will detennined only by the maxim
of helping itself serVe as a "very adequate type for the morally good" be reconciled
with the other statement (28 [I 15]) that "though elsewhere natural laws make every-
thing harmonious. if one here attributed the universality of law to this maxim there
would be the extreme opposite of harmony, the most arrant conflict, and the com-
plete annihilation of the maxim itself and its purpose"? Yes, if we consider the words
"If this will made itself into a universal law," for then the freedom of each is re-
stricted by that of the others (d. 34-35 [123-24]). That such a conception of the
mutual restriction of wills, if put into practice, would necessarily lead to the greatest
happiness is assened in Critique of Pure Reason, A 809 = B 837. where it is called
an "Idea of reason." This is the Idea of a realm of ends, a legitimate moral motive,
since our maxims would be legislative for such a realm. But the Idea of Wliversal
happiness is only the type of such a realm. In the quotation from p. 18 (115) it may
be supposed that Kant is referring to the universalization of each man's selfish maxim
as a law of nature and not to the maxim as it must be modified and restricted if it is
not to lead to "arrant conflict," while the sentence on p. 70 (162) refers to the maxim
as thus refined and restricted.
[61 Practical Concepts and Judgment
benevolent action is the desire for benefits from the recipients of my
helpfulness. This is not only a misinterpretation of what Kant said
and meant; it is a very poor foundation even for enlightened selfishness,
human gratitude being the delicate flower that it is. The question is
not What do I stand to gain by being altruistic? The question is Can
I, standing frequently in need of help from others in carrying out my
own purposes, consistently will a maxim to be universal according to
which I could not in principle expect or demand such help? It is not
that by acting unkindly I cut myself off from their kindness; this
mayor may not happen. It is a question of whether, among all my
maxims, I can include one which, made universal, would thwart my
efforts to carry out the terms of the others. If my will is good, the
expected Of hoped-for help from others is not the motive of my action;
but if my will is rational, I must consider the consistency of .willing
both my private purposes and a state of affairs in which the likelihood
of satisfying them would be reduced to nothing.
The type of moral judgment, thus interpreted, does not require
uniformity of actions but only mutual interaction under a common
maxim. This interpretation is certainly favored by the wording in the
Critique which requires the universalizability of the maxim but allows
"variety in the rule" (20 [106]), but is not consistent, in its full im-
plications, with the wording of the Typic (69 [[ 61 ]), which requires
the universalizability of the actions themselves. Only against the latter
formulation is the criticism of HegeF8 and others valid that some
actions, e.g., almsgiving, cannot be universalized. As a rule, of course,
"Give alms" cannot be universalized; there must be some people who
cannot give alms if anyone is to give them. As a principle, however, of
helping others in distress, this maxim of imperfect obligation can be
uni versalized.
At the end of this chapter, Kant fully develops the meaning of the
type as nature considered as a realm of ends to be made actual by our
actions in it. Nature herself, and not merely her law, is considered as
a type or symbol but not as the realm of moral ends. Accordingly,
Kant warns against the error of Wolff and others who confused the
realm of nature and the realm of ends in their doctrine of perfection. 79
This error is empiricism in morals,80 which is inevitably heteronomy
78 Lectures on the History of Philosophy, trans. Haldane and Simson (l8c)S> , III,
¢<>.
79Cf. Vernilnftige Gedancken von der Menschen Tbun und Lassen, \ 137.
80 Critique of Practical Reason, 70,94 (161, 188); Critique of Pure Reason, A 474 =
B 501.
and thus destructive of the purity of morals. And he warns also against
taking a type of the realm of ends as if it were a schema of a tran-
scendent intuition and thus confusing the mere ideal thought of a
realm of ends with an alleged intuition of an actual, ready-made, realm
of ends. 81 In either direction, theory of what is (either of the world
of phenomena, in the former, or of the transcendent world, in the
latter) pre-empts the place of autonomous theory of what ought to be.
All these theories ascribe to judgment a role that can properly be
filled only by reason, which alone can supply the universal principle
needed by judgment in the guidance of conduct.
81 Critique of Practical Reason, 85--86 (179). This is fanaticism (Schwiirmerei) in
morals and is related to mysticism (ibid.,71 [163]).
x
The COCOTranscendental Deduction" of the Principle
of Pure Practical Reason; Commentary on
Section 7 andAk., 42-50 (Abbott, 131-40)
§ I. INTRODUCfION
It has been noted that the structure of Kant's exposition in §§ 1-6 and
in chapter ii is analytical and problematical. These pages inquire into
what would be the form of the law if the hypothesis is correct that
reason of itself alone can give an a priori practical law. While there
is a categorical finality in Kant's statements about the character of
practical reason as such, he is proceeding only analytically in §§ 1-6
with respect to pure practical reason. He is asking only: If pure reason
is practical, what is its law and what is the nature of the will that can
obey this law?
Nevetheless, a man who believes as finnly as Kant did that the moral
law is a law of pure reason cannot keep this belief entirely under
cover. There are several places where his belief in the reality of pure
practical reason is obvious, before he has reached the place in his
treatise where these assertions are systematically justified. 1 Even in the
Critique of Pure Reason, which does not pretend to show that freedom
is real but only that it is possible, he often breaks through this logical
reticence and comes out with assertions such as this: "I am justified
r
in making the assumption that moral laws a priori can determine the
will] in that I can appeal not only to the proofs employed by the
most enlightened moralists but to the moral judgment of every man,
in so far as he makes the effort to think such a law clearIy."2 The
"moral judgment of every man" is the true starting point of the Kantian
moral philosophy, and this fact should not be forgotten, however
synthetic, rather than analytical, the structure of the second Critique
is meant to be.
1 E.g., Critique of Practical Reason, 19.27,29 (1°5,115,117).
"Transcendental Deduction" of the Principle I6S
Kant's readiness to appeal to what the ordinary man thinks in moral
matters does not strike the twentieth-century reader as being starding,
for to us this is the only place to begin. But the sharpness of the reply
to Titte13 shows that there was a spirit of novelty (at least in Con-
tinental philosophy) in going into moral problems from this direction,
instead of in the more usual way of erecting a world system that might
subsequently be used to explain, or often to explain away, ordinary
moral phenomena. In this procedure, Kant is indebted to Rousseau·
and to the conscientiousness of pietism, which manifested itself in the
profound but simple-minded righteousness of his parents, whom he
openly revered all his life. Yet this moral consciousness may be il-
lusory; duty may be a "vain delusion and a chimerical concept."5 We
do not know that true virtue can be found anywhere in the world,
even if the concept is valid; and no examples can prove that it does
exist. 6
The tone of Kant's argument suddenly changes in § 7. He no longer
attempts to restrict himself to hypothetical statements about what
would be true if pure reason were practical, but boldly asserts: "Pure
reason is practical of itself alone, and it gives (to man) a universal
law, which we call the moral law."7 To show the contrast between
the analysis and the assertion, compare the following two passages:
Sometimes we find, or at least believe we find, that the ideas of reason have
in actual fact proved their causality in respect to the actions of men, as appear-
ances. S
3 Critique of Practical Reason, 8 n. (93 n.); cf. Critique of Pure Reason, A 831 =
B 859: "Do you really require that a mode of knowledge that concerns all men
should transcend the common understanding and should only be revealed to you
by philosophers? Precisely what you find fault with is the best confirmation of the
correctness of these assertions."
4 "By inclination I am an inquirer. I feel a consuming thirst for knowledge, the
unrest which goes with desire to progress in it, and satisfaction in every advance in
it. There was a time when I believed this constituted the honor of humanity, and I
despised the people, who know nothing. Rousseau corrected me in this. This blind-
ing prejudice disappeared and I learned to honor man. I would find myself more
useless than the common laborer if I did not believe that this attitude of mine [as
an inquirer] can give worth to all others in establishing the rights of mankind" (XX,
44)·
5 Foundations of tbe Metaphysics of Mora/s,402 (18).
6/bid·,40 7 (23-24).
7 Critique of Practical Reason, 31 (120). Here appears the confusion between law
and imperative which we have already noticed occasionally. Kant should have said:
"It gives a universal law which we call the 'moral law,' valid for all rational beings,
and to man it issues a corresponding categorical imperative."
166 "Transcendental Deduction" of the Principle
One need only analyze the sentence which men pass upon the lawfulness of
their actions to see in every case that their reason, incorruptible and self-con-
strained, in every action holds up the maxim of the will to the pure will, i.e.,
to itself regarded as a priori practical []l (Il 0 ) J.
What is it that authorizes Kant to make this change in the status of
the hypothesis that pure reason can be practical? A change in mood
does not of itself constitute a step in argument.
There are two reasons for it: the alleged "fact of pure reason" and
the somewhat equivocally titled "deduction" of the principle.
z. THE "FACT OF PURE REASON"
§
What was previously only a methodological standpoint, the assump-
tion of moral consciousness, now functions as an actual premise of
the argument, in spite of Kant's having acknowledged that it might be
illusory. He says:
The consciousness of this fundamental law may be called a fact of reason,
since one cannot ferret it out from any antecedent data of reason and since
I the law] forces itself upon us as a synthetic proposition a priori based on
no pure or empirical intuition [3 I (no)].
In order to regard this law without any misinterpretation as given, one must
note that it is not an empirical fact but the sole fact of pure reason [3 I (110)].
This Analytic proves that pure reason can be practical, i.e., that of itself
and independently of everything empirical it can determine the will. This it
does through a fact wherein pure reason shows itself actually to be practical.
This fact is autonomy in the principle of morality.1I
The moral law is given, as an apodictically certain fact. as it were, of pure
reason, a fact of which we are a priori conscious. 10
II
•
Critique of Practical Reason, 42 (131). In the next paragraph, this fact is said
to be identical with the consciousness of the freedom of the will.
10 Ibid., 47 (136). Note that this sentence says that it is "as it were a fact." On
pp.6, 31,42,4] (91,110,131, IJ2) Kant calls it a "fact"; and on pp. 47,55.91, and 104
(136, 145. 18 5, 199) it is called a "fact, as it were," or some similar qualification is
made. (Abbott omitted this qualification on his pp. 136 and J85.) This has suggested
to some that Kant was uncomfortable with this unusual use of the word "fact" and
became cautious after its earlier use. But though this is the most plausible interpre-
tation of the difference, his caution departed when he had finished the CTitiq~, for
we find the following: Critique of ludy;m.ent, § 91 (V, 468 [po-liD-freedom is the
fact; Metllphysik deT Sitten, Rechtslehre \ 6 (VI, 251)-practical law of freedom
is a fact; VorllTbeiten ZUT Tugendlehre, XXIII, J78-"It is res facti that this law is
in us and is indeed the highest"; Opus postumum, XXI, zI-the categorical impera-
tive is the fact of moral-practical reason.
Critique of PUTe Reason, B xxii and xviii, speaks of "practical data of reason" and
practical principles as "a priori data of reason." Reflexion 720J (i.e., Duisburg Frag-
ment 6) introduces the concept, perhaps for the first time, but without the word
"fact." But the moral law is here compared to a prior.i representations of space and
time as givens,
"Transcendental Deduction" of the Principle 167
With respect to these famous but obscure passages, we must ask
two questions: (a) What is the alleged fact of reason? and (b) Is it
a valid foundation for that which Kant attempts to build upon it?
a) \Vhat is the fact of pure reason? Kant himself does not seem to
have made up his mind on the best way of expressing it. The text
shows the following meanings: in the first quotation, it is consciousness
of the law; in the second and fourth quotations, it is the moral law
itself; in the third quotation, it is autonomy. Since Kant identified
freedom as autonomy with the moral law (33 [122]), perhaps the
second and third may be allowed to stand as one. But a prima facie
distinction exists between "consciousness of the moral law," which can
certainly be said to exist as a fact (whether we wish to call it a "fact
of reason" or not), and the law itself, of which we are conscious
(whose "factuality" is sub judice).
If this prima facie distinction is finally valid, Kant's argument re-
fuses to move, except in a circle. For everyone will grant that the
"fact" in the first sense exists, but it does not imply the "fact" in the
second of the senses. Yet it is the second of the facts that is essential
to Kant's argument.
b) Is it really a "fact" in the sense required? That the fact in the
second of the senses admittedly cannot be explained is, in itself, no
reason to reject it.ll Some facts must be unexplained; but why this
one, instead of some other that might be incompatible with it? Perhaps
we have a fundamental intuition or insight into its truth? I think that
Kant undoubtedly believed that we do, though he would not have
called it an "intuition"; but fundamental insights and intuitions do not
seem to be any different from the kind of evidence that Kant has
already admitted might be wrong. An appeal to insight Of intuition
is a confession of failure to find an argument or premise from which
some truth can be derived and an unwillingness to surrender it in
spite of that. In principle, some kind of intuition is indubitably nec-
essary, but that does not mean that some particular intuition is either
necessary or indubitable; it may indeed be wrong. Perhaps the moral
law is the kind of fact that must be assumed if we are to explain and
render intelligible our moral experience? 12 But if this is the meaning
11 Critique of Practical Reason, 3 J, 47 (HO, J 36).
12 The Deduction proper does not strictly follow the second of these paths but
adduces an independent warrant for the principle from the need of theoretical reason
to assume freedom. Hence it falls outside the scope of the censure of this sentence
(cf. below, pp. 173 f.). That the moral law is a "fact" in the sense in which, say,
mathematics and natural science are "facts" and therefore subject to the same kind
of justification is the interpretation of Hermann Cohen, Kll7lts Begrondung der
Ethik, p. 224. and his Marburg followers.
here, it hardly deserves the name of fact but only of assumption, for
the experience it is supposed to organize is itself sub judice and could,
perhaps, be equally well organized in terms of some other assumptions,
e.g., those of psychoanalysis.
Thus if we permit the prima facie distinction between the two al-
leged facts to stand as a fundamental duality, Kant cannot make a
transition from the undisputed fact (that we are conscious of a moral
law13 ) to the disputed fact (that there is a law that can come only
from pure practical reason). Yet it is the latter whose factuality is to
be shown if the moral law is to be justified.
I believe, however, that this duality of meaning of "fact" does not
represent Kant's premises properly. There are two intimations of this.
First, there is his assertion that this fact (whatever it is) of pure reason
is the sole fact; and I do not think Kant means that it just happens to be
the only fact of pure reason. I think there is something about this fact
that makes him regard it as a priori unique. Second, there is the dif-
ficult and obscure hint that was never developed but is the clue to
the interpretation I shall present here. "Freedom and unconditional
practical law reciprocally imply each other," Kant says (29 [117])'
"I do not here ask whether they are actually different, instead of an
unconditional law being merely the self-consciousness of a pure prac-
tical reason, and thus identical with the positive concept of freedom/'
i.e., with the concept of autonomy.
To explore this possibility, we need to draw still another distinction
in the meaning of "fact of pure reason." "Fact of pure reason" may
mean a fact known by pure reason as its object, modo directo. Or it
may mean the fact that there is pure reason, known by reason re-
flexively. These may be distinguished as "fact for pure reason" and
"fact of pure reason."
When we think of the moral law as a fact for pure reason, i.e., as
an object of a particular and unique insight or intuition, the very
uniqueness of this fact is at least a ground of suspicion. The first
Critique showed that there are not any facts known by pure reason
without sensuous intuition. Facts are given to reason only by and
through intuition. Should there be any fact "for" pure reason, it
would be only a fact "as it were." If he is speaking of the fact "of"
pure reason, on the other hand, this is not subject to the suspicion we
13 If it should be argued that "to be conscious of" is like "to know" and therefore
must refer to a true proposition or actual object, then even the first of the facts is not
indisputable. It should then be stated: "We believe that there is a moral law," and
this states a fact but does not imply, even in ordinary speech, that there is a moral
law,
feel about the fact "for" reason. Kant's point is that in any willing
there is a principle which is purely rational,14 and "if pure reason is
actually practical, it will show its reality and that of its concepts in
action."I:!
Only a law which is given by reason itself to reason itself could
be known a priori by pure reason and be a fact for pure reason. The
moral law expresses nothing else than the autonomy of reason (33
[ 12 2]); it is a fact for pure reason only inasmuch as it is the expression
of the fact of pure reason, i.e., of the fact that pure reason can be
practical. That is why the moral law is the sole fact of pure reason
and for pure reason.
If this seems to be a rather tricky argument, let us turn to the
moral phenomenon which it represents. A moral principle is not bind-
ing upon a person who is ignorant of the principle or law. On the
other hand, if a person believes that an imperative is valid for him,
then it is in so far forth valid for him, and he shows that reason is
practical even in the awareness of trus aspect of a valid claim. This
is true whether the imperative expresses a claim that is in fact valid
or not. Only a being with an a priori concept of normativity could
even make a mistake about this. To argue against it is to appeal to
normative grounds and is as ridiculous as to attempt to prove by reason
that there is no reason (12 [97]). "I say that every being that cannot
act otherwise than under the Idea of freedom is thereby really free in
a practical respect."16 But the Idea of freedom is expressed in the
moral law; hence to be conscious of moral constraint, i.e., of the law-
this is the fact of pure reason-ipso facto validates the practical claim
of a moral law, which is the fact for pure reason. It is indeed a very odd
kind of fact, and Kant is right in calling it a "fact, as it were."
To recapitulate: moral consciousness, consciousness of duty, is an
undisputed fact. Prima facie, it does not justify the assertion that duty
is not an illusory and chimerical concept. (If I believe, for instance,
that there is a God and that duty is determined by the will of God,
and if there is in fact no God, then it may remain a fact that I feel the
14 This is connnned by the analysis of even hypothetical imperatives, chap. vi, § 6.
But Kant goes even further and says that we become conscious of the moral law
whenever we construct maxims for the will. By this I think he means that in every
decision on a policy of life there is a putative rationality in the rules subsumed under
the principle, which, if fully elaborated, would require the rationality of the motive
as well as that of the means to its satisfaction (cf. Critique of Practical Reason, Z9
[1171).
15 Critique of Practical Reason, 3 (87). Cf. Critique of Pure Reason, B 430-31.
which must have been written only shortly before the second Critique.
16 Foundations of the Metaphysics of Morals, 448 n. (67 n.).
call of duty, but it is not a fact that duty is an objectively valid con-
straint.) But because the moral law-the fact for pure reason-expresses
nothing but the lawgiving of reason itself, the fact of pure reason is
reflected in the fact for pure reason. For a being who thinks that there
is some obligation, there is some valid law. The metaphysical deduction
has shown what this law must be; it is the fact for pure reason.
S 3. THE DEDUCTION
In the Critique of Pure Reason, Kant says: "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,
they entitle the deduction."17 This is, at least in part, the sense in which
Kant uses the word "deduction" here. It is not a disputed fact that
men use the concept of causation and that they believe themselves to
be under legitimate moral constraint. But the question of the right to
use these concepts is a different one from that of the fact of their use.
A legal deduction is one in which a jurist makes a syllogism: "X is
right because X is A, and the Constitution says, 'A is right.''' But a
Kantian transcendental deduction does not have this neat structure.
There are not any principles implanted in men's minds from which
the principles of causality or morality could be shown to be logical
consequences and therefore permissible or necessary. Such a principle
would not be a bit more obvious than the concept of cause or ought
itself and would be subject to the same challenge that Hume brought
to the principle of causation. If we attempt to deduce, in this sense,
the moral law from some other fact for reason, such as man's natural
desire for happiness, that which is thereby derived is not a necessary
law. There are no better-established principles from which moral
principles can be syllogistically deduced.
The process of transcendental deduction is not that of linear in-
ference from a premise to its logical consequence. It is a process of
taking some body of alleged fact (e.g., mathematics or science) which
has been challenged and showing (a) what its necessary presupposi-
tions are and (b) what the consequences of denying these presup-
positions are. Thus, for example, the objective validity of mathematics
had been challenged by those who held that its propositions were
based upon experience and so lacked necessity (Hume in the Treatise)
or were analytical and so lacked objective application (Hume in the
Enquiry). The Critique of Pure Reason attempted to show that, if
mathematics is empirical, not only is it uncertain but the space which
it treats of can be neither finite nor infinite. That is, an empiricistic

mathematics is self-contradictory. Kant then formulates another theory
as to what space is, in which the antinomy can be avoided, and then
shows that one of its consequences is the necessity of mathematics.
This deduction is not undertaken for the sake of mathematicians in
their own work, for they had got along very well (in Kant's opinion)
with quite erroneous views of space. It was performed to keep them
from drawing unwarranted conclusions concerning the nature and
limits of their science and to prevent skeptics from deriving spurious
aid and comfort from erroneous opinions about mathematics.1s Simi-
larly, Kant attempts to answer Hume's challenge to the concept of
causation. But he does not do so "by taking for granted that which
Hume doubted"19 in order to initiate a syllogistic linear inference to
a conclusion that Hume rejected. He does so by trying to show that
if causation is only a subjective association of ideas, sanctioned by
custom, then even the data from which we might make an induction
to an emaciated "causal" generalization would not be available, since
only in a causal system can we distinguish between an objective order
of events from which an induction might be made and a subjective
sequence of representations which is too variable to support an in-
duction. 2o That is, only if some consequence of the causal conception
is already available, can Hume's reconstruction of the principle on his
own grounds be undertaken.
In its broad outlines, Kant's procedure here is comparable to that
in the first Critique, though, strangely enough, Kant so emphasizes the
differences that he denies, in spite of the title of the section, that there
can be a deduction of the principle of pure practical reason (47 [136]).
He does not do this, however, until he has already availed himself of
the "fact of pure reason" and asserted that the principle needs no
deduction.
Before describing what Kant actually does in the remainder of this
section, which seems prima facie to be sadly misnamed, let us consider
what we might reasonably have expected him to say in a deduction.
We should expect him to introduce here a notion very prevalent now,
that of "moral experience" as a realm to be analyzed, articulated, and
established. 21 A critical regression upon its presuppositions would lead
him to one or more synthetic a priori propositions. Their justification
would not lie in a claim that they are "firmly established in themselves"
1S/bid., A 87 =B Uo-1I. 19 Prolegomena, IV, 158 (Beck,6).

20 Critique of Pure Reason, A I¢ = B 141.
21 Cf. ibid., A 807 = B 8}5.
but by proof that they are principles without which the experience
in question, as the prius, would be unintelligible. If it be argued that
this would be a petitio principii, at least the answer could be returned
that in any argument one assumes the universe of discourse of the
premises. The argument would have made no headway against a critic
who stubbornly refused to concern himself with moral phenomena,
just as the Critique of Pure Reason would make no dent in the armor
of a silent skeptic who refused to assert or deny that 7 + 5 = 12.
"Moral experience" here would have the same status that "possible
experience" had in the first Critique, and both these realms must be
assumed at least problematically if epistemological or ethical analysis
is to have anything to work upon.
Yet this is not Kant's way here. The moral principle neither has nor
needs a deduction. To understand the peculiarity of the method of the
Critique of Practical Reason, we must return for a moment to the
Foundations of the Metaphysics of Morals.
In the Foundations, Kant's argument takes somewhat the form ex-
pected. The concepts of an absolutely good will and a universally
legislating will are synthetically related to each other. If necessarily
related, they must be related through some third pure cognition (like
pure intuition in the theoretical Critique). The third cognition is
furnished by the positive concept of freedom and is the idea of an
intelligible world as the archetype for the sensuous world insofar as
it is subject to our will. The intelligible or supersensuous world is
nothing else than the world of nature considered under the autonomy
of pure practical reason (43, 44 [I J2, I 33 ]). This seems to be as close
as Kant came to the notion of a "moral world" or a system of moral
experience as the justification of the moral principle.
In the Critique, by having said that the principle needs no deduction,
he apparently stands the argument of the Foundations on its head. He
uses the moral law, the fact of reason, as the prius to deduce something
else, namely, freedom, which is its ratio essendi.
The argument, in spite of Kant's denial that it is a deduction of the
moral law, is formally like the deduction of any other synthetic a pri-
ori principle in the first Critique. The concept of freedom is called
upon to playa role analogous to that of intuition. If there were an
intuition of freedom, the parallelism of the two arguments would be
perfect; but there is not. To show that an Idea of reason (freedom)
can be a substitute for intuition in a deduction requires us once again
to recall the abstract structure of the deduction in the first Critique,
for the notion we are examining will otherwise seem completely bizarre
and incredible.
The deduction of the categories requires that the objects of knowl-
edge to which the categories apply be given in possible experience. It
was necessary, therefore, to begin with a study of the way in which
they are given (intuition) and then to follow with a study of the
concepts and judgments by which the objects of intuition are thought.
In the second Critique, however, we cannot begin with intuitions but
must begin with principles which are the "given." Instead of holding
our concepts to intuitions, as theoretical knowledge requires, we must
sharply separate them and hold the concepts up to principles. In the
course of the elimination of the inruitively or empirically given (30
[II 8]), the paths of the two Critique's diverge. When we eliminate
the empirical material in the former, we are left with a pure intuitive
form; when we do so in the latter, we are left only with the mere
form of "ought." Since synthetic theoretical judgments are possible
only by reference to the pure form of intuition, which, as it were,
bridges the gap between the concepts of their subjects and predicates,
it would seem that no synthetic pure judgments would be possible in
moral philosophy at all. Yet they are possible and are actually given
as the "fact for pure reason."
"A good will [sc. a pure practical reason] has as its maxims only
universal laws" is claimed to be a synthetic a priori judgment, for "by
analysis of the concept of an absolutely good will, that property of
the maxim cannot be found."22 How, then, can it be confirmed? Not
by finding an intuition (which would be the obvious step in theoretical
philosophy) but by adducing some substitute for intuition. 23 It must
be purely intellectual, and therefore a moral sense will not suffice. It
must be a priori, for otherwise the synthetic judgment would not be
a priori. And it must, like intuition, have an independent warrant;
that is, it must not be just a product of the thought it is to justify.
This third thing, this substitute for intuition, is the Idea of freedom.
And freedom is not given at all! Only its Idea is given.
Here the argument takes a truly astonishing turn. Since the synthetic
a priori judgment that one expected to see Kant deduce needs no de-
duction and can have none, it is used for the deduction of the Idea
of freedom itself. Fortunately, however, the "deduction" is not a linear
inference; and, while the moral law serves as a ground for the deduc-
22 Foundations of the Metaphysics of MOTals, 447 (66).
23 Kant does not use the expression "substitute for intuition." But that freedom
has this function is made perfectly clear in the analogies he draws between the
methods of justifying a priori synthetic theoretical judgments (by means of in-
tuition) and that of justifying a priori practical judgments; the analogy is best drawn
in Opus postumum, XXI, 410--11.
tion of freedom, the concept of freedom is made to serve also as the
"credential" of the moral law. It is such a credential because of the
independent warrant that this concept has. Only this independent
warrant makes it possible for Kant to break out of the circle of using
freedom to establish the moral law and the moral law to establish
freedom. The independent warrant of the moral law, which is the
ratio cognoscendi of freedom, is the fact of pure reason. The inde-
pendent warrant of the concept of freedom, the ratio essendi of the
law, is found in its theoretical use:
This kind of credential for the moral law, namely, that it is itself demon-
strated [0 be the principle of the deduction of freedom as a causality of pure
reason, is a sufficient substitute for any a priori justification, since theoretical
reason had to assume at least the possibility of freedom in order to solve one
ofits own needs [48 (137); italics added).
In order to solve one of its own needs, a theoretical need. This is the
need to avoid an irresolvable contradition or incompleteness in theo-
retical reason itself: "[Speculative] reason showed freedom to be pos-
sible only in order that its supposed impossibility might not endanger
reason's very being and plunge it into an abyss of skepticism" (3 [87]).
This refers us back to a peculiarity in the solution of the third antin-
omy. Not only did that, like all the antinomies, have to be resolved
in order to avoid blatant self-contradiction; it had to be resolved by
showing that both thesis and antithesis were true. If the thesis were
false and the palm of victory given to the antithesis, the contradiction
would, of course, have been resolved, but theoretical reason's interest2'
in thinking a connected world would have been thwarted. Reasoning
demands a totality of conditions, and if a cause which is not also an
effect were shown to be impossible, this demand could not be met
in either the phenomenal or the noumenal world. The concept of
freedom could not then even be a valid regulative Idea of reason as
we push our thought from the conditioned back to the conditions and
conditions of conditions. Theoretical reason, in that event, would be
in as hopeless a plight as if the direct logical contradiction had not
been removed at all.
Theoretical reason thinks that freedom is possible not merely because
it can discover no valid logical evidence against it; it requires us to
think of freedom as possible. 25 But this alone does not show that we
know it to be real. Any evidence that would justify our claim to know
24 Critique of Pure Reason, A 474 = B 502-3.
25 Speculative reason must assume freedom only if it wishes to complete its func-
tion in speculation, and this is an arbitrary need, hypothetical and without the status
of law (Critique of Practical Reason, 5 [89]). Though a problematic concept, it is
absolutely indispensable to speculative reason (ibid·,7 [93]).
that it is real would fix "the keystone of [he whole arch of the system
of pure reason and even of speculative reason." This evidence is that
of pure practical reason, which substantiates transcendental freedom
"in the absolute sense needed by speculative reason in its use t)f the
concept of causality" (3 [87]). The concepts of pure practical reason,
such as freedom, are accordingly "not like the props and buttresses
which usually have to be put behind a hastily erected building, but are
rather true members making the structure of the system plain, and
let the concepts, which 'were previously thought of only in a prob-
lematic way, be clearly seen as real" (7 [92]).
The Critique of Practical Reason, therefore, does not begin with
a mere permit to occupy the "empty space" left by the first Critique,
with no indication of how it is to be filled. The space, empty for
knowledge, has a shape determined by the demands all around it for
knowledge of what can indeed be known. The vacant place, into
which the concept of natural causality cannot enter, is suited for only
one kind of tenant: "Pure practical reason now fills this vacant place
with a definite law of causality .... This is the moral law."26
What, we may ask at the end of this circuitous route, has been
gained for the principle of pure practical reason? The fundamental
principle, already asserted as a "fact," is not left a naked and isolated
assertion or an assertion surrounded by a closed, circular, and empty
system. It is supported in that it is precisely of the form required if
the dialectic of theoretical reason is not to be irresolvable.
The Critique of Pure Reason said that if an irresolvable conflict
bet\veen our knowledge of nature and our moral principles arose, it
would be the latter that would have to be surrendered. 27 But, recog-
nizing the essentiality of the concept of freedom, it also said that if
the antinomy could not be solved, both would be jeopardized. 28 Now,
because the Idea of freedom was required but not confirmed by the-
oretical reason, yet is confirmed practically through the fact of pure
reason, there is no danger of conflict. Rather, there is mutual support,
which far surpasses the evidential value of mere consistency. The
independent warrant of the concept of freedom-to wit, that it is
needed also by theoretical reason-makes it serve as a systematic cre-
dential for the reality of pure practical reason.
26 Critique of Practical Reason, 49 (139); Critique of Pure Reason, A 288-89 =

=
B 344-45; A 255 :::::: B 310; A 259 B 315; Foundations of the Metaphysics of Morals,
462 (83), etc. This is one of Kant's favorite metaphors, and there are many imagi-
native variations upon it.
27 Critique of Pure Reason, B xxix, A 536 = B 564.
28 Ibid., A 543 =B 571.
XI
Freedom
§ I. INTRODUCTION
Discussions of freedom are so frequent in Kant's works that the full
compass of the concept and its attendant problems cannot well be sur-
veyed in a running commentary on passages taken seriatim. In this
chapter, therefore, I shall attempt to examine Kant's ideas on freedom,
following an order which will be made clear in the second section.
Some of the difficulties in interpreting the Critique become more
manageable when we realize that its central doctrine of freedom of the
will involves two different concepts of freedom and two different con-
cepts of the will. Each of these pairs of concepts had a long and trou-
bled history before Kant, and each was used in his own earlier work.
His most important contribution in the second Critique was to show
that what is sound in each involves the other. But he did not help the
reader see that this is what he was doing, because he did not first estab-
lish, or even define, the two pairs of concepts and then bring them to-
gether; the unwary reader may not realize that Kant was using two sets
of concepts, because he shifts from one to the other without notice and
his language does not often indicate directly which he is using. Fur-
thermore, it is by no means certain that he was himself at all times
aware of the duality of his concepts and of the problem he was work-
mgon.
We must take the confusing fabric he wove, trace out the various
threads which form two patterns, and then, if at all, try to find a single
larger pattern composed of both. This program of work is summarized
in § 2 and occupies §§ 3-12, and it should be judged only in the light
of these later sections; §§ 13 and 14 discuss two points which are inde-
pendent of the major hypothesis of this chapter.
§ 2. TWO CO~CEPTIONS OF WILL AND FREEDOM
In the Critique of Practical Reason, the concept of will, which has

freedom as its attribute, is equivocal. Theories of freedom of the will
which do not seem to be consistent with one another are presented
side by side, but they are actually theories about different things and
17 6
Freedom 177
answer different questions. The Critique is the meeting place of two
different, but not explicitly distinguished, conceptions of the will and its
freedom, one of which comes chiefly from the Critique of Pure Reaso1l,
the other of which is a heritage from the Foundations. The two con-
ceptions are explicitly distinguished only 1n the later works, after their
interdependence has been shown in the Critique of Practical Reason.
\Ve must see the Critique, therefore, as a bridge where the tangled
paths of the earlier works converge and then for the first time clearly
separate on the other side.
From the Critique of Pure Reason there is inherited the concept of
freedom as spontaneity, the faculty of initiating a new causal series in
time. From the Foundations there is taken the concept of freedom as
autonomy, as lawgiving, and hence as independence from any pregiven
law. The two faculties are generally called by one name, "will," and
discussed under the name of one problem, that of "freedom of the
will." Only later did Kant give the reader any help in distinguishing
between them, when he "officially" called the former faculty Willkur 1
and the latter Wille. He had often before used these words, sometimes
to indicate a tacit distinction and sometimes apparently interchange-
ably; but at no time had he intentionally and consistently restricted
himself to one of the topics to the exclusion of the other. Even in the
Metaphysics of Morals, where the attempt is made to draw a good dis-
tinction and to avoid confusing them, he does not often succeed in
keeping discussion of one of them from interrupting discussion of the
other.
1 There is so much dispute about the exact meaning and proper translation of
Willkiir that I shall frequently leave it in Gennan. Born translated it as arbitrium,
following Kant's own suggestion, and thus distinguished it from voluntas. Bami
translated it as a1 bitre and Picavet as libre cboix; on the French words d. Khodoss,
Kant: La Philosophie pratique, Glossary, p. 141. Capra's translation is libero arbitrio.
Abbott used will or choice or elective will; Beck added free will and faculty of
choice to this imposing Jist. The tenn Willkiir was used by Wolff, Verniinftige Ge-
dancken von Gatt, der Welt und der See/e des Menscben, § 5l9: "Freedom of the
soul is the faculty of the soul to choose, through its own Willkiir, between two
equally possible things the one that pleases it more". the "Erstes Register" (Glossary)
of this work translates Willkiir as spontaneitas, and Baumeister, Pbilosophia defini-
tiva (1768), § 911, following Wolff, in tum defines spontaneitas: "est principium,
sese ad agendum detenninandi intrinsecum." A distinction suggesting that between
Wille and Willkiir is drawn by Wolff, Psycb%gia empirica, § 881: "The very act
of willing [actus volend;] is called volitio in opposition to voluntas, since the latter
denotes a faculty of mind or a power or possibility of eliciting that act." Similarly,
Baumgarten, Metapbysica, § 6<)0: "A rational act of desiring something is a volitio.
Therefore I have a faculty of willing [vo/endi] or a will [voluntas]." Baumeister,
op. cit., § 893: "Volido est ipse actus volendi ...." Volitio does not exist in classical
Latin in this sense.
178 Freedom
His formal definitions of will are drawn up as if will were a faculty
or combination of faculties which can be observed and of which we
have a direct awareness. Will is the faculty of determining our causal-
ity through a conception of rules (32 [120]), and, since for the deriva-
tion of an action from a rule or law reason is required, will is nothing
but practical reason. It is the relationship between understanding and
the faculty of desire (55 [145 D. The faculty which makes a rule of
reason the efficient cause of an action through which the object is to
be made real is will; the will is never determined directly by the ob-
ject or our conception of it, but always by a rule of reason (60 [r 5 I]).
That the will is, in this sense, free from direct sensuous necessitation is
an empirical fact. 2 This conception of will, as a faculty of desire
guided by a rule of reason taken as a maxim, later becomes more spe-
cifically Willkiir, the faculty of choosing an object which is left in-
completely determined by the maxim itself. It has, therefore, an incen-
tive (Triebfeder) for action in addition to the law, while Wille has no
Triebfeder. 3 Willkiir mayor may not be free, according to the kind of
law it puts into the maxim or the degree to which the maxim and not
the momentary representation of the object determines the action. It
does not give rise to laws but only to maxims, but it may, and when
moral it must, make laws its maxims. 4
In contrast to this, there is a concept of will not as the direct deter-
miner of action but as the lawgiver to the maxims which will determine
action. In this sense, Kant says, not quite accurately, that laws deter-
mine what ought to happen and maxims determine what does happen. 5
But the point made is sound enough: reason is necessary to the formu-
lation of a law, but a maxim determines behavior directly. In the for-
mulation of a law, we have to do with the real use and not with the
mere logical use of reason;6 by the "real use" is meant the establishment
of an a priori synthetic proposition, and by the "logical use" is meant
merely the inferring of actions from a rule. Pure practical reason has
2 Critique of Pure Reason, A 801 = B 830, there called WiIlkiir. It is illustrated,
but not defined, by reference to the faculty of choice known empirically (Meta-
physik der Sitten, VI, 216).
3 Vorarbeiten zur Tugendlehre, XXIII, 378.
4 Ibid., 383; Metaphysik der Sitten, VI, 216.
5 Foundationr of the Metaphysics of Morals, 420 n. (38 n.).
8 On the general distinction between the real and the logical use of reason, cf.
above, pp. 7, 75. An empirical practical reason is always merely logical in its use.
Kant draws the distinction in Critique of Pure Reason, A 800 = B 818, by calling the
logical use the "regulative use" of pure practical reason and contrasting it with the
"constitutive use" by which reason gives the law.
Freedom 179
nothing to do with the logical derivation of actions from given rules.
There is little or no verbal justification in calling such a pure practical
reason a "will" at all. But it makes perfectly good sense to speak of it
as determining the will (WillkiLr) and of its doing so freely, independ-
ently of sensuous conditions.
If practical reason determines the will (W iUkur), then we can say
that the latter is free in the psychological or comparative sense (9 6
[190]), even though there may be a natural law connecting the con-
ception of practical reason with the action and even if conceiving this
law is itself a naturally caused event in the inner life and even if the law
is a practical translation of a natural law. But to think that this is free-
dom in the sense needed by ethics seemed to Kant to be a "miserable
subterfuge." If all the causes of action are internal to man, not external,
if they are intellectual and not sensuous, and if the laws of their con-
nection with action are psychic and not physical (97 [190-<) I]), still
the corresponding concept of freedom, regarded as adequate by Kant's
predecessors and by himself in the Nova dilucidatio (1755), 7 is inade-
quate to the needs of ethics.
A new conception of freedom is called for. A law which is given for
moral obedience must not be a law of the connection of means to a
desired end and hence what I have called the "practical translation of
a natural law." It must be a law given by reason to a nature to be made
real, not one taken from a nature already realized. A will or Willkur
which can obey such a law must be independent of the mechanism of
nature, in which all connections are among phenomena, for a law
which demands absolute and not contingent obedience must be purely
formal, commanding by virtue of its form which is known by reason,
and not its phenomenal content. This independence of the mechanism
of nature is "freedom in the strictest sense" or transcendental freedom,
whose logical possibility was established in the first Critique (29 [116]).
But what of the origin of the law itself? Kant's most important dis-
covery is that the law is not a mere restriction on freedom but is itself
a product of freedom. Precisely this conception marks the chief ad-
vance of the second over the first Critique. This is the Copernican Rev-
olution in moral philosophy. The Critique of Pure Reason saw reason
as that which set bounds to a freedom which is itself without law,S but
it did not show how it does so; it established neither the provenance of
the law nor the mechanism by which it is effective.
The law is a product of the freedom of Wille as pure practical rea-
7 N O'IJa dilucidatio, Proposition ix (England, 2. 3I).

S Critique of Pure Reason, A 569 = B 597.
180 Freedom
son, not of lVillkur. (Though one must complain sometimes that Kant
writes Wille when Willkiir would be correct, I do not believe he ever
used Willkur to refer to Wille as pure practically legislative reason.)
We cannot say that the actions of Wille are free, because Wille does
not act. 9 It gives only a law for the submission of Willkur, which does
act. Yet it is free in that its decree follows from its own nature. It does
not mediate laws of nature to WiLIkur bent upon the satisfaction of
some arbitrary purpose; that is the function of practical reason in its
logical use. It does not counsel, but commands, and it commands as a
principal, not as an agent. Through submission to it, Willkur supple-
ments its negative freedom with a positive freedom which comes from
submission to its own idealized nature as purely rational will. Using a
political metaphor, as he so often did in speaking of the realms and ter-
ritories of the legislation of reason, Kant says it is autonomous, free in
itself,lO i.e., free in the positive sense. Willkur participates in this au-
tonomy to the degree that its negative freedom vis-a.-vis nature is exer-
cised in adherence to the law of pure practical reason. Pure practical
reason spontaneously creates an Idea of a natura archetypa,ll and Will-
kur, taking this as its object, can become an efficient cause of giving to
the world of nature the fOfm of such an intelligible world.
But we must never suppose that there are two faculties related to
each other in some external, coercive way. There is only one, but it has
prima facie two kinds of freedom, though one of them will eventually
be shown to be the perfection or logical form of the other. Willkur is
fully spontaneous only when its action is governed by a rule given by
pure practical reason, which is its legislative office. It is very hard to
avoid speaking as if there were two faculties without falling into the
opposite error of failing to distinguish between the two roles and the
two meanings of freedom. But unless we are to make Kant more diffi-
cult and obscure than he already is, we must be on our guard against
both an oversimple identification of function and a "two-faculty
theory."
Freedom in the positive sense is not so fraught with problems as that
in the negative sense. If we could presuppose freedom, Kant says, the
law would follow analytically from it (3 I [120]), and a will (in this
9 Metaphysik der Sitten, VI, 226.
10 Votarbeiten zur Tugendlebre, XXIII, 383. Its autonomy is contrasted with the
heteronomy of Willkiir; the freedom of Willkiir, which is independence of material
of desire, is freedom in the negative sense, while the legislation, or autonomy, of
Wille is freedom in the positive sense (Critique of Practical Reason, 33 [l2.z]).
11 Critique of Practical Reason, 43 (132); d. Critique of Pure Reason, A 548 =
B 576.
Freedom 181
sense) and a free will are identical. The freedom of this will is demon-
strated by the fact that there are absolute obligations, which could not
be so if all the laws of practical reason were empirical, i.e., if practical
reason were not pure, i.e., autonomous. Though the two conceptions
of will and freedom are inextricably bound together, the metaphysical
difficulties are found mostly on the side of freedom in the negative
sense, where it comes into contact with the natural necessity of the
phenomenal world.
§ 3. KANT'S RECAPITULATION OF THE ARGUMENT OF

THE FIRST "CRITIQUE"
The "Critical Examination of the Analytic of Pure Practical Reason"

and "Of the Right of Pure Reason to an Extension in Its Practical Use
Which Is Not Possible to It in Its Speculative Use" give a brief resume
of the contributions which the Critique of Pure Reason made to the
theory of freedom. We may well use these sections as a preliminary
guide through what Kant himself called the "thorny paths" of that
work.
As in the Prolegomena, where he credits Hume with having awak-
ened him from his "dogmatic slumber," Hume appears also here as the
chief instigator of Kant's critical labors. Hume, working on the princi-
ple that there can be no legitimate idea that does not arise from impres-
sions, failed to find an impression of necessary connection and thus
concluded that "the concept of cause [i.e., of necessary synthetic con-
nection of events in time] has been acquired surreptitiously and illegiti-
mately" or, more drastically, "it can never be acquired or certified be-
cause it demands a connection in itself void, chimerical, and untenable
before reason, a connection to which no object could ever correspond"
(51 [141]; d. 56 [14 6 ]).
Granting the validity of Hume's inference from its premises, Kant
finds the only error to lie in Hume's premise that objects of experience
are things in themselves (53 [143]). It perhaps sounds strange to at-
tribute this premise to Hume, who is generally thought of as a phenom-
enalist or subjectivist with reference to the objects of knowledge. But
if we understand Kant correctly, we see that this is indeed Hume's
premise, though it can be less startlingly formulated in non-Kantian
terminology. Kant means that Hume believed that the objects of
knowledge, though called "impressions," are known as they are and in
the order in which they are given, without our actively participating
in their generation and synthesis. But of things as they are (whether
they be impressions or Lockean substances), we cannot understand
why, if one is given, another must also be given. The concept of causal-
182 Freedom
ity arises neither from reason (53 [143]), for the connection between
a cause and an effect is not one whose denial is self-contradictory, nor
from experience, for connections known only in this way are not
know 1 to be necessary. If Hume is correct, it is a "bastard of the imag-
inatior' impregnated with experience."12
The first Critique, however, with its Copernican Revolution, saved
the necessary connection among events by showing that, though it is
necessary for knowledge of objects, it gives knowledge only of objects
of experience; for they are only phenomena synthesized by rules, not
things in themselves, of which we have no knowledge. The rules for
the synthesis are necessary in a way in which Hume's synthesis by cus-
tom is not. (Since the rule is not derived from knowledge of objects
[experience] but, on the contrary, makes it possible, the rule must have
its seat and origin in pure understanding.) Because 0'£ its purely intel-
lectual origin, it would appear that the concept and rule are applicable
also to things which are not objects of possible experience but lie be-
yond its boundaries (54 [144]). And this is correct, but with a very
important proviso that distinguishes Kant from his rationalistic prede-
cessors. If the concept were empirical in its origin, it could not even be
used to think of objects beyond experience; but it does not follow that,
since it is not empirical in origin, it can be used in order to know them.
The restriction of its use in thought falls away, but a Hume-like re-
striction on its use in knowing remains.
That which made causal knowledge of nature possible is precisely
that which is lacking in our thought about things beyond experience,
to wit, intuition. Intuition, givenness to sense, is indispensable if there
is to be any synthesis of concepts. All necessary connection between
pure concepts is analytical; but concepts can refer to intuitions, and
thus, through reference to a necessary third thing, can be brought into
necessary, but synthetic, relation to one another. Our intuition is sen-
suous, i.e., it presents things to us only as they affect and appear to us,
not as they are in themselves. But when we try to know things as they
are in themselves and apply the categories to them, the intuitive con-
dition is missing because we have no intellectual intuition (31 [120]).
This lack, however, does not prevent us from using the categories to
think of things as they are in themselves. Such thought may even be
valid, but it is not a cognition of things in themselves. If there is any
reason why this application of categories should be made, then the
purely non-empirical origin of the categories enables us to make the
application, which would not be possible if Hume had been right about
the origin of these concepts (55 [145]).
12 Prolegomena, IV, 157-58 (Beck, 5.6).
Freedom 183
At this point, Kant's recapitulation diverges in a very important way
from the actual argument of the Critique of Pure Reason and from his
more accurate estimate of it elsewhere in the Critique of Practical Rea-
5011.13 For he tells us here that it is not a theoretical purpose but only a
practical purpose that makes necessary the application of the concepts
beyond the limits of empirical knowledge. 14 For a full development of
this important difference, we must turn to the first Critique itself.
§ 4. FREEDOM AS A THEORETICAL IDEA
A category freed from the limitations to possible experience and
handed over to reason for use in a complete synthesis of all experience
is called an "Idea."1:> All synthesis of representations by the under-
standing is only partial; but reason demands a total and unconditioned
synthesis, i.e., reason's Ideas always have reference to a whole which is
unconditioned but which contains all conditions of its parts (107
[202]). The principle of reason is that if the conditioned is given, the
entire sum of its conditions must also be given, and therefore the abso-
lutely unconditioned, which is either this sum or some member of it,
must be given. In carrying Out the inquiry into the unconditioned,
there are as many Ideas as there are categories of conditionality. In this
section, however, we shall concern ourselves only with the category of
causality and its corresponding Idea.
In respect to the causal dependence of one thing upon another, we
apply the principle just enunciated that if the conditioned is given, the
unconditioned must likewise be given, else the conditioned, which can
occur only when the totality of its conditions is given, would not oc-
13 E.g., Critique of Practical Reason, 3,48 (87, 137) and as already elaborated in
chap. x, § 3.
14 Still, Kant says, not only is the theoretical purpose subordinate to the practical,
but all purposes of reason are ultimately practical (ibid., III [218]; Critique of Pure
=
Reason, A 816 B 844). But the point here is that a theoretical purpose is served
by a concept of reason, not that the theoretical purposes themselves are subordinate
to the practical.
13 Critique of Pure Reason, A 40<) = B 435-36. This is the theory of the "cosmo-
logical Ideas" and differs markedly from the more general theory (A 299 = B 356)
which finds the origin of the Ideas solely in the category of relation and in the fonns
of syllogism. The two theories exist side by side in the Critique, and Kemp Smith
(Commentary on Kant'r Critique of Pure Reason, p. 478) argues that the former is an
"early" theory. The footnote added in the second edition, B 395, suggests the pre-
eminence of the deduction of the Ideas from syllogism. In both theories, fortunately,
the Ideas of God, the soul, and freedom have a place, and, interestingly enough, the
place of the treatment of freedom is the same whichever of the two theories is used
-it arises in the third antinomy, corresponding to the category of causation and
to the hypothetical syllogism.
184 Freedom
cur. In searching for this unconditioned, or at least in forming a clear
concept of it, even though it is the concept of something that cannot
be given directly in any single experience, we have two alternatives be-
fore us: (I) We can suppose that the series of conditions is infinite, so
that no member of the series is unconditioned, while the series as a
whole is unconditioned. (2) We can suppose that the series is finite and
that there is an unconditioned member (the first member) in it. The
first is equivalent to supposing that natural causation, the relation of
one temporal event (cause) to another (effect), is the only kind of
causation; and, since no event in time can be found that is not also the
effect of an earlier one, we commit ourselves to the doctrine of an in-
finite series of events as the condition of any given event. The second
is equivalent to assuming that the natural causation is not the only kind,
since in an infinite series the unconditioned cannot be found. And this
is equivalent to assuming another kind of causation, a "causality of
freedom," i.e., a causality that is absolutely spontaneous, "whereby a
series of appearances, which proceeds in accordance with laws of na-
ture, begins of itseJf."16
Neither of these is an arbitrary supposition. In spite of their contra-
riety, each must be supposed, and thus theoretical reason necessarily
falls into an antinomy. The opposition between the two in the antinomy
is not just a curiosity of philosophy but is an inescapable opposition
between two not-to-be-gainsaid interests of mind. It is an opposition
that must be resolvable by the instrumentality of reason, since it is pro-
duced by reason; and it is not an opposition that philosophy could
calmly accept with resignation, because its own interests are so deeply
involved in each side and in their reconciliation. 17
Let us recount briefly the demonstration of each of the conflicting
theorems. The thesis is this: "Causality in accordance with the laws of
nature is not the only causality from which the appearances of the
world can one and all be derived. To explain these appearances it is
necessary to assume that there is another causality, that of freedom."
The proof is largely a repetition of the Aristotelian-Thomistic proof
of the impossibility of an infinite series of causes and hence of the ne-
cessity of a first cause, i.e., of a cause that is not itself an effect and
hence is free in the sense defined. That is, in a series of conditions and
conditions of conditions, there is never a first condition; but the law of
nature is that nothing occurs without a condition that is a priori suffi-
cient. Hence, granting that something does occur, the law of nature is
self-contradictory when taken in unlimited generality. Therefore, nat-
16 Critique of Pure Reason, A 446 ~ B 474.
17 Ibid., A,.so = B 508; cf. Critique of Practical Reason, 3 (88).
Freedom 18S
ural causation, causation under the law of nature, is not the only kind,
etc.
The antithesis is as follows: "There is no freedom; everything in the
world takes place solely in accordance with the laws of nature." The
proof is this: If there is a spontaneous cause or an absolute beginning in
the natural causal series, the later members of the series are independent
of the earlier, and thereby the "unity of experience," which depends
upon the lawfulness of events in time, is made impossible, and thereby
no criterion of empirical truth or objectivity is possible. Is
The thesis is ascribed to dogmatists (e.g., Plato) and the antithesis to
empiricists (e.g., Epicurus) .19 In general, the moral and religious inter-
.
est is on the side of the thesis, which is the interest of pure reason;20 the
interest of science in its unending search for causes within experience
is in the antithesis. The latter interest is not, indeed, incompatible with
the moral interest except when empiricism itself becomes dogmatic,
that is, when it is not content with the unending investigation of phe-
nomena but extends its claims and methods into metaphysics and there-
by causes "irreparable injury ... to the practical interests of reason."21
The former practical interest is not necessarily incompatible with the
scientific, but there is the danger that, while "supplying excellent prac-
tical principles, it permits reason to indulge in ideal explanations of nat-
ural appearances ... to the neglect of physical investigation."!!2 There
is, however, one clear advantage on the side of the thesis: the antitheses
of all the antinomies "render the completion of the edifice of knowl-
edge quite impossible .... The architectonic interest of reason, the de-
mand not for empirical but for pure a priori unity of reason, forms a
natural recommendation for the assertion of the thesis."23 The interest
of speculative, and not merely of practical, reason therefore lies on the
side of freedom.24 But if it were not for the antinomy, metaphysics
18 Critique of Pure Reason, A 451 = B 479, remark on the antithesis.
19 Ibid., A 466 =B 494, A 471 =B 499.
20 This is not, however, the whole story. The real interest of morality is not on
the side of dogmatism but on that of freedom, which the dogmatists defended (d.
Critique of Practical Reason, 146 ff. [144 ff.J and below, chap. xiii, § 7). Let there be
no guilt by association in moral philosophy.
22/bid., A 471 = B 500. 23 Ibid., A 474 = B 502-3.
24 This is the sense in which Kant says that reason in its theoretical employment
requires the concept of freedom (Critique of Practical Reason, 3,7,48 [87,93, 137]).
One might say that reason's first interest is the removal of the antinomy, but Kant
does not call this its "interest" but says it is merely the condition of there being any
reaSon (ibid., 120 [116]).
186 Freedom
would be only an extension of physics, i.e., the dogmatization of the
antithesis. The antinomy which prevents this is, therefore, the "most
fortunate perplexity" into which pure reason can fa11.25
§ 5. RESOLUTION OF THE THIRD ANTINOMY
The principle which generated the antinomy was as follows: If the
conditioned is given, the entire series of its conditions is given; the con-
ditioned is given, therefore the unconditioned is given. But, Kant says,
there is a fallacy here, in that the major premise takes "conditioned" in
the sense of a pure category, while the minor takes it in the empirical
sense of a concept of understanding applied to appearances. But, in ap-
pearance, not all conditions are given as phenomena; they are not ge-
geben, but aufgegebe1l-not presented as a gift but assigned as a task.
The conditioned within the phenomena is the prescribed object of in-
quiry. The antithesis is correct when "conditioned" is taken in the
sense of a temporal condition of a phenomenon, for there is no first
phenomenon. On the other hand, the thesis is correct when "condi-
tioned" is taken in the sense of the major premise, which does not re-
strict it to the tempora1. 26 If "conditioned" were taken in the same
sense in the two premises, the antinomy would be irresolvable, i.e., we
could show that the conditions would have to form both an infinite
homogeneous series and a finite heterogeneous series (95 [ 18 9]). But
now we can see that the antithesis not only can, but mUst, be true of
phenomena in time, while the thesis can and must be true of things in
themselves in their relation to phenomena. Kant then says that the an-
tinomy and its resolution afford indirect proof of the ideality of ap-
pearances, the Transcendental Aesthetic having given the direct proof.27
The Idea of a totality of conditions, therefore, is not a constitutive
Idea in experience for which we could find a corresponding experience
25 Ibid., 107 (203); cf. Critique of Pure Reason, A 464 == B 492.
26 Critique of Pure Reason, A 499 f. = B 527 f.
=
27 Ibid., A 506 B 534. Kant wrote Garve (September 21, 1798 ; XII, 257) that the
discovery of the antinomy was what led to the conclusion that space and time were
only forms of appearance (cf. also Critique of Practical Reason, 10 7 [203]). It should
be noted. however, that the proof of the antithesis presupposes the validity of the
Kandan doctrine of the phenomenality of nature, and especially that of the second
Analogy of Experience. The proof of the thesis, on the contrary, is a typical ration-
alistic argument that does not depend upon any exclusively Kantian doctrines. The
antinomy alone cannot, therefore, be used as a premise for the phenomenality of the
natural order, but it may well have suggested it to Kant. The fact that the proof of
the thesis is independent of the Kantian doctrine of space and time is of considerable
importance in evaluating the use that Kant made. or could have made, of his doctrine
of freedom (cf. below, ~ 8).
Freedom 187
of an object. It is a regulative Idea or a "rule prescribing a regress in
the series of given appearances, and forbidding [reason] to bring the
regress to a close by treating anything it may arrive at [in experience]
as absolutely unconditioned."28 This we may call the "Regulative Idea
of Causal Mechanism," and by it we are enjoined from introducing
free causes into our study of nature but are directed to find the condi-
tion of every phenomenon in still other phenomena. If, however, we
regard the phenomena as things in themselves, this regulative Idea must
be regarded as constitutive, and we run into the antinomy in such a
way that neither freedom 110r natural causation can be saved.:!9
Kant next attempts to show that if there is any reason to assert the
existence of free causality, this does not involve any contradiction with
natural mechanism. He does so by asserting that the effects of free cau-
sality would be in the phenomenal series and thus in the order of natu-
ral mechanism. Every appearance is the appearance of a reality; its ap-
pearance is connected with other appearances under the causal law of
nature and is predictable with certainty. But in its relation to that which
is not appearance, i.e., the noumenon, and not a member of the tem-
poral series, it is an effect of a freely acting cause, where freedom is
defined as the power of being a cause without being an effect. Hence,
in principle, every event in the world is a product of both natural and
free causation. We do not understand it in its latter relation; all our
knowledge is knowledge of the connections of phenomena among
themselves. We cannot apply the category of causation to things in
themselves so as to have knowledge of them; but we can apply the
category by anal ogy 30 to the relation of noumena to phenomena and
think of the former as a free cause of the latter without infringing on
the principle of mechanical causation so far as our possible knowledge
is concerned.
In doing this, however, we must be careful not to claim more than
we can perform. Kant hedges in this right to think freedom. We .;:an-
not know the free causality of a thing in itself;31 we cannot think of
the thing in itsdf as acting at various times and places (for it is not in
28 Critique of Pure Reason, A 509 = B 537; d. also Critique of Practical Reason,
48 (13 8).
29 Critique of Pure Reason, A 543 = B 57J. Causality in nature cannot: be saved
because we cannot detennine a synrhetic necessary connection except insofar as it
is prescribed by theoretical reason, and theoretical reason can prescribe only for the
realm of phenomena. ("Theoretical reason" is here used in a broad sense, and in-
cludes also the understanding and intuition.)
30 Ibid., B 431-32.
3I/bid., A 540 = B 568; Critique of Practical Reason, 133. 134 (13J f.).
188 Freedom
space and time);32 we cannot use our concept of freedom as a supple-
ment to our concept of natural causation when we are ignorant of how
the latter should be applied in a particular case; and we cannot expect
the course of nature to be any different when we admit free causes
from what it would be, were they denied. 33 All that Kant says he has
accomplished is to show that "causality through freedom is at least not
incompatible with nature" and that his intention was not to establish
either the (real) possibility or the reality of freedom. 34 Freedom, there-
fore, is only a problematic concept. .
§ 6. PRACTICAL FREEDOM
At this point it is well to consider a peculiarity of the proof of the
thesis of the third antinomy. I mentioned earlier that Kant's argument
is an adaptation of the Aristotelian argument for the existence of a first
cause. What, then, is contributed by it to the establishment of freedom?
Kant says, in the Observation on the Thesis:
The necessity of a first beginning, due to freedom, of a series of appear-
ances we have demonstrated only in so far as it is required to make an origin
of the world conceivable.... But since the power to spontaneously begin a
series in time is thereby proved, it is now also permissible for us to admit
within the course of the world different series as capable in their causality
of beginning of themselves, and so to attribute to their substances a power of
acting from freedom. 3 :>
One may well say, then, that the demonstration, if successful, accom-
plishes entirely too much. It seems to justify the concept of freedom,
if anywhere, then everywhere. Every phenomenon has the two dimen-
sions of relations, one to previous phenomena, one to noumena. The
second dimension or relation is not what is meant by freedom in any
interesting sense, because it is indiscriminately universal. Freedom as a
universal predicate is of no interest. The concept of noumenal causa-
tion is empty unless we could know either (a) the substance in ques-
tion or (b) a noumenallaw of its activity. In the case of a stone, for
instance, we know neither, and therefore we have to be content-and
are easily contented-with knowledge of the phenomenal relations of
phenomenal stones. In the case of one's own self, however, though we
do not know ourselves except as appearances, we do know the law of
32 Critique of Pure Reason, A 541= B s6c); Critique of Practical Reason, 99 (192).
33 Critique of Pure Reason, A 550 = B 578; cf. also Idea for a Universal History,
Introduction (VIII, 17-18).
34 Critique of Pure Reason, A 448 =B 476; cf. Critique of Practical Reason, 3 (87).
35 Critique of Pure Reason, A 449-50 = B 477-78.
Freedom 189
the noumenal action, for that is nothing other than the moral law. Kant
appeals to noumenal causation, therefore, only where there is some rea-
son to go beyond the phenomenal causation, and these are found only
in human volition; even when morality is not in question, there is this
occasion for use of the concept of noumenal causation:
In lifeless, or merely animal nature, we find no ground for thinking that
any faculty is conditioned otherwise than in a merely sensible [mechanical,
phenomenal] manner. Man, however, who knows all the rest of nature solely
through the senses, knows himself also through pure apperception, and this
indeed in acts and inner determinations which he cannot regard as impres-
sions of the senses. 36
Now the "interest of pure reason," which Kant said at the beginning
is on the side of the thesis, comes into its own. The practical concept
of freedom, he says, is based on the transcendental Idea of freedom, and
without it cannot stand. 37 Freedom in the practical sense is the inde-
pendence of Willkur from coercion through sensuous impulses. In man,
there i:; a power of self-determination which is independent of nature.
Practical freedom presupposes that, although something has not hap-
pened, it ought to have happened, i.e., that the cause in the field of ap-
pearance was not so determining that it necessarily excluded a causal-
ity of our will. Though everything that we might will might be due,
indirectly, to sensuous impulse, the impulses and all phenomena under
the law of nature cannot give rise to the concept of "ought," which
entails a concept of free causation and not of natural causation. "No
matter how many natural grounds or how many sensuous impulses may
impel me to will, they can never give rise to the 'ougbt,' but only to a
willing which, while very far from being necessary, is always condi-
tioned; and the 'ought' pronounced by reason confronts such willing
with a limit and an encl,-nay more, forbids or authorizes it."3S The
thought of "ought" is impossible if all laws are natural laws;39 the
thought of "ought" implies the thought of a free "can," and if pure
reason is actually effective in the control of conduct, then there is free
causation in the transcendental as well as in the practical sense. There-
by, also, transcendental freedom ceases to be an all-embracing and
hence empty concept.
Kant is insistent, in both Critique's, on the necessity of transcenden-
36 Ibid .• A 546 = B 574; d. Critique of Practical Reason, 49.99.100 (139. [93. 194>-
37 Critique of Pure Reason, A 534 == B 561.
3S/bid.• A 548 = B 576.
=
39/bid.• A 547 B 575: "We cannot say of anything in nature that it ought to be
other than what it actually is in all its temporal relations."
190 Freedom
tal freedom if practical freedom is to be real. 40 There is an empirical
freedom, which Kant calls "comparative," which is not absolute or
transcendental spontaneity but one which is empirically found in some
of our acts, e.g., those in which we exercise what common sense calls
self-control, but not in others, e.g., in acts springing from rage or lust.
In comparative freedom, the determining causes are internal to the
agent and not compulsion from without. This is the freedom of a pro-
jectile, which, once started on its trajectory, follows it necessarily but
from its internal momentum. Even if the internal causes are intellectu-
ally thought reasons and not impulses, this freedom is still only the
freedom of a well-run machine like a dock. 41 Empiricists (and here
Kant is no doubt thinking of writers like Ulrich and Schulz) think this
comparative concept of freedom is an adequate condition for eth-
ics; but this is a mistake. The concept of freedom is the "stumbling
block of all empiricists" (7 [93]), whose own concept of freedom is a
"wretched subterfuge" (96 [189]).
Immediately this question arises: Does not Kant's doctrine, as over
against that of the empiricists, infringe on the mechanism of nature?
And if it does not, is it one bit more adequate to ethics than the em-
piricists' concept is? Kant says freedom in his concept does not in-
40 In asserting that inward rational detennination, which can be empirically ex-

hibited, is not sufficient to morality, Kant is disagreeing with almost all his predeces-
sors and contemporaries and with his own earlier view (Nova dilucidatio) that
such acts are free because they express the activity and not the passiVity of reason.
Section i of the Canon of Pure Reason is often regarded as having assened that
practical freedom is independent of transcendental freedom and as therefore being
in this respect out of harmony with the "critical teaching" and hence as "early."
Though the Canon was probably one of the earliest pans of the Critique of Pure
Reason to be written, I beg to differ with this interpretation. The context should
be sufficiently explanatory of the apparent discrepancy between Kant's statements
for us not to have to suppose that he changed his mind on this central point during
the course of writing the book. By a "canon" Kant means "the sum-total of the
a priori principles of the correct employment of cenain faculties of knowledge"
(A 796 = B 824). The last paragraph of Section i does not say that practical freedom
could stand if transcendental freedom were not real; it says merely that this question
does not concern us in the practical field or in a canon where we "demand of
reason nothing but the rule of conduct" and do not require that this rule be shown
to be irreducible to a law of nature. "This problem does not come within the prov-
ince of reason in its practical employment." This does not say that it is of no con-
cern to a critical examination of practical reason; it says only that it is a theoretical,
not a practical, problem. This is repeated even in the Foundations of the Metaphysics
of Morals, 448, n. 456 (67, n. 76), and in the review of Schulz's Sittenlehre.
41 Critique of Practical Reason, (}f>-en, 98, 101 (H)lr-I, 195). The allusion to Leibniz'
automaton spirituale has reference to Theodicy, H 51,4°3. Kant perhaps knew of
Vaucanson's automata through La Mettrie, L'Ho'Mme machine.
Freedom 19 1
fringe on the laws of nature, and yet it is a basis for moral imputation.
To show this, he draws a distinction between the intelligible character,
the noumenal or transcendental subject, which is a causa noumenon,
and the empirical character, which is only its appearance in space and
time.4 2 The former is free, the latter is under the law of nature. The
former is thought (not known) as an unchanging substance in which
nothing happens (because it is not in time); the latter is its unrolling in
time under causal conditions of nature. We infer the nature of the for-
mer from that of the latter; we punish or reward the latter in the belief
that it is a manifestation of the freedom of the former. The freedom of
the empirical character is at first to be understood only negatively, as
not being necessitated by things in nature. The freedom of the former
is positive, for it originates a series of events in the world which would
not have happened, had the intelligible character been different. 43 Ev-
ery event in human conduct can be seen, therefore, in two ways: as a
necessary consequence of preceding events and as directly determined
by the intelligible character.
§ 7. CRITICISM OF THIS DOCTRINE
This is hard doctrine, and Kant admits that judgments under it "seem
at first glance to conflict with equity" (99 [193]). But is it only "at
first glance"? Why should a man repent the wrong he has done (98
[ 192])? How Can we hold a man responsible for his actions and yet
say, at the same time, that "before ever they have happened, they are
one and all predetermined in the empirical character?"44 That if we
knew all the empirical facts and the natural laws of their connection so
42 Critique of Practical Reason, 97 (Igo); Critique of Pure Reason, A 545 == B 573;
Reflexion 5608. Critique of Practical Reason, 1)8, 100 (191, (94), however, speaks of
our consciousness of having freely created our empirical character, and in this re-
spect it seems hardly to fall under natural law or to fit into the reality-appearance
rubric.
43 Critique of Pure Reason, A 554 ~ B 582, A 556:::: B 584; d. the very similar
passage in Critique of Practical Reason, 99 (193)' Hence it must not be simply as-
sumed that the intelligible-empirical distinction coincides with that between Wille
and WilIkiir. The intelligible character acts timelessly, manifesting its acts in the
conduct of the empirical character. The Wille does not act but legislates for (is a
law for) the Willkiir. Kant also speaks of personality (intelligible personality) as
an archetype the respect for which constitutes the motive for the Willkur (ibid.,
87 [180]). The freedom of Wille is distinguished from that of W illkur as libertas
noumenon from libertaJ phaenomel1on; yet the latter is not to be defined empirically
(d. Metapbysik der Sitten, VI, 216; and Opus postumum, XXI, 470), for this gives
only a comparative or psychological sense of freedom. Kant's statements are so
cryptic that it is hard to know whether he is entirely consistent or not.
44 Critique of Pure Reason, A 553 :::: B 581.
192 Freedom
that "his future conduct could be predicted with as great a certainty as
the occurrence of an eclipse," we could nevertheless still assert that the
man is free" (99 (193))? This requires us to accept a secularized ver-
sion of the classical theological quandary of man's freedom and God's
foresight, and it seems no more intelligible in structure and equitable
in outcome than that hoary mystery.
If by "freedom" we mean noumenal causation and assert that we
know no noumena, then there is no justifiable way, in the study of phe-
nomena, to decide that it is permissible in application to some but not
others of them to use the concept of freedom. The uniformity of hu-
man actions is, in principle, as great as that of the solar system; there
is no reason to regard statements about the freedom of the former as
having any empirical consequences. If the possession of noumenal free-
dom makes a difference to the uniformity of nature, then there is no
uniformity; if it does not, to call it "freedom" is a vain pretension.
There seems to me to be only one way out of the dilemma. There are
faint suggestions of it in two widely separated works of Kant, but he
never fully developed them. He does not seem to have felt the paradox
in his own views that all his critics and most of his disciples felt.
The first suggestion is this: Instead of regarding the world as consist-
ing of two realms-a phenomenal under one set of laws and a noumenal
under another-can we not think of one world under two aspects, the
aspects to be defined methodologically with reference to the purposes
we have in holding these two perspectives on a common world, and
not ontologically? Kant intimates such a two-aspect theory, instead of
the more commonly known "two-world theory," in his contrast be-
tween the observing, theoretical attitude and the acting, practical atti-
tude 45 and in his statement that "supersensuous nature is nothing else
than nature under the autonomy of pure practical reason" (43 [132))'
The other suggestion, necessary to the full development of the pre-
ceding one, is his conclusion in the third Critique 46 that the distinction
between natural and moral law is dependent on the peculiar nature of
our understanding. We can read these passages as a suggestion to re-
gard the two kinds of laws as co-ordinate, not one as subordinate to the
other in constitutive authority in experience. The only evidence we
have that Kant entertained such a view is in § 70 of the Critique of
Judgment, where he speaks of the thesis of the complete mechanical
determination of nature, in its antinomic relation to that of teleological
causation, as a regulative Idea even in respect to nature. It would be
easy to extend this to the antinomic relation between freedom and nat-
45 Critique of Pure Reason, A 550 ::::: B 578; cf. chap. iii, above.
46 Critique of Judy;ment, § § 70 and 76.
Freedom 193
ural causation. The Idea is expressed as a maxim: "All production of
material things and their forms must be judged to be possible according
to merely mechanical laws."
Had Kant said, in the solution to the third antinomy in the first Cri-
tique, that the tfue meaning of the antithesis is that it is a maxim for pro-
cedure and not a constitutive principle of nature, then we would have
had two maxims: "Always (in science) search for mechanical causes
and allow no non-natural causes to enter into the explanation of natural
phenomena," and "Always (in ethics) act as if the maxim of the will
were a sufficient determining ground of the conduct to be executed or
judged." Neither of these is a declarative a priori statement; they tell
us, rather, what we must do in order to be a spectator or an actof, but
one cannot be both at the same time and with respect to the same item
of conductY
It may well be that we would, under these rules, sometimes hold a
man responsible for actions that he could not have avoided doing, for
human freedom is far more limited, I think, than Kant held it to be. In
this event, we are simply unjust judges. It may well be that we give an
abstract, schematic causal explanation to some event which did not, in
fact (though we shall never possibly know it), have a sufficient natural
cause; in this event, we are dogmatic (unavoidably so) in our scientific
work. But the alternative to sometimes being unjust is that of always
being unjust when we hold a man responsible for any of his actions
when, if Kant is correct, none of them could have been left undone in
the course of nature and history as constituted by natural law.
The solution proposed here involves reading back into the Transcen-
dental Analytic of the first Critique some of the conclusions of the
Transcendental Dialectic. Specifically, it requires that the sharp distinc-
tion between constitutive category and regulative Idea be given up,
that even the categories be regarded as devices for the regulation of
experience and not as structures necessarily given in a fixed constitu-
tion of our experience of nature, and that the Analogies of Experience,
which Kant called "regulative" in a very modest sense,48 be reinter-
preted as regulative in the full sense of the Dialectic. 49
47 Cf. above, chap. iii. The second of the maxims need not be interpreted in the
manner of the fictionalists, i.e., on the assumption that the "as if' intrOduces a con-
trary-to-fact condition. There is no more need to assume that the latter is any more
fictional than the fonner or that the former is constitutive while the latter is merely
regulative.
48 Critique of Pure Reason, A J 79 = B 222.
40 Such an interpretation of the Transcendental Analytic, in fact, can be recom-
mended on purely epistemological grounds, though it no doubt distorts Kant's own
estimate of the relation of the functions of understanding and of reason.
194 Freedom
If we undertake to make these revisions, we are allowed to regard
the moral or practical realm as a perspective of the realm of experience,
which, through other regulative ideas (categories), is seen as the realm
of nature. We will then no longer have to think of science as dealing
with appearances (in some ontoiogically pejorative sense) and morals
as dealing with noumena (in some epistemologically pejorative sense).
Both can keep their a priori structures intact; both will claim to cover
all the relevant experience;5o but each will be carried out for different
purposes and only occasionally will come into conflict with each other
-a conflict to be settled by moral scrutiny sensitive to all the myriad
facts of life-instead of invariably doing so, as, according to the ortho-
dox Kantian theory, they must. That one of these realms is a limiting
case of the other, that the categories of one of them can be derived
from those of the other, is a view, classically developed by Fichte and
strongly represented in both idealism and pragmatism, which goes far
beyond anything Kant said or probably would accept, though the germ
of it is present in the third Critique.
§ 8. FREEDOM AS AN ACTOR-CONCEPT; SPONTANEITY
That we have a right and even find it necessary to make use of the
concept of freedom in our actions, regardless of how the antinomy is
resolved and whether freedom is compatible with natural causation or
not-in short, that the actor in a moral situation must act as if he were
free and thereby shoulder all the responsibilities that he would have if
it were theoretically proved that he is free-is shown by Kant in several
ways. It is not shown, however, by appealing to what Kant calls com-
parative freedom, i.e., empirically observed relative independence from
outward stimuli and inward impulse through the exercise of intelligent
foresight. It is shown only through an elaboration of the inward phe-
nomenon of choice and a regression upon its conditions.
In the second edition of the Critique of Pure Reason there emerges
a conception of one's own existence which is not present, at least so
explicitly, in the first edition. It is the conception that we have a direct
experience of our own spontaneous activity as a substance.1I1 This ex-
50 Cf. Critique of Judgment, Introduction, II: The realm of theoretical reason and
that of practical reason do not limit each other in their legislation, but they per-
petually do so in the world of sense.
III Critique of Pure Reason. B 157-58 n.: "I cannot detennine [i.e., know categor-
ially] my existence as that of a self-acting being; all that I can do is represent to
myself the spontaneity of my thought. that is, of the determination; and my existence
is still determinable only sensibly, that is, as the existence of an appearance. Sut it
is owing to this spontaneity that I entitle myself an intelligence." Similarly, Founda-
tions of the Metaphysics of Morals, 451 (71); Critique of Practical Reason, 56 (J47).
Freedom 195
perience is neither a sensuous intuition:;:? nor an abstract thought. Kant
never tells us what its epistemic character is, but that it occurs is a fact
to which the epistemology of the Critique of Pure Reason does scant
justice.
Such self-awareness of one's spontaneity and the attendant presup-
position of freedom is found even in the act of theoretical thinking,
though we can use this conception of the self as a thinking substance in
no theoretical explanation of the inner life. 53
But the clearest evidence of one's spontaneous freedom-sometimes
said to be the only evidence 54-is one's awareness of obligation, which
is a necessitation of a wholly different kind from all natural necessita-
tion and which produces a unique kind of feeling (92 [185]). It has
this revelatory function when, in fact, the moral law is not obeyed
but only acknowledged. 55 As wholly different from natural determina-
tion, it cannot be understood in theoretical terms. The freedom that
can be understood, he says in his review of Ulrich, is of no use in eth-
52 When Kant, in the precritical period, believed that there was an intellectual in-
tuition, it was to this faculty that he ascribed consciousness of the self (see Reflex-
ionen 4228, 4336, 600 I). With the denial of intellectual intuition to man, no so-to-
speak official position is taken on knowledge of one's own spontaneity, but the ex-
perience is not denied. On the whole question, with a collection of sources, d.
Heimsoeth, Studien zur Philosophie Immanuel Kants, pp. Z45 ff., and Ingeborg
Heidemann, Spontaneitiit und Zeitlichkeit, pp. 173 ff.
113 The theoretical uselessness of this conception is the main point in the Paralo-
gisms of the first Critique; even there the practical importance of the concept is
recognized and preserved (B 431-32). Cf. Foundations of the Metaphysics of
Morals, 448,451-52 (67,70-71), on the awareness of spontaneity. In Besprechung
von Schulz's Sittenlehre (1785), VIII, 13, Kant remarks that the determinist in meta-
physics claims freedom for himself in the conduct of his own thinking and a for-
tiori in his action. Cf. also Beantwortung der Frage: 1st es eine Erfahrung dass
wir den ken? (Cassirer ed., IV, 519-10), where it is denied that the awareness of
thinking is an Erfahrung, and it is called merely "transcendental consciousness";
similar! y Critique of Pure Reason, A 117 n.
114 Critique of Pure Reason, B xxxiii, B 430-F; Critique of Practical Reason, 42
(13 1 ) •
55 The pair of examples in the Critique of Practical Reason, 30 (1I8), brings out
this contrast between natural necessitation, which is effective, and a moral necessita-
tion, which mayor may not be effective. The first example, concerning the man
who says his lust is uncontrollable, resembles one in Rousseau: "Let us suppose the
maddest of men, the man who has his senses least under control; let him see the
preparations for his death; let him realize that he will cenainly die in torment a
quarter of an hour later; not only would he from that time forward be able to resist
temptation, he would even find it easy to do so" (Emile, Book IV ["Everyman" ed.,
pp. 289-9°)). The second example anticipates one to be used in the Methodology
(155-5 6 [154])'
196 Freedom
ics, and the freedom needed in ethics cannot be understood. 56 All that
we can do is to comprehend its incomprehensibility::'· and accept it as
guaranteed by the fact of pure reason, the moral law which reveals it
to US. 58
With this argument, that the consciousness of the moral law itself
proves the reality of freedom,59 we are brought to a somewhat differ-
ent conception of freedom, which we must now examine.
§ 9. FREEDOM AS SUPREME LEGISLATION; AUTONOMY
Let us suppose that the Critique of Pure Reason has proved that it is
not self-contradictory to say that there is in man a causa noumenon or
a faculty of initiating a new causal series in the world. This faculty
would be called "freedom in the negative sense"60 or freedom from
nature. But freedom is not lawless caprice, any more than it is lawless
in its effects found in nature and historv.
-'
But what limits freedom and
renders it lawful? The Critique of Pure Reason says that reason places
a limit on freedom,61 but how it does so is left unexplored and unex-
plained.
To answer this question is one purpose of the Foundations of the
Metaphysics of Morals. The problem is to determine a law that the will
can obey without losing its freedom through that very act of obeisance
to law. The will can "obey" laws of nature in the sense of intentionally
using them, in the form of means-end statements, in seeking men's goals
in nature. But to be free even from the importunities of the desire for
happiness-a goal given by man's natural existence-requires that the
law of reason be not such a law borrowed from nature and hence, in-
directly, leaving action under the domination of nature. The law must
be given by reason. Just as the will (Willkur) considered as a faculty
in man may be free in the negative sense of spontaneous activity, prac-
tical reason is spontaneou:; in the sense of giving law instead of subject-
56 Ak., VIII, 458.
58 Cf. above, chap. x, and Critique of Practical Reason, 31 (no).
59 It is likewise taken as a route to the consciousness of self, independent of em-
pirical conditions, and thus serves to supplement the failures of rational psychology
as a doctrine of the soul, though without making any contribution to theoretical
knowledge (cf. Critique of Pure Reason, B 43<>-31; Critique of Practical Reason,
105-6 [zoo-zol]).
60 Ibid., A 553 = B 581; Critique of Practical Reason, n 5. 6, 8 (z9. 33 [116-17,
Ill]).
61 Critique of Pure Reason, A 569 = B 597. It exercises this control even in issuing
hypothetical imperatives, i.e., in its logical as well as in its real use (ibid., A 54B =
B 576).
Freedom 197
ing itself to an alien law. It gives a law to Willkur which it has freely
legislated-legislated by the necessity of its own nature. The faculty of
lawgiving is will in the sense in which pure practical reason is will, and
its legislation is "freedom in the positive sense," or autonomy. In the
choice of this language, Kant, as so often, thinks in political metaphors
and analogies, and especially in terms of Rousseau's political theory.
This meaning of freedom is quite different from the one we have
been examining heretofore. Here it is not a question of whether the
acting subject can initiate a new causal series, but it concerns the
source of the law which the subject follows in such an initiation. The
consciousness of the moral law, and not obedience to it, is the evidence
of freedom in the positive sense (47 [136]). Freedom and uncondi-
tional law reciprocally imply each other; the latter is the ratio cogno-
scendi of the former, the former the ratio essendi of the latter (4 n.
[88 n.]). Freedom analytically pertains to reason. 82 The moral law
reveals to me a "marvelous faculty," a purely intellectual principle
for the determination of my existence. 83
Here we may appropriately notice one of Kant's best-known argu-
ments for freedom, in which he says that a being that cannot act except
under the Idea of freedom is really free in a practical respect and is
obligated by the laws which follow from that Idea, regardless of
whether we can prove theoretically that he is free or not. 64 Which
freedom is warranted by this argument? Not the freedom of spon-
taneous action, for consciousness of this could be illusory. Everyone,
including the fatalist or determinist, acknowledges, with Mephistoph-
eles, that "Ou glaubst zu schieben"; yet the fatalist or determinist adds,
"aber du wirst geschoben." The argument applies to freedom in the
sense of obligation-creation, not obligation-execution. For we are not
directly obligated to do anything, to initiate a specific causal series
in time. We are obligated directly and unexceptionably only to take
up and act upon a certain maxim and to reject the maxims and actions
which are incompatible with it, and even this thought shows a legisla-
tion of reason which is not derived from nature, since nature presents
no "ought." To think one's self obligated to do something is to be
really obligated to do something (though we may not be, in fact,
obligated to do precisely that which we thought we were obligated
62 Critique of Practical Reason, 31 (120): the moral law "would be analytic if the
freedom of the will were presupposed." Cf. Foundations of the Metaphysics of
Morals, 420 n., 426 (37 n., 44).
83 Critique of Pure Reason, A 431-32'
84 Foundations of the Metaphysics of Morals, 448 n. (67 n.).
198 Freedor.n
to do) .65 The thought of obligation is, therefore, self-guaranteeing
in a way in which the feeling of spontaneity is not self-guaranteeing.
Hence one can say that freedom in the sense of legislation is real, even
though he admits that in the entire history of mankind no free action
may ever have been performed.
§ 10. THE SYNTHESIS OF THE TWO CONCEPTIONS
Kant regrettably did not single out these two conceptions and then
formally show their relationship to each other. He deals with them
together, without clearly showing, at any moment, which it is that
he is talking about. Had he followed a truly synthetic method, he
would have set up the two concepts and then related them. As it is,
however, we have to find the relationship between them by an analysis
of the complex mixture. Fortunately, it is not difficult to do so, and
Kant himself does it, all but explicitly, in § 6 (29-30 [116-17]). I
shall now paraphrase this passage, using the two conceptions of will
and of freedom, in order to show the synthetic or bridging function
of this important section.
Granted that human beings have a free will (Willkur)66 in the
practical and phenomenological sense, as a faculty of spontaneously
initiating a new causal series in nature, what is the law of its action?
A law must have a content, which may be taken from nature, and a
form, the form of universality, which can be prescribed only by rea-
son. 61 If the Willkur obeys a law because of its content, it can be free
in the practical sense, for will shows its freedom even in obeisance to
hypothetical imperatives and, indeed, in actions which are evil. But if it
is to be determined necessarily, i.e., irrespective of the desires which
are the material cause of its willing, as it is if there is duty, it must be
determined not by the content but by the form of law. The form of
law is universality, fitness for universal legislation. In giving such a
law, reason is not responding to the promptings of nature. It is there-
fore a spontaneous legislator and is free.
Hence a free, i.e., spontaneous, WillkuT, when it is good, is deter~
mined by a free, i.e., autonomous, Wille, or pure practical reason,
which gives it a law. 68 It can obey only this law without jeopardy to
85 A prima facie obligation can be abrogated only by another prima facie obliga-
tion or real obligation. It cannot be simply denied, leaving a kind of moral vacuum.
66 Kant should have used the word Willkur in this section, but he did not.
61 This is true of both natural and moral laws (cf. Critique of Practical Reason,
16 [1141>.
68 Foundations of the Metaphysics of Morals, 455 (74): The good will is a law
for the bad will as a member of the world of sense. Unfortunately, Kant used Wille
for each in this passage.
Freedom 199
its freedom. Indeed, it gains in freedom, by now being an autonomous
as well as a spontaneous will. Thus is added to the negative concept
of freedom (spontaneous independence of foreign laws) the positive
concept of freedom (autonomous self-legislation). There is a loss only
of lawless freedom: "Where the moral law speaks there is no longer,
objectively, a free choice [Wahl] as regards what is to be done."68
The Willkur can obey the law of Wille without losing its own
freedom only because they are not two faculties externally related to
each other. They are two aspects of practical reason, differing as the
legislative and executive functions. The former function binds the
latter; the former is the pure form of the latter. It gives a law that the
latter 'Would obey, were it a holy will, i.e., if the Willkur fully realized
its potentialities of purity. The law is found not by seeking something
outside the Willkur but by a regression upon the conditions of its
full freedom, conditions that are not actually realized in the natural
man. Thus, Kant explicitly says, we find Wille by a regression upon
the conditions of Willkur.70
That the will of man-because of the fact that it is not a simple
notion and has two distinguishable aspects-can be both obligation-
creating and obligation-executing is one of the most dramatic theses
in Kant's philosophy, as dramatic as, and analogous to, the Copernican
Revolution in his theoretical philosophy. So long as the origin of law
was ascribed to nature or experience or God, no matter how rigoristic
and certain it might be, obedience to it had to be heteronomous and
restrictive of freedom. That the moral law was rational and was to be
discovered a priori and was in some way binding on man quite apart
from the rewards that were believed to accompany obedience to it-
all this was well-known doctrine in the German academic philosophy
of Kant's time. But, because of the abstract ontology of perfection,
no specific law seemed derivable from the putative sources of the moral
law, and obedience to it was always in danger of being explained only
eudaemonistically. Because Kant discovered the law in the concept of
rational will (not abstract perfection or "will in general") 71 and was
able to derive its formula from the concept of its source and because
this source was an idealization of the will as a faculty of spontaneity
in man, Kant did not have to seek any outward motivation for obedi-
ence to it. Rational personality as lawgiving expressed an "is" which
is ipso facto an "ought" for partially rational beings; the law and its
69 Critique of Judgment, ~5 (V, 110 [45]).

70 Metapbysik der Sitten, VI, 111; cf. Critique of Practical Reason, 30 (u8).
100 Freedom
imperative and the conditions necessary to obedience to it have a com-
mon source which Kant's predecessors never found-indeed, never even
sought. They were c011$equently never able to convert their formalistic
ethics into a practical doctrine without jeopardizing or destroying the
alleged formality of the principle.
The doctrine of autonomy was anticipated only by Rousseau, for
only Rousseau saw the essential connection between law and freedom,
while others in the eighteenth century saw law only as a restriction
on freedom. Though Rousseau worked out their essential connection
only in politics and had his doctrine there adopted with little change
by Kant, the doctrine of self-government through law by free citizens
is deepened into a moral and metaphysical doctrine by Kant. With
Rousseau, Kant can then say that obedience to a law that one has
himself prescribed is the only real freedom. 72
Obedience and prescription are and remain, however, quite different
functions. Can our human will be autonomous as well as spontaneous?
The question can be answered affirmatively on several grounds. First,
it cannot be spontaneous without being autonomous,73 unless it is to be
lawless and, accordingly, useless to morals as well as incompatible with
science. Second, reasoning in the opposite direction, it can be spon-
taneous because, under autonomy, it ought to be.74 If the awareness
of our duty placed impossible demands upon us, as it would if our
WillkuT were not potentially free, then the thought of duty would be
illusory. How do we know that it is not? Only because of the positive
evidence of the fact of pure reason 75 and the resolution of the third
antinomy-the latter making it possible, the former making it actual.
Third, Kant believes that the empirical nature of man is such that
arguments from human nature which would show autonomy to be
impossible can be shown to be false.
72 "Obedience to a law which one has prescribed to himself is freedom" (Rousseau,
The Social Contrllct, Book I, chap. viii).
73 Even the evil will is autonomous in the sense that the person who does moral
evil freely incorporates an incentive into his maxim and makes it (what it is not in
itself) a rule in accordance with which he will conduct himself (Religion within
the Limits of Reason Alone, VII, 24 [Greene and Hudson, 19]; d. below, S 11).
7. One of Kant's most famous "statements"-"Thou canst because thou shouldst"
-does not exist in his writings in this neat fonn (ct. David Baumgardt, "Legendary
Quotations and the Lack of References," jou1"11al of tbe History of Ideas, VII
[H)461. 99""102, and L. W. Kahn, "Legendary Quotations," ibid., VIII [1947],
u6). But statements that express this inference less succinctly abound, e.g., Critique
of Practical Reason, 30 (118-19); Critique of Pure Reason, A 807 = B 835; UbeT
den Gemeinspruch • . •• VIn. 187; Metaphysik deT Sitten, VI, 380; Streit deT
Fakultiiten, VII, 43-44; Vorlesungen fiber Metaphysik (Kowalewski cd.), p. 600;
Opus postummn, XXI. 16.
TIS This is discussed in chap. i of the Analytic and in chap. xii of this commentary.
Freedom 201
It has often been objected that there are two paradoxes in Kant's
ethics: (I) Kant is, or is reputed to be, "individualistic" in his ethics,
while the moral person is, for him, only an abstraction that is meant
in some way to dominate and restrict the inditidual person. (2) Kant's
ethics is so autonomous that the social or universalistic aspect of
morality is left unfounded because it is an outward restriction on the
freedom he insists upon. 76
Quite apart from the fact that these two alleged paradoxes cancel
each other out, it is possible to show that neither is valid and that both
arise from a misinterpretation of the point we are now discussing.
Each, if fully expanded, would entail the supposition that in Kant's
doctrine there are two wills, and which paradox is drawn depends
upon the critic's belief concerning which is the more important to
Kant (which happens always to be the one that is less important to
the critic-for such are the ways of philosophical polemics). (I) If the
Wille or pure practical reason as an abstract epistemic or moral concept
is emphasized, then the individual human W illkuT is restricted and is
not free. (2) If the legislation is thought of as issuing from an individual
WillkuT, it is not possible to see how the laws issued will meet the
requirements of social universality and harmony.77
But there are not two wills. There is one will with its formal condi-
tion, which is universally valid reason, and its material condition arising
from the specificity of its involvement in the world at particular times
and places. And the two paradoxes are not paradoxes of Kant's ethics
so much as manifestations of the human predicament in which we find
in ourselves individualized manifestations of universal mandates and
injunctions. Man is the only being in the world that not only is a
manifestation of some universals but ought to be an instance of others;
he is an individual that gives no valid laws to others that he does not
lay upon himself, that gives no privileges to himself that he does not
allow to others. Had the Kantian teaching missed what is true in these
paradoxes, it would have been less true and less responsive to the
paradoxical aspect of human life itself, in which man finds himself
neither a brute nor a god, neither a mere particular nor a mere
universal.
But, of all the misinterpretations of Kant, perhaps none is more ob-
76 It is noteworthy that the same "paradoxes" have been found in Rousseau. Was
Rousseau the father of an impersonal fascism or of democratic individualism run
wild? An affinnative answer to either of these questions shows a misunderstanding
of Rousseau analogous to the misunderstanding of Kant explored here.
77 Cf. George Santayana, Egotism in German Philosophy (1916; zd ed., New York,
1940 ),Pp· 5~51.
101 Freedom
viously wrong or more widely accepted than the accusation that Kant
represents something vaguely called "Prussian philosophy," in which
blind obedience to law is so esteemed as an absolute virtue that neither
political nor moral freedom is allowed to be more than a name. The
accusation is too ridiculous to deserve serious consideration on its own
merits and should be refuted only because of its widespread accept-
ance. 78 But the refutation is easy: it not only separates two wills within
the individual but puts each in a different person, attributing rights to
one and duties to the other. 79 It forgets that all moral discipline is, for
Kant, self-discipline and that self-discipline, while not the whole of
morality, is a necessary condition of it. It forgets that Kant taught that
all just government is self-government. The same man, by virtue of
the same faculty in its positive and negative use, is both subject and
legislator in the realm of ends and in the just state.
§ I I. SUMMARY
Let us pause and summarize the results of our argument, results
anticipated in § 2 above, and relate it to some other concepts in Kant's
ethics.
We have distinguished two meanings of will: Wille as practical
reason, the legislative function, and WillkUr, as the executive faculty
of man.
The freedom of the former is autonomy; it gives a law to WillkuT.
This law is determined by the nature of Wille and not by anything
else in the world, including human nature or the will of God. The
moral law is a synthetic a priori statement of what a WillkuT would
necessarily do if it were exclusively rational; it is a law or imperative
of duty for a Willkur which does not do by nature what the law
78 John Dewey is not responsible for the extremes of this view, for he acknowl·
edged Kant's "individualism" and held only that such "Prussianism" arose because
"the two worlds of Kant were too far away from each other" and could be con-
nected only through the remnants of the idealistic theory of history and the state
(cf. German Philosophy and Politics [New York, 19[51, p. 1%2, and the defense of
Kant by Julius Ebbinghaus, "Interpretation and Misinterpretation of the Categorical
Imperative," Philosophical QUaTterly. IV [[954], 97-108. On the latter see also K.
Kolenda, "Professor Ebbinghaus' Interpretation of the Categorical Imperative,"
"Philosophical Quarterly, V [[955],74-77).
79 Perpetual Peace, VIII, 348 n. (Beck l [-I 2 n.): "With regard to the most sub·
lime reason in the world that I can think of with the exception of God (say, the
great Aeon). when I do my duty in my post as he does in his, there is no reason,
under the law of equality, why obedience to duty should fall only to me and the
right to command only to him." God is excepted only because of the inapplicability
of the concept of duty to a holy Willi but the same law applies both to man and
to God.
Freedom l03
requires. In addition to its real use in discovering or formulating the

moral law, practical reason also has a merely logical use in the deriva-
tion of rules of actions either from the moral law or, in the case of
prudence, from human desires and the laws of nature.
The freedom of the latter is spontaneity, the faculty of initiating a
causal series in nature. It can exercise this (negative) freedom in one
of two ways: (I) It can take the law of pure practical reason as the
limiting condition on its maxims, out of respect for law or the rational
personality that decrees it. It is then a good will, acting out of duty.
If it did so without any internal obstacles, but by its own nature, it
would be a holy will; struggling, as it does, against the sensuous im-
pulses, it is at most a virtuous will. (2) It can take some other principle
(maxim) posited or adopted by practical reason, in its logical use, as
its formal principle. If it does this, there are two possibilities: (a) It
may take a principle which is not opposed to the legislation of reason,
and then it is a legal and sometimes a prudent will. (b) It may take a
principle that is opposed to the law of pure practical reason, and then
it is an evil will. In either case, it is a free will.
The Willkur, however, can fail to exercise its freedom or realize its
potentiality of being free in a negative sense. Then it gives way to the
importunities of sense and is a will in name only, really being an
arbitrium brutum. 8o The pure practical reason, on the other hand,
cannot fail to be free 8l and autonomous, however little effective it may
be in its control of Willkur. However depraved WillkuT may be, it
still hears the "heavenly voice" of pure practical reason, so that even
the most hardened criminal trembles before its tribunal (35, 80 [[24,
17 2 ]).
§ 12. MORAL EVIL
Consider the following dilemma which has embarrassed many de-
fenders of Kant, who have accepted the conclusion that a good will is
a free will. If there is evil, it must be a result of a failure to be free.
Therefore, either there is no moral evil, all evil being natural and
therefore not imputable to human responsibility, or goodness of will
is not equivalent to moral freedom. It is therefore concluded that
when Kant asserted the existence of radical evil in human nature, in
Religion within the Limits of Reason Alone, he was diverging from
his own critical doctrine. Depending upon the reader's standpoint, this
80 Critique of Pure Reason, A 534 =B ;6z.
81 Foundations of the Metaphysics of Morals, 448 (67). Even theoretical reason
is free in this sense, for "we cannot conceive of a reason which consciously responds
to a bidding from the outside with respect to its judgments."
104 Freedom
is regarded either as a sign of Kant's robust sense of the hard fact of
moral evil, which he admitted no matter what it did to the neatness
of his system, or as a regrettable failure to remain consistent with his
own high teaching.
But the alleged contrariety between the teachings of the Religion
and of the Critique disappears in the light of our analysis. Kant speaks
of the reality of moral evil in the immediate context of the most ex-
treme statement of the causal predictability of all human action (100
[193]), in a place where freedom to do both moral and immoral actions
seems to be put to its most extreme test. He says, of an evil man, that
his "actions by the uniformity of conduct exhibit a natural connec-
tion. But the latter does not render the vicious quality of the will
necessary, for this quality is the consequence of freely assumed evil
and unchangeable principles. This fact makes it only the more ob-
jectionable and culpable."
This is consistent with the fuller treatment in the Religion, and both
are consistent with the summary given above in § I I. The Religion tells
us that by the "nature of man" is understood the subjective ground
of decision (of Willkur) which is independent of determination by
the impulses, for otherwise there would be neither moral good nor
moral evil. The subjective ground must itself be an act (Aktus) of
freedom; it cannot be other than a "rule, which the Willkur makes
for itself for the use of its freedom." The evil lies, therefore, not in a
failure under the conditions of nature to exercise freedom but in a
maxim that is freely adopted and is in opposition to the maxim (not
of the WillkuT, as are the impulses) of the pure practical reason. 82
This propensity to take some other maxim than the moral law as the
governing principle of action can no more be explained than the
opposing disposition to take the moral law. Both are there as predis-
positions to, not as causally determinative of, free acts which can in
no way be explained.
So long as there is a WillkuT responsive to the determinations of
desire or other natural impulses, there is the constant danger of think-
ing that anything that seduces the lVillkuT from obedience to the
moral law ipso facto destroys its freedom, so that the resulting evil is
not morally imputable. "But the freedom of Willkiir," he says, "has
the peculiarity that it can be determined to action through no incen-
tive except insofar as man has taken this incentive up into his maxim,
i.e., made it a general rule by which he will act; only in this way can
any incentive, whatever it may be, coexist with the absolute spontaneity
82 Religion within the Limits of Reason Alone, VI, 21 (Greene and Hudson. I,),
Freedom 105
of WilikuT, i.e., with freedom."83 The disposition to evil (or to good)

is itself, as Kant says in the Critique, something freely chosen. 84
It is very easy to see how the clear separation of Wille from WillkuT
provides a secure place for imputable evil, so much so that one could
construct, from § I I, the Kantian theory of evil in its main outlines,
even if we did not have the evidence from the Religion. When WillkuT,
in the exercise of its spontaneity, acts on maxims which are incompati-
ble with the moral law, it does moral evil. It is only the freedom of the
will as pure practical reason that is analytically connected with morali-
ty, but this will does nothing but issue orders which mayor may not
be obeyed. It neither sins nor does virtuous actions,85 because it does
not act at all. Only because it was believed that there was one function
of will and one kind of freedom was it erroneously thought that the
Critique identified free and moral acts. When Kant spoke of moral
evil, therefore, it was natural that he should be thought to have fallen
into serious inconsistency.
§ 13. FREEDOM AND CREATION
The relation of freedom to metaphysical or theological necessity is
perhaps a less pressing one nowadays than that of its relation to physical
necessity. But this was not so obviously the case in Kant's day, and
some of his earliest discussions of freedom, in the Nova dilucidatio,
concerned the ancient problem of reconciling human freedom with
God's foresight and predestination. The classical problem of predesti-
nation in its relation to morals and its bearing upon the problem of evil
fascinated Kant as a young man, and he occasionally returned to it in
83lbid., 14 (19). So far, therefore, from denying that imputable evil exists, the
doctrine holds that all actions are imputable; for, so long as there is a Willkiir at all.
it is acting by maxims which are either moral or immoral. Hence there are no
morally neutral acts (ibid., 14 [10]). When a Willkiir without choice surrenders to
the forces of blind impulse, we should not call it WillkUr at all, since we do not, in
these circumstances, impute responsibility to the person.
84lbid., 1) (ZI); Critique of Practical Reason, 100 (194).
85 Kant says "freedom of choice [Wahl] of maxims of action" is "absolute spon-
taneity" or libertas noumenon, and from it. "one can never adduce a reason why
anything against the law occurs," while freedom of WillkUr, subjective freedom or
libertas pbaenomenon, is freedom in respect to choice [Wahl] of the lawful or the
unlawful (Opus postumum, XXI, 470). If one adds to the statement about absolute
spontaneity that one can never adduce a reason why anything happens in accord-
ance with the law, since such transcendental freedom has no explanatory value, we
have here the distinction between what Henry Sidgwick has called "neutral free-
dom" and "moral freedom" ("The Kantian Conception of Free Will," Mind, XIII
[1888],405- 11 ) •
206 Freedom
his later works. so The special problem of the Critique concerns the
possibility of freedom in a world in which the realm of nature and
the realm of grace are connected in the Leibnizian manner, a manner
dangerously close to the Kantian. If freedom and God's omniscience
are not compatible in Leibniz, then they would not seem to be com-
patible in Kant either. It is this, rather than the Calvinistic form of the
puzzle, that occasions Kant's discussion here.
He argues that if God is the cause of men's actions through the
original creation of man's substance, then only comparative freedom
exists, and morality is impossible. But though Kant is willing to grant
the premise of the creation of noumena, he denies that the inference
is valid. It is invalid because the syllogism contains four terms, "causa-
tion" and "creation" not being equivalent. It would be valid only if
things in themselves were temporal, and if they are temporal, then
even an attempt like that of Mendelssohn87 to make God non-temporal
will not save freedom. For, whatever might be the nature of God, if
things in themselves are in time, God's creation of them is a temporal
act and restricted to the conditions of time, which conflicts both with
God's theological and metaphysical predicates and with the initiation
of a new causal series in time.
But since things in themselves and, a fortiori, God are not in space
and time, the relation between God and things in themselves cannot
be a causal relation, though we have to think of it by analogy with
causation. The causal relation holds only among phenomena. If it were
ontoiogically real and God's creation were itself causal, then "man
would be a marionette or an automaton" and only Spinozism would
remain. 88 But because Kant has, on other grounds, denied the tem-
porality of things in themselves, the relation of God to them and,
a fortiori, to their temporal appearances,89 is not one of cause.
Kant's view may be summarized by saying not only that a syllogism
whose major premise contains the term "creation" and whose minor
86 E.g., Critique of Practical Reason, 100-103 (194-96); Religion within the Limits
of Reason Alone, VI, 144 (Greene and Hudson, 135); Metapbysik der Sitten, VI,
180 n.; Perpetual Peace, VIII, 361 n. (Beck, 24 n.); incidentally in Critique of Pure
Reason, A 206 = B lSI-P. The relation of historical detenninism to freedom in his-
tory is discussed in Idea for a Universal History and in Streit der Fakultiiten, VII,
41. The relation between the problem of freedom with respect to God's foresight
and that of freedom in a Laplacean universe is briefly but suggestively handled in
T. D. Weldon's Kant's Critique of Pure Reason (ld ed.; 1958), pp. 110-11.
87 In his Morgenstunden; cf. Critique of Practical Reason, 101 (195).
88 Critique of Practical Reason, 101-1 (1g6); on Spinozism in its relation to fanati-
cism, see What Is Orientation in Thinking? VIII, 143 (Beck,301).
89 Cf. Perpetual Peace, VIII, 3610. (Beck, 24-25 n.).
Freedom 207
contains "causation" is formally invalid but that if "X creates Y" and
"The appearances of Y cause the appearances of Z," we can say
nothing of the relation of X to the appearances of Z.
But I do not see how this meets the issue. For X is also the creator
of Z. Since the appearances of Y and Z could not be different from
what they are without a different noumenon underlying each90 and
since each is created by X, it is not easy to see how X can be let off
so easily and exculpated for the appearances of Z.
Kant himself seems to have felt that his argument here is not entirely
convincing and to have put it forward only faute de mieux. The last
paragraph of this section (103 [197]) can be read only as evidence of
such dissatisfaction.91 As a puzzle from theology and classical ration-
alism, I suspect that the problem did not interest Kant at this time as
much as it had earlier. The relation of God to the world is not an
object of theoretical knowledge but only one of practical belief. This
particular aspect of the relation does not have positive moral con-
sequences which make it necessary for Kant to take a stand on it, and
such puzzles are dismissed in the Strife of the Faculties.
§ 14. FREEDOM AS A POSTULATE
Freedom is the only one of the Ideas of pure reason that we can
know. It is proved by the apodi~tic law of practical reason. All other
Ideas gain reality (i.e., are known to have objects) only through their
connection with it. 9 :! These Ideas are those of God and the immortality
of the soul. They are called "postulates" because they are dependent
upon the need of human reason to establish the possibility of the
highest good; they are not directly necessary to morality or revealed in
the fact of pure reason, and they cannot be theoretically justified.
It is a little surprising, therefore, to find Kant calling freedom in
the positive sense also a postulate of pure practical reason. 93 This is to
be explained in three ways. First, there is the wide latitude that Kant
90 =
Critique of Pure Reason, A 556 B 584.
91 One of the first commentators on the Critique (Brastberger, Untersuchungen
tiber Kants Kritik der praktischen Vernrmft, p. 156), who was almost sentimentally
sympathetic to and uncritical in his acclaim of Kant's book on most points, rejects
Kant's solution to this problem or, rather, confesses that he cannot find what Kant's
solution is, Crusius, who in so many matters concerning freedom anticipated Kant,
declared the problem to be insoluble.
92 Critique of Practical Reason, 4 (88). The postulates "give objective reality to
the Ideas of speculative reason in general" (ibid., 131 [229]). In this sense freedom
may be, and is, considered a postulate.
93 Ibid" 131 (ZZ9). Elsewhere he says that it is freedom in the negative sense which
is a postulate (cf. Zum ewigen Frieden in der PhiJosophie, VIII, 418).
208 Freedom
permitted himself in using the word "postulate." The variety of its
meanings will be discussed below in chapter xiii.
Second, there is the specific relation of freedom to the summum
bonum rather than to the moral law itself. With reference to the
highest good as an object of a necessary human need, whatever it is
necessary to assume in order to achieve it may be considered a pos-
tulate. Though the Analytic has given better reasons to assert freedom
than that it is necessary to the highest good, when the highest good
is the topic of discussion, whatever we have to assume in order to
show its real possibility is to that extent a postulate. We cannot con-
clude that the doctrine of the Analytic is in any way abandoned or
even modified because what was there "deduced" is here "postulated."
The doctrine of freedom as a postulate is, in fact, earlier than the
deduction of freedom. In the Dialectic of the second Critique we hear
the last echo of a doctrine developed in the Critique of Pure Reason
which put God, freedom, and immortality on exactly the same level.94
Third, there is a difference in the meaning of freedom as a condition
of the moral law and as a condition of the summum bonum. Freedom
in the latter sense is an object of faith, not a scibile;95 it is the faith
( Vertrauen) in the achievability of the summum bonum, i.e., the belief
in virtue (Glaube an die Tugend) as adequate to achieve the highest
good. In this sense, freedom is not mere autonomy but "autarchy of
will."96
94 Critique of Pure Reason, B 395 n. (not in A). Albert Schweitzer (Die Religions-
philosopbie Kants, chap. ii and p. 134) argued that the original plan of the second
Critique was that it would continue the doctrine of the cosmological Ideas of the
first, and in the order of their derivation in the Dialectic but that Kant then dis-
covered the special prerogative of the Idea of freedom. This caused him to take that
Idea from the Dialectic as planned and put it into the Analytic.
95 Critique of Judgment, § 91.
96 Fortschritte der Metaphysik, XX, 295.
XII
The ("~esthetic"
of Pure Practical Reason; Com-
mentary on Analytic, Chapter III; Part
of Dialectic; and Methodology
§ I. THE PSYCHOLOGICAL AND THE ETHICAL PROBLEM
Kant has repeatedly assened the necessity of distinguishing between
the subjective, psychological factors in conduct, which can be learned
through observation, and the objective a priori law or norm of morali-
ty, which can be discovered only by pure reason. The Foundations of
the Metaphysics of Morals tried to show that this distinction is im-
plicit in ordinary moral consciousness, however unsophisticated it may
be and however little self-conscious discipline of reason may be found
in it. In that work Kant attempted to give this distinction a definitive
form so that the a priori principle could be discerned in its simple
purity. The Critique of Practical Reason, in following a synthetic or
progressive method, does not begin with the consciousness of duty in
order to show that it has features which cannot be exhaustively ex-
plained in empirical psychological terms, but with a delineation of the
formal and material elements in all willing. Only at the end of chap-
ter i of the Analytic is the statement of the supreme principle of pure
practical reason attained. It is not based upon any empirical data given
by or about human nature; it is a law for rational beings generally.
It is, moreover, a norm or a regulative principle for men and for any
other rational beings in the world, if there are any, who are not pure
rational beings. That is, it is a priori descriptive of a fully rational
being and a regulative concept for a partially rational being; hence it
is meant to be applied to human beings.
"Apply to" may mean two different things. It may mean that this
law is applied as a norm in judging the conduct of men, being a
standard of comparison between what is and what ought to be. And it
may mean that one renders obedience to it, so that it is the supreme
factor in the determination of choice and the conduct issuing from it.
The former of these meanings is explored chiefly in the Metaphysics
209
210 The "Aesthetic" of Pure Practical Reason
of Morals, where it lends itself to a full casuistical treatment. Except
incidentally, it is only the second of the two which is studied in the
Critique of Practical Reason. There are two separate but closely con-
nected discussions of it, one in the chapter "The Incentives of Pure
Practical Reason" and the other in the Methodology. Some corollary
problems are discussed in the chapter on the resolution of the antinomy
arising from the Idea of the summum bonum.
The problem of these chapters and sections is a singularly difficult
one for Kant, because in the rest of his treatment of the moral prin-
ciple he is writing about rational beings in general, and the path from
a priori knowledge to a posteriori application is always (not for Kant
alone) a hard one; it is one that must be trodden with exceeding care,
for Kant realized here and in his other works that a mistake in the
analysis of the a posteriori factors could cast doubt on the validity
of the a priori features of experience that were his chief concern. In
none of his works is this transition made in such a manner as to give
complete satisfaction to Kant's critics, and their reluctance to accept
his psychology has favored a skepticism toward his theses in pure
philosophy. It is, nevertheless, a passage that must be made if pure
reason is to be practical.
Man is a rational being, but he is also a being of need, impulse, and
sense. He is, or may be, a free agent, but he is also a part of the
mechanism of nature. His self is noumenal, but he also has an empirical
character that sets him off from all other rational beings and all other
empirical objects. How, then, can we describe his relation, actual and
ideal, to the pure moral law? How can the moral law or his con-
sciousness of it be a determining factor in his conduct, as it is if he can
properly claim to be a moral agent?
Before venturing upon Kant's actual answer to this question, let us
try to become clear as to what would constitute, in his terms, a satis-
factory answer to it. Kant says that this question is equivalent to that
of how the will can be free, and to that question, like all others con-
cerning fundamental powers or faculties, no answer can be given. We
cannot find any higher principle that will throw light on such facts
as the sensuous character of our intuition, the practicality of reason,
the fact that our reason is discursive rather than intuitive, and the
freedom of the will. But we can go this far: we can explain their
inexplicability and show that attempts to explain them away are not
successful. Though it displaces the inexplicable mystery of man by
only one step, the Critique Of Practical Reason does attempt an ex-
planation, in psychological terms, of how the knowledge of the moral
The "Aesthetic" of Pure Practical Reason 2 It
law can be effective in the determination of conduct. "We shall not

have to show a priori why the moral law supplies an incentive," because
we cannot, "but rather what it effects (or, better, must effect) in the
mind so far as it is an incentive" (72 [165]). It is essential that this
mystery be removed from the phenomenological surface, as it were,
for the thing is so puzzling that doubts of its reality can have the
actual effect of reducing the effectiveness of this incentive; and if this
incentive is reduced or destroyed, there is only legality-contingent
accordance with the requirements of moral law-but no morality.
And, as practical moral teachers, we must know what steps to take to
render this incentive effective and sufficient. This is possible only if
we know the mechanism of its effectiveness and in this way remove
doubts that it can be effective-doubts that led Hume, for example, to
regard reason as the slave of the passions instead of their master, as
Kant believed it to be.
Kant has another reason for giving an extensive analysis of the moral
disposition and intention, within which the consciousness of the law
is effective. The proper object of moral judgment is not the law, but
man. A man is a moral agent only if he meets two demands, viz., his
actions must accord with the law, and they must be undertaken be-
cause the law demands them. That is, an action is moral because it is
regarded as duty. We can never understand the subjective phenomena
of morality or have a firm basis for either moral imputation or moral
decision if we do not understand the way in which the moral law
enters into consciousness and makes a clear demand upon our allegiance.
If we cannot understand this, we leave the door open to those who
would give simpler explanations that have the advantage of being
readily comprehended but the disadvantage of being theoretically
wrong and damaging to morality itself.1
The justification for calling the discussion of these topics "the aes-
thetic" is found in the analogies that Kant points out between the
analytical parts of the first two Critique's.2 The division of the Cri-
tique's is parallel, but the parts are in reverse order. The first begins
with the sensuous givenness of representations, proceeds to study their
synthesis into the consciousness of objects, and ascends, finally, to the
regulative principles by which consciousness of objects is synthesized
into experience of a world. The second begins with the principles
1 This practical function of the philosopher, in whose "subtle investigations the
public takes no interest," is alluded to in Critique of Practical Reason, JoB (1°3),
and in the last sentence of the book. 163 (Z62).
2/bid., 16,90 (IOZ, 183); cf. above, chap. iv, § 7.
(chap. i), applies them to objects of practical reason (chap. ii), and,
finally, comes to "their application to the subject and its sensuous
faculty" in the chapter on "the aesthetic of pure practical reason, if
I may be allowed to use, on the basis of analogy, [this term] which
[is] not entirely suitable" (90 [183]). (Kant remarks that even the
term "Logic" is not entirely suitable either.)
We may well consider for a moment the extent to which this term
is and is not suitable. It is suitable insofar as the word "aesthetic" has
general reference to the sensibility,3 in this case to feeling as a mode
of sense.' But it would be a mistake to expect very much similarity
between the two Aesthetics. In the first Critique, there is a Tran-
scendental Aesthetic having to do with the sensible conditions of
a priori synthetic knowledge. Practical philosophy, however, has been
excluded from transcendental philosophy,5 and Kant is not here con-
cerned with the manner in which the objects of practical reason are
given to knowledge. He is, rather, involved in a study of how the
subject, as a sensuous being, is affected not by the objects of pure
practical reason but by its principles.
§ 2. A DIFFERENT BUT RELATED PROBLEM

It might appear that the question of this chapter has already been
answered as well as Kant can answer it. For, it may be said, we are
here r::oncerned with the relation of the intelligible to the sensible
nature of man, and Kant has already said that this noumenal-phenom-
enal distinction is the kernel of the doctrine of transcendental freedom.
Though the psychological-ethical and the phenomenal-noumenal
problems are intimately related, they are not entirely the same. The
doctrine of transcendental freedom was believed by Kant to be nec-
essary to that of moral or practical freedom, but not sufficient to it.
The problem of tf,e Aesthetic is not, therefore, solved by resolving
the third antinomy. Here Kant is concerned with an immanent phe-
nomenological problem, not with one in the ontology of appearance
and reality. The question is not directly How could a law of the in-
telligible world be reflected in the determination of events in the
phenomenal? It is, rather, How can a being in the phenomenal world,
through his knowledge of the law of the intelligible, control his con-
duct so that this law does in fact become effective? To answer this
question we need to know as much about man as about the law; and
a treatment of transcendental freedom cannot give us this information.
3 Critique of Pure Reason, A = B 35.
21 4 Cf. above, chap. vii, § 3.
IS Critique of Pure Reason, A 15 = B 19; d. above, p. 9 f.

The "Aesthetic" of Pure Practical Reason 213
§ 3. SOME EARLIER ATTEMPTS AT AN ANSWER
Prior to the Foundations of the Metaphysics of Morals, Kant had

considered the problem before us, but not until the full maturity of
the critical doctrine had been reached did he appreciate the singular
difficulty of any solution.
In the Prize Essay, The Distinctness of the Principles of Natural
Theology and Morals, and under the influence of both \Volff and
the English moralists, Kant stated the formal principle, "Do the most
perfect thing that can be done by you," but adds, against the ration-
alists, that no particular obligation can be derived from this 6 "except
when indemonstrable material principles of practical knowledge" are
added. 7 There is, he then believed, an unanalyzable "feeling for the
good" which supplies these material principles, and, though he did not
say so explicitly, it is quite clear that he believed the moving force in
morality to come from the moral feelings and not from the formal
principle itself. Yet this essay ends on an unexpectedly tentative note:
"It is still to be settled whether it is simply the cognitive faculty or
whether it is feeling (the primary inner ground of the appetitive
faculty) which decides the basic principles of practical philosophy."
It is as if Kant, even then, was not secure in his conviction that feeling
is an independent factor in morality.
In the Observations on the Feeling of the Beautiful and the Sublime,
published the same year, Kant seems to have decided in favor of feel-
ing as the source of moral ideas. After setting forth the necessity of
universal principles in morality (since a merely good disposition is not
enough), he says of the principles required: "They are not speculative
rules, but the consciousness of a feeling which lives in every human
breast, and which extends much farther than to the particular ground~
of sympathy and complaisance. I believe I cover it when I say it is the
feeling for the beauty and worth of human nature." He then explains
how this feeling and its associated "broadened inclination" can control
our good-natured drives and give rise to the "beauty of virtue."8
This is eudaemonism, and the Dreams of a Spirit-Seer attempts a
"pneumatological" explanation9 of these more extensive and compre-
hensive social feelings, which he calls "moral," through a mechanism
of the community of rational beings under psychical laws that are
compared to Newton's laws which govern the unity of physical nature.
6 Cf. also Critique of Practical Reason, 41 (130), and above. pp. 106 f.
7 Prize Essay, II, 299 (Beck, 284).
8 Beobachtungen uber das Gefuhl des Schiinen und Erhabenen, II, 217.
9 Triiume eines Geistersehers, II, 330, 335; cE. above. p. 159. n. 68.
This is a speculative metaphysics of moral feeling, in which meta-
physics is only a hypothetical extension of the type of explanation
valid within empirical experience. But if we remember what "meta-
physics" meant to Kant before and after 1770 and the change which
came in his conception of its scope and method about this time,10 we
shall not be surprised at the developing intellectualism present in a
very significant statement that Kant made in commenting upon an
ethical essay by his former pupil, Marcus Herz. In 177 3 he wrote:
"The concept of morality must please in the highest degree, must have
moving power; and though it is indeed intellectual, it must have direct
relation to the basic incentives of the wi11."l1
This view, not further developed at that time and, indeed, so briefly
indicated as to make its full interpretation hazardous, is probably the
first intimation of the doctrine of the critical writings of the 1780's.
In the intervening period, Kant seems to have been so torn between
two incompatible doctrines of the moral motive (if, indeed, he had
time to think of them at all) that when we come to the Critique of
Pure Reason we find him espousing each of them on a single page.
Thus the necessity of rewards and punishments as a spur to morality is
asserted in a passage which sounds like a religious eudaemonism:
It is necessary that the whole course of our life be subjected to moral
maxims, but it is impossible that this should happen unless reason connects
with the moral law, which is a mere Idea [and not yet a maxim], an opera-
tive cause which determines for such conduct as is in accord with the moral
law an outcome that is in exact accordance with our supreme ends [i.e.,
happiness]. Thus without a God and a future world invisible to us now but
hoped for, the glorious Ideas of morality are indeed objects of approval and
admiration but not springs [Triebfeder] of purpose and of action. 12
Yet almost immediately and with no obvious consistency, he denies
that the prospect of future happiness makes the moral disposition
possible. The moral disposition is one of worthiness to be happy, to
which the hope of happiness is added by these "postulates"; but the
desire for happiness does not generate the moral disposition. 13 The
10 Cf.above, p' 7 fr.
11 To Herz, end of 1773 (X, I'H)' The Vorlesungen iiber Metaphysik ([Politz
ed., 1821], p. 242), of uncertain date, speaks of the moral feeling as that through
which the motives of the understanding gain moving power. The student's lecture
notes, however, even if accurately datable, cannot be depended upon to be accurate
where any subtle issue is at stake, as every professor will ruefully acknowledge.
12 =
Critique of Pure Reason, A 812-13 B 840-41.
13 In a similar passage in the Refiexions on his essay on the sublime and beautiful,
Kant has recourse to the mysterious working of future rewards and punishments on
The "Aesthetic H of Pure Practical Reaso1l 2 15
second of these views is the one that is to appear again and again, and
the religious eudaemonism of the quoted passage is transcended, never
to be asserted again. Still we look in vain in the first Critique for an
analysis of the desire to be worthy of happiness, which is a truly moral
and not eudaemonistic desire. Without such an analysis, the view of
the Critique of Pure Reason seems to be an incompatible mixture.
For an analysis of this desire, we must turn to the celebrated Duis-
burg Fragment 6, written, I believe, just after the first Critique. H
Here the desire for happiness is still fundamental, though here again
Kant recognizes that such a desire does not eo ipso generate virtue.
But happiness is here given a moral definition, under which it can be
desired morally. Happiness is well-being only insofar as it is de-
pendent upon our choice and not upon accidental circumstances; it
is, accordingly, nothing but "well-ordered freedom." Well-ordered
freedom, the exercise of which occasions self-satisfaction, is the unity
of all actions under general laws and is equivalent to morality. Hence
Kant concludes that morality is a necessary condition of happiness
and makes it possible, but does not have happiness as its purpose. It is
the a priori form of happiness, the content being contingent upon
circumstance. Virtue and only virtue can bring about happiness, but
it does not necessarily do so, and, even when it does, that is not the
source of its worth. But to feel that we are the authors of a state of
being worthy of happiness (i.e., to have its a priori condition) is itself
a positive feeling of self-contentment, and this constitutes the human
worth of morality and is a necessary factor in happiness. The inner
applause is a sufficient motive and is an "intellectual pleasure" in the
enjoyment of freedom. l :!
This fragment constitutes a transition from a dualistic to a monistic
theory of the moral motive. Previously, except perhaps in the one
sentence of the letter to Herz, the cognitive and conative factors had
been kept separate, with the consequence that pure reason could not
consistently be asserted to be intrinsically practical. In Fragment 6
the moral disposition which can remain pure even under their influence. He says:
"If, however, in some supernatural way there is some influence on the purity of our
nature, future rewards no longer have the character of motives" (XX, 18). This is
no longer an expedient that Kant could take in the Critique, though the thought
remains in the estimate of the purity of Christian morals and his rejection of a
hedonistic interpretation of Christian motives (113 n. [110 n.1).
14 Reflexion 7102 (Ak., XIX, 17Crth; Reicke, 1,9-(6).
15 The term "intellectual pleasure" is later rejected (Critique of Practical Reason,
117 [ 21 3]).
216 The "Aesthetic H of Pure Practical Reason
the conative element is itself intel1ectualized 16 and formalized, and
moral happiness, i.e., a happiness in morality and not a happiness re-
sulting from it, becomes the goal. But when, in the next few years,
Kant formulated the doctrine of autonomy, even moral happiness be-
came a corollary effect and not a motive as a separate and distinct fac-
tor in moral purpose. Weare now ready to study this last and most
momentous change in Kant's views.
§ 4. INCENTIVES
Triebfeder is the generic name for the dynamic or conative factor

in willing,l7 It may be mere impulse, as in an animal, or an interest,IS
in which the representation of some law, natural or moral, guides ac-
tion. If our interest is in the object of an action, Kant has already
shown that the incentive must be a sensuous desire under the principle
of self-love (22 [1081). If we are interested in the action directly, it
is because we find satisfaction in the law of the action, and the ques-
tion is: How can a sensuous being take an interest in a pure rational
principle? Or How can a rational principle itself be the incentive for
a sensuous being?
In the Foundations, Kant distinguished between "incentive" and
"objective ground of volition" or motive, as between the desire for
subjective ends and that for ends valid for every rational being. lo
Consequently, he said there that the moral imperative excludes from
the legislation all admixture of any interest as an incentive. By this he
means that our incentives are not the author of any moral law, and
this is a view from which he never subsequently departed. But the
ascription of incentives only to the sensuous side of our nature, with
the resulting contrast between incentive and motive, is abandoned;
and though, in a sense still to be explained, all incentives are subjective,
they need not be private and sensuous. Indeed, even in the Founda-
tions, the Idea of the intelligible world (the moral realm) is called the
"incentive,"2.() showing that the privacy of incentives is not essential
to the concept. The subjectivity of incentive can thus mean two
16 Josef Bohatec (Die Religionsphilosophie Kants, p. 141) refers to the doctrine of

Fragment 6 as "intellectual eudaemonism."
17 Cf. the discussion of the proper translation of Triebfeder, above, p. 90. n. Z.
18 Critique of Practical Reason, 79 (172): "From the concept of an incentive
comes that of an interest." Cf. also Foundations of tbe Metaphysics of Morals, 459
n. (So n.); Vorarbeiten ZUT Tugendlehre, XXIII, 378; Metapbysik der Sitten, VI. 212.
19 Foundations of the Metaphysics of Moralr, 4 2 7 (45).
2olbid.,¥>z (83).
things: (a) it has a reference to the private, personal motivations of

the subject, which are based on sensuous impulse, which is, at most,
guided by reason, and (b) it refers to the workings of the moral
principle, which is itself objective, upon the constitution of the human
subject, and this working is the incentive which is obviously subject-
conditioned as well as objectively determined. This sense of "subjec-
tive" is retained in the Critique, where, we are told, an incentive is
"a subjective determining ground of the will whose reason does not
by its nature necessarily conform to the objective law" (72 [164])'
Here "subjective" merely means located in and thus in part depending
upon the constitution of the subject, without implying that this sub-
jectivity is indicative of dependence upon personal differences in
sensuous desires. There is, therefore, no contradiction in saying that
the objective law must be the incentive, though its role as incentive is
subjective and presupposes "the sensuousness and hence the finitude of
such beings" as man.21
The relation of the moral law to the incentive is therefore a formal
distinction analogous to that between the law and the categorical im-
perative or the law and duty. The imperative is the law as it appears
to a finite being whose will does not obey the law by nature; the
imperative is the law, but in the imperative mood. Similarly, for finite
sensuous beings the law appears as a constraint or necessitation, i.e.,
as duty, because its "is" is our "ought."22 Contrariwise, the moral law
is a dictate of pure practical reason, which, as such, has no incentive. 23
Hence incentives arise only in a being affected by sense not because
the incentive is subjective in any pejorative meaning but because a
non-sensuous rational being would, by nature and without an incentive,
execute the moral law.
§ 5. DESIRE AND PLEASURE

We must, therefore, know something of the sensuous nature of man
in order to understand the way in which the moral law, objective
though it is, can become a subjective incentive. For this, we must turn
to Kant's psychology and to the "transcendental definitions,"24 which,
21 Critique of Practical Reason, 76 (168). The statement "The law itself must be
the incentive" is not quite accurate and must be qualified (cf. below, § 7).
22 Foundations of the Metaphysics of Morals, 449, 455 (68,75).
23 Vorarbeiten zur Tugendlehre, XXIII, 378.
24 Erste Einleitung in die Kr;tik der Urteilskraft, XX. 230 n.; Critique of Practical
Reason,9 n. (94 n.).
he says, are all that he needs to "borrow from psychology" for this
Critique.
That man is a living being-indeed, that there are men in the world-
affected by desires and susceptible to pleasure is an empirical fact. The
relation between pleasure and desire is to be determined empirically
when desire is given an empirical content (as hunger, thirst, and the
like). It is essential that the psychologist not give definitions which
are implicit hypotheses about matters which can be settled only em-
piricalIy.25 They must be defined in such a way as to leave open for
empirical determination the relations which subsist between the defined
feelings. They must be defined, he says not quite accurately,26 only
"in terms belonging to the pure understanding, i.e., categories, which
contain nothing empirical." There are three such definitions:
Life is the faculty of a being by which it acts according to the laws of the
faculty of desire. The faculty of desire is the faculty such a being has of
causing, through its representations, the reality of the objects of these ideas.
Pleasure is the representation of the agreement of an object or an action with
the subjective conditions of life [9n. (94n.)].
We have already mentioned the distinction between the lower and
the higher faculty of desire. 27 The latter is not one of experiencing a
subjective need which will produce pleasure only indirectly through
success in bringing an object of desire into existence, but it is the
faculty of desire whose object is one that can be represented only by
pure reason and caused by virtue of this representation. That is, an
object of the higher faculty of desire is not a thing considered
materialiter but a thing considered formally as under a law and pleasing
only because of its conformity to and generation through the law.
Since only reason can represent a law, reason is the higher faculty of
desire, and its a priori principle is a purposiveness which is an impera-
tive for all actual purposing. 28
The question of the present chapter may therefore be phrased thus:
How can reason be the higher faculty of desire and not merely provide
norms for judgment, post facto?
2:; This is in conformity to Kant's theory of the philosophical uses and dangers of
definition (cf. my "Kant's Theory of Definition," Philosophical Review, LXV
[1956],17HI).
26 He means not categories, but predicables (cf. Critique of Pure Reason, A 8z =
B (08). But even this is not correct. Pleasure is an empirical, not a transcendental,
predicate (ibid., A 343 = B 401).
27 Cf. above, chap. vii, § 4.
28 Critique of Practical Reason, 24-Z5 (112); Erste Einleitung in die Kritik der
Urteilskraft, XX, 146.
The "Aesthetic" of Pure Practical Reason Z 19
§ 6. THE GENESIS OF RESPECT

We are now ready to deal systematically with the early parts of
chapter iii. Unfortunately, this is the most repetitious and least weIl-
organized chapter in the book, and there are a number of infelicitous
expressions that have seemed to many critics to be evidence of serious
inconsistencies. Nevertheless, the main points are made again and again,
sometimes in almost identical words and as conclusions from similar
arguments, so that there is little occasion for any serious misunder-
standing or for doubt as to what Kant means.
If morality is genuine, the moral law must directly determine our
choice; this we are told in the first sentence (7 I (1641). If it does not,
some feeling or desire which is independent of the law must be the
cause, and the action can at best be merely legal. In genuine morality,
the moral law must be the incentive.
Why and how the moral law is such an incentive is the unanswerable
and perhaps meaningless question: Why ought I be moral? But the
question here is not one of the authority of the moral law; it is a
question of what are the conditions, in a being like man, that make it
possible for him to take an interest in the law or to have the law as
his incentive. And this question, Kant thinks, can be answered a priori
with the help of his transcendental definitions.
Our inclinations are not, by nature, law-abiding. The moral law,
therefore, is felt as a constraint on them in a being possessed of a prac-
tical reason. Some inclinations it thwarts by disciplining them into a
coherent system like that described in Fragment 6; thus selfishness is
disciplined into rational self-Iove. 29 But one kind of inclination30 is ab-
solutely opposed to the moral law; it is inclination to regard one's own
subjective maxims and interests as having the authority of law. This is
self-conceit or moral arrogance, and it is absolutely incompatible with
morality. The vision of the moral law before man strikes down his self-
conceit and humbles his arrogance. It produces a feeling which is pain,
under the transcendental definition of pain. What humiliates us is an
object of respect. Hence we respect the moral law and have a feeling
of respect for it even when we do not obey it.
The moral law occasions also a positive feeling; for the interest or
29 Rational self-love mayor may not be moral. In any event, it is not a basis for a
direct duty (d. Critique of Practical Reason, 93 [187], and Religion within the
Limits of Reason Alone, VI, 45 n. [Greene and Hudson, 41 n.]).
30 CE. above, chap. vii, § 7, for a discussion of the two kinds of selfishness. It is
hardly appropriate to refer to either as an inclination. It is rather a system of in-
clinations under a maxim. Butler rightly called it a "sentiment" in contrast to a
"particular passion."
purpose of reason is the reign of moral law, and whatever accords with
this interest will necessarily be associated with a feeling of pleasure or
satisfaction. In the chapter before us, Kant never says explicitly that
there is a pleasurable component in respect, and, indeed, he generally
makes it appear more unpleasant than pleasant. But elsewhere he calls
respect "negative pleasure"Bl and declares it to be the ground of a feel-
ing of pleasure or satisfaction (116 [213]). But though there is at most
an analogon of pleasure, both positive and negative effects on feeling
are recognized in respect: not only are we struck down by the majesty
of the moral law, we are also raised up by it to a height and destination
that nothing in our natural talents can sustain. To avoid the notion that
respect is a sensuous feeling passi vely experienced, Kant seems hesitant
not only to call it pleasure hut even to call it feeling; but it produces an
effect on conduct that is analogous to that produced by pleasure or
pain. To call it simply pleasure or pain would obscure the unique phe-
nomenological features of respect (88 [18 I ]).
The dissimilarity of rational and empirical grounds of determination is
made recognizable through the resistance of a practically legislating reason
to all interfering inclinations, which is shown in a peculiar kind of feeling
which does not precede the legislation of practical reason but which is, on
the contrary, first occasioned by it, as a compulsion. That is, it is revealed
through the feeling of respect of a kind that no man has for any inclinations
whatever, but which he may feel for the law alone [91 (185)].
But the enjoyment is unmistakable, and the only danger in recognizing

it is that of subreption, of falsely locating it so as to think of it as the
cause instead of as the effect of the law's determination of the will.
We must recognize the inward contentment as a mode of pleasure
which necessarily accompanies obedience to the law but must guard
against thinking of it as the determining ground or object of the action.
The sublimity of the moral law is more than a metaphor for Kant.
Not only does he use the language of the aesthetics of the sublime in
describing the moral law, but he gives an analogous interpretation of
the origins of the feelings of sublimity and respect. In both there is
humiliation or thwarting of our sensuous nature (our perceptual fac-
ulty and imagination in the sublime. our feeling of worth in respect)
which occasions a pain which, in turn, is transmuted into a kind of ela-
tion by the discovery in ourselves of a power superior to that which
has humiliated us (in the sublime) and superior also to that power in us
which has been humiliated (in feelings of both respect and of the sub-
lime). But whereas a subreption necessarily occurs in the sublime feel-
31 Cf. Critique of Judgment, § 13.

The "Aesthetic" of Pure Practical Reason U I
ings, so that we attribute to the object a sublimity which actually ex-

ists only in ourselves, the feeling of respect is directed to a law, which
is a law of our own freedom, self-imposed and not imposed upon us
from without, and to the persons, ourselves or others, who embody this
law. Hence respect for the law and respect for our personality are not
distinct and even competing feelings, as are the two feelings which
merge in our experience of the sublime. 32
§ 7. AN ALLEGED INCONSISTENCY
Serious inconsistencies have been alleged to be present in Kant's dis-
cussion of the incentive to morality. He states:
The moral law should directly determine the will. If the determination of
the will occurs in accordance with the moral law but only by means of a
feeling of any kind whatsoever which must be presupposed in order that the
law may become a determining ground of the will bl (164)]
we have to do with legality, not morality. These are the opening
sentences of chapter iii and must, because of their position, be given
pre-eminent importance and weight.
Then he tells us that respect for the law is the incentive for moral-
ity,33 that respect is not the incentive to morality but is morality itself
(76 [168]), and that the law is the incentive (72 [164])'
It is regrettable that Kant was not more careful; though, had he been
so, the race of Kant-commentators would have been unemployed. But
it is unjust to do more than complain of his carelessness here and to
allege serious and insurmountable inconsistencies where the develop-
ment of the argument itself makes sometimes one and sometimes an-
other choice of words appropriate, though the statements taken out of
their context certainly seem incompatible.
In spite of what Kant says, the law itself is not the incentive. A law
is just not the sort of thing that can be an incentive. At most, con-
sciousness of a law can be an incentive. If the law itself were a deter-
minant of conduct, without the intervention of consciousness (which
32 Cf. Critique of Practical Reason, 86-87 (180); Metaphysik der Sitten, VI, 402-
3; Critique of Judgment, § 27. He had previously (Beobachtungen uber das Gefuhl
des Schonen und Erhabenen, 2. Abschnitt) distinguished among virtues those which
are sublime from those which are beautiful, obviously following Burke, Inquiry into
the Origin of Our Ideas of the Sublime and Beautiful, Part III, Sees. X and XI. On
the long history of these aesthetic concepts in morals before Kant and in Kant's
early works, see Giorgio Tonelli, "Kant: dall'estetica metolfisica all'estetica psico-
empiric a," Memorie della Academia delle Scienze di Torino (1955), pp. 77-4ZI,
passim (use the exhaustive index).
33 Critique of Practical Reason, 78 (171); at 85 (178): "respect for duty,"
means, for us men, also feeling), it would not be a practical law, and
men would not be free agents. Kant frequently speaks of laws' doing
things, when he means that the world is such that the laws describe it
(natural laws) or when he means that consciousness of the law will
induce a rational being to do what the law demands. u
Hence we must ask: What is the nature of the consciousness of the
law, such that it can be the incentive? If the Critique leaves any room
for doubt on this, it is removed by the Metaphysics of Morals. There
Kant says: "The respect for the law, which is, subjectively, called the
moral feeling, is identical with the consciousness of one's duo/."35 We
do not, therefore, have a theoretical knowledge of our duty and of our
will at potential variance with it, with the subsequent addition of the
feeling of respect. To know one's duty may not be, as Socrates be-
lieved, to do it. But to know what the law requires and to feel respect
for it (if not actually to obey it)36 are, for Kant, identical; even knowl-
edge of a moral law is not aseptically theoretical. Hence a false con-
trast between "law" and "respect" falls away when we read for the
former "consciousness of law," which is the only mode in which a
practical law can be effective and yet be distinct from a natural law.
But, then, what of the opening sentences? The second sentence is
often read as if Kant, when he wrote it, thought that the law must di-
rectly determine the will without any feeling intervening between the
law and the decision-a view at variance with the remainder of the
chapter. Yet he did not believe this even when he wrote this sentence.
For though the word "presupposed" is there precisely to guard against
another error (that of the moral-sense or moral-feeling school), it
should also guard against the opposite error of supposing that the de-
termination must or can be without any subsequent feeling. The de-
termination of the will by law does not require any feeling which must
be presupposed; that means that it does not require any feeling inde-
pendent of, or prior to, the consciousness of the law. The feeling of
respect, like any other feeling, need not be presupposed as something
in the natural man awaiting an object which can be found only in the
law; the feeling of respect presupposes the law and our consciousness
3. Cf. ibid., 29 ("7), the suggestion that the law is only the self-consciousness of
practical reason.
36 Even the sinner, who transgresses the law, shows his reverence for it by trem-
bling before its majesty (Critique of Practical Reason, 80 (171)). Cf. Religion within
the Limits of Reason Alone, VI, 46-47 (Greene and Hudson, 41-42) on the re-
establishment of the effectiveness of respect for the moral law, "which we have
never been able to lose."
of it (or is, indeed, equivalent to the latter). The first sentence, there-
fore is not incompatible with, but anticipates, the later statement: "In
the subject there is no antecedent feeling tending to morality."31
§ 8. THE MORAL FEELING

Though Kant made use of the concept of moral feeling in his pre-
critical works, the general trend of his teaching appeared to be an
elimination of all feeling from morality. It is hard to know whether the
desiccated picture of Kant's personality as that of a reiner Verstandes-
mensch has led people to ignore his positive theory of the moral feel-
ing or whether a caricature of his ethics as otherworldly has created
the false picture of the man; but the two conceptions usually go to-
gether, and they are both wrong.
Kant banished feeling from a positive position in ethics only insofar
as feeling is sensuously effected and, even as moral feeling, is put in the
false position of being the source of the rational principle. Thus he
criticizes Epicurus (I 15 [2 II ]) for finding a pleasure in morality with-
out giving a definition of morality in terms other than those of pleas-
ure; and he regards all theories of moral sense or moral feeling as being
necessarily subsumed under the principle of self-love (38 [128]),
since all pleasure, whatever its origin, is of one and the same kind (23
[ I JO]). He held, moreover, that any theory of moral feeling which
gave it a founding role necessarily led to subjectivism and relativism; a
man, he said, could legitimately excuse himself from a moral duty by
saying that he did not like the action demanded, if feeling is the source
of moral approbation. as
Nevertheless, moral feeling has a place in Kant's ethics, as is made
clear in the chapter before us; but it is an effect of the consciousness of
the moral law on sensuous feeling, not a feeling given prior to the dis-
covery of what the law demands of us in principle and not a mystical
feeling whose source would be in some otherworldly inspiration.
The moral feeling has two components, because it is, stripped to its
essentials, the same as respect (75, 80 [168, 173]), though, as we shall
see, it contains some other feelings that are corollary to respect. But
37 Critique of Practical Reason, 75 (167.168); similarly, 23-24,117 (110, 2I3). The

same point is emphasized in Critique of Judgment, ~ 29, Remark (VI, 27 J [II J]).
But compare the statement at Critique of Practical Reason, 25 (112): "Reason deter-
mines the will in a practical law directly, not through an intervening feeling of
pleasure or displeasure, even if this pleasure is taken in the law itself." It is difficult
to reconcile this statement with the others. for it suggests that pure practical reason
can determine the action without the mechanism explored in this chapter.
a8 Lectures 011 Ethics, J3. 37.
whereas in his discussions under the term "respect" Kant emphasizes
the humbling of the self which is essential to the aggrandizement of the
law, under the term "moral feeling" he puts more weight upon the en-
joyment of the spontaneity and freedom which is felt in the moral dis-
position.
Before discussing this positive joy in morality, which is not explicitly
included under "moral feeling" (though only the word is lacking),39
we must turn for a moment to another description of moral feeling and
then consider the concepts of duty and virtue.
In the Metaphysics of Morals Kant lists the predispositions, "the
subjective conditions of susceptibility to the concept of duty" by vir-
tue of which man can be obligated, even though the consciousness of
them, as an influence on the mind, can only follow upon the moral
Iaw. 40 One of these is moral feeling, and here Kant's account of it varies
markedly from that in the Critique.
The moral feeling, he says, "is the susceptibility to pleasure or pain
merely from the consciousness of the agreement or conflict of our ac-
tion with the law of duty."41 And, he continues in a sentence difficult
to translate, all determination of the will (Willkur) proceeds (a) from
the representation of the possible action (b) through the feeling of
pleasure or pain (through taking an interest in the action or its effect)
(c) to the act. The aesthetic condition, the feeling, is either pathologi-
calor moral: the former if the pleasure precedes the representation of
the law, the latter if it follows it and is, as it were, pleasure in the law.
The differences between this statement and those in the Critique arise
from the fact that Kant is talking of two different things under the
same name, under a single name that had had a long history of applica-
tion to each. In the Critique he is discussing the feeling that a rational
sensuous being has in the face of recognized duty; in the Metaphysics
he is discussing not a feeling as a phenomenological state of conscious-
ness, but as a potentiality, an Empfiinglichkeit. Naturally, this must
precede, logically or temporally, the actual feeling of respect. But this
39 That is, in the principal discussion, Critique of Practical Reason, 115-18 (ZII-
15).
40 Metaphyrik der Sitten, VI, 399. The others are love of one's fellows, conscience,
and respect (for self). The Critique of ]udfSment, § 88, gives another listing of
"mental dispositions that make for duty," viz., gratitude, obedience, and humility
(submission to deserved chastisement), and in § 91 speaks of gratitude to and ven-
eration for the unknown cause, arising from the admiration of nature and producing
an effect on moral feeling. A similar thought is expressed in Critique of Practical
Reason, 160 (z58), in the anecdote about Leibniz.
41 Similar definitions are in Religion within the Limits of Rearon Alone, VI, 27
(Greene and Hudson, 23) and in Lecturer on Ethicr, p. 44.
statement does not conflict, except verbally, with the statement that
there is "no antecedent feeling tending to morality"; it is as if one were
to say in English, "A man must have feeling before he can have a feel-
ing"-perhaps a not very lucid way of saying a simple thing, but cer-
tainly no evidence of fundamental confusion in the speaker's mind.
Yet, in spite of this, there is a difference between the analyses of the
actual moral feeling (i.e., the feeling that is felt, and not the potential-
ity for it) as given in the Critique and as suggested in the Metaphysics.
The Metaphysics of Morals says that the moral feeling is either pleas-
ure or pain, pleasure if there is felt accordance with the law and pain if
there is not. The Critique, on the other hand, says that there is some-
thing analogous to both pleasure and pain even if the action does con-
form to the law, because even in conformity there is the pain of humil-
iation arising from the thwarting or striking-down of inclination.
But there is no real inconsistency here. The Metaphysics of Morals
makes a sharp contrast between pathological and moral pleasure and
pain and discusses only the latter. Now there is no moral pain in obedi-
ence to the law; there is at most a pain arising from the thwarting of
pathological feeling. But, because of the difference drawn between the
QVo kinds of feelings, Kant can discuss one without the other in the
later work. In the former, he discusses them together and thinks of
pathological feeling as the material cause of moral feeling. "Sensuous
feeling," he says (75 [168]), "which is the basis of all our inclinations,
is the condition of the particular feeling we call respect." In the Meta-
physics of Morals this relationship is neglected, and therefore there is
missing this peculiar mixture of pleasure and plain. That he is discussing
moral feeling, as actual, in a slightly different sense in the two works is
shown by the fact that in the Metaphysics moral feeling is not equated
with respect, while in the Critique it is.
§ 9. DUTy f2 AND PERSONALITY

In the Foundations,f3 Kant formulates the definition: "Duty is the
necessity of an action done from respect for law." We are at last in a
t2The form of the celebrated apostrophe to duty (86 [18o}) may have been sug-
gested by Rousseau's apostrophe to conscience in Emile, which Kant admired,
On Kant's text see H. Romundt, "Vorschlag zu einer Anderung des Textes von
der Kritik der praktischen Vernunft,' Kant-Studien, xm (1908), 313-140 and re-
plies by B. Bauch and P. Natorp, ibid., pp. 315-16. Kant made little use of this pop-
ular trope. since it was hardly in keeping with the Sachlichkeit of his style. There is
an apostrophe to sincerity in Religion within the Limits of Reason Alone, VI, 190 n.
(Greene and Hudson, 178n.).
tS Foundations of the Metaphysics of Morals, 400 (16).
position to understand the full meaning of this in all its ramifications,
when Kant says: "Respect for law is not the incentive to morality, but
is morality itself" (76 [168]). Morality is a disposition or settled habit
to act in accord with law because it is law; and one who respects law,
in the sense described here, is one who acts from duty. Duty is the ne-
cessitation or constraint to an action by a law in the consciousness and
conduct of a person who does not, by nature, necessarily act as the law
requires. A non-sensuous rational being would not have to have any in-
centive (72 [164]) to such obedience, and therefore the concept of
duty is not applicable to such a being; such a being would have a holy
will. The "ought" of the moral imperative is an "is" for such a being. 44
But man, as a sensuous being, is "under the discipline of reason" (82
[175]) and is not endowed with a holy will. His will is at best virtuous.
Even so, man is a person and, as such, is holy. He is an end-setting
being and therefore an end in himself. He has dignity and is not like a
thing having only a price. Personality-that attribute which distin-
guishes rational beings from things-is freedom from the mechanism of
nature through a capacity to be subject to laws given by itself (auton-
omy). An empirical self, having only negative freedom, is "subject to
his own personality so far as he belongs to the intelligible world" (87
[ 180]). Personality is sublime, and whenever we respect any man, we
respect the law of the intelligible world which he more or less ade-
quately represents in the phenomenal world (77-78 [170]). Categorical
imperatives for the empirical self are laws of the intelligible self; the
"should" of the former is the "would" of the latter. While man, em-
pirically regarded, is "unholy enough," personality and humanity in
him are holy.
With this explanation of the provenance of duty, there is a signif-
icant introduction of the concept of personality. Whereas Kant had
emphasized, in the first Critique, the impossibility of a rational psychol-
ogy of the soul as substance and had made the self only an appearance
-an emphasis continued in the second Critique-there is here and else-
where a transition to a more metaphysical interpretation of the self not
only as thought but as directly experienced as a being noumenal in
character. 411 The phenomenalistic interpretation of the self. though it is,
as it were, the official Kantian doctrine, shows signs here and there of
being transcended in a metaphysical personalism and activism, since the
spontaneity felt in self-consciousness separates man from nature. While
a "metaphysics" in the classical and not in the Kantian sense seems to
u Ibid., 449, 455 (68,75); d. Critique of Judgment, § 76.
45A full study of this, with a collection of relevant passages, is given in Heim-
5Oeth's Studien zur Philosophie Immanuel Kll1lts, pp. 227 fl.
be suggested by this, it is not a theoretical but a practical-dogmatic

metaphysics which is involved:
Meanwhile, we may still retain the concept of personality ... in so far as
it is merely transcendental, that is, concerns the unity of the subject, other-
wise unknown to us, in the determinations of which there is a thorough-
going connection through apperception. Taken in this way, the concept is
necessary for practical employment and is sufficient for such use; but we can
never parade it as an extension of our self-knowledge through pure reason.46
Again and again it is asserted that no theoretical use of the concept can
be made, e.g., in theoretically valid arguments for immortality. But the
evidence of spontaneity and our knowledge of the law of the self as
noumenon certainly give a richer conception of personality than that
of the transcendental unity of apperception, though the propositions in
which it functions are all practical, not theoretical.
Personality, then, is not a category; it is an Idea of reason, and per-
sonality is not a given. We are persons, but no finite sensuous being is
fully adequate to the Idea of personality. In human na!Ure, considered
empirically, we find at most only a "predisposition to personality,"
which is the capacity for respecting the moral law and making it suffi-
cient incentive for the will. 47 When this predisposition is strengthened
through practice and becomes actual and effective, there is a state of
virtue and goodness of character; a good man in the empirical world is
one whose law is derived from, and is followed out of respect for, the
Idea I>f his personality in the intelligible world. Even in an evil man,
one who voluntarily embraces other maxims than those conforming to
the moral law, the predisposition to personality is not lost; it has only
been rendered ineffective by a free choice against the demands of the
moral law. Moral evil is voluntarily going against the demands of one's
own personality as pure practical, legislative, reason.
§ 10. VIRTUE
Virtue is a "naturally acquired faculty of a non-holy will" (33
[121]). It is th" moral disposition in conflict or, as Abbott eloquently
translates it, "the moral disposition militant."48 Without the two poles
of sensuous inclination and pure rational principle, it could not arise.
46 Critique of Pure Reason, A 365, absent from B; cf. B 431-32, absent from A.
41 Religion within the Limits of Reason Alone, VI, 27~28 (Greene and Hudson,
22-13). This is identified with moral feeling in the sense of the Metaphysik der
Sitten, VI, 399. instead of in the sense of the second Critique (cf. above, p. 224).
48 Critique of Practical Reason, 84 (t 78) • Abbott's translation here aptly preserves
the military symbolism of the page.
We stand under the discipline of reason; we are conscripts in the moral
host, taking orders, and not volunteers, acting from our individual
merit. We are members of the realm of ends, but we are subjects in it
and not, at first, sovereigns of it (82 [175])' In man, virtue is always
defective, and virtue itself, like personality, is a mere Idea to which no
perfectly adequate empirical representation can be found. 49
In many of Kant's examples of virtue, the necessity of acting out of
respect for law and not from inclination is presented as if the action
had to take place without, or even always against, inclination. But this
is not what the examples were chosen to illustrate, and this is no part
of Kant's theory; even acting from inclination is not in itself evil, but
only the taking into one's maxim of action and insubordination of sensu-
ous incentives to rational incentives-this insubordination itself being
free and spontaneous and a matter of principle-is the root of moral
eviPo
The purpose of the examples of virtuous action can be understood
only in the light of their polemical or pedagogical use. An act in which
a man did what duty required of him and to which his inclination also
pointed might easily appear-and certainly would appear to a man with
Kant's skeptical hatred of cant-to be an action done in accordance
with law because of his inclination. Even where inclination seems to be
against the law, the mystery of the human heart is so deep that we can
never be sure that we will not, upon more penetrating self-analysis,
come upon "the dear self" as the real determinant. 11l But where we find
the dear self in this role, there is no morality; hence the best illustra-
tions of genuine morality will be found where the dear self is obviously
opposed, though that does not mean, as we shall soon see, that the best
morality is one in which there must be active and painful self-abnega-
tion. Virtue does not entail a renunciation of happiness, but only a
willingness "to take no account of happiness when duty is in question"
(93 [I 86]).
But those who read Kant's examples and fail to understand the text
that they were meant to illustrate have always regarded his ethics as
singularly repellent and have thought that if they represent typical and
most estimable virtue, then it is an ethics which they must reject. 52 In-
clinations and feelings, however, have a legitimate place in Kant's ethics
which is not brought out by these examples. They may even be seen as
491bid., 127 n. (214 n.); Critique of Pure Reason, A 315 = B 372; Critique of Judg-
ment, § 57, Remark 1.
ISO Religion within the Limits of Reason Alone, VI, 36 (Greene and Hudson, 3x)'
lSI Foundations of the Metaphysics of Morals, 407 (24).
1S2 Cf. above, p. 120, n. 20.
necessary corollaries of true virtue, since without inward resistance

there is holiness of will but no virtue.~3 But human nature is not left, in
the practice of virtue, with what Hegel called the "undigested lump of
sour virtue in the stomach"; the rational element can be strengthened
until it brings about a reformation of the natural man. lS4 What would
be morally indifferent or even evil before this reformation 5lS can be-
come a mark of the most complete virtue.
§ I I. MORAL SATISFACfION
We said above, in § 7, that in moral action we find a satisfaction in
the experience of our own autonomous spontaneity and that this is the
positive element in respect as the moral feeling. 56 Man is humiliated by
his vision of the law, before whose majesty even the boldest sinner
trembles (80 [I 72]); but he is elevated in his consciousness that this
humiliation is the mark of his higher vocation, for it is humiliation of
himself by himself, of his heteronomous natural being by his autono-
mous intelligible being (cf. 76 [1691 on Fontanelle).
In morality, therefore, in the consciousness of determination by law,
there is a feeling of joy, since reason's interest is being furthered (116
[z I 2]). Freedom itself, as the exercise of moral will, thus becomes sub-
ject to enjoyment, which is not happiness but essential to it. lS1 This joy,
however, cannot define virtue, since it arises only from a prior con-
sciousness of virtue, which, formally and abstractly considered, has
113 A holy will has no incentives, no duty, and no virtue; it is only a standard or
Idea by which we judge human incentives, duty, and virtue. Kant observes that the
Stoic sage was believed to be so perfect that he was "elevated above duties though
he propounded duties to others," but that this error arose only because the Stoics
failed to conceive the law "in the same purity and rigor as does the precept of the
Gospel" (Critique of Practical Reason, 117 n. [114 n.J).
5f On "sweet and sour duty," cf. Metaphysik der Sitten, VI, 377, 391, and Lectures
on Ethics, p. 199.
illS Even the inclinations refined to civility have nothing stable or of absolute
wonh in them: "Everything good that does not grow from its connection with a
morally good disposition is nothing but empty show and glittering misery" (Idea
for a Universal History, Seventh Thesis, VIII, 16). In such judgments, Kant shows
some of the passion and often uses some of the words of Rousseau.
116 The Critique of Judgment, § 29, General Remark (V. 171 [III-U», emphasizes
the positive side of the moral feeling more than the second Critique does. In fact,
the discussion of se1f-contentment in the second does not directly identify it as a
component in the moral feeling, and the whole treatment of it is not found in the
chapter on the incentives but is introduced later into the chapter on the summum
bonum. Nevertheless, it is cleat that the joy in morality is a part of it, the humilia-
tion of natural man before the law being a necessary precondition.
51 Critique of Practical Reason, 118 (liS) and Fragment 6 (Reftexion 7201).
230 The" A esthetic" of Pure Practical Reason
nothing to do with happiness itself. It was the fundamental logical error
of the eudaemonists to recommend virtue as a source of happiness with-
in one's self, arising from self-approbation predicated upon conscious-
ness of virtue, without giving a criterion for approbation so that con-
formity to it could indeed be a source of joy. But the error of Epicurus
and the moral-sense theory (38, 116 [128, 21 2]), already pointed out
by Hutcheson and Price,II8 should not cause us to commit the opposite
error of denying a positive joy to be possible in moral conduct.
Hume59 had complained of the lack of a name for this feeling of
moral well-being, and Kant, as if directly answering Hume, says:
Do we not have a word to denote a satisfaction with existence, an analogue
of happiness which necessarily accompanies the consciousness of virtue, and
which does not indicate a gratification, as "happiness!) does? We do, and
this word is "self-contentment," which in its real meaning refers only to
negative satisfaction with existence in which one is conscious of needing
nothing. Freedom and the consciousness of freedom, as a capacity for fol-
lowing the moral law with an unyielding disposition, is independence from
inclinations, at least as motives determining (though not as affecting) our
desiring; and, so far as I am conscious of freedom in obeying my moral
maxims, it is the exclusive source of an unchanging contentment necessarily
connected with it and resting on no particular feeling. 60
That this feeling is the stronger the more purely the law is presented.
is a fundamental thesis in Kant's educational theory.61 Other impure
motives may well work in the direction of legally correct action (81
[174]), but they may cloud the vision of the man for the only thing
that can bring him to morality and lasting equanimity. Thus, Kant
continued:
Even the feeling of sympathy and warm-hearted fellow feeling, when pre-
ceding the consideration of what is duty and serving as a determining
ground, is burdensome even to right-thinking persons, confusing their con-
sidered maxims and creating the wish to be free of them and subject only
to law-giving reason. 62
:>8 Cf. above, p. (06.
119 Enquiry concerning the Principles of Morals, Appendix IV (ed. Selby-Bigge
[zd ed.J, p. 3(4).
60 Critique of Practical Reason, ((7 (zr4); on contentment d. Wolff, Verniinf-
tige Gedancken 'Clon Gott, der Welt und der Seele des Menschen, § 463.
61 Critique of Practical Reason, I~Z, 1~7 (25t, 255); d. Foundations of the Meta-
physics of Mora/s, 41()-u (27); Uber den Gemeinspruch •.. , VIII, 288.
62 Critique of Practical Reason, 1I8 (214). italics added. Similarly, Foundations
of the Metaphysics of Morais, 428 ('J6). But contrast the remarks in Religion within
the Limits of Reason Alone, VI, 58 (Greene and Hudson, 51) on the vanity and
sin of wishing to be free from all inclinations.
The "Aesthetic" of Pure Practical Reason 13 I
It was no doubt in reaction to this that Friedrich Schiller wrote:
In the Kantian moral philosophy the idea of duty is presented with a hard-
ness which frightens away all the graces and which could mislead an obtuse
mind to seek moral perfection on the path of a dour and monkish asceticism.
However much the great philosopher tried to avoid this misunderstanding,
which must have been disturbing to him above all, with his serene and free
mind, I still think that he himself gave obvious occasion to it through the
sharp and shrill contrast between the two principles which work on the will
of man-an occasion which perhaps can hardly be avoided in his theory.63
To this Kant replied in a passage in the Religion which is not in any
way inconsistent with the teaching of the Critique, though, if it had
been said in the Critique, the "occasion" would not have been so ob-
ViOUS:
Now if one asks, What is the aesthetic character, the temperament, so to
speak, of virtue, whether courageous and hence joyous or fear-ridden and
dejected, the answer is hardly necessary. This latter slavish frame of mind
can never occur without a hidden hatred of the law. And a heart which is
happy in the performance of its duty (not merely complacent in the recogni-
tion thereof) is a mark of genuineness in the virtuous spirit-a genuineness
even in piety, which does not consist in the self-inflicted torment of a re-
pentant sinner (a very ambiguous state of mind, which ordinarii y is nothing
but inward regret at having infringed upon the rules of prudence 64 ), but
rather in the fum resolve to do better in the future. This resolve, then, en-
couraged by good progress, must needs beget a joyous frame of mind, with-
out which man is never certain of having really attained a love for the good,
i.e., of having incorporated it into his maxim.6Ii
Let this lay the ghost of an old error. But it has often been said that
the Kantian doctrine is "false to human nature," in that it neglects or
denies the role of spontaneous feeling precisely because it puts them in
63 0 ber Anmut und Wurde. Note that Schiller says that this is a misunderstanding
of Kant's views and that Kant regarded Schiller as a philosophical ally, not an op-
ponent (Cf. Religion . .. , VI, 13 n., first sentence [Greene and Hudson, 19 n., sec-
ond sentence]). Schiller's famous verse can be regarded only as a joke and not as
presenting Schiller's real views of the Kantian philosophy. There Was a point that
actually divided them, however, though it is not clearly brought out in their Ausein-
andersetzung, to wit, whether the collaboration of rational and sensuous elements is
essential to vinue (Schiller) or, at most, one of the duties (Kant, Metaphysik der
Sitten, Tugendlehre. S 48 [VI, 473]-to unite the virtues and the graces is a duty).
For perhaps the best full study of the Kant-Schiller controversy and the extensive
literature that has grown from it see Hans Reiner, Pflicbt 'lind Neigung (Meisen-
heim, 1951), pp. z8-49.
64 On repentance see Critique of Practical Reason, 98 (19Z).
6Ii Religion within the Limits of Reason Alone, VI, 13 n. (Greene and Hudson,
(9 n .).
second place, after the rational determination. 66 It therefore seems to
make impossible, as well as undesirable, demands upon the natural man.
Not, perhaps, transcendentally impossible demands, because of Kant's
doctrine that "thou canst, because thou shouldst," but demands that
living men cannot be reasonably expected to acknowledge.
There are three answers to this. One is that it is an irrelevant criti-
cism, since for Kant an ethical theory is not to be judged by the facts
of human nature (about which, indeed, we know far too little and into
which we surreptitiously introduce unexamined pessimistic or optimis-
tic moral judgments according to our taste).67 Second, as a corollary to
this, we must remember that such judgments of human frailty are not
alien to Kant himself. His derogatory judgments of the moral pros-
pects of human nature and society are, in fact, sharper than those of
either Aristotle or Dewey, two other great proponents of reasonable
morality.68 And, third, Kant saw the need, for us men, of supplement-
ing the pure moral motive with others more natural to man, in spite of
his warnings against this in the Critique, where he was examining pure
practical reason. Love is the most important of these motives:
When it is a question not merely of the representation of duty [which is
the task of the Critique J but of its execution, if one asks about the subjective
ground of action on which, if presupposed, one can at least expect what man
will do and not what on the objective ground he ought to do, it is love which
66 Dr. Johnson was no great philosopher, not even a competent one. But he was a
man of solid understanding whose testimony on a point of moral judgment carries
a weight it would not have had if it had issued from a degree of philosophical so-
phistication he neither had nor admired; thus does the wise layman sometimes serve
as the impartial observer in philosophical disputes. Boswell writes: .. 'Sir,' said Mr.
Johnson, 'I can lay but little stress upon that instinctive, that constitutiona~ goodness
that is not founded upon principle. I grant you that such a man may be a very good
member of society. I can conceive him placed in such a situation that he is not much
tempted to deviate from what is right; and so, as goodness is most eligible when
there is not some strong enticement to transgress its precepts, I can conceive him
doing no harm. But if such a man stood in need of money, I should not like to trust
him. And even now, I should not trust Mr. Dempster with young ladies, for there
is always a temptation" (London Journal, July 11, (763). The same is given in
slightly altered fonn in the Life ("Modern Library" ed.), p. 268; cf. also Price,
Rewew of the Principal Questions of Morals, :ed. D. D. Raphael (Oxford. (948),
p.I9I.
67 Uber den Gemeimpruch ... , passim, but especially VIII, 276-77.
68 Foundations of the Metaphysics of Morals, 407 (23); Idea fOT a Universal His-
tory, Sixth Thesis (VIII, 23): "Out of such crooked lumber as man is made of,
nothing truly straight can be builded." Similarly, Religion within the Limits of
ReMon Alone, VI, 100 (Greene and Hudson, 92),
is an indispensable supplement to the imperfection of human nature, as a free
assumption of the will of another under one's own maxim. 69
This love cannot be commanded; it is not a duty, because it is self-
contradictory to require of a man that he do something out of inclina-
tion. 70 It must, therefore, be distinguished from the love which can be,
and in the gospel is, commanded, which is practical love or the ready
willingness to fulfil our total obligation to others. Kant, with his con-
viction that the true kernel of all religious truth is moral, regards this as
equivalent to the love commanded by Christianity.71
§ n. MORAL EDUCATION
By what course is the human race brought to a state in which the
moral disposition may rise and flourish? How can the child, in a civi-
lized state, be brought from innocence to moral maturity? These two
questions are closely related~ and though in the Critique Kant deals
only with the second, the first was very much on his mind during the
years immediately preceding the writing of it.
In the Age of Enlightenment, religion as the moral teacher of man-
kind was gradually pushed from its pre-eminent position. Though Kant
esteemed religion as the moral teacher of mankind, religion was, for
him, only its teacher, not its source. The great Enlightenment philoso-
phers sought a generally naturalistic origin of morality, of which reli-
gion might be seen as only a vehicle. Lessing, Herder, Kant, and Schil-
ler were at one in attempting a natural history of freedom, and, with it,
of morality, though for Kant, at least, neither natural history nor di-
vine revelation was adequate to explain its peculiar form and unique
authority. These attempts led to their philosophies of history, in which
the education of the human race is the chief theme, appearing even in
the titles of some of their books.
69 Das Ende alJer Dinge, VIII, 337. A similar office. is assigned to sympathy, which
is "an impulse, planted in us by nature, to do that which the representation of duty
alone would not be able to accomplish" (Metaphysik der Sitten, Tugendlehre, § 35).
But in moral education it is bener to try to weaken the inclinations than to ally them
with consciousness of duty, an alliance which may indeed make for better actions
but will not produce better men (Reflexion 6722; cf. Critique of Practical Reason,
88 [182]). Even to let other motives co-operate with the moral law is risky (ibid.,
72 [.64]). The difference in these passages cannot be denied; but the emphasis upon
a kind of moral synergism is more obvious in the practical than in the purely theo-
retical investigations of morals.
70 We cannot he commanded to do anything from inclination Critique of Prac-
tical Reason, 83 [l76J and Das Ende aller Dinge, loco cit.
11 On practical love cf. Metaphysik der Sitten, VI, 449, and below, p. 243, n. [3,
That the human race must advance itself is clearly and sharply stated
by Kant:
Nature has willed that man should, by himself, produce everything that
goes beyond the mechanical ordering of his animal existence, and that he
should partake of no other happiness or perfection than that which he him-
self, independently of instinct, has created by his own reason.
It is
just as if she had willed that, if man ever did advance from the lowest
barbarity to the highest skill and mental perfection, and thereby worked him-
self up to happiness (so far as it is possible on earth), he alone should have
the credit and should have only himself to thank-exactly as if she aimed
more at his rational self-esteem than at his well-being. 72
There is a natural predisposition to morality in man, but not natural
morality. From a state of natural innocence, men fell into evil; history
is his gradual working himself out of it. The Conjectural Beginning of
Human History treats the Book of Genesis as an allegory of this fall
and its consequences: "The history of nature begins from the good, for
it is the work of God; the history of freedom begins from evil, for it
is the work of man."73 The Idea for a Universal History traces the his-
tory of man's social arrangements, which are his measures for control-
ling the evils naturally arising from his "unsocial sociability" or need
for, but antagonism to, his fellows. As Heraclitus said, conflict is the
origin of all things; and the product is civilization and ordered state, in
which man assumes the role of unselfishness, respect for others, and de-
cency-a role which fools nobody and which is called "permitted moral
pretence."74 From it, however, true morality can arise when reason,
gradually perfected as an instrument, becomes aware of its higher voca-
tion. Thence a.rises the genuine interest in morality, and thereafter
"moral pretence" must be combated. 75
Assuming that genuine moral ideas have a certain currency in soci-
ety, how is the child brought to recognize them and respect them?
This is the problem of moral education proper. It is, for Kant, a unique
and irreducible division of pedagogy, not merely because of its method,
72 Idea for a Universal History, Third Thesis, VIII, 1,)-10.
73Mutmasslicher Anfang der Menschengeschichte, VIII, 115, an obvious para-

phrase of the opening sentence of Emile.
74 Anthropologie in pragmatischer Hinsicht, § 14 (VII, IF-53): "All human virtue
in society is counterfeit; only a child takes it for real money. But it is better to have
counterfeit coins than to have nothing in circulation; and finally they can be ex-
changed for hard cash, though only at considerable discount."
i5 Critique of Pure Reason, A 748 = B 776.
but because of Kant's moral egalitarianism. Whereas the philosophers
of the Enlightenment generally were convinced that sound ethics had
to be based upon knowledge and that progress in knowledge was the
precondition and almost the guarantor of moral progress but, as prac-
tical men, were willing to tolerate a philosophically untenable popular
morality (based on religion) as a kind of Interims-Ethik, Kant sharply
rejected both their premises and their conclusion. Kant, more than any
other philosopher of his age, respected the "ordinary moral conscious-
ness" of the ordinary man; under the influence of his early pietism and
of Rousseau, he came to regard the unshakable moral convictions of the
simple and humble as the proper starting point for philosophical analy-
sis; and philosophy, so far from being the moral teacher of mankind
(8 n. [93 n.]), is given the task of defending it from its outward ene-
mies-the philosophers of heteronomous ethics-and its internal dangers
-moral fanaticism and mysticism. 76
In this egalitarian context, even for children, education for morality
is successful only when conducted in the Socratic manner. Strictly
speaking, moral education is perhaps impossible, since morality is a
product of a sudden inward revolution in the manner of willing17 and
each act must be regarded as if it were an entirely fresh beginning. 7s
But for us men, in whom virtue is always defective, goodness is only a
continuous striving after the good,79 and this striving can be stimulated
and guided.
The "Methodology of Pure Practical Reason" describes what Kant
believed to be the most effective measures in this stimulation and guid-
ance. Fundamental to all of them is his conviction that the only incen-
tive to morality is respect for the moral law and that the more purely
this law is presented, the stronger is the incentive.so It is a mistake,
therefore, to recommend virtue on any grounds other than its intrinsic
worth, for if it is recommended for its utility, there will always be
smart people who can find other surer means to their advantage. s1
Equally to be avoided is "sentimental" education which inspires the
child to high-flown fancies and thoughts of merit instead of duty.82
76 Critique of Practical Reason, 70-71, 163 (162-63, 262); Foundations of the Meta-
physics of Morals, 405 (22).
77 Religion within the Limits of Reason Alone, VI, 47-48 (Greene and Hudson,
43-44)·
78lbid.,41 (3 6).
80lbid·,48 (44); Critique of Practical Reason, 157, 158, 159 (255,256,257).
81 Cf. UbeT den Gemeinspruch ... , VIII, 288.
82 Critique of Practical Reason, 86 ([79). Kant, himself a novel-reader. warns

against romancing (ibid., 155, 157 [253,255], Reflexion 7236).
Instead of persuasion and a vain attempt at instilling a moral disposi-
tion as if it were plasma that could be injected into the child's mind
and heart, Kant offers the Socratic method for bringing to light the
"darkly thought metaphysics of morals which dwells in the reasonable-
ness of each man."83 He then proposes a moral catechism, in the man-
ner of Rousseau's Emile, in which the child's mind is sharpened to the
intrinsic difference between moral dignity and the price of things. 84 He
proposed, as a practical measure, that educators should collect histori-
cal examples for such analysis; his own example is of efforts to suborn
a witness in the trial of Anne Boleyn.85 When the child's judgment is
made keen to the essentials of the action under examination, he begins
to take an interest in the law of reason by which his analytical powers
are exercised. Thus he is prepared for the final stage of his education.
Finally, the teacher will call the pupil's attention to the "purity of will
by a vivid exhibition of the moral disposition" in the examples. 86 In this
way the child is to be brought to experience the contentment and self-
respect that can spring only from virtue, and his "heart is freed from a
burden which has secretly pressed upon it" (161 [259]).
The Critique of Practical Reason, "which is only preliminary," does
not deal with the education for specific duties. This is a casuistical
problem that Kant takes up in the Metaphysics of Morals, the Lectures
on Ethics, and the Lectures on Pedagogy.
84 Examples are given ibid., VI, 480 ff., and in the Vorarbeiten to the same work,
XXIII,4 1 3-[5·
85 Warnings against examples in morality are warnings against the danger of
hysteron proteron as well as against heteronomy and Sturm und Drang (d. Foun-
dations of the Metaphysics of Morals, 408'"'9 [25], etc.). But they are useful in moral
education, because they are more effective than abstract precepts (Amhropologie,
§ 75)·
86 Critique of Practical Reason, 160 (259). For this "vivid exhibition" Kant quotes
(158-59 [257]) Juvenal Satires viii. 79-84, evidently one of his favorite passages, for
he used it again in Religion within the Limits of Reason Alone, VI, 49 n. (Greene
and Hudson, 45 n.) and in Metaphysik der Sitten, VI, 334.
Part III
XIII
The Dialectic of Pure Practical Reason; Com-

mentary on Dialectic, Chapters I and II
(except Sees. IV and V)
§ I. WHAT DIALECTIC IS
Both theoretical and practical reason have a dialectic and on the same
grounds, viz., as reason they seek the unconditioned for all that is
conditioned, but they cannot find it as an object of knowledge, though,
in default of a critique, it appears that they have found it (107 [101]).
Dialectic is the exposure of the illusion that the unconditioned, as re-
quired by reason, is an object of some definite and specific cognition.
The Critique of Pure Reason gives a full account of what dialectic
is. General logic, Kant tells us there, is sometimes employed as if it
were an organon for the production of knowledge instead of as a
canon for the institution and evaluation of inferences among cogni-
tions which have their source in experience. 1 As such an organon, it is
an art of giving to ignorance and sophistry the appearance of truth.
But instruction in such an art is beneath the dignity of philosophy;
and dialectic, he says, has come to mean the criticism of the 'illusions
arising from the sophistical art.
Transcendental logic likewise produces illusion when the categories
of the understanding, which give knowledge only when related to
experience, are used as an organon for knowing things in general,
whether given in experience or not. 2 The transcendental illusions which
arise in this way are not arbitrary and intentional, like those produced
by the sophistical dialectician for forensic purposes, but arise necessari-
ly from the inescapable, yet insatiable, requirements of thought. The
conditions of all judgment are the categories, but out where experience
cannot reach they are mere Ideas, rules, or maxims for the employment
of our reason. Unchecked by experience, they inevitably appear to be
21bid., A 63 =B 88.
139
240 The Dialectic of Pure Practical Reason
objective truths of the highest kind, since no experience can ever
refute them. They seem to be true, that is, until their illusoriness is
exposed by criticism, which prevents them from deceiving us, though
it cannot prevent them from arising, any more than a knowledge of the
laws of light can prevent our seeing optical illusions.3
The master method of exposing dialectical illusion is what Kant
calls the "skeptical method," which consists in "watching, or rather
provoking, a conflict of assertions, not for the purpose of deciding
between them, but of investigating whether the object of controversy
is not perhaps a deceptive appearance."4 The aim of the skeptical
method is not skepticism but certainty, since, by exposing illusion
where it exists, it will prevent it from infecting the entire organism
of reason with dubiety. By restricting the pretensions of theoretical
knowledge, it opens the way for the practical use of reason, and the
antinomy, which destroys speculative or theoretical metaphysics about
the supersensuous, is thereby the "most fortunate perplexity" into
which pure reason could ever fall (107 [203]).
§ 2. THE DIALECTICAL ILLUSIONS OF PURE PRACTICAL REASON
One might well be surprised that there is a dialectic of pure prac-
tical reason. In the first Critique Kant says that moral philosophy does
not lend itself to the skeptical method, since it can "present its prin-
ciples, together with their practical consequences, one and all in con-
creto, in what are at least possible experiences, and the misunderstand-
ing due to abstraction is avoided."1i While the going beyond sense
experience is the occasion for the theoretical dialectic, leading theory
into a "chaos of uncertainty, obscurity, and instability," the power of
practical judgment shows itself at its best where everything sensuous
is excluded. 6 There is no need to see whether pure practical reason,
like theoretical reason, "presumptuously overreaches itself"; it does
not "lose itself among unattainable objects or contradictory concepts,"
and pure practical reason stands in need of no critique (3, 16 [87, 102]).
Yet we are promised in the Introduction a Dialectic as the exhibition
and resolution of illusion in the judgments of practical reason (16
[ 102]). And this does not refer to the illusion that practical reason
in its empirical function overreaches itself and usurps the place of pure
practical reason; 7 for the section which fulfils this promise is clearly
3 Ibid., A 297 = B 353-54. 4 Ibid., A 423 = B 45 1• II Ibid., A 425 = B 453.
7 This is the sense of "dialectic" in the Foundatitms (405 [21]): from the conflict
of reason with the senses a "natural dialectic arises, a propensity to argue against the
stem laws of duty."
The Dialectic of Pure Practical Reason Z41
entitled "Dialectic of Pure Practical Reason." In this it seems to be

in open conflict with the first paragraph of the Preface and the
Introduction.
But practical reason is not a reason by which we know anything;
its reality is shown by an act, and, so far as the question of volition
is concerned, it always has "objective reality." We may say that, so
far as pure practical reason is practical, it has no dialectic and creates
no illusions because it issues no declarative statements; it only requires,
inspires, guides, and judges actions which may not be real but are
always possible. Still, so far as it is reason, it seeks the unconditioned
condition for its actions and judgments and decisions, but in doing
this it is theoretical reason employing "practical data."8 Kant himself
says as much in formulating the ultimate questions of philosophy.
"What can I know?" is a purely speculative question. "What ought
I to do?" is a purely practical one. "What may I hope?" is "at once
practical and theoretical, in such a fashion that the practical serves as
a clue that leads us to answer the theoretical question, and when this
is followed out, to the speculative question."9
While the Dialectic of Pure Practical Reason is not restricted to the
dialectical illusions arising from hope based upon the performance of
duty, this constitutes its principal problem. All the discussions in it
fit the formula that the practical as such is not dialectical and that the
skeptical method is not needed in practical philosophy as such. The
illusions are theoretical illusions about morality, not moral illusions.
The moral illusion, which is heteronomy, has already been exposed in
the Analytic (109 [205]).
Because the illusions to be exposed are theoretical, we cannot expect
so much novelty here or advance beyond the first Critique as we found
in the Analytic. Most of the problems have already been discussed in
the Dialectic of the earlier Critique, though to some degree with a
different outcome.
Explicitly, there is one illusion arising from the fact that practical
reason is reason and therefore seeks the unconditioned. It may seek the
unconditioned as the determining ground of the will in perfection or
the will of God (41 [[29-30]). This error has been dealt with in the
Analytic. Or, when the true unconditioned condition of morality (the
moral law) is given, it may seek the unconditioned as the totality of
the obj ect of pure practical reason in the concept of the highest good
and seek to know it theoretically (108 [203]). This is resolved in the
Dialectic.
8 Critique of Pure Reason, B xxii. 9 Ibid., A 805 =B 833.

There is, however, a far more important illusion that concerns Kant
in the Dialectic. This is the illusion that there is a necessary conflict
between theoretical and practical reason (I2 1 [2 I 8D and that the first
Critique forbade him to do what he does in the second. A major
part of the Dialectic is devoted to exposing this apparent conflict of
reason with itself.
§ 3. THE CONCEPT OF THE HIGHEST GOOD

The concept of the highest good (summum bonum) is the concept
of the unconditioned for the practically conditioned, i.e., the concept
of a supreme end which unites all other ends. Without it, there could
be no system of ends. As a putatively individual, concrete state of
affairs completely determined ~y its definition, it is called an Ideal
and not a mere Idea or concept.10 As a system of ends, it includes not
merely the form of the will but also its objects. While the Analytic,
in the doctrine of the moral good, taught that the moral good is the
sole object of pure practical reason, the Dialectic does not abstract
from all the diverse purposes of will but defines the condition under
which they can and must be synthesized in a single system. In doing
this, it requires the introduction of non-moral goods under a moral
condition, and the concept of this systematic connection of distinct
kinds of good is that of an a priori synthetic connection of morality
as bonum supremum with the totality of other goods (summed up as
"happiness"ll) into a whole called the bonum consllmmatum or the
complete good.
There are three important questions to ask about the summum
bonum interpreted as bonum consummatum: (I) Is it the determining
ground of the moral will? (2) Is there a moral necessity (duty) to
seek and to promote it? (3) How is it possible? We shall deal with
the first two questions in this section.
I. Is the highest good the determining ground of the moral will?
Repeatedly Kant says that it is not; yet, even so, his answer is not so
clear and unequivocal as one might wish. Consider the two paragraphs
at the end of chapter i of the Dialectic (IO()-IO [204-5 D. The first
tells us that only the moral law is the determining ground of the moral
10 Ibid., A 810 = B 838, A 840 =B 868.

11 The highest good is identified in other ways, too. It is the Kingdom of G<ld
(u8 [:ZZ4]), the intelligible world (131 [Abbott, 130, mistranslates this passage and
does not indicate the identification of the summum bonum with the intelligible
world]), the existence of natural beings under the moral law (Critique of /udJS11Zent,
§ 86), the moral vocation of man (Critique of Pure Reason, A 840 = B 868). These
need not be discussed separately.
Tbe Dialectic of Pure Practical Reason 143
will; any other makes for heteronomy. The second tells us that "the
concept of it and the Idea of its existence as possible through our
practical reason are likewise the detennining ground of the pure will."
But this is only because the moral law is included and thought in the
concept. That is, it is only the law as a necessary component of the
highest good that is the determining ground. 12
This is certainly an inept way of making one point twice; it means
that the highest good is not an independent determining ground of
the will in addition to or in place of one of its components. It cannot
be supplementary to it and leave the purity of will undefiled; but Kant
is unwilling to draw this conclusion in its full force. For, he tells us,
the moral will must have an object as well as a form, and, because of
the finite and sensible nature of man, the concept of the possibility of
the highest good is necessary to the moral disposition, but not to the
definition of duty.ls But if the "possibility of the highest good" means
anything more than its necessary condition, it is to that extent in-
compatible with what he has said earlier and more consistently about
the lawful form of the maxim itself being the object of the moral Will.14
12 Critique of Practical Reason, 109 (204-5). Since the object is not defined inde-
pendently of the moral law, having the highest good as an object does not ipso facto
entail heteronomy (ibid., 64 [I 56]). But it would entail heteronomy if it were a de-
termining ground independently of the moral law. Hence the happiness included
in the summum bonum may be the object, but it cannot be the determining ground,
although it might yet be said that the summum bonum is the determining gr<?und
inasmuch as it contains the law as its own condition. This is insisted upon in Uber
den Gemeinsprucb ... , VIII, 280 n.
IS Thus it is not possible to subject the whole course of life to moral maxims unless
reason connects with the moral law an outcome in exact conformity to our supreme
= =
ends (Critique of Pure Reason, A 812 B 840; d. A 811 B 839). The will requires
the conditions of the highest good as "necessary conditions for obedience to the
precept" of the moral law (Critique of Practical Reason, 131 [229]). "The subjective
effect of the law .•• the intention to promote the practically possible highest good,
at least presupposes that the latter is possible" (ibid., 143 [%41]). The law, which
causes only respect and does not recognize the need of man "to love an object,"
is widened at the behoof of this need so as to take up the final monl goal of reason
into its determining ground, "the synthetic enlargement of the concept of the law
taking place through [reference to] the natural character of man as a being of needs
who cannot be indifferent to the results of his actions" (Religion within the Limits
of Reason Alone, 6 n. [Greene and Hudson, 6 n.]). "Reason needs to assume the
highest good and its conditions to prevent the highest good and, consequendy, all
morality from being regarded as a mere ideal. which would be the case if the highest
good. the Idea of which necessarily accompanies morality, never existed" (Wbat
Is Orientation in Tbinking? VIII, 139 [Beck, 199]). Yet that the highest good is
necessary to give "firmness and effect to the moral disposition" is denied in UbeT
den Gemeinsprucb ... , YIn, %79.
14 Cf. above, pp. 133 ff.
Kant simply cannot have it both ways. He cannot say that the
highest good is a motive for the pure will, and then say that it is so
only under the human limitation that man must have an object which
is not exclusively moral (for there is nothing moral in happiness except
insofar as its condition is worthiness to be happy, and even then the
moral value lies in the worthiness, not in the enjoyment). The theory
of the Analytic requires him to deny that the concept of the highest
good provides an autonomous motive. While the hope for the highest
good may in fact be a necessary incentive to do that to which the
concept of duty would not move man, it is clear that to admit the
latter human-all-too-human-fact into the determination of conduct
in accord with moral norms is to surrender autonomy.
The existence or even the possibility of the summum bonum, there-
fore, cannot be held, in consistency with his settled views, to be logi-
cally or ethically necessary as a motive to genuine morality. At most,
the hope for it may be psychologically necessary to a semblance of
morality, perhaps to the highest semblance of it which can be attained
by man; belief in its possibility may be a legitimate accompaniment
of morality which is pure and autonomous. But whether such a "may"
could be changed to an "is"-this is an empirical question, and no
consistent answer to it seems to be found in the various parts of Kant's
works.
2. What is the relation of the highest good to the moral law with
respect to practice? Kant tells us that we are commanded by reason
to seek to realize the highest good. If the highest good were not pos-
sible, the moral law would be null and void. 111 It is for this reason that
it seems to him to be so essential to show the highest good to be
possible.
Yet Kant is almost casual in introducing his readers to this command
of reason. None of the formulations of the categorical imperative
have had this content. In the Metaphysics of Morals, where he will be
directly concerned with what the law requires of us, the highest good,
as he developed the concept here, is not among the "ends which are
also duties." And it is easy to see why this command of reason is not
fully expounded: it does not exist.
Or at least it does not exist as a separate command, independent of
the categorical imperative, which is developed without this concept.
For suppose I do all in my power-which is all any moral decree can
demand of me-to promote the highest good, what am I to do? Simply
act out of respect for the law, which I already knew. I can do abso-
lutely nothing else toward apportioning happiness in accordance with
15 Critique of Practical Reason, 1]4 (:uo). This is strongly denied elsewhere, e.g.,
ibid., 14Z-4J (z4r); Critique of Judgment, § 87 (V. 451 [301]).
The Dialectic of Pure Practical Reason 145
desert-that is the task of a moral governor of the universe, not of a
laborer in the vineyard. It is not my task; my task is to realize the one
condition of the summum bonum which is within my power (143 n.
[242 n.]); it is seriously misleading to say that there is a command to
seek the highest good which is different fr'im the command to fulfil
the requirements of duty.
The two questions (p. 242) must therefore be answered in the nega-
tive. The highest good is a synthetic concept, but all the moral con-
sequences drawn from it (as motive, as object) are drawn from one
of its members (bonum supremum), not from both (bonum con-
summatum).
The truth of the matter is that the concept of the highest good is
not a practical concept at all, but a dialectical Ideal of reason. It is not
important in Kant's philosophy for any practical consequences it might
have, for it has none except those drawn from the concept of bonum
supremum. It is important for the architectonic purpose of reason in
uniting under one Idea the two legislations of reason,18 the theoretical
and the practical, in a practical-dogmatic metaphysics wholly distinct
from the metaphysics of morals. Reason cannot tolerate a chaos of ends;
it demands the a priori synthesis of them into a system. An impartial
observer cannot approve of a disparity between happiness and worthi-
ness to be happy, but neither in nature nor in the moral law can Kant
find anything except a contingent connection between them. If, there-
fore, we are to conceive of a system of ends, as reason requires for
its own satisfaction (and not for obedience to a law which speaks with
commanding authority long before its credentials are presented), then
we must suppose that the highest good is possible. But we must not
allow ourselves to be deceived, as I believe Kant was, into thinking its
possibility is directly necessary to morality or that we have a moral
duty to promote it, distinct from our duty as determined by the form
and not by the content or object of the moral law.
§ 4. THE ANTINOMY OF PURE PRACTICAL REASON

If the reader accepts the conclusions that I have just drawn, he can
hardly attach to an antinomy in the concept of the highest good the
importance which Kant professes to find in it. Believing, however
inconsistently, that the moral law stands or falls with the possibility
or impossibility of the highest good,17 Kant gives a quite exaggerated
18 The assumption of the highest good is definitely assigned to theoretical reason

=
at Critique of Pure Rearon, A 80C) B 837, and Critique of Judgment, U 83-84; d.
also Critique of Pure Reason, A 839 = B 867.
17 Critique of Pure Reason, A 811 = B 839; Critique of Practiclll Reason, J 14 (ZIO),
importance to the antinomy that he believes he finds. He goes so far
as to say that the exposure of the "self-contradictions of the pure prac-
tical reason" compels us to undertake a critique of it (109 [204]),
assigning a role to this antinomy that we have seen the theoretical
antinomy actually had in the birth of the first. 18 This is no doubt a
mere fafon de parler, and when we examine the antinomy we shall
find that it is really quite a poor thing, wholly unable to carry this
great historical and systematic burden. We shall also find, regrettably,
that Kant's usual high-quality workmanship is not much in evidence
in the discussion of the antinomy.
I shall, first, give a summary presentation of Kant's argument and
then see what revisions it needs in order to make it actually conform
to his intentions.
The summum bonum is not a simple concept, as the Stoics and
Epicureans believed in identifying the motives to virtue with those
directed to the attainment of happiness. It is a synthetic concept with
two independent components,19 not a concept whose predicate is found
by analysis. But it is an a priori concept, since reason requires a nec-
essary connection between the two members, viz., happiness in propor-
tion to worthiness to be happy.20
As a synthetic concept, the connection between the components
must be that of ground and consequent. There are two possibilities:
thesis-the desire for happiness must be the necessary and sufficient
ground of or motive to morality; antithesis-the maxim of virtue must
be the necessary and sufficient condition of happiness.21 The first is
absolutely impossible, as shown in the Analytic. The second is impos-
sible in the world as we know it in experience, for here happiness
depends upon circumstances and prudence, not upon mere intention
and purity of heart. Any connection between them in this world is
contingent and accidental, and this does not satisfy reason's demand
for a necessary connection. Hence the connection is neither analytic
nor synthetic a priori, and the concept of the highest good is therefore
18 Cf. above, p. 15, n. 4. In Prolegomena. § sza, however, he asserts the unique-
ness of the theoretical antinomies.
19 Both Stoics and Epicureans denied this, differing only in respect to what was
the analysandum (Critique of Practical Reason, II z [z08]).
20 Ibid., J I I, 113 (z07, 20<). I wish to correct the error of having omitted the word
"pure" in both editions of my translation. The first sentence of the section beginning
on p. 113 (z07) should read: " •.• assumed by a pure practical reason ...."
211bid., 113 (209). Note that the identification of these two antinomic proposi-
tions as thesis and antitheses is not made by Kant himself and is a matter of dispute,
as pointed out later in the text.
"impossible before reason." With its exposure, the moral law, which
commands that we pursue it as a necessary object, is discovered to be
nugatory.
The second of the possibilities, however, is not absolutely false. It
is false only under the assumption that the connection is one that
must be found in accord with the laws of nature; and even then it is
only subjectively false, i.e., our reason cannot conceive how it would
be possible, but it cannot show that it is impossible (145 [244])' Just
as the antinomy of freedom was resolved by assigning the truth of the
thesis to man's supersensible nature, here it is resolved by assigning
the connection between the two components to an intelligible world.
Thus the possibility of the highest good is established,22 if we assume
that there is an intelligible world governed by a moral being with
requisite power to apportion happiness in accordance with virtue.
Thus Kant. But it should be obvious that we do not have here an
antinomy in any strict sense. First, in spite of his statement (10<) [204])
that he is concerned with the "self-contradiction of pure practical
reason," the two propositions are not contradictories. Second, each
does not have an independent warrant;23 one of them is false on its
face. The whole antinomy is devised and artificial. Third, its resolution
is not what one would expect from the proffered analogy to the third
antinomy. For there the antithesis is vindicated in the phenomenal
world and the thesis in the noumenal; here the thesis is not vindicated
at all.
Even so, it is possible to reformulate the antinomy so that it better
conforms to Kant's purposes. I shall give two such statements:
I. THESIS: The maxim of virtue must be the cause of happiness.
ANTITHESIS: The maxim of virtue is not the efficient cause of happiness;
happiness can result only from successful use of knowledge of the laws
of nature.
22]bid., 119 (11 5). The argument has been interrupted by a discourse on the con-
nection between happiness and virtue not mediated by the intelligible world (pp.
115-19 [ZII-I5]). These paragraphs might well be set apart from the rest and called
a "Remark." We have already commented upon them in chap. xii. The connection
between virtue and happiness in the highest good is made possible in an intelligible
world of perfect beings without the mediation of God, according to the Critique of
=
Pure Reason, A 810 B 838, where it is called a "system of self-rewarding morality."
But for us men this is not sufficient, or, rather, we are not sufficient to it.
23 Each proposition in an antinomy is "not merely in itself free from contradiction
but finds conditions of its necessity in the very nature of reason-only that, un-
fortunately, the assertion of the opposite has, on its side, grounds that are just as valid
and necessary" (ibid., A 4z I = B 449).
This is a real antinomy, since the propositions are contradictory,
not contrary; each expresses an inescapable interest of reason (moral
and theoretical); and each is a true formula for one of these interests.
Furthermore, the resolution of this antinomy conforms to that of the
third antinomy in the first Critique. The antithesis is and remains
(unfortunately for the virtuous) true of the sensuous world; the wind
is not tempered to the shorn lamb, and the rain falls on the just and
the unjust alike. The thesis is false in "a system of nature which is
merely the obj ect of the senses" (II 5 [21 I]). But it may be true if
there is an intelligible world in which there is a moral government
distributing happiness (in the sensuous world, under natural condi-
tions)24 according to laws holding in the intelligible world (laws of
the moral intentions of persons). Hence the pursuit of virtue does not
rule out the hope, based upon the possibility, of a proportionate
happiness.
Another statement of the antinomy 25 is possible, and it has the merit
of keeping the summum bonum in the center of attention throughout.
II. THESIS: The summum bonum is possible. Proof: The rnorallaw requires it.
ANTITHESIS: The summum bonum is not possible. Proof: The connection
between virtue and happiness is neither analytical nor synthetic a priori
nor empirically given.
RESOLUTION: The antithesis is true of the sensible world if the laws of
nature have excusive sovereignty; the thesis may be true of the intelli-
gible world because the synthetic connection of virtue (as ground) to
happiness (as consequent) is not absolutely impossible.
In either formulation we are enabled to see that the concept of the
highest good is not impossible. Kant, believing that it is morally nec-
essary, then turns to those conditions which would make it actual
if superadded to morality, which is the only one of the conditions
that lies within our control and competence.26
24 Kant says that the effect must be in the sensuous world. He therefore does not
base the postulate of inunortality upon the need for happiness in another world
as the locus of this reward. But in this life the adjustment is not made. The four
paragraphs which I have suggested should be set apart as a "Remark on a Self-re-
warding System of Morality" do not speak of the moral contentment in this life as
being happiness. This would suggest that the happiness in the summum bonum must
be that of an afterlife, not under a system of nature. So, even before we are ready
to take up the postulate of immortality, we run into an obscurity, the first of many
we shall find in this notion (cf. below, chap. xiv, ~ 4).
25 Slightly modified from a proposal by August Messer, Kants Ethik, p. 88.
26The twofold division of his task, mentioned on p. 1I0 (205), seems to refer to
this; but this division did not actually serve to articulate the succeeding sections
of the Dialectic. Perhaps Kant had intended a more extensive discussion of the postu-
late of freedom prior to the introduction of the other two postulates; then this
division would apply.
§ 5. THE PRIMACY OF PURE PRACTICAL OVER THEORETICAL REASON
The question is: Have we a right to use concepts of reason, and
to assert objects for them, which are beyond the sphere of knowledge
marked off for and by theoretical reason? It is not sufficient to point
out merely that the interests of practice extend beyond those of theory.
Theory alone does not decide between the thesis and the antithesis of
the third antinomy, and theory would leave us in a state of vacillation
which could be ended, if theory were the only use of reason, by a deci-
sion forced upon us by the needs of practice. 27 This decision would in-
deed be a decision in favor of the thesis. The failure of theoretical reason
to decide between thesis and antithesis would pennit us to employ the
requisite "faith" in the claims of dogmatism but could not give faith
the "form of science," i.e., show its systematic rational strucrure. 28
There is not now, after the Analytic's demonstration of the function
of pure practical reason, any remnant of arbitrariness or irrationality
left in the decision that must be made. We do not have the contrast
between "reason" and "faith" but that between "theoretical reason"
and "rational faith," i.e., an a priori faculty that goes beyond theory.
A contrast exists, but not necessarily an opposition, because (a) there
is only one reason with two interests, not two reasons with opposing
interests (12 I [2 I 7] ), and (b) the practical does not require us to go
against anything that the thtoretical reason has established or can
establish. Put in classical theological terms: the things of faith are be-
yond the things of theoretical reason but not beyond the things of
reason simpliciter; and there are no valid things of faith that are against
the things of theoretical reason.
Theoretical reason pushes inquiry to the farthest limits in striving
to reach the unconditioned. It cannot do this by attaining knowledge
of the unconditioned, but its interest is whatever furthers the exercise
of this function. 29 Its true interest, therefore, is not in some special
cognition, or definition, of the unconditioned but solely in the restric-
tion of speculative folly. The interest of practical reason is the "de-
termination of the will with respect to the final and perfect end" (120
27 Critique of Pure Rellson, A 475 = B S03.
281bid., A 470::;:: B 498.
29 There is an inconsistency in Kant's use of the word "interest," He identifies
interest with (a) the end of the use of a faculty and (b) a principle under which
its exercise is advanced. Thus, of theoretical reason, a is "the knowledge of objects
up to the highest a priori principles," and b is the establishment of limits which
cannot be breached by "monstrosities of reason" which interfere with its orderly
progress. Its interest in sense b is the restriction of speculative folly (Critique of
Practical Reason. no, IZ J [117]). The interest of theoretical reason appears to the
dogmatic metaphysician to be iI; to the critical metaphysician like Kant, it is b.
2. 50 The Dialectic of Pure Practical Reason
[216]), and whatever shows this end to be possible furthers the prac-
tical interest.
The two interests may superficially appear to be in conflict, but
they are not opposed to each other if the interest of theoretical reason
is not some end which it cannot achieve (knowledge of things in
themselves) but only the guaranty of the orderly progress of science,
and if the interest of practical reason is not presented as cognition
but as determination of actions. Then judgments about what furthers
the latter interest can be compatible with the former interest. These
judgments will not be cognitive judgments known theoretically to be
true; they will be only postulates.
The Critique of Pure Reason itself prepares the way for the assertion
of the primacy of the practical over the theoretical interest of reason.
The interest of speculative reason in the Ideas needed by pure practical
reason is very small;30 yet the Ideas needed by speculative reason for
its own purpose are the same as those needed by pure practical rea-
son;31 and theoretical reason can present them only as problematic
concepts.S2 Practical reason, if there were a conflict between it and
theoretical reason, would enter the contest with the "right of posses-
sion" on its own side.ss In the absence of opposition, the question
concerns solely which maxim should be followed: to refuse stubborn-
ly to admit any non-empirical objects (the canon of Epicurus) or to
"admit certain theoretical positions" inseparably bound with a priori
practical principles but going "beyond any possible insight of the
speculative reason (although not contradictory to it)" (120 [:u6]).
The doctrine of primacy prevents what is a difference from becoming
an incompatibility, by establishing an order of subordination instead
of co-ordination. And in this subordination the practical has primacy
because "every interest is ultimately practical."3~
If theoretical reason can be assured that these "theoretical positions"
are required by the practical interest and exceed, but do not conflict
with, its own interest, this primacy is assured. Kant claims that the
requirement is shown by the resolution of the antinomy of practical
reason and the necessity of the conception of the summum bonum.
Sections 6-9 of this chapter of the Critique are devoted to the establish-
ment of the absence of conflict.
30 Critique of Pure Reason, A 74)8 = B 816. =
31/bid., A 7¢ B 814.
32/bid., A 796 = B 814: "Reason has a presentiment of objects which possess a
great interest for it. But when it follows the path of pure speculation. in order to
approach them. they fly before it."
s3/bid., A 777 = B 805.
=
IU Ibid., A 816 B 844; Critique of Practical Reason, 111 (u8).
The Dialectic of Pure Practical Reason 15 I
§ 6. THE POSTULATES OF PURE PRACflCAL REASON 35
Kant warns us at the beginning36 against confusing "postulate" as he

uses this word in the Critique with the usage of the mathematician.
Since the mathematician does not now mean what Kant says he meant
then, his explanation, meant to clarify matters, succeeds (through no
fault of his own) in making them only more obscure. It is therefore
necessary to speak more fully of mathematical postulates.
In mathematics, a postulate was, for Kant and his contemporaries,31
an indemonstrable practical (i.e., technically practical) proposition
giving a rule for the synthesis of an object in intuition, when the
possibility of the object is known a priori. Mathematical proof is by
means of constructions of objects in intuition, and postulates are the
a priori rules for such construction. For example, the following is
called a postulate, what is postulated being the possibility of the action
of constructing a circle: "Around the center C, with radius CA or
diameter AB, to describe a circle."38 The use of the word "postulate"
in more recent mathematics differs from that in Kant's theory of
mathematics and, in fact, closely resembles the meaning that it has in
his moral philosophy. This is because recent mathematics has become
"deductive" instead of "constructive," and a postulate is an assumption
made in order that some mathematical purpose may be achieved. Thus
instead of the postulate being, in the eighteenth-century manner, "To
construct a parallel to a line through a point," it is now the assumption
"Through a point external to a line in a plane, one and only one
parallel line can exist." A postulate now differs from an axiom, if at
all, only in not being intuitively certain and resembles it in being
theoretical instead of practical. With the general acceptance of a game
theory of mathematics, the difference between axiom and postulate is
only verbal. For Kant, on the other hand, axioms differed from postu-
lates in being both intuitively certain and theoretical rather than prac-
tical. No philosophical proposition was, for him, axiomatic. 39
In philosophy, however, a postulate is an assertion of the possibility
35 The specific postulates will be discussed in chap. xiv.
36 Critique of Practical Reason, II n. (96 n.). Cf. Critique of Pure Reason, A 234
:::: B 287: "In mathematics a postulate means the practical proposition which contains
nothing save the synthesis through which we first give ourselves an object and gen·
erate its concept...." Also Vorlesungen fiber Logik, § 38, and Reflexion 4370.
31 Similarly, Wolff, Philosophia rationalis sive logica, § 269, and Baumgarten,
Acroasis logica, § 250.
38 CE. Gottfried Manin, Arithmetik und Kombinatorik be; Kant (Itzehoe, 1938),
p. 19; d. also p. 21.
39 Critique of Pure Reason, A 731 = B 760; Prize Essay, II, 178 ff. (Beck, z64 tf.).
or acruality40 of an object as a corollary to the acknowledgment of a
necessary law. The practical proposition is the law (or rather its
associated imperative); the postulate is a theoretical proposition, but
it is not a proposition that is theoretically (i.e., apodictically) certain.
Kant likewise discriminates between hypothesis and postulate. 41 Giv-
en some X, whether merely actual or necessary, Y is postulated if Y
is known a priori to be the only condition under which X is known to
be even possible; or Y is hypothesized if it is merely assumed as an
explanation for X.42 Sometimes Kant writes as if the difference lies in
the nature of the purposes involved in the explanation of X. If X is
necessary and Y is a priori essential to it, then Y is postulated; if knowl-
edge of X is an arbitary (theoretical) purpose, Y is hypothesized, even
though it is necessary to it.43 This is Kant's meaning in saying that we
must postulate, e.g., the existence of God because morality (the "X"
of our schema) is necessary and God is necessary to it, while the
existence of God is only hypothesized in natural theology or specula-
tive metaphysics-not because the concept is any less essential there
but because the purpose is arbitrary. Moreover, a hypothesis is a
species of cognition which can be raised to certainty by increase in
knowledge (e.g., a scientific hypothesis) or, if it is based only on a
subjective need of reason (as is the hypothesis of God's existence in
natural theology), it is a mere hypothesis that cannot possibly be
raised to the dignity of knowledge, though it claims to be knowledge.
A postulate, on the other hand, is not subject either to supplementation
or correction by new facts, nor does it, when properly understood,
claim to be an objective cognition. Still, it is "in degree inferior to no
cognition even though in kind it is wholly different."44
A postulate in the Critique of Practical Reason, then, is "a theoretical
proposition which is not as such demonstrable, but which is an in-
separable corollary of an unconditionally valid practical law" (122
[219]). The law demands an action but not the acceptance of any
40 Each is asserted in Critique of Practical Reason, 134 (231-32); d. also below,
p. 259, nn. 1 and 2.
41 Also between a presupposition and a postulate: "At some future time we shall
show that the moral laws do not merely presuppose the existence of a supreme
being, but also, as themselves in a different connection absolutely necessary, justify
us in postulating it" (Critique of Pure Reason, A 634 = B 662). Similarly, Critique
of Practical Reason, 141 (240).
42 Ibid., A 633 = B 661. Accordingly, a postulate is called a "necessary hypothesis"
in Critique of Practical Reason, lIn. (96 n.).
43 Critique of Practical Reason, 1 I n. (¢ n.); What Is Orientation in Thinking?
VIII, J4I (Beck,300-30I).
4. What Is Orientation in Thinking? VITI, 141 (Beck, 301).
The Dialectic of Pure Practical Reason 2.53
theoretical corollary. We accept the corollary not because it is com-
manded but because, in putting ourselves under the moral law, "the
principle which determines our judgment, while subjectively a need,
is the ground of a maxim of moral assent, as a means of promoting
that which is objectively (practically) necessary."45 That is, I am
commanded to seek the highest good; I can do so only under the
condition, objectively, that the highest good is possible and, subjective-
ly, that I believe it is possible; but I am not commanded to believe
anything (144 [242]). I am commanded to do something I could not
do and would not recognize that I ought to do, unless I held, without
objective proof, that certain things are possible.
If, however, as we have argued in § 3, the command to seek the
summum bonum is merely the command to do my duty, the close
connection of the moral law with the postulates (with the exception
of that of freedom) cannot be established. While Kant thinks that
the connection is a rational one, it is, so to speak, subjectively rational,
like the need to assume the existence of God in speculative meta-
physics (as an object of a hypothesis). It has only "practical reality,u
which means merely "validity for practice" and not for knowledge.
In fact, however, even this is to claim too much, for it is only because
of the natural character of man that he needs any supplement (in the
form of belief) for the performance of his duty.46 In the light of this
fact, it is hard to agree that Kant has met the difficult issue posed by
Wizenmann (143 n. [242 n.]).
45 Critique of Practical Reason, 146 (144). In Zum ewigen Frieden in der Philoso-
phie (VIII, 418 n.; not to be confused with Perpetual Peace) it is said that a postulate
is an imperative, commanding a cenain maxim and not directed to an object. While
this conforms to Kant's statement that the law (meaning the imperative) is a prac-
tical postulate (Critique of Practical Reason, 46 [135], denied at 132 [lZ9]), it is not
possible to reconcile this isolated statement with the full teaching of the Critique.
Indeed, it is not possible even to reconcile this note with the teaching in the body
of the essay in which it is found, according to which a postulate is a practically
necessary theoretical proposition about an unknowable object.
46 "All men could have enough [incentive to the fulfilment of their duty] if they
would only hold themselves [as they ought] merely to the precept of pure reason
in the law. What do they need to know of the outcome of their moral actions,
which the course of the world will bring about? For them it is enough that they do
their duty, though it be that this life is all and that in it happiness and wonh never
come together. But because of the inescapable limitations of man ...." (Religion
within the Limits of Reason Alone, VI, 6 n. [Greene and Hudson, 6 n.]). See Foun-
dations of the Metaphysics of Morals, 411-Il (28), where it is stated that moral
philosophy abstracts, even more than theoretical philosophy, from the peculiar
nature of human reason. The doctrine of the postulates is obviously at variance with
this; the statement in the Foundations concerns the formulation of the law, not that
of duty and the postulates.
254 The Dialectic of Pure Prllctical Reason
The need of reason to believe in the existence of a highest good and
to postulate the existence of its conditions does indeed arise from "an
objective determining ground of the will," but only because of "in-
escapable human limitations." It is therefore not a "need of pure
reason" but a need of the all-too-human reason. Yet this criticism of
Kant's reply to Wizenmann does not consign Kant a place among the
pragmatistic proponents of a "will to believe," since James, as a typical
example, held the "right to believe" to spring from a "will to believe"
precisely because pure reason could not, for him, be practical. The
pragmatic need to believe rests on a failure of reason to demonstrate
what we men must believe to be true if we are to be moral. In this
predicament the pragmatist claims that we have the right to "consult
our passional nature," whereas Kant says we consult reason, which is
more than a theoretical cognitive faculty. Hence Kant has, or claims
to have, a rational criterion of and motive for belief, which is lacking
to the pragmatist. But, if my argument is correct, there can be no need
of pure reason for the postulates, not because pure reason may not
have some needs but because its need is exhausted in issuing the moral
command. If reason is taken in the broader sense of concern with
the promotion of the good under human limitations, the difference
between his view and that of the pragmatists becomes chiefly verbal,
and Kant answered Wizenmann no better than James answered his
critics.
But let us return to Kant's text. 47 Propositions which were prob-
lematical for speculative reason are now shown to have objects (134
[232 D, since, if they did not, their non-existence would render the
law null and void.4 8 In willing to do what duty requires, I do not know
that there is a God, though I am, in a literal sense, morally certain of
it; it is not certain, but I am certain. 49 In willing to do what duty
requires, I act as if there were a God, or "I will that there be a God."50
I am not thereby enabled to have any insight into the nature of God
or the other conditions of the highest good; no synthetic propositions
47 The argument of chap. ii, S vii, is a continuation of chap. i, S ii of the Analytic
(50-57 (140-47])·
48 "I inevitably believe in the existence of God and in a future life, since my
moral principles would [otherwise1 be themselves overthrown, and I cannot dis-
claim them without becoming abhorrent in my own eyes" (Critique of Pure
Reason, A 828 = B 856).
49 Ibid., A 829 = B 857.
50 Critique of Practical Reason, 143 (241). It is on this point that the fictionalist
interpretation of Kant is based. But Kant does not meet the entrance requirements
of the fictionalist school: the existence of God is not theoretically false and to be
accepted only as a practical fiction.
can be made about them, for no intuition occurs to connect my con-
cepts with their objects. Theoretical reason gains nothing by this
addition, for it cannot make any use of the concepts which practical
reason hands to it beyond the use it already had for them as regulative
concepts (13 6 [233-34]). Hence the moral argument does not breach
the walls that theoretical reason so solidly built between what can
and what cannot be known. The task of theoretical reason with respect
to these objects is and remains merely negative, a police function: 51
to prevent fanaticism, to purify them of anthropomorphic elements
which would sensualize them and destroy even their moral function
(141 [239]), and to ward off dogmatic attacks on them. 52
§ 7. BELIEF AND IGNORANCE

The epistemic mode in which a postulate is held is a "faith of pure
practical reason," "moral belief," or "rational belief."53 Through this
epistemic mode, Kant encompassed within philosophy what had previ-
ously been an epistemic resource outside of, or hostile to, philosophy.
The faith-reason contrast before Kant was parallel to that between
revealed theology and philosophy. With Kant, several kinds of faith
are distinguished, and one of them is tamed and domesticated into
philosophy.
Faith or belief, however rational it is, remains contrasted with
knowledge. Knowledge is assent on grounds that are both objectively
and subjectively sufficient; mere opinion is assent on grounds that are
neither objectively nor subjectively sufficient; but faith is assent on
grounds that are subjectively sufficient in spite of being objectively
insufficient. 54 The subjectively sufficient grounds may be need in
speculation (doctrinal belief), revelation and ecclesiastical dogma
(historical belief), need for the purposes of some arbitrary use of
skill (pragmatic belief), or reason's needs for the satisfaction of duty
(rational belief, a belief of pure reason). The former are grounds of
theoretical beliefs; the latter, of practical beliefs.55
51 Critique of Pure Reason, B xxv.
52 Ibid., A 829 = B 857.
53 Critique of Practical Reason, 126, 144, 146 (223, 242, 244); Critique of Pure
=
Reason, A 828 B 856; What Is Orientation in Thinking? VIII, 141 (Beck, 300) i
Foundations of the Metaphysics of Morals, 462 (83),
54 Critique of Pure Reason, A 821 = B 850'
55 Ibid., A 825 =
B 853; What Is Orientation in Thinking,? VIII, 141 (Beck,300).
That the distinction between doctrinal and rational belief may have arisen in Kant's
response to Hamann's version of Hume is suggested by Philip Merlan, "Hamann
et les Dialogues de Hume," Ret/ue de metaphysique et de morale, LIX (1954),
28 5-89.
The faith of pure reason is needed to "orient ourselves"116 in the
"empty space" of thought beyond experience, in that "wide and
stormy ocean, the native home of illusion, where many a fogbank and
many a swiftly moving iceberg give the deceptive appearance of
farther shores."117 It must be a faith in objects of practical reason,
since theoretical reason, in default of intuition, gives no point of
orientation. Subjective needs, which vary from man to m:m, have no
place in the determination of these objects, for they all concern objects
of experience whereof knowledge can be attained. In this, Wizenmann
was correct (143 n. [242 n.]). But to be subjective does not mean to
be arbitrary and contingent; "subjective" means dependent upon the
nature of the subject, and this can be interpreted either a priori or a
posteriori. Subjective needs require satisfaction and may be validly
satisfied after knowledge has reached the end of its tether; that is one
reason why it is so important to determine a priori the difference
between a mere boundary and a limit of knowledge. 58 Only when we
are sure that theory has been brought to a stalemate, can the resource
of faith be used.
This faith cannot be commanded, and it is not essential to have it
in order to be mora1. 59 But to have it is to enjoy a positive moral
worth,60 and morality, though compatible with "doubtful faith," is
not compatible with a dogmatic unbelief which would deny possibility
to the object of pure practical reason. 61 But in no sense does faith
precede morals as premise, and whatever faith contains as a surplus
beyond what is needed for moral exercise is superstition and fanaticism.
The history of religion is the history of the gradual replacement of
historical faith by its kernel of rational faith, which the Christian
religion has contained, but in a masked way.62
But is not faith a very paltry substitute for the kind of knowledge
that the first Critique showed to be impossible? Not at all, Kant
116 What Is Orientation in Thinking? VIII, 136 n. (Beck, 296 n.) i cf. Critique of
=
Pure Reason, A 475 B 503.
117 Critique of Pure Reason, A 236 = B 295i cf. also A 395-¢.
118 Prolegomena, ~ 59.
119 Critique of Judgment, § 87 (V,451 (3031), But one who does not have it must
surrender a purpose he ought to have. Spinoza is Kant's example of such a truncated
moral philosopher.
60 Fortschritte der Metaphysik, XX, 298.
61 Critique of Judgment, S 91 (V, 471 [325]).

62 Religion within the Limits of Reason Alone, VI, 116, 118, etc. (Greene and
Hudson. 106, 109, etc.); MutmassIicher Anfang der Menschengeschichte, passim;
Streit der Fakultiiten, VII, 38-47.
ans·wers. \Vhile in the precritical period he did indeed believe knowl-
edge of the supersensible was both possible and compatible with needs
of morals (though not essential to them), in the critical period he not
only denied that it was possible because of the dialectical illusions but
asserted this to be a victory of reason, and not merely of practical
reason. For if knowledge of the supersensible were possible, it would
be merely an extension of theoretical knowledge and would carry into
the supersensible realm the procedures and conclusions of science in
the world of nature; and this would be at most a quantitative ag-
grandizement of reason but would exclude a more extensive, but quite
different, employment of reason-the practical. In order to give reason
full scope for its development, therefore, Kant says, "I have ... found
it necessary to deny knowledge in order to make room for faith."63
The price paid, however, for saving morals from the dogmatism of
metaphysics is ignorance, as the other side of faith. But suppose it were
possible to have knowledge of the supersensuous and suppose, further-
more, that the objects I knew there were not, in fact, objects under
laws of phenomena. Suppose, that is, that it were possible to establish
a rational, speculative metaphysics which would not be a foundation
for "all that unbelief, always very dogmatic, which wars with morali-
ty";64 suppose that we had the "elevated knowledge" of God, freedom.
and immortality that some boasted of and thereby gained Kant's
withering contempt 65-what then?
A theme celebrated in the Faust tradition66 is elaborated in the last
section of the Dialectic. 67
Granted all these impossible suppositions: this knowledge, if the
accession to it did not bring with it a change in our "whole nature,"
would change rational faith into pragmatic faith. and virtue would be
impossible (146 [245])' "Most actions conforming to the law would
be done from fear, few would be done from hope, none from duty.
The moral worth of actions, on which alone the worth of the person
and even of the world depends in the eyes of supreme wisdom, would
not exist at all. The conduct of man, so long as his nature remained
63 Critique of Pure Reason, B xxx. Similarly, Coleridge: the existence of God
"could not be intellectually more evident without being morally less effective; with-
out counteracting its own end by sacrificing the life of faith to the world mecha-
nism of a worthless because compulsory assent" (Biographia iiteraria, chap. x).
64 Critique of Pure Reason, B xxx.
65 Critique of Practical Reason, 5 (89-90); cf. Prolegomena, IV, 373 n. (Beck,
122 n.).
66 Especially Chamisso's Faust (1803).
61 Cf. also Crjtique of Pure Reason, A 476 =B 504; A 80J =B 829.
2S8 The Dialectic of Pure Practical Reason
as it now is, would be changed into mere mechanism, where, as in a
puppet show, everything would gesticulate nicely but no life would
be found in the figures" (147 [245], italics added).
Kant said, in the first Critique, that the moral interest was on the
side of dogmatism (the side of the theses). That the thesis should be
true is indeed the interest of pure practical reason; but that it should
be known to be true conflicts with the moral interest under the human
limitations. Kant said, as a young man, long before his moral theory
was developed, "It is necessary that one be convinced of the existence
of God; but not so necessary that one should prove it."68 Now the
old man discovers the deepest moral root of this Faust-like apprehen-
sion of the danger of knowledge of secrets of the other world-secrets
he himself had played with when he wrote his Dreams of a Spirit-Seer.
In the absence of knowledge, respect for law allows us a conjecture
or a glimpse into the mystery;' but "God and eternity in their awful
majesty" do not stand "unceasingly before our eyes" so as to blind
us to our precariously human condition and to obliterate that state of
mind in which we imperfect beings, naturally bent upon happiness,
can exercise our autonomy. "It is therefore good that we do not know
there is a God, but only believe it."69 Against the Enlightenment's
exaggeration of the saving power of knowledge, which was a securali-
zation of pre-Enlightenment exaggeration of the faith in and knowl-
edge of God, Kant stands between Rousseau and Goethe: "Ein guter
Mensch in seinem dunklen Drange / 1st sich des rechten Weges wohlbe-
wusst."70 Other knowledge of the supersensible he does not need arid
should not have.
68 Einzig miiglicher Beweisgrund .. . , end (11,163)'
69 Refiexion 4996; d. Critique of Practical Reason, 148 (146).
70 FQUst, PCQlog im Himmel.
XIV
The Postulates of Pure Practical Reason;

Commentary on Dialectic, Chapter II,
Sections IV and V; and Conclusion
§ I. INTRODUCTION
In chapters xi and Xlll it was mentioned that Kant uses the word
"postulate" rather loosely and that his list of practical postulates varies
from place to place. There is some diversity even within the Critique
of Practical Reason. The possibilityl and the actualit y 2 of God, free-
dom, and immortality; the possibility of the summum bonum;3 and the
moral law itself4 are named as postulates. But, in spite of this diversity,
Kant's meaning and intention in the Critique are clear: the three
postulates, officially and deliberately so called, are the actual freedom
of the will, the immortality of the soul, and the existence of God.
We have already discussed the concept of freedom and the sense
in which freedom is a postulate. 5 There remains, therefore, in this
chapter to discuss only the two remaining postulates. We shall begin,
however, by examining once again the concept of postulate in the
light of how a postulate is "demonstrated," then examine the two
postulates themselves, and conclude with a discussion of the relation
between ethics and religion. The very last section comments upon
Kant's conclusion to the whole work, which, though it follows the
Methodology, follows smoothly upon the Dialectic.
1 Critique of Practical Reason, 134 (132): " ..• Through the practical law, which
requires the existence of the highest good possible in the world, there is postulated
the possibility of those objects of pure speculative reason,"
2 Ibid., next sentence: "Practical reason inexorably requires the existence of those
objects for the possibility of its practically and absolutely necessary object, the
highest good." Ibid., 125 (221): " .•. The postulate of the possibility of a highest
derived good (the best world) is at the same time the postulate of the reality of a
highest original good, namely, the existence of God" (all italics supplied).
3 Cf. the just preceding quotation.
4 Ibid., 46 (135); denied at 131 (229). 5 In chap. xi, ~ 14.
260 The Postulates of Pure Practical Reason
§ .:. IHI'. SfRL'CrCRE OF A 1\lORAL ARGUME;\" r
Kant does not use the words "moral argument" in the second Cri-
tique, but it is the name by which he refers to this argument in the
third. Although it is not an entirely suitable name, it has become so
widely accepted that it is useless to try to reject it; but we must try
to be very clear in deciding what a "moral argument" is; else we shall
find the term so vague that it covers some arguments that are logically
quite different from the argument first called a "moral" one.
Ic is universally recognized that a value-judgment or a practical
proposition containing an "ought" cannot be derived by an argument
from premises which are exclusively factual. Enthymemes in which
this is done always require a hidden value-premise, which is omitted
simply because it is thought to be obvious. Only a philosopher would
object to an argument that said, "You ought to take an aspirin because
you have a headache." But even he would make himself tedious if he
insisted upon completing such syllogisms. Still, such an argument is
not valid unless the value-premise is supplied.
A moral argument is one in which a factual conclusion is reached
from a value-premise, either immediately or taken in conjunction with
a factual premise. 6 Such an argument likewise is formally invalid. It
cannot be expected that a philosopher of Kant's skill and stature should
be guilty of such a sophomoric blunder; yet the ordinary interpretation
of the moral argument attributes this nonsense to him.
Italicizing two words in Kant's definition of a postulate shows him
to be aware of the danger of this kind of error, though whether he
finally avoided it in its most subtle and insidious form is not a matter
that can be decided by examination of a single sentence of his text. He
says a postulate is "a theoretical proposition which is not as such
demonstrable, but which is an inseparable corollary of an uncondi-
tionally valid practical law" (122 [219])' We must now examine this
"as such."
Whether a proposition is theoretical or practical depends ultimately
upon its function, not upon its formal structure or content, though we
have the means, syntactic or semantic, to make these functions clear.
"A causes B" is a theoretical proposition; "Since you wish to have B,
do A" is a practical proposition; "A is a means to B" has some of the
characteristics of both. Just as we can move from theoretical to prac-
tical propositions by supplying a practical premise (e.g., "If you want
X, you ought to do what will lead to X"), it would appear that we
e If the value-tenn is the middle, then, of course, a factual conclusion can be

validly reached from two value-premises.
The Portulater of Pure Practical Reason 261
might move from a practical to a theoretical proposition by supplying

another premise. The only premise that would serve, however, would
be a value-premise, giving a syllogism in which the value-term was
the middle. We have already seen that the moral law is a theoretical
proposition describing a priori how a pure rational being would act
and that the categorical imperative expresses this law in imperative
form (as a practical proposition) for rational beings whose will is not
pure. We can easily state a syllogism from which the theoretical state-
ment about a pure rational being can be derived from the practical
imperative for us: The maxim I ought to act upon is one that a ra-
tional being would act upon; the maxim I ought to act upon is Y;
therefore, Y is a maxim for a purely rational being.
In the Dialectic's discussion of the postulates, however, the problem
is not simply one of syntactics and formal logic. \Ve come to the
following situation: Given a practical proposition, Kant argues that it
can be valid, even for practice, only if a theoretical proposition is
assumed and if this theoretical proposition is known to be neither
demonstrable nor refutable as such. The necessity of practically ac-
ceding to the practical proposition, therefore, is substituted for the
unavailable theoretical evidence for the theoretical proposition, and
the theoretical proposition is asserted on moral grounds which are
epistemologically (not psychologically) subjective, that is, on grounds
which are found in a "need of reason" and not on grounds of objective
evidence. The theoretical proposition supported in this way is a
postulate. It is valid for no theoretical purpose; and, inasmuch as all
the grounds for holding it are practical, all the consequences drawn
from it must likewise be restricted to their practical bearing.
Such a derivation of the theoretical proposition contributes nothing
to our knowledge of tne truth of the proposition, if truth be taken as a
theoretical (i.e., cognitive) predicate. Kant, as we shall see, does not
always remain within the precisely valid limits of his "proof." But he
tries.
Such an argument, he correctly says, does not permit us to compre-
hend how the Ideas of God, freedom, and immortality (the Ideas of the
objects of the postulates) are possible or are to be theoretically deter-
mined; "All that is comprehended is that [they are] postulated through
the moral law and for its sake" (133 [23 I ] ). If he is correct, I cannot
render obedience to the moral imperative to seek the summum bonum
unless I believe that it is possible, and I can believe it possible only if I
believe in the existence of these objects. This belief is not one growing
out of a mere subjective wish for the highest good, a wish I might hap-
pen not to feel; for then the imperative to seek it would be hypotheti-
cal. Nor is the belief a hypothesis for any theoretical purpose which I
confidently hold for some theoretical purpose only because I need it
for this purpose and know that it cannot be refuted. It is, he says, a be·
lief that I cannot renounce and at the same time maintain my allegiance
to moral law, an allegiance I cannot surrender without becoming con-
temptible and hateful in my own sight.
This "cannot" is meant to refer to a transcendental impossibility-
not a logical one because the relation of the moral law to the existence
of God is not analytical, and not a psychological one, for many a man
has in fact achieved virtue in a world in which he found no comfort
and support. I use the word "transcendental" in the sense of the Cri-
tique of Pure Reason, to refer to the conditions of a priori synthetic
propositions. 7 But the quotation just given shows that the transcenden-
tal connection is not between the moral law I accept and the objective
conditions of the summum bonum; it is not that the acknowledgment
of the moral law either requires the existence of God or justifies a the-
oretical knowledge of God's existence; it requires only that I believe in
God's existence. Even if God does not exist but I only believe that he
exists, the practical consequences for obedience to the moral law are
the same. A postulate, therefore, does not have to be known to be true,
it does not even have to be true, for it to serve its practical function.
Kant always recognizes the first of these denials. But though he does
not derive knowledge of their truth in his moral arguments, he does
seem to think that he derives their truth.
If Kant's theory of postulates reduced to the very modest form' in
which the necessity of believing was asserted (or even proved) with-
out raising any claim to the truth of the belief (except insofar as one
cannot believe what one knows to be false), it would be of little philo-
sophical interest. It would be interesting in a phenomenology of mor-
als, but practical reason would not be justified by such an argument in
occupying the "empty space" left by theory.
But all that Kant's argument, if otherwise valid, legitimately entails
7 This is obviously wider than Kant's meaning of "transcendental," which is

generally restricted to the justification of a priori synthetic knowledge, and accord-
ingly it is denied that morals belong in transcendental philosophy (Critique of PUTe
=
Reason A 15 B 29). But there seems to be no other word that indicates the sphere
of these synthetic a priori theoretical judgments which are not pans of the structure
of theoretical knowledge as such. Kant himself points out the resemblance between
his moral philosophy and transcendental logic. He says that Wolff's general prac-
tical philosophy is to his metaphysics of morals as general logic is to transcendental
logic (Foundations of the Metaphysics of Morals, 390 [6]). The moral argument for
the existence of God is referred to as a ''transcendental argument" in Critique of
Pure Reason, A s89 = B 617; similarly in Critique of Practical Reason, 113 (209).
The Postulates of Pure Practical Reason 263
is the necessity of making certain postulations as acts, and not the truth
of the postulates thus made. But Kant does not, as I have said, restrict
his conclusions in this way. He soon regards the proof of the necessity
of the postulations as contributing, though only very little, to theory.8
For, he tells us, the practical argument "compels knowledge [i.e., the-
ory] to concede that there are such objects without more exactly de-
fining them." Theory thus gains an "accession," but not an accession
that it can use in the further extension of its realm by making synthetic
a priori judgments about them; still the Ideas, which, for theoretical
reason, were objectively empty ("without objects"), are now shown
to "have objects." Since it cannot extend its knowledge of them, the
work of speculative reason on this gift from the practical is negative,
"not broadening, but purifying." But the objects are there; they are
asserted; the category of existence is applied. 9 All the disarming cau-
tions about how little is gained by theory remind one of the young
lady who excused herself for having an illegitimate baby by insisting
that the baby was very small.
How, then, does Kant justify moving from "the postulates must be
asserted" to "the postulates have objects"? He does so through the doc-
trine of the primacy of pure practical reason ([ [9 ff. [2 I 6-18] ). This
doctrine arises, as we have seen, from the conviction that we have not
two reasons, but one reason with two interests; that these interests are
not incompatible; and that the two interests point in the same direction,
while only one of the routes of approach to the unconditioned can ac-
tually reach its destination with a proof without flaw.
The doctrine of primacy is Kant's closest anticipation of the later
coherence theory of truth. If we cut off theoretical from practical rea-
son and ascribe only to the former the privilege of establishing truths,
8 The practical-dogmatic metaphysics which gains this accession is a part of
theory, yet its propositions are valid only from a practical point of view; thus it
differs both from speculative metaphysics (which is not valid at all) and from the
metaphysics of morals (which is a pan of practical philosophy). Needless to say.
Kant never worked out the details of such a hybrid of theory and practice. but its
foundation lies in the point we are here examining. The moral argument for the
existence of God, he says (FoTtschT;tte deT Metaphys;k, XX, 305), is "sufficient to
ground a theory of the supersensuous, but only as a practical-dogmatic transition
to it. It is not really a proof of God's existence absolutely (simpliciteT) , but only in
a cenain respect (secundum quid), namely with respect to the final goal that the
moral person has and ought to have. It is related to the rationality of assuming such
a being, where the person has a right to pennit an influence on his decisions to come
from an Idea that he has made according to moral principles, exactly as if he had
made it in accordance with a given object."
9 All quotations and allusions in this paragraph refer to Critique of PTactical
Reason, 135 (232-33); similar assenions are also on the following page.
we find (a) that we cannot establish a coherent self-dependent system
and (b) that if the doctrine of postulates is left in the state of justify-
ing only the process of postulation as a practical act but not the postu-
lates as true, an entire area of human experience, completely rational in
itself, is likewise left unfounded in any theory about the world. If, on
the contrary, we say with Kant that the two can be taken together,
with the inadequacy of the one complemented by the primacy of the
Dther, we are on the way to agreeing with the later idealists in their
methodology that whatever we have to believe in order to "save the
appearances" is, to that extent, true, whether the "appearances" in
question are empiricotheoretical or practical. Kant's Sacblichkeit or
modesty is shown in his refusal to make any speculative use of the ob-
jects posited in this way; but I think there can be little doubt that he
regarded his argument as an argument for that which was posited in
this way and not merely as an argument for the necessity of this
positing.
§ 3. IDEAS AND POSTULATES
The Critique of Pure Reason contains two quite different accounts

of transcendental Ideas. According to one, the categories, when not
schematized and thus putatively applied to things in themselves, are
Ideas, Le., concepts which have no adequate object in experience,. Ac-
cording to the other, the Ideas are an original possession of reason, be-
ing its principles for the systematizing of experience, which can be
undertaken only if there are concepts of unconditioned conditions. In
the latter theory, the logical origin of the Ideas is to be found not in
the table of judgments, which is our clue to the categories, but in the
types of syllogisms that reason can use. In each theory, however, the
Ideas have this in common: they are a priori necessary to reason, but,
as they can have no objects of intuition corresponding to them, they
are necessarily dialectical if taken as more than regulative. The dialec-
tic is resolved, yet the necessity of the Ideas is saved by showing that
they are maxims necessary to the conduct of thought, not by showing
that they really do have objects. Whether they have objects remains an
irresolvable problem for speculative reason; but speculative reason does
not need, for its own interest, to solve this problem. It suffices specula-
tive reason only to know that the Ideas cannot be shown to be false as
cognitions and that they can be shown to be necessary as maxims.
It is the second of the two theories of Ideas that is important in
Kant's practical philosophy; for the purpose of the treatment of the
postulates is to show that the three Ideas derived from the second of
the theories do have objects. The three Ideas in the second theory are
these: the Idea of the absolute unity of the subject of experience (the
soul), the Idea of the absolute unity of the series of conditions of ap-
pearances (the world), and the Idea of the absolute unity of the condi-
tion of all things in general (God) .10 Speculative metaphysics is the
study of these three Ideas and the attempt to define and prove the real-
ity of each of these objects. The Critique of Pure Reason examines the
proofs of speculative metaphysics concerning each. The proofs of the
first are listed in the chapter on Paralogisms, those of the second in that
on the Antinomy, and those of the third in that on the Ideal of Pure
Reason. l l Corresponding to these Ideas, we accordingly find in the
Critique of Practical Reason three postulates. Thereby Kant aims to
show that theoretical and pure practical reason point to the same ob-
jects, but, whereas they "fiy before it when it follows the path of pure
speculation,"12 they can be definitely grasped if pursued on the path of
the practical.
§ 4. THE IMMORTALITY OF THE SOUL
The doctrine of the immortality of the soul was always taken seri-
ously by Kant. But he never had much faith in theoretical arguments
for it. Even when he was still traditionalistic in his rationalism, the
premise for immortality was found in the "inco~plete harmony be-
tween morality and its consequences in the world"; but he did not have
a crudely eudaemonistic theory, for he held that it was better to base
the belief on immortality upon the moral disposition rather than to base
the moral disposition on a hope for future rewards. IS
In the Critique of Pure Reason, Kant discovered that all synthetic
a priori judgments about the soul are proved only in syllogisms which
contain paralogisms. His argument, though long and complex, can be
stated in its essentials very simply. The thought of the self in the tran-
scendental unity of apperception, the "1" in the" 'I think' that must be
able to accompany all of my representations," is the thought of a sim-
ple substance. But the soul, though thought of as a substance, is not
given under the only condition whereby a substance can be known,
viz., in time. Hence the judgment that the self is a simple substance,
from which its immortality was to be derived, is not a valid synthetic
a priori judgment, and no synthetic predicate can be proved of the
10 Critique of Pure Reason, A 334 = B 391; B 395 n.; cf. Critique of Judgment,
§ 9 1 (V, 473 f) z5J)·
11 In a slightly different context, they occur, respectively, as the theses of the
second, third, and fourth antinomies.
12 Critique of Pure Reason, A 796 == B 8Z4.
13 Triiume eines Geistersehers, 11,373.
soul. The illusion that it can be proved arises because the paralogism is
not noticed whereby the major premise has to do with the pure con-
cept of substance while in the minor premise and conclusion substance
is thought of as a mode of existence in time. 14
This theoretical argument, Kant wittily remarks, "so stands on the
point of a hair that even the schools keep it from toppling only so long
as they keep it spinning around like a top."15 But the failure of the the-
oretical argument does not in the least prejudice "the right, nay, the
necessity, of postulating a future life in accordance with the principles
of the practical employment of reason."16 Indeed, the arguments which
are "serviceable for the world at large" are unaffected by this criticism,
and the failure of speculative arguments is itself a valuable "hint" to
turn from vain speculation in these matters to practical concerns.17
"The proofs serviceable to the world at large," as seen in the Cri-
tique of Pure Reason, are not the proofs finally acceptable to Kant in
the second Critique. There are two~ First, there is the analogical argu-
ment that every organ is suitable for its proper function and that our
"human endowments" so far transcend the limits of utility in this
"present life" that its proper function and destiny cannot be restricted
to its corporeal existence.t 8 The other is the postulation of the immor-
tality of the soul as directly necessary to the summum bonum, without
which "the glorious ideas of morality are indeed objects of approval
and estimation, but not springs of purpose and action."19 The first is an
example of the teleological argument that Kant respected and recom-
mended, even though he showed it to be theoretically invalid;20 we
H The paralogism is most succinctly explained in Critique of Pure Reason, B 410-
II. Another refutation of the argument for immortality from the simplicity of the
soul is directed against Mendelssohn at B 414 if.; still another is found in the second
Antinomy, especially A 443 = B 471.
15 Ihid., B 424. 16 Ibid. 17 Ibid., B 421.
IS Ibid., B 425-26; d. A 827 = B 855. This is a doctrinal belief, not a moral belief,
because its premise is a theoretical judgment about man's (moral) nature. It differs
subtly but essentially from the moral argument of the second Critique. I have not
been able to find any anticipation, in other writers, of the moral argument proper;
but the doctrinal belief based upon man's moral nature, which was Platonic in in-
spiration, was widely accepted in the eighteenth century. It was formulated by
Addison (Spectator, No. I I I [1711]), who said he did not remember having seen
this, the "strongest" of the arguments for immortality, presented elsewhere; by
Platner (Philosophische Apborismen, H 1176-79 [1782]), by Mendelssohn (Pbaedon
[1769]), by Crusius (Anweisung vernunftig zu leben, Part I, Thelematologie, H 118-
2 I [1744]), and by Lessing (Eniebung des Menschengeschlecbtes, H 79 if. [1780]),
all of whom had been read by Kant.
19 Critique of Pure Reason, A 813 = B 841; d. A 811 = B 839-
20 Ibid., A 613 = B 651.
shall meet it again later. The second is an argument which is incom-
patible with the doctrine of autonomy and disappears in deserved
oblivion. 21
Elsewhere in the first Critique, in a part written just before the Cri-
tique of Practical Reason, there is a remark that anticipates the doctrine
of the second Critique more clearly than the other two arguments do.
In the Preface to the second edition, he says that the belief in immor-
tality is based on a "notable characteristic of our nature, never to be
capable of being satisfied by what is temporal (as insufficient for the
capacities of its whole calling)."22 This seems to be an intimation of the
moral argument that is to follow in a few months; but it is so much an
obiter dictum and so closely related to the theoretical hypothesis or
doctrinal belief mentioned in the previous paragraph that it may not be
justified to see more than an obscure germ of the later argument in it.
But the argument of the Critique of Practical Reason is a model of
economical lucidity. We shall restate it in numbered sentences in order
to facilitate comment.
I. The highest good is a necessary object of the will.
2. Holiness, or complete fitness of intentions to the moral law, is a
necessary condition of the highest good.
3. Holiness cannot be found in a sensuous rational being.
4. It can be reached only in an endless progress, and, since holiness
is required, such endless progress toward it is the true object of the
will.
5. Such progress can be endless only if the personality of the rational
being endures endlessly.
6. The highest good can be made real, therefore, only on "the sup-
position of the immortality of the soul" (122 [218-19])'
First, I must comment upon statement 6. Kant says that the highest
good is possible only "on the supposition of the immortality of the
soul." This might mean that it is practically possible only to a being
that supposes, i.e., postulates, that it is immortal. Or it might mean that
it is practically possible only if the soul is immortal. In the light of what
we have said above about the postulates and of Kant's own statement
of the postulate that "I will ... that my existence in this world be also
an existence in a pure world of the understanding" (143 [241]), one
might well prefer the second interpretation. As so often, here we have
to do only with Kant's fafon de parler; often, throughout his works,
he uses such constructions as "the concept of x" when the argument
21 Cf. Metapbysik der Sitten, VI,490.

22 Critique of Pure ReIlS071, B xxxii.
z68 The Postulates of Pure Practical Reason
permits or even requires simply "x." But in this passage, I think it is
clear that he means something of both of the alternatives. We shall
therefore reformulate statement 6 to read: "The highest good, there-
fore, can be made real only if the soul is immortal; and a moral being
that acknowledges the law must suppose that it is immortal."
We have already commented extensively on statement 1. 23 But our
criticisms of it have not touched the nervus probandi of this argument.
For here, as shown in sentence 4, Kant is not in the least concerned, or
at least not directly concerned, with the second component of the
highest good. He is here dealing only with virtue and its perfection, and
the argument would seemingly be unaffected by the complete omis-
sion of the second component. Why, then, does he bring in the con-
ception of the summum bonum in statements I and 2, only to omit it
in sentence 4?
I hesitate to suggest what seems to me to be the only reason, as it is
almost incredible that Kant could have made such. a mistake. But in de-
fault of more credible explanations, I must propose one that does not
seem to have been noticed by other critics of the postulates.
Kant has, I believe, confused the "supreme condition" (bonum su-
premum)24 of the summum bonum, which is virtue, with the supreme
perfection of virtue, as though his definition of the summum bonum
were the conjunction of perfect happiness with the perfection of vir-
tue (i.e., holiness). This conception I shall call the "maximal concep-
tion of the summum bonum":25 perfect blessedness under the condi-
tion of holiness. If the maximal conception is substituted for the one
that Kant formally defined (110 [206]), to wit, the conception that
happiness in various degrees is apportioned by the degree of virtue-a
conception I shall call the "juridical conception of the summum bo-
num"-the premise necessary for the argument is supplied. Only be-
cause Kant thought of the maximal conception at this point is he able
to suppose that the necessity of the summum bonum entail., the neces-
23 Above, chap. xiii, S 3.

24 Critique of Practical Reason, 110 (206). Note the contradiction between this
and the second sentence of 122 (218).
25lbid., 129 (227): greatest happiness combined in the most perfect proportion
with the highest degree of moral perfection. Similarly, V orlesungen iiber M eta-
pbysik (Heinze ed.), p. 712: "Two things constitute the summum bonum-the
moral perfection of the person and the physical perfection of his state." In Critique
=
of Pure Reason, A 810 B 839, the summum bonum is called an "Ideal of reason"
and not a mere Idea or concept, and Ideal being "the concept of an individual
object completely determined through the mere Idea"; hence, as completely de-
termi,ned by the definition, the supreme good is there taken in its maximal sense.
Tbe Postulates of Pure Practical Reason 269
sity of holiness, which in turn is said to entail immortality. That he be-
gan with the juridical conception is shown by his statement that, grant-
ing the existence of God, the necessary relation of morality to happiness
is possible in the sensuous world (I 15 [21 I]); that he changed to the
maximal conception is shown in the statement in the section on Immor-
tality that "the Infinite Being" inexorably requires holiness "in order to
be true to His justice in the share He assigns to each in the highest
good" (123 [219]).
Let us try to clear the air of this erroneous conception of the sum-
mum bonum by reformulating statements 1 and 2 into (I): There is a
necessary moral imperative, "Be ye perfect." Premise 3 may stand as a
sad commentary on human nature. Premise 4 suggests that perfection
may be either (a) a state or (b) an endless progress toward a state, and
that a and b, while different for us men, may be identical for an intel-
lectual non-temporal intuition such as we may suppose God has. Prem-
ise 5 tells us that (4, b) is possible if man's duration is without end.
We may now restate the whole argument.
I. The moral imperative requires of man that he be perfect (holy).
II. No sensuous being can be in a state of perfection (holiness).
III. But in the eyes of God, continuous and unending progress toward
perfection is equivalent to perfection as a state in the distributive
justice of the summum bonum.
IV. This progress is possible only if the soul is immortal.
V. Therefore, the moral imperative (I) can be fulfilled only if the
soul is immortal; and a being that acknowledges this imperative
must suppose himself to be immortal.
Against this reformulation, two objections can be made. The first is
that it presupposes the existence of God in Premise III, which -is based
on the second and weaker of the two components of the summum
bonum. Premises I and II are compatible (assuming that Premise I is
true, and a moral law is valid) only on the assumption III of an intui-
tive intelligence, which Kant ascribes to God; whereas God is postu-
lated only to mediate the two components of the summum bonum,
which is not, in fact, achieved in the sensuous world.
The second objection concerns the formulation of Premise I itself.
Would it not be far simpler, we may ask, to let "perfection" mean
what it does ordinarily mean, namely, a state, and then modify the al-
leged imperative to read: "Seek perfection"? 'Ve are told to seek the
Kingdom of God, not to settle in it. This is indeed all that is clearly
demanded of me, and the imperative that is so formulated demands it
without in the least being "degraded from its holiness, by being made
out as lenient (indulgent) and thus compliant to our convenience."26
This imperative and obedience to it would meet the putative commands
of God-if it can be shown that there is a God-without the speculative
conception that what for us is an infinite progress -in time is for him
a state.
Only through the substitution of the maximal for the juridical con-
ception of the swrrmum bonum, can we say that a state of holiness must
be attainable, even in infinite progress. Neither the Kantian text nor
the Christian doctrine, which Kant is here rationalizing, nor the voice
of duty itself requires the maximal conception.
What, then, is left of the postulate? Only a hope. "On the basis of
[one's] previous progress from the worse to the morally better, and of
the immutability of intention which thus becomes known to him, he
may hope for a further uninterrupted continuance of this progress,
however long his existence may last, even beyond this life."27 This hope
need not in the least jeopardize the autonomy of the moral disposition.
If we assume that the moral condition with which life ends is un-
changed on entry into another life, it is wise to act as if there is a future
life.28 The moral disposition may inevitably engender the belief in it.
But it is not a duty to believe in it;29 and, if my analysis is correct, Kant
has given no very good reason why one should believe in it.
It suffices, in conclusion, to point out two puzzles which are present
in Kant's theory of immortality. Presumably the soul upon death, if it
is indeed immortal, is no longer a denizen of the world of space and
time; if we say it never was, as when Kant says the acts of pure prac-
tical reason are not in time (98--<)9 [191--<)2]), we only multiply the
difficulty of the problem I am about to mention. If, I say, the soul is no
longer under the temporal condition, it is not possible to understand
what is meant by "continuous and unending progress." We can con-
ceive of substances only under the temporal condition, and our lan-
guage must be adapted to this temporal condition, even though it is
recognized that this is not justified. If we say the soul is immortal and
mean that it is eternal, we may add, "Of course our intellect can con-
ceive of eternity only as infinite temporal duration; but eternity is not
26 Critique of Practical Reason, 122 (219). I take this occasion to correct an error
in the first sentence of the paragraph here referred to, as it occurs in the New York
edition (p. 117) of my translation. For "toward" read "to reach."
27 Critique of Practical Reason, 123 (119-20); the hope includes a hope of blessed-
ness (ibid., 123 n., 117 n., 1 28-z9 [1I0 n., 1I4 n., Z25-26]).
28 Das Ende aller Dinge, VIII, 330.
29 Critique of Practical Reason, I ¢ (244) and, in its dogmatic theological form,

Religion within the Limits of Reason Alone, VI;1Z8 n. (Greene and Hudson, 119 n.).
Tbe Postulates of Pure Practical Reason 27 I
a quantum of time, but is atemporal." Very well; but we must be re-
minded that the premise for the eternity of the soul includes the idea
of continuous change, which is a temporal and not an eternal mode.
The other puzzle concerns the happiness of the immortal soul. This
puzzle is of less moment in the interpretation of the argument for im-
mortality than in that of the argument for the existence of God, since
the need for happiness as a part of the surllmum bonum enters as a con-
dition only in the latter argument. Nevertheless, it is difficult to frame
a conception of the happiness or bliss of a being no longer affected by
desire, even if, in virtue, "the determination of the will directly by
reason alone is the ground of the feeling of pleasure" (I 16 [2 I 3]). It
has always been difficult to render a conception of an afterlife both in-
telligible and concretely attractive, even on hedonistic principles. And
when these principles are denied or the conditions of their application
are withdrawn from the soul in its immortality, it seems impossible to
speak of happiness as a reward for virtue or as a component of the
highest good for beings no longer affected by sensuous desire.
But I doubt very much that these objections to the conception of
life after death would have troubled Kant. He was not concerned with
any theoretical determination of the supersensuous, because it would
be impossible on theoretical grounds and empty of practical signif-
Icance.
§ 5. THE EXISTENCE OF GOD
In The Only Possible Premise for a DemonstratioN of the Existence

of God (1764) Kant proposed an ontological argument quite different
from the Cartesian form, and in Part II of that work he supplemented
the ontological with a teleological argument. The ontological argument
in the Cartesian form was refuted in the Critique of Pure Reason, but
he did not consider at any later time the specific kind of ontological
argument that he himself had invented; instead of being refuted, it just
died a natural death along with the rest of speculative rationalistic
metaphysics, for it suffered from the fundamental defect of all Onto-
logical arguments, viz., that it is not possible to know the existence of
anything from mere concepts (139 [236]).
In the Critique of Pure Reason God is called the "Ideal of Pure Rea-
son," not just an Idea; for God is thought of as a single individual sub-
stance and the ground of the existence and unity of all things in gen-
eral. Three possible theoretical proofs are distinguished: the ontological,
or proof of existence from the concept of perfection; the cosmological,
or proof of the existence of a first cause from the existence of the
world; and the physicoteleological, or proof from the empirical evi-
dence of intelligent design in nature. These are the only possible
proofs, for they have as their respective premises a mere concept, a
concept of existence, and a concept of a specific existence. The Struc-
ture of the refutation can be briefly described. The ontological argu-
ment is invalid because "God exists" is a synthetic proposition and
therefore cannot be proved without intuition, but intuition of a super-
sensuous being is necessarily lacking. The other arguments, however,
while they, as it were, lead up to the concept of God, cannot realize it
without surreptitiously introducing the ontological form of argument.
Hence no theoretical proof of God's existence is possible.
The teleological argument, however, is worthy of respect and re-
mains subjectively persuasive even when its formal fallacy is revealed.
It remains useful when its concept of God as the Author of design in
nature is used simply as a regulative Idea for anticipating the orderli-
ness of nature. Its use as a regulative maxim can always benefit science
in the comprehension of nature and never injure it except by misuse,
even when the facts discovered by its use do not bear out the expecta-
tion of teleological unity. 30 But if God as the source of order in nature
is taken as an object of knowledge, as in the pretensions of natural the-
ology, the intere~t of theoretical reason is infringed, since the concept
of God, now taken anthropomorphically, is used "to impose ends upon
nature, forcibly and dictatorially, instead of [helping us to pursue] the
more reasonable course of searching for them by the path of physical
investigation"-though this investigation is guided by the regulative
Idea that the variety of nature is to be explained under a minimum of
laws, which is rendered intelligible on the assumption of a design. 31
Like the other Ideas, therefore, the Idea of God properly understood
is a problematic one that cannot be asserted by theoretical reason bur
need not be asserted for the purposes of theory. Theory can use it, pro-
vided only that its possibility is guaranteed. 32 But the needs of practic:ll
30 Critique of Pure Reason, A 688 ;::: B 716.
31/bid., A 6<)2 = B 720; Critique of Judgment, Introduction, V.
32 Cf. above, p. 259, nn. I and 2. Kant uses the word "possibility" in two senses,
and thus sometimes says that the Critique of Pure Reason shows the Ideas to be
possible and at others that it could not show them to be possible (Critique of Pure
=
Reason, A 558 B 586; Critique of Practical Reason, 4, 13 2, [45 [88, H9, 243]).
"Possible" in the first usage means "logically possible," i.e., thinkable without con-
tradiction or thinkable through the (unschematized) categories; in the second, it
means "really possible," i.e., possible in the sense of the first Postulate of Empirical
Thought in General (Critique of Pure Reason, A 1I8 = B 266), or having some
connection with a given structure of actual experience. The sense in which the
existence of God is logically possible is the same as that in which, for instance, the
inhabitants of another planet are logically possible. But the real possibilities are dif-
reason build on this possibility an assertion. The first attempt to justify
this has already taken place in the first Critique.
The reader of this commentary has already had the passage in which
this attempt is made called several times to his attention; only a brief
reminder of it is needed here. The SU1111n1.tm bonum is not possible un-
less the soul is immortal and God exists-the former because in this
world happiness and worthiness to be happy are not matched, and the
latter because they cannot be brought into unison without a supernatu-
ral agency. The summum bonum is a necessary object of the will; hence
it is necessary to postulate it along with its conditions. But though Kant
says it is not the desire for happiness, even happiness in accord with
virtue, that is the motive to morality, he nevertheless says that without
the summum bonum and these conditions of it, "the glorious ideas of
morality are indeed objects of admiration and approval, but not springs
of purpose and action."33 We postulate the existence of God "in order
that through such agency effect may be given" to the moral laws. 34
The element of reward is no longer a principal premise for the argu-
ment for immortality in the second Critique, but Kant does not omit it
from the later argument for the existence of God. 3 :> Nevertheless, with
ferent, for the existence of God is not connected with intuitions but only with con-
cepts. To be really possible in the senSe meant in the second Critique is to be (0)
logically possible and (b) related necessarily to some other fact (viz., the moral)
whose reality is given. Thus the fact of pure reason is the practical corollary of in-
tuition (cf. above, pp. 152 f., 172 f.) in converting mere concepts of the logically
possible into cognitions that the logically possible is really possible (cf. Critique of
Practical Reason, 66 [157]).
33 Critique of Pure Reason, A 813 ::::: B 841.
34 Ibid., A 818 = B 846, A 589 :::::: B 617, A 634 :::::: B 662; Reflexionen 61 10,6236.
35 Only in Religion within the Limits of Reason Alone, VI, 101 (Greene and
Hudson, 92) does Kant seem to have been able to argue for the existence of God
exclusively from the first component of the summum bonum. The highest moral
good, he there tells us, cannot be achieved by an isolated individual, but only in a
moral commonwealth. The establishment of this moral commonwealth cannot be
the work of man, who can only make himself worthy of membership in it; its actual
establishment can be only by the grace of God. Virtue is worthiness to this grace
(which now replaces worthiness to be happy), and the summum bonum is the
Kingdom of God to be established by God. Hence the moral command to seek the
Kingdom of God requires that such a moral legislator exist or justifies the belief
and hope independently of the rewards it may contain. The last time the moral
argument is given in its classical fonn, involving both the components of the
summum bonum is in Critique of judg;mem, § 87, where it is presented with only
one modification. Here the natural desire of man for happiness is not taken as a
prior condition for the definition of the summum bonum, but the existence of God
is taken as the condition under which a man may morally .set before himself any
the development of the theory of autonomy, there is a decisive modi-
fication of the peculiar feature of the argument of the first Critique.
The belief in the existence of God is based upon the putative necessity
of the second element in the summum bonum (124 [220]); but he now
succeeds in explaining more fully how it can be present and not func-
tion as a motive-a matter that made the conclusion in the first Critique
seem heteronomous, as pointed out above. Now the reason for assum-
ing the Slmrmum bonum and its conditions is not that man naturally
desires happiness and happiness is a component of the summum bonum;
it is that the moral command would not be just ineffective, it 'Would be
null and 'Void, if it commanded the impossible, and the summum bonum
would be impossible (so far as human mind can comprehend it) (145
[244]) if God did not exist.
We are now ready to summarize Kant's argument for this postulate.
I. Happiness is the condition of a rational being in the world in whose
whole existence everything goes according to wish and will.
2. Man's will is not the cause of nature and does not bring nature
into complete harmony with the principles of his will.
3. There is, therefore, no ground in the moral law (or in nature) for
expecting a necessary connection between the morality and happiness
of men.
4. But such a connection, in the concept of the summum bonum, is
postulated in the command that we ought to seek the summum bonum.
5. The highest good must, therefore, be possible.
6. Therefore, a cause adequate to it must be postulated.
7. Such a cause must be the Author of nature, acting through under-
standing and will. Such a being is God (IZ 5 [n 1]).
Let us begin our examination of this argument with statement 3. The
disproportion of worthiness to be happy and actual happiness can be
and often has been taken as evidence against the existence of God, at
least in the sense of statement 7. How then does Kant manage to build
this premise of his opponents into an argument for the existence of
purpose as necessary. even if the realization of it is not within his power. This pur-
pose is not happiness desired and subsequently restricted to the moral condition of
wo~~hiness. but happiness only insofar as it is morally earned. A similar thought is
in U beT den Gemeinsprucb ... , VIII. 180 n.
Though Kant supplemented the moral argument in various ways. as we have seen.
he did not renounce it (d. George A. Schrader. "Kant's Presumed Repudiation of
the Moral Argument in the Opus postumum," Philosophy. XXVI [l95d. 218-41).
That the "new" doctrines of the Opus postumum can be found side by side with the
moral argument is shown also by Walter Reinhard. UbeT das Verha/mit von Sitt-
lichkeit una Religion bei Kant (Bern: Haupt, 19%7).
God? The answer is found not in any alleged theoretical fact or hy-
pothesis of the kind ordinarily used to "solve" the problem of natural
evil, but by the practical premise (4), viz., "Seek to realize the highest
good."
We have already commented extensively on this alleged impera-
tive. 36 We have argued that, as an imperative, it is a command only that
we seek virtue, let the eschatological chips fall where they may. But
Kant regards the second component of the summum bonum as essential
because he holds an ideal of the rationality of morals. This is described
at the beginning of chapter ii of the Dialectic. Happiness, he says there,
is required in the summum bonum "not merely in the partial eyes of a
person who makes himself his end but even in the judgment of an im-
partial reason, which impartially regards persons in the world as ends-
in-themselves." (Here we meet Kant's version of the English philoso-
phers' "disinterested observer.") "For to be in need of happiness and
also worthy of it and yet not partake of it could not be in accordance
with the complete volition of an omnipotent rational being, if we as-
sume such only for the sake of argument."31
This seems innocent enough; but notice that it completely displaces
premise 4, the practical clause of which made the argument of the sec-
ond Critique a moral argument. 3S The alleged command to seek to
establish the summum bonum now contributes nothing to the concep-
tion of the distribution of happiness in accord with worthiness. The
argument based on this conception of the summum bonum as rational
is a revision of the teleological argument, which is purely theoretical.
It is not, of course, a physicoteleological argument, but only an anal-
ogy to it. It is a teleological argument, based not on the moral com-
mand in question but on the moral phenomenon as requiring a designer
for the adjustment of twO disparate things to each other. This is not
made clear in the Critique of Practical Reason, where the moral argu-
ment is compared with the physicoteleological to the detriment and cen-
sure of the latter (138 [236]). But it dominates the final theological
sections of the Critique of Judgment, and the moral argument of the
36 Cf. above, pp. 244-45, 153.
37 Critique of Practical Rells01l, J 10 (206). The practical consequences of this im-

partiality is the conception of each man as an end-in-himself; this is drawn at [31-32
( 12 9).
38 Kant has not begged the question by this statement, as it might appear. The
conception of God here is not used to establish the concept of the surmnum bonum.
The conception of an impartial observer is the essential one; it follows that if there
were an impartial observer with the requisite power, the SU1mJ'lUm bonum would
be made real.
second Critique had already been assigned to the obscurity of a diffi-
cult footnote by the time Kant wrote the treatise on religion. The ex-
plicit formulation of the new teleological argument is given in the
Metapbysik der Sittel1, in the sample of moral catechism Kant works
out. The passage is as follows:
TFACHER: \Vhen we are conscious of a good and acti\'e will through
which we hold ourselves to be worthy (or at least not unworthy) of hap-
piness, can we base on it the certain hope of partaking of this happiness?
PCPlL: Ko, not on that alone . . . . Our happiness remains only a wish
that cannot be a hope unless some other power is added.
TEACHER: Has reason perhaps grounds in itself to believe in God, i.e., to
assume as real a power that apportions happiness according to desert, or-
dering nature and ruling the world with supreme wisdom~
PUPIL: Yes; for we sec in the works of nature which we can judge such
an extensive and profound wisdom that we cannot explain It except through
the inexpressibly great art of a Creator; and frolll this we have reason to
promise ourselves a not less wise government as respects the moral order,
which is the highest ornament of the world; to promise ourseh'es that if we
do not nuke ourselves umvorthy of happiness through trespassing against
our duty, we may also hope to participate in it.:m
This passage is especially noteworthy. There is no discussion of the
SU11munn bonum in the Metaphysik der Sitte7l. The proof of God's ex-
istence is an argun1ent from design, pure and ~imple. As a theoretical
argument, it "always deserves to be mentioned with respect,"40 but it
is neither theoretically coercive nor independent of the ontological ar-
gument. Yet it is, if my analysis is correct, the hidden sense of the
moral argument, since the practical premise of the moral argument-
the command to seek the whole Sll1ll1ml111 bonum-merely calls atten-
tion to the internal heterogeneity of this concept without placing us
under any obligation to seek the second of the components.
The shift from the practical to the theoretical argument, however,
contributes nothing to the theoretical fruitfulne~s of the concept of
God; it may indeed lessen it. For theoretical arguments, as shown in
the first Critique, could lead, if they were valid, only to the cosmologi-
cal concept of a cause, which is less than what is meant by God, or
they enrich the concept with anthropomorphic predicates, reasoning
by analogy.41 These analogies, however, never lead to the superlative~
39 Metapbysik der Sit ten, VI, 482; d. also Critique of Judgment, § 91 (V, 479
IHz]), where we are told that the teleological argument is a desirable confirmation
of the moral.
40 Critique of Pure Reason, A 623-24 ::::: B 65 I-51.
41/bid., A 817::::: B 845; Critique of Practical Reason, 140 (23 8).
demanded by the concept of God. 4 :.! If the anthropomorphic elements
in the physicoteleological conception are removed, nothing is left of
the concept of God but the mere name.
The moral argument, paradoxical though it may seem, leads to a less
anthropomorphic conception of God than that of natural theology.
For all the predicates essentially attributed to God are predicates which
define merely a rational being endowed with a will (131 n. [22 8 n.1),
and these concepts are not empirical, psychological concepts. We do
not need to consider the peculiar nature of human understanding or
will-that the former is discursive and that the latter is sensuously af-
fected-in the definition of moral personality, but only the canonical
relation of the former to the latter. 43 This relation is the starting point
for the (moral) argument for the existence of God, and this is the only
essential content that the concept of God has. The only conclusions
that can be drawn from the concept concern these two in their inter-
rela!:ions; they are all moral in their import. If we try to "sensualize"
the conception of God, we weaken its moral force by mixing empirical
concepts drawn from human nature with the purely rational concepts
of a moral being in general.
The only valid theology, therefore, is moral theology; God is a con-
cept belonging not to physics (or its extension, metaphysics) but to
morals (138 [236]). Until the moral motive was explored, metaphysics,
based on the study of nature, did not need a rational theology, and no
trace of one is found before moral consideration generated it.H Simi-
larly, the ultimate teleology of the world is moral, not natural. The
final purpose of creation is moral; it is the summum bonum (130 [22 8]).
Men serve and glorify God-in the figurative sense, the only one Kant
will permit-by respect and obedience to his command.
The theological discussion in the second Critique ends here without
developing the most interesting philosophical and ethical conclusions
which are implicit in it. The theoretical interest in the concept of God
is to find an absolute and sufficient ground for the unity of the world,
a cause of all causes, and a purpose of all purposes. If we examine this
42 Critique of Practical Reason, 1]8-39 (236-37)-an obvious echo of Hume's

Dialogues concerning Natural Religion. The necessary anthropomorphism of the
physicoteleological conception of God is described in Critique of Pure Reason,
A 6cn-7oo =::: B 725-28.
43 Critique of Practical Rea$On, 137 (234); d. ibid., 57 (t47), and Foundations of
tbe Metaphysics of Morals, 4tt-J2, 425 (28,43),
44 The dc>velopment is placed about the time of Anaxagoras in Critique of Prac-
tical Reason, 140 (238), but it seems to be attributed to Christianity in Critique 0/
Pure Reason, A 817::::: B 845-
doctrine a moment, the transition from the penultimate to the last para-
graph of Section V will not appear so sudden. Kant says: "It follows
that, in the order of ends man (and every rational creature) is an end
in hirr.self."45 But how it follows is explained only in the Critique of
Judgment.
The third Critique argues that the purposive order of nature must be
judged as itself without a purpose, unless we can find in it something
that is an absolute purpose, under which all other purposes can be sub-
sumed. This absolute purpose must be an autonomous will, for a good
will is that by which a being can have an absolute (not relative) worth
and, in reference to it, the world itself can have a final purpose, i.e., be
a system consonant with and therefofe admissible by reason.46
The purposes in the phenomenal world, under the law of nature, can
thus be synthesized only by the Idea of an intelligible world in which
the final purpose is a moral one, the existence of rational beings in a
realm of ends. 47
Our conception of purpose and natural mechanism is dependent upon
the fact that we possess a discursive understanding; if we had an intui-
tive understanding, these two conceptions could be constitutively and
not merely regulatively synthesized. In the regulative Idea of a com-
plete teleological order, which is possible only if there is moral auton-
omy as an end in itself, the two legislations of reason-the theoretical
and the practical, or nature and the realm of ends-are at last shown to
be compatible with each other. Only in such a world, in which we sup-
pose that there is a legislation for nature by a moral governor, can the
necessary connection between the two elements of the summum bonum
be thought. Without this conception, we would have to give up this
conception of the SU1'llmum bonum or suppose that it is brought about
only beyond the realm of nature or that it remains a miraculous event
in a world whose lawful constitution has nothing to do with moral
law.4s
It is the concept of the summum bonum as the final purpose of the
world with its corollary concept of God that finally bridges the gap
between nature and morals. Through these concepts Kant believes that
he is enabled to approach most nearly the goal of a single system of
45 Critique of Practical Reason, 131 (129). It has quite other premises, from which
it follows much more clearly and convincingly, in Foundations of the Metaphysics
of Morals, 4 2 9 (47).
48 Critique of Judgment, § 86 (V,44 2-43 [293-94]).
47 Ibid., and Uber den Gebrauch teleogischer Prinzipien, VIII, 182 f.
48 Cf. Critique of Pure Reason, A 816 = B 844.
Tbe Postulates of Pw'e Practical Reason 179
philosophy4!J and to show that practical and theoretical reason are fi-
nally one and the same, The system that brings them into unison, how-
ever, exists only for reflective judgment, i.e., as a guiding maxim for
the systematization of experience, not as a determinative principle from
which specific natural a.nd moral consequences can be drawn. If it were
not a merely regulative principle for judgment, we should have the
double absurdity of a theological physics and a theological morality.:iO
§ 6. MORALITY AND RELIGION

There is no such thing as a theological morality, i.e., a system of
moral rules derived from knowledge of God. There are three reasons
for this. First. we do not have the knowledge. Second, if we did have
it and used it as a moral premise, the autonomy of morals would be
destroyed. Third, moral laws are not dependent upon any lawgiver, as
if a difference in the nature of God (or the non-existence of God)
would make any difference in the determination of duty. Theological
morals commits a hysteron proteron; for our entire concept of God, so
far as it is valid, grows out of our moral conceptions. ln Actually, of
course, this is not true, for theological belief is originally historical, not
rational,52 It is based on revelation or alleged revelation and is never
pure but contains historical and psychological accidents. But it con-
tains a hidden kernel of pure rational belief, which Kant uncovers in
his treatise on religion, in the Strife of the Faculties and the Conjectural
Beginning of Human History.
Religion is "the recognition of all duties as divine commands, not as
sanctions, i.e., arbitrary and contingent ordinances of a foreign will,
but as essential laws of any free will as such."53 They can be recognized
as commands only under the assumption of a legislator, who need not
be the author of the law but only thought of, in religion, as the source
of our obligation to obey it. There are no duties to God,54 certainly no
<co Ibid .• A 840 =B 868.
50 Critique of Judgment, § 91. last paragraph of book.
51 Foundations of the Metaphysics of Morals, 408-<,> (25); Critique of Prllctical
Reason, 139 (237).
52 Kant to Jacobi. August 3D, 1789 (XI, 76): "If the gospel had not previously
taught universal moral laws in their purity, pure reason would not have been able
to comprehend them in such perfection; but since they are given, one can now
convince anyone of their correctness and validity merely by reason."
53 Critique of Practical Reason, 129 (226); cf. Critique of Judgment, § 91 (V. 481
[334J); Metaphysik der Sitten, VI. 487; Streit der Fakultiiten, vn.
36.
54 Against \Volff. who had divided duties into those toward self. toward others,
and toward God.
duty to believe in his existence; but regarding all duties as if they were
divine commands of a morally perfect and omnipotent legislator for
both morals and nature connects the moral disposition with the hope
for the S1I7Jn1ll17JZ b07lum and adds, to respect for the law, dimensions
of love of and adoration for God. 55
This is not a wholly new and adventitious accommodation of Kant's
moral ideas to an inherited religious tradition. His attitude toward
morality, from the beginning, had a religious humility and single-
minded ness, and he is here only claiming that his moral system already
contains the essential elements of religion. Religion, properly under-
stood, is nothing but the recognition of the holiness of morals, to the
defense of which the whole of his ethical work had been devoted
from the beginning.
Nevertheless, the definition of religion acknowledges a dimension
of moral law that was taken from it by the Copernican Revolution in
ethics. Sittengesetz and A'loralgesetz were relatively new words in
German philosophy when Kant wrote; and prior to Kant they meant
a law for morality drawn from the will of God, as a supplement to
natural and positive law. Kant secularized the conception of moral law,
against the theonomic doctrines of both the Wolffians (e.g., Baum-
garten) and the critics of \Volff (especially Crusius), and thereby
erected what has been called "the first non-theological philosophical
ethics since Thomasius."56
The definition of religion given here does not add any new substance
or authority or content to the moral law. It is not a transformation
of the ethical position that Kant has just gained against theonomy
but a restriction, rather, of the broader claims of religion itself to have
dogmas and sources of insight disconnected from the moral. The defini-
tion of religion must be taken quite literally: it is a definition of re-
ligion, not a surreptitious modification of the concept of morals. 57
55 Metapbysik der Sittm, VI, 227; Religion within the Limits of Reason Alone,
VI, 6 n. (Greene and Hudson, 6 n.). Kiesewetter, one of Kant's most enthusiastic
(yet singularly instructive) followers, would not have regarded this as an addition.
He wrote Kant, March 3, J790 (XI, 137): "I am convinced ... that the fundamental
principle of your moral system is perfectly harmonious with the Christian religion,
perhaps even that if Christ had heard and understood you he would have said, 'In-
deed, that is what I intended to say in speaking of the love [of] God.'''
56 By Herbert Spiegelberg, Gesetz. und Sittengesetz. (Zurich and Leipzig: Niehaus,
1935), p. 2 sz. This book gives a valuable account of the development of the concept
of moral law; few studies, I think, show Kant's originality in concepts and termi-
nology better than this account of the slow development of the notion of moral Jaw
and its differentiation from natural and divine law.
57 Streit der Fakultiiten, VII, 42. "Love God" and "love thy neighbor" alike have
only ethical content (Critique of Practical Reason, 83 [175-76]). Aside from the fact
The Postulates of Pure Practical Reason 28,
Christianity, the only true religion/is is also the only one that con-
tains a true moral theory. It is not heteronomous, because it commands
that we seek holiness apart from the motive for divine reward. And
it does not base the knowledge of what the law commands on the
acceptance of any historical dogma. But, while it is as pure as the
Stoic conception, it is more realistic, in not permitting us to think
that holiness (or wisdom) is humanly possible without the grace of
God. And without this confession of human impotency, the moral
ideal is secretly degraded to a level achievable by natural man. In the
face of the sublimity of the moral law, however, humility and not
Stoic pride is the only adequate response (127 n. [224 n.] ).
§ 7. THE TWO AWESOME THINGS
Natural theology, arising from the contemplation of nature, is sup-
plemented by contemplating that which is not nature; not by con-
templating God, of whom we have only a "conjecture" and whom
we do not see in his "awful majesty," but by harkening to the "heaven-
ly voice of duty" in us.
The two realms of reason, the theoretical and the practical, are
brought into poignant juxtaposition in the celebrated conclusion: "Two
things fill the mind with ever new and increasing admiration and awe,
the oftener and the more steadily we reflect on them: the starry
heavens above me and the moral law within me" (16 I [260]). These
two had often before been represented together by Kant,59 but never
that Kant's theology was moral and that of the deists was natural, this connection of
morals and religion, whereby the latter contributes nothing to the content of the
former, is common to both. Thus Kant says (Streit der Fakultiiten, VII, 36) that
there is no material difference (difference in object) between morality and religion,
but only a formal difference; and Tindal characteristically wrote: "Acting accord-
ing to the reason of things considered in themselves [is morality; religion is] acting
according to the same reason considered as a rule of God" (Cbristianity as Old as
Creation; quoted from Leslie Stephen, History of English Thought in the Eight-
eenth Century, I, '44).
58 Perpetual Peace, VIII, 367 n. (Beck, 31 n.): There can be only one true religion
because there is only one true morality. Similarly, Religion within tbe Limits of
Reason Alone, VI, 107 (Greene and Hudson, 98), and Streit der Fakultaten, VII, 36:
there is only one religion, of which Christianity is the plainest (schliclneste) form.
59 Allgemeine Naturgescbicbte und Tbeorie des Hi1'1mle/s, concluding paragraph;
Beobachtungen tiber das Gefiibl des Schon en und Erhabenen, II, 208-<); Traume
eines Geistersehers, II, 332; Einzig 1110glicher Beweisgrund ... , II, 141. A similar
conjunction in Seneca, one of Kant's favorite authors, has been pointed out by
Vaihinger in Ad Heft,·iam 111atrem de comolatione cap. viii (cf. "Ein beriihmtes
Kantwort bei Seneca?" Kant-Studien, II [1898], 491 ....()3); but, as Vaihinger in-
dicated, the Seneca passage is more like those in Kant's precritical writings than
that in the Critique.
with the grand simplicity of this passage. That the two were deeply
connected in Kant's own life of feeling, no doubt having first been
joined by Kant's mother,60 is sufficiently attested by the early biog-
raphers who knew him personally.
But there is a marked progress in this passage, not only stylistically
but also philosophically; it arises from the fact that, when writing the
earlier ones, Kant had not yet succeeded in sharply distinguishing be-
tween the legislations of the two realms. In his earlier works he had
thought of the starry heavens as a possible abode of moral beings of a
higher order than man or as the place of our souls after death. All that
is now transcended in the complementary conceptions of an unmoral
astronomy and a non-naturalistic ethics; all that remains of the older
conception of the relation of the moral law to the natural law is the
Typic. The stark contrast between the two, not some simple harmony
hazarded between them, gives force to their bold contexture. What
had previously been thought of as a synthesis from the standpoint of
the physical object-the evidence of natural teleology, the natural
origin of the moral disposition-he now sees as a synthesis of contrasts
within man's own rational nature. The starry heavens seem sublime
because man first feels reduced by them to impotency, only to rise
above them again when he knows that his rational nature, which
comprehends them, is not subdued but heightened by the magnitude
and power revealed in them. He erroneously attributes a sublimity to
nature which actually belongs only to his own rational being;61 the
sublimity ascribed to nature is a clue to his own superiority to nature,
for we convert "respect for the idea of humanity in our own subject
into respect for the object."62 What the starry heavens awake in us
only indirectly is produced directly by the contemplation of the moral
law, sublime in itself, and of the moral agent who embodies the law;
the humility thus induced in man is itself sublime. 63
There is no tendency remaining now to use natural concepts in the
articulation and elaboration of moral ideas or to confuse two things,
to the detriment of each. But the two things are not set before us
merely as a literary peroration. Even on the last pages Kant extracts
60 In the famous statement on his reverence for his mother (jachmann, Immanuel
Kant geschiJdert in Briefen an einen Freund [1804], neunter Brief), he speaks of her
as having "planted and nourished the seed of the good" and "opened [his] heart to
the impressions of nature," the same combination as in the present apothegm.
61 Critique of /ud?;mem, § 16 (especially V, 25 6 [95]).
62 Ibid., § 27 (V,lS7 [96]).
s31hid., § 18 (V, 264 [103]); Critique of Practical Reason, 117 (213); Beobach-
tung en uber das Ge[iibl des Scbonen und Erhabenen, II, 215.
a lesson of use to moral philosophy by drawing an analogy from the
history of astronomy.
"Admiration and respect," he says, "can indeed excite to inquiry, but
they cannot supply the want of if' (162 [261]). The consequence of
passive admiration and an admixture of human wants in subsequent
astronomical investigation was astrology.64 A like undisciplined elab-
oration of the "noblest attribute of human nature" led, in morals and
religion, to fanaticism and superstition. With the perfection of the
method of scientific investigation in astronomy and physics, there was
brought forth "a clear and henceforth unchangeable insight into the
structure of the world" (163 [26 I]). Perhaps even more important, it
gave a model to philosophy, to set it on the secure path of science. 6s
This method he now recommends in the investigation of morals. "Sci-
ence," in the sense of critically and methodically directed inquiry, is
"the narrow gate that leads to wisdom."
The Critique of Practical Reason is meant to be the key to that
narrow gate.
64 He has drawn a similar analogy in Prolegomena, IV, 366 (Beck, 115): "Critique
stands in the same relation to the common metaphysics of the school as ... astron-
omy to the astrology of the fortune-teller."
65 Cf. Critique of Practical Reason, 108 (2°3); Critique of Pure Reason, B xxiv.
Bibliography
Bibliog;raphy
I. TEXTS OF "KRITIK DER PRAKTISCHEN VERNUNFT"

Critik der practischen Vernunft von Immanuel Kant. Riga. bey Johann
Friedrich Hartknoch, 1788 [sic]. Pp. 292. Zweyte, vierte, fiinfte, und sechste
AufIage[n], 1792, 1797, 1808, 1827. (Third edition not published; second
edition was a double printing.) Reprint: Neueste mit einem Register
vermehrte Auflage. Gratz [sic], 1796. Other reprints: Frankfort and
Leipzig, 1791, 1795, 1803·
ROSENKRANZ, KARL, and SCHUBERT, F. W. In Vol. VIII of their edition of
Kaut's works. Leipzig, 1838.
HARTENSTEIN, G. In Vol. IV of his edition of Kant's works. Leipzig, 1838
(in Vol. V of the edition of 1867).
KIRCHMANN, J. H. VON. In the series "Philosophische Bibliothek." Berlin,
1869, 1870, 1882, 1897.
KEHRBACH, KARL. In the series "Reclams Universalbibliothek." Leipzig,
[ 1878 ].
VORLANDER, KARL. In the series "Philosophische Bibliothek." Leipzig, 1906;
9th ed., 1929.
NATORP, PAUL. In Kants Gesanmlelte Schriften, Vol. V. Berlin: Konigliche
Preussische Akademie der Wissenschaften, 1908, 1913.
KELLERMANN, BENZION. In Immanuel Kants Werke. Edited by Ernst
Cassirer et al. Vol. V. Berlin: Bruno Cassirer, 1914, 1922.
GROSS, F. In "Grossherzog Wilhelm Ernst Ausgabe" of Kants Sammtliche
Werke. Leipzig: Inselverlag, 1920.
WEISCHEDEL, W. In Kants Werke. Vol. IV. Leipzig: Inselverlag, 1957.
II. TRANSLATIONS OF THE "CRITIQUE OF PRACTICAL REASON"

Kant's Critique of Practical Reason and Other Works on the Theory of
Ethics. Translated with a memoir of Kant by THOMAS KINGSMILL ABBOTT.
London: Longmans, Green, 1873; 6th ed., 1909; reprinted, 1954.
Kant's Critique of Practical Reason and Other U'ritings in MoraL Philos-
ophy. Translated and edited, with an Introduction, by LEWIS WHITE
BECK. Chicago: University of Chicago Press, 1949.
Immanuel Kant: Critique of PracticaL Reason. Translated, with an Introduc-
tion, by LEWIS WHITE BECK. New York: Liberal Arts Press, [1956].
lmmanuelis Kanti; Critica rationis practicae. In Opera ad phiLosophiam
critic am Latine vertit Fredericus GottLob Born, voLumen tertium. Leipzig:
Impensis Engelhard Ben. Schwickerti, 1797.
288 Bibliography
French translations by JULES BARN[ (Paris, 1848), FRAN~OIS PICAVET
(Paris, 1888; 2d ed., 1902; reprinted with new Introduction by FERDINAND
ALQUlE [Paris: Presses Universitaires de France, 1949]), and J. GlBELIN
(Paris: Vrin, 194,).
Italian translation by FRANCESCO CAPRA (Bari, 1909; 7th ed., Florence,
1955', with Introduction by EUGENIO GARIN).
HI. TRANSLATIONS OF OTHER WORKS BY KANT AS CITED IN COMMENTARY
(IN CHRONOLOGICAL ORDER)
Principiorum primorum cognitionis metaphy sicae nova dilucidatio (1755).
Cited as "Nova dilucidatio." F. E. ENGLAND, Kant's Conception of God,
Appendix, pp. 213-52' London: Allen & Unwin, 19 2 9.
The Distinctness of the Principles of Natural Theology and Morals (1764).
Also called "Prize Essay." L. W. BECK, Kant's Critique of Practical Rea-
son and Other Writings in Moral Philosophy (Chicago, 1949),PP. 261-8,.
The Forms and Principles of the Sensible and Intelligible Worl4s (1770).
Cited as "Inaugural Dissertation." JOHN HANDY SIDE, Kant's Inaugural
Dissertation and Early Writings on Space, pp. 35-83. Chicago and Lon-
don: Open Court, 191.9.
Lectures on Ethics (ca. 1780). Translated by LOUIS INFIELD. New York:
Century Co., [1930].
Critique of Pure Reason (1781, 1787). Translated by NORMAN KEMP SMITH.
London and New York: Macmillan Co., 1929.
Prolegomena to Any Future Metaphysics (1783). Translated by LEWIS W.
BECK. New York: Liberal Arts Press, 1951.
Foundations of the Metaphysics of Morals (178,). In ABBOTT, op. cit.
Abbott's title is Fundamental Principles of the Metaphysics of Morals.
What Is Orientation in Thinking? (1786). In BECK, Kant's Critique of
Practical Reason and Other 'Writings in Moral Philosophy, pp. 293-305.
Critique of Judgment (1790). Translated by J. H. BERNARD (1892). New
York: Hafner, 1951.
Religion within the Limits of Reason Alone (1793). Translated by THEO-
OORE M. GREENE and HOYT H. HUDSON. Chicago: Open Court, 1934.
Perpetual Peace (1795). Translated by LEW[S WHITE BECK. New York:
Liberal Arts Press, [1957]. (This translation, with Akademie pagination,
is included also in BECK, Kant's Critique of Practical Reason ... , pp.
306-45, but citations are to the 1957 edition.)
IV. STUDIES OF KANT
ADICKES, ERICH. "Korrekturen und Konjekmren zu Kants ethischer Schrift,"
Kant-Studien, V (1901), 207-[4.
ASTER, ERNST VON. "Band V und VI der Akademie Ausgabe," Kant-Studien,
XIV (1909), 468-76. (Review of the Natorp edition of the Kritik der
praktischen Vernunft.)
ATTISANI, ADELCHI. MetodQ attivo e metodo speculativo nella metodica
della ragion pratica di E. Kant. Messina: Sessa, 1951.
Bibliography 189
BALLAUF, THEODOR. Verniinftiger Wille und gliiubige Liebe. Jnterp-retationen
zu Kants und Pestalozzis JVerk. Meisenheim: Anton Hein, 1957.
BARNI, JULES. Examen des Fonde11lents de la mhaphysique des m(J!urs et
de la Critique de la raison pratique. Paris: Ladrange, 185 I.
BECK, LEWIS WHITE. "Apodictic Imperatives," Kant-Studien, XLIX (1957),
7- 2 4.
---. "Les deux concepts kantiens de vouloir dans leur contexte politique,"
Annales de philosophie politique, IV (1962), II 9-37.
BENDAVID, LAZARUS. Vorlesungen uber die Critik der p-raktischen Vernunft.
Vienna: Stahel, 1796.
BOHATEC, JOSEF. Die Religionspbilosophie Kants in der "Religion innberhalb
der Grenzen der blossen Vernunft" mit hesonderer Beriicksichtigung
ihrer theologisch-dog1llatischen Quellen. Hamburg: Hoffman & Campe,
193 8.
BRASTBERGER, GEBHARDT ULRICH. Untersuchungen uber Kants Kritik der
praktischen Vernunft. Tiibingen: Cotta, 1792.
CAIRO, EDWARD. The Critical Philosophy of Immanuel Kant. 2 vols. London
and New York: Macmillan Co., 1899 and 1909.
CASSIRER, ERNST. Rousseau, Kant, Goethe. Translated by JAMES GUTMAN,
P. O. KRISTELLER, and J. H. RANDALL, JR. Princeton: Princeton University
Press, 1947.
COHEN, HERMANN. Kants Begriindung der Ethik. Berlin: Diimmler, 18n.
COOK, WEBS1ER. Tbe Ethics of Bishop Butler and Immanuel Kant. Ann
Arbor: University of Michigan, r888.
DELBos, VICTOR. La Philosophie pratique de Kant. Paris: Alcan, 1905; 2d ed.,
19 26 .
DORING, A. "Kants Lehre vom hochsten Gut," Kant-Studien, IV (r898),
94- 101 .
DUNCAN, A. R. C. Practical Reason and Morality: A Study of Immanuel
Kant's Foundations for the Metaphysics of Morals. Edinburgh: Nelson,
1957·
EISLER, RUDOLF. Kant-Lexikon. Berlin: Mittler, 1930.
ENGLAND, F. E. Kant's Conception of God. London: Allen & Unwin, 1929.
EUCKEN, RUDOLF. "Uber Bildnisse und Gleichnisse bei Kant," Zeitschrift
fiir Philosophie und philosophische Kritik, LXXXIII (1883), 161--93.
FACKENHEIM, EMIL. "Kant's Concept of History," Kant-Studien, XL VIII
(1957), 39 1 --9 8.
FERVERS, KARL. Die Beziehung zwischen Gefiihl und Willen bei Tetens und
Kant. Diss., Bonn, 1925.
FOUILLEE, ALFRED. "La raison pure pratique cloit elle etre critiquee?" Revue
philosophique, LIX (1905), 1-33.
GAHRINGER, ROBERT E. "The Metaphysical Aspect of Kant's Moral Philos-
ophy," Ethics, XLIV (1954), 277--91.
GELFERT, JOHANNES. Der Pflichtbegrilf bei W ollf und anderen Philosophen
der deutschen Aufklarung mit Rucksicht auf Kant. Diss., Leipzig, 1907.
290 Bibliogtapby
GRASSI, LEONARDO. Preludi storico-attualistici alla Critica della ragion pratica.
Catania: Crisafulli, 1943.
GUEROULT, M. "Canon de la raison pure et Critique de la raison pratique,"
Revue internationale de pbilosopbie, VIII (1954), 333-57.
GUREWITSCH, ARON. 2ur Geschichte des Achtungsbegriffes und zur Theorie
des sittlichen Gefiihls. Diss., Wlirzburg, 1897.
HAEGERSTROM, AXEL. Kams Ethik im Verhaltnis zu seinen erkenntnistheo-
retischen Grundgedanken systematisch dargestellt. Uppsala: Almqvist &
Wiksell; Leipzig: Harrassowitz, 1902.
HAEZRAHI, PEPITA. "The Avowed and Unavowed Sources of Kant's Theory
of Ethics," Ethics, LXII (1952), 157-68.
HEGLER, ALFRED. Die Psychologie in Kants Ethik. Freiburg: J. C. B. Mohr,
189 1 •
HEIDEMANN, INGEBORG. Spontaneitat und Zeitlichkeit: Ein Problem deT
Kritik der reinen Vernunft. Erganzungsheft der Kant-Studien, No. 75.
Cologne, 1958.
---. Untersuchungen zur Kantkritik Max Schelers. Diss., Cologne, 1955.
HE[MSOETH, HE[NZ. Studien zur Philosophie Immanuel Kants: Metaphys-
ische Urspriinge und ontologische Grundlagen. Cologne: Balduin Pick,
195 6 .
HENR[CH, D. "Hutcheson und Kant," Kant-Studien, XLIX (1957), 49-69.
HOFFDING, HARALD. "Rousseaus Einfluss auf die definitive Form der
kantischen Ethik," Kant-Studien, II (1898), I 1-21.
JODL, FRIEDRICH. Geschicbte der Ethik als phiJosophischer Wissenschaft.
Stuttgart and Berlin: Cotta, 1920.
JONES, WILLIAM T. Morality and Freedom in the Pbilosophy of Kant. Lon-
don: Oxford University Press, 1940.
JOSTEN, CLARA. Christian Wolffs Grundlegung der praktischen Philosophie.
Leipzig: Meiner, [931.
KAUBLER, BRUNO. Der Begriff der Triebfeder in Kants Ethik. Diss., Leipzig,
19 17.
KHODOSS, CLAUDE (ed.). Kant: La Raison pratique. Paris: Presses Univer-
sitaires de France, 1956. (A very useful collection of texts, arranged by
subject matter and with glossary.)
KIESEWETTER, JOHANN GOTTFRIED CHRISTIAN. 0 ber den ersten Grundsatz
der Moralphilosophie ... nebst einer Abha1ldlung vom Herrn Prof. Jacob
iiber die Freiheit des Willens. 2 vols. Berlin: Carl Massdorff, 1790,1791.
KNOX, T. M. "Hegel's Attitude to Kant's Ethics." Kant-Studien, XLIX
(1957),70-81.
KRUEGER, GERHARD. Philosophie und Moral in der kantischen Kritik. Tiibin-
gen: J. c.B. Mohr, 1931.
LINDSAY, A. D. Kant. London: Oxford University Press, 1934.
LIPPMANN, E. O. VON. "Zu: 'Zwei Dinge erfiillen das Gemiit . . . " " Kant-
Studien, XXXIV (1929),251)-61; XXV (1930),409-10.
LORENTZ, P. <lUber die Aufstellung von Postulaten als philosophische
Methode bei Kant," Philosopbische Monatshefte, XXIX (1893),412-33.
Bibliograpby 29'
MARTY, FR. "La Typique du jugement pratique pur: la morale kantienne
et son application aux cas particuliers," ArciJives de philosophie, No. [
([935), pp. 56-8 7.
MELLIN, GEORG SAMUEL ALBERT. Encyclopiidisches Worterbuch der kriti-
schen Pbilosophie. 5 vols. in [0. Ziillichau and Leipzig, [797.
---. Kunstsprache der kritischen Philosophie oder Sammlung aller Kunst-
'Worter derselben. Leipzig: Fromann, '798.
---. Marginalien und Register zu Kants Kritik der Erkenntnisvermogen.
II. Kritik der praktischen Vernunft. Ziillichau, [795. Neuherausgegeben
von LUDWIG GOLDSCHMITT. Gotha: Thienemann, '9°1.
MENZER, PAUL. "Der Entwicklungsgang der kantischen Ethik in den Jahren
1760-[785," Kant-Studien, II (1897), 290-321; III (1898),4°-104.
MESSER, AUGUST. Kants Ethik: Eine Einfuhrung in ibre HauptprobJeme und
Beitriige zu deren Losung. Leipzig: Veit, 1904.
---. K07mnentar zu Kants ethischen und religionsphilosophischen Haupt-
schriften. Leipzig: Meiner, 1919.
MICHAELIS, CHRISTIAN FRIEDRICH. Uber die sittliche Natur und Bestimmung
des Menschen: Ein Versuch zur Erliiuterung iiber I. Kants Kritik der
praktischen Vernunft. 2 vols. Leipzig: Iohann Gottlob Beigang, 1796,
1797·
MORITZ, MANFRED. Studien zum Pflichtbegriff in Kants kritischer Ethik.
The Hague: Nijhoff, '951.
NAHM, MILTON C. " 'Sublimity' and the 'Moral Law' in Kant's Philosophy,"
Kant-Studien, XLVIII (1957), 502-14'
NECK lEN, FERDINAND. Die Lehre vom Gefuhl in Kants kritiscber Ethik.
Diss., Leipzig, 1938.
PATON, H. J. The Categorical Imperative: A Study in Kant's Moral Phi/os-
o ph y. London: Hutchinson [ [946]; Chicago: University of Chicago
Press, 1948.
---. In Defence of Reason. London: Hutchinson, 1951. (See especially
"Can Reason Be Practical?" [pp. I 17-56] and "Kant's Idea of the Good"
[pp. 157-77]·)
PAULSEN, FRIEDRICH. Immanuel Kant, His Life and Doctrine. Translated by
J. E. CREIGHTON and ALBERT LEFEVRE. New York: Charles Scribner's
Sons, 1902.
PEACH, BERNARD. "Common Sense and Practical Reason in Reid and Kant,"
Sophia (Padova), XXIV (1956),66-71.
PORTER, NOAH. Kant's Ethics: A Critical Exposition. Chicago: Griggs, 1886;
2d ed., 1894.
ROSENTHAL, GERTRUD. "Kants Bestimmung des Erziehungsziels," Archiv fiir
die Geschichte der Philosoph ie, XXXVII (1926), 65-74.
Ross, SIR [WILLIAM] DAVID. Kant's Ethical Theory: A Commentary on the
Grundlegung zur Metaphysik der Sitten. Oxford: Clarendon Press, 1954.
RUGE, ARNOLD. Die Deduktion der praktiscben und der moralischen Freiheit
aus den Prinzipien der kantischen MoraJlehre. Hab.-Schrift, Heidelberg,
19 10.
191 Bibliography
ScHILPP, PAUL ARTHUR. Kant's Pre-critical Ethics. Chicago and Evanston:
Northwestern University Press, 1938.
SCHMIDT, KARL. Beitriige zur Entwicklung der Kant'schen Ethik. Marburg:
Elwert, 1900.
ScHWEITZER, ALBERT. Die Religionsphilosophie Kants in der Kritik der
reinen Vernunft bis ZUT Religion innerhalb der Grenzen der bJossen
Vernunft. Freiburg: J. C. B. Mohr, 1899.
SEIDEL, ARTHUR. Tetens Einfluss auf die kritische Philosophie Kants. Diss.,
Leipzig, 1931.
SNELL, FRIEDRICH WILHELM DANIEL. Menon oder Versuch in Gespriichen
die vornehmsten Punkte aus Kants Kritik der praktischen Vernunft zu
erliiutern. Mannheim: Schwan & Gotz, 1789; 2d ed., rev., 1796.
STANGE, CARL. Die Ethik Kants: Zur Einfii.hrung in die Kritik der prakti-
schen Vernunft. Leipzig: Dietrich, 1920.
TEALE, A. E. Kantian Ethics. New York: Oxford University Press, 1951.
UNGER, RUDOLF. "'Der bestirnte Himmel uber mir': Zur geistesgeschicht-
lichen Deutung eines Kant-Wones." In Festscf.?rift zur zweiten Jahr-
hundertfeier seines Geburtstages [in Konigsberg], pp. 239-63. Leipzig:
Dietrich, 1924.
VIDARI, GIOVANNI. "Sguardo introduttivo alIa 'Critica della ragion pratica,' ,.
Rivista di filosofia, XV (1924),223-31.
DE VLEESCHAUWER, H. J. La Deduction transcendentale dans l'a:uvre de
Kant. Paris: Leroux, 1937. (See especially Vol. III.)
VORLANDER, KARL. Immanuel Kant, der Mann und das Werk. 2 vols. Leipzig
Meiner, 1924.
WEBB, C. c. J. Kant's Philosophy of Religion. Oxford: Clarendon Press,
19 26 .
WHITTEMORE, ROBERT. "The Metaphysics of the Seven Formulations of the
Moral Argument," Tulane Studies in Philosophy, III (1954), 133-61.
WILLE, E. "Konjekturen zu Kants Kritik der praktischen Vernunft," Kant-
Studien, VIII (19°3),467-71.
WITTE, JOHANNES. Beitriige zum Verstiindnis Kants. Berlin: Mecklenburg
Verlag, 1874. (Contains one of the very few discussions of the categories
in the second Critique, pp. 92 ff.)
WUNDT, MAX. Kant als Metaphysiker: Ein Beitrag ZUr Geschichte der
deutschen Philosophie im acbtzehnten /ahrbundert. Stuttgart: Enke, 1924.
ZILlAN, ERICH. Die Ideen in Kants theoretischer und praktischer Philosophie.
Diss., Konigsberg, 1927.
ZWANZIGER, JOHANN CHRISTIAN. Commentar ii.ber Herrn Professor Kants
Kritik der praktischen Vernunft; nebst einem Sendschreiben an den
gelehrten Herrn Censor. Leipzig: Hischer, 1794.
I have not been able to find a copy of Theodor Gottlieb Ratze's Bei/age
zu Kants "Kritik der praktischen Vernunft" (Chemnitz, 1794) or of C. T.
Michaelis' Zur Entstehung der Kritik der praktiscben Vernunft (Berlin,
(893). The former is listed in Benjamin Rand's Bibliography of Philosophy,
Bibliography 293
Psychology, and Cognate Subjects (Vol. III of Baldwin's Dictionary of
Philosopby and Psycbology [New York: Macmillan Co., 1905]), Part I,
p. 3 ro. The latter is listed in Friedrich U berwegs Grundriss der Geschichte
deT Philosoph ie, 12th ed., Vol. III, p. 716; but there is reason to believe this
entry is an error.
While this book was in press, and hence too late for me to make use of
them, there appeared two very instructive studies by John R. Silber of topics
in the Critique of Practical Reason: "Kant's Conception of the Highest Good
as Immanent and Transcendent" (Philosopbical Review, LXVIIl [October,
1959], 469-<)2) and "The Copernican Revolution in Ethics: The Good
Re-examined" (Kant-Studien, LI [I95~0], 85-101). They are especially
relevant to the issues examined in this Commentary, chap. ix, § 3, and chap.
xiii, § 3, respectively.
Index of Passages Cited from UCritique
of Practical Reason"
AKADUIIE ABBOTT AKADEMIE ABBOTT
EDITION" TRANSLATrON COMMENTARY EDITION· TRANSLATION COMMENTAIIY
3 10. 43. 44. 4 6• 169. 114 78• 13 0

174. 175. 18 3. 18 5. I 15 99. 160. 164
188. 240 115 98. 146. 160. 161
3 88 16. 27, 46 • 184 116 98, 122
4 88 27. 28, 59. 197. 207. 116 HZ. 196. 198
27 2 "7 H2. 164. 168. 169.
4 28.48 196. 198. 2ZZ
5 46 • 58. 174, 257 30 118 Ill, 173. '99. 200
5 40, 4 6 , 58, 257 30 "9 840 200
6 27n, 46 31 120 54. '40 , 15 2 • 165.
6 16, 58. 59. 166 166. 167. 18o, 182.
7 16. 19.57.59. 175 196,197
7 '74. 18 5. 190 32 110 166. 178
8 46• 5" 52. 57. 112. JZ III 107
154. 165. 235 33 111 'J.17
8 94 45. 46 • 53. 59. 60 33 IZZ 92,122. 167. 169. 196
9 94 90. 21 7. 218 34 12 3 96.102,115.118,1;0.
9 95 60 135. 16,
I[ ¢ 139. 149. 25 1• 252 48. 161. 203
12 97 60.169 91, 116
13 99 21 116
13 100 60. 1°4 ([6
14 100 59
101 32,44.46,49.92.113 128 66.100. 103.105.106.
15
16 102 44, ~. 540 66. 12 7. 223. 230
211.24° 12 9 66. 105. 107
10 5 77. 164 12 9 103. 104
106 79. 162 12 9 104, 241
107 13° 108, 2'3. 241
75. 84.95. 1[7
I; I 140. 166. 195
108 75. 8 3
108 97. 216 1]1 134. 161. 166. 17 2 •
ISo. 192
I DC} 38f• 93. 9B
109 39. 93. 105 17 2
110 93,95. 22 3 46
1[0 223 1'3
I l l , 253, 259
III 39. lOS
liZ 218 166. t67. '7 1 • 197
liZ 1)8. Zl8. ZZ 3 46, '74. 18 3, 18 5
18 7
1'3 40 • 80,99
114 83. 157. (1)8 '75, 188
181
• Akademie pagination included In ZI
Beck translations. 61, 18t. 182
295
296 Indexes
AKAD£HI£ ABBOTT AKADEHIE ABBOTT
EDITION· TRANSLATION COHHENTAIlY EJ)lnON* TRANSLATION COHHENTAlIY
18z 163. 1 35
38, 166, 178, 18z 221,225
181 191, 2H, 226
195 2Z0
254 46 , 233
119. 130• 13 1 , 137. 52
IS? H,212
90 55,67, 211
1}1 55, 66, 7 8, 81, 12 7.
13 0 128
36 , 13 0 , 13 2 , 1<tH, 17 8 47. 55.69
13S 66, 166
IH 195,218,220
60, 1Z7 66, 120, 141
60, 1%7 100, 107, 115, 120,
60 IH. 228
60 93 18 7 21 9
243 94 18S 162
13 6 , 137 95 189 186
39,137, 13 8 , 139 96 189 190
14 1, 14 2, 149, 273 96 190 179, 190
1%6, 141, 14z, 144, 97 190 179, 190, 19 1
15 2 97 19 1 179
146 98 19 1 190, 19 1, z7 0
15 6 9B 192 116, 191, 231, z70
113, 1%9, 158 99 192 188,270
158, 16z 99 193 30, 188, 191, 192
15 8 , 235 100 193 20 4
157, 161, 162, 235 100 194 188, 191, 205,206
235 101 195 190. z06
16 3, z3S 101 196 206
219, ZZI 102 196 206
217, %21,126, 233 10 3 197 207
211 104 199 25. 166
100 105 200 19<>
100 106 200 19<>
ZZ3 106 101 16
ZZ3, :n5 10 7 202 18 3
117. 221, 126 107 20 3 25, 186. 240
129 108 20 3 211, 283
:n6 109 204 z42. 246 , 247
126 109 20S 24 1 , 24 2 , 243
22[ 110 20 5 242, 248
35.9 1 110 206 268,275
203. :u6, 221, 229 III 20 7 246
223 112 20S 246
230 113 209 246, 26z
50, 1%5, 226, 228 114 210 244, 245
280 115 211 223, 224, 248, 269
2B,28o 115 212 2240 247
138,228 I16 2IZ 101, 2%4, 229, 23 0 ,
I2S. 135, 211 247, 27 1
163 u6 21) 220, :ni, 2-47
Indexes 2.97
AItADEIllE AaJlOTT A~DE"IE A.IIOTT
EDITION- TaANSLATION CO.... ENTARY EDlnON- TaANStATION COIlJOIIYARY
117 U) liS. 12). 1240 147. 1)6 134 lSS. 16)

181 137 134 177
117 214 106, 114. 130. 247 13S 136 175. 177
118 114 12 40 1)0.247 139 1)6 17 1
118 115 u4, 129. 147 139 137 177. 279
119 115 147 140 138 176• 171
119 u6 163 141 239 ISS
IZO 116 ISS. 249. 150• 16 3 142 140 151
120 '117 149. 16] 142 241 244
1Z1 21 7 47. 249. 26 3 143 141 243. 1440 2540 267
IU 218 183. 24 1• 250. 26) 143 242 28. 245.2S6
1%1 u8 267. 168 144 142 2B.2SS
122 21 9 252. 260. 167. 270 145 243 17 2
113 21 9 169. 270 145 144 247,274
113 210 us. 170 146 144 185. 255. 1,0
1%4 210 274 146 245 185. 157
1%5 121 259. 174 147 245 185, 25 8
126 Z13 155 147 246 185
1%7 224 u8. u9. 270. 281 14IJ 146 185. 2SS
u8 224 241 151 249 46
151 25° 103
u8 uS 27° 153 25 1 141• 2)0
u8 226 17° 154 25 2 no
1%9 216 270,279 ISS 253 235
119 u7 268 155 254 195
1]1 218 277 156 154 195
1]1 229 175. 27 8 157 255 23 0 , 235
13 2 lZ9 207. 243. 2n, 259, ISS 25 6 135
158 257 2]6
27 2• 175 257 2JS,236
131 142 159
23° 160 158 101, 124
133 13 1 187, 261 160 259 23 6
134 13 1 18 7. 151 161 259 23 6
134 23 2 187, 1SZ. 254 161 260 17. 53. 281
135 23 1 259. 163 162 261 28)
135 233 26 3 16] 161 18)
1)6 233 155. 163 16] 262 211, 235
Index of Names and Subjects
Note.- The titles of Kant's works are not cited on the mere occasion of a refer-
ence to them or of a quotation from them; they are cited only where there is Some
discussion of the work in question or where the occurrence of a particular tenn
in the particular work is of importance.
Abbott, Thomas Kingsmill, viii, xi, 90 n., A priori, 10, 60, 65, 118 n.
91 n., 91 n., 166 n., 177 n., 1%7, 14% n. theoretical and practical, 84
Abel, Jacob Friedrich, 57 Aquinas, St. Thomas, 37 n., 184
Act, 111 ff., 119, 141 Architectonic, viii fl., IS, 55 f., 126, 185
universalizability of, 159 n., 160, 161 Aristotle, 37,41, 135, 1840 188, 231
Action, 1340 153 Astrology, 283
Activism, 116 Aufgegeben, 186
Actor, 30 ff., 193 ff. Autarchy of will, 208
Addison, Joseph, 166 n. Autonomy, II, 14, 59n., 102, 1Z1f1.,
Aesthetic 151 n., 166, 168, 180, 196 fl., 216,267,
in Critique of Practical Reason, 56, 27 0 ,174
67 f., 69, chap. xii Aversion, faculty of, 90 n.
transcendental, 56, 67, 186, 111 Awe, 181 f.
Aesthetic concepts, 120, 2:z I n., 131 j cf.
also Beauty; Sublimity Bad will, 198 n.; cf. also Evil
Alleged Right To Lie, On an, 80 Baeumler, Alfred, vii
Almsgiving, 161 Balguy, John, 41 n.
Altruism, 100 Bami, Jules, vii n., 154 n., 177 n.
Analogies of experience, 186 n., 193 Bauch, Bruno, 12 5 n.
Analytic, 65 f. Baumeister, Friedrich Christian, 94 n.,
of practical reason, 53, 56, 66 ff., 78, 111 177 n.
of pure practical reason, 80, 1Z7 Baumgardt, David, 100 n.
transcendental, 56, 65 f., 69, 193, 1 II Baumgarten, Alexander Gottlob, 37 n.,
Anaxagoras, 177 n. 90 n., 94 n., 106 n., 131 n., 177 n.,
Anderson, Georg, 14 n. 151 n., 280
Announcement of Lectures for Winter Beauty, 91, 106, 132 n., 113
Semester 1765-66, 6 n. Beck, L. W., viii, 85 n., 103 n., 110 n.,
Anthropology, 7, 30 n., 90 177 n., 119 n.
Anthropology from a Pragmatic Point Bedingung, 81
of View, 54 Belief, 107, 1SS f., 166 n., 270; cf. also
Anthropomorphism, 155, 176 f. Faith
Antinomy, 15, 156 Bendavid, Lazarus, 150 n.
of practical reason, 16, 245 ff. Bering, Johann, 13,57 n.
of theoretical reason, 48, 140, 147, 165 Bernacker, Karl, 94 n.
third, 14 ff., 1740 183 n., 184, 186 fl., Bernard, J. H., xi
188 fl., 193, 194 fl., 200, 211, 247 fl., Bestimmung, 78, 81
149 Bestimmungsgrund, 78, 95
A posteriori, 20, 84, 110 Bewegungsgrund, 91 n.
Apperception, 189 Biester, Johann Erich, 57 n.
transcendental unity of, 138, 127, 265 Bohatec, J osei, 116 n.
Appet;tus, 37, 94 Boleyn, Anne, 136
Applied ethics, 10 n. Bonifas, 131, 132,149 n.
"Apply to," meaning of, 101) Bonum consummatum, 24 2 , 245
Indexes 299
Bomnn supremum, z4z, 245, z68 "Critique," meaning of term, 9, 43 if.
Bom, Friedrich Gottlob, J 3 n., 91 n., Critique of Judgment, 49 f., 191 f., ZZ9 n.,
J77 n . Z75 f.
Boswell, James, Z3Z n. "Critique of Moral Taste" [projected
Brastberger, Gebhardt Ulrich, Z07 n. book],6n.
Broad, C. D., JlI n. "Critique of Pure Practical Reason"
Brockdorff, Cay von, 60 n. [projected book], I I if., 4Z if.
Burke, Edmund, ZZJ D. Critique of Pure Reason, chap. ii, 43 f.,
Burthogge, Richard, 37 n. 51 if.
Butler, Joseph, 94 n., [OJ f., U9D. canon of, [0 if., 190 n.
classification of sciences in, 53
Caird, Edward, 4B composition of, 3,8, 13 if., SS
Calvinism, zoo deductions in, 69, 109 f., 170 if.
Canon of pure reason, 10 I., [4 ff., z4t dialectic of, 139f.
J9O n • freedom in, 16 if., J77, [S[ if.
Canz, Israel Gottlob, 6 n. God in, 171 if.
Capra, Francesco, 91 n., 177 n. immortality in, 165 f.
Cassirer, H. W., vii maxims in, 83 n., 193
Catechism, z36, 276 moral philosophy in, JO n., [64
Categories of practical reason, tz9, motives in, Z14 ff.
136 if.; cf. also under specific cate- organization of, SS f.
gories primacy of practical reason in, 250
Category, ll, Z40 109 f., 13 6 ff., 143, 145, relation of, to Critique of Practical
193, z18, 139. z64 Reason, 46 if., 56 if.
Causa noumen07l, 187. 189. 191, 19z, 196 to Metaphysics of Morals, J I
Causality, ZI, 13, z5 f., 60, JZ9. J37. 14z. to Prolegomena, 52
ISS, 171, 175, 181 if.• 183 n .• 186 if., to "System of Philosophy," 54 n.
z06f. schematism in, 157
Chamisso, Adelbert, Z57 n. self in, 194
Character. intelligible and empirical. 13, theory of categories as regulative max-
30 n., 191, zz6 ims, 193
Cheselden, William, 59 n. two theories of Ideas in, 183 n., 164
Choice. chap. iii, 131, 134t 194 Crusius, Christian August, 38 D., 104, 107,
Christianity, ZIS, 229 n., z33, 156, 170, J2 3 n., 107 n., z66 n., z80
177 n., 180n., 281 Cumberland, Richard, 106 n.
Cicero, Marcus Tullius, [32 n. Custom, II4 f.
Clarke. Samuel, 103 n.
Cohen, Hermann, vii, 167 n. Decorum, 107, 119 n.
Coleridge, Samuel Taylor, 257 n. Deduction, II, 170
Condition (Bedingung), 81 and exposition, I J I
Condition (Zustand) , 97, 147 f. in Foundations of the Metaphysics of
Conjectural Beginning of Human His- Morals, 110. 171
tory, 234, z79 metaphysical, 69, 70, 1<>9 if., n6, 153
Conscience, 107 n., ZZ4 n. of principle of pure practical reason,
Constant, Benjamin, 80 chaps. ix and x, 16, H, 109 ff., 1 n,
Contentment, 106 f., no, 119 n., 130, 136, JJ6, 116, 170 if.
248n. subjective, 69
''Copernican Revolution," ZI, tz40 J79, transcendental, 69, 109 f., I z6
182, J99 Definition, 118 D.
Copernicus, Nicholas, ZI transcendental, 53, ZI7 f., ZI9
Cosmological argument, Z7J f. Deism, z81 n.
Counsel of prudence, S5, 1[6 Dempster, George, 231 D.
Creation, 105 if. Design, idea of, 27%, 276
"Credential of the moral law," 46 n .. Desire, Ion., 38,53, 7[ if., 81, 90 if., [35,
174 if. 117 if.
300 Indexes
Desire-Continued 151 f., I66if., I¢, 200
lower and higher faculties of. 38 f .. Faith, 26, :z8, 249. 255 f.; ct.
also Belief
94ft'· Fanaticism, 2)5, 255 f., :z83
rational. 139 Faust, 157 f.
Desired and desirable, the. 133. 138 f. Feder, Johann Georg. 57, 5B n., 59 n., 60
Detennination (Bestim1mmg). 78. 81. 121 Feeling,s. 15,68,93, 2Zl; cf. also Moral
Dewey. John. 201 n .. 132 Feeling
Dialectic Fichte, Johann Gottlieb, 54 n., 194
in Foundations of the Metaphysics of Fictionalism, 193 n., 254 n.
Morals, 140 n. First Cause. 188
of pure practical reason. 16, 440 46. Flatt, Johann Friedrich, 59
208. chaps. xiii and xiv Fontanelle, Bernard Le Bovier de, 229
transcendental, 14. 56. 12 7. 193, 139 f. Fonn, 116, 198
Dignity, 226, 13S and content or matter, ¢, 134
Doctrine of right, 53 Fonnalisrn, 118 n.
Dognna~m, 185, 249. 255. 25 8 Fonnula, 5B n.
Dreams of a Spim-Seer, 113, 158 Foundations of the Metaphysics of
Dualism, 26, 191 Morals
Duncan, A. R. C .. viii, 11 n. autonomy in, I¢
Duncan-Jones, Austin, 111 n. classification of heteronomous princi-
Duty,9, 10 n., 41 n., 46, So. 5... 113, u8 n., ples in, 103
140 n .. ISO, 198. 217, 212, 215 if.; cf. deduction in, 110, 171
also Obligation dialectic in, 240 n.
perfect and imperfect. 1140 147 f •• examples in, 110 n., 161 f.
150 if.• 161 freedom in, 177
prima facie. 198 n. imperatives in, 84 ff.
incentive and motive in, 90 n., 216
Earp. C. B., x Kant's replies to reviews of, 57 if.
Ebbinghaus, Julius, 202 n. law and maxim in, 81
Education. 104 f .. 1]3 on limits of practical philosophy, 49
Egoism, 97, roo if.; ct. also Self-love; organization of, 51 f., 84, I II, 209
Selfishness personality in, 126
Empiricism, S' 8 n., 44t 60 f., 162, 18S, t8c} relation of
"Empty space." 175, 25 6,262 to Critique of Practical Reason,
End which is also a duty, 107, 244 vii f., Ion., I I f., 14f., 17, 43, 48
End-in-itself, 216. 275 n., 178 to Metaphysics of Motals, 54
Ends Fragment Six, H, 16,51, 166n., 115
realm of, 99, 145 f., 160 if., 227, 178 Frederick the Great, 15
system of, 242, 245 Freedom, I l f., 16f., 26ff., p, 46n., 48,
Enlightenment, 233, 235, 2sB 70, 141 if., 145. chap. xi, 110, 114
Enthalten, 78 categories of, 1)8, 144 ft'.
Enthusiasm. us, 163 n. comparative, 179, 190, 191 n., 194, 106
Epicurus, 104, 105, 185, 223, 230, 146, 250 as "credential of moral law," 174 f.
Epimenides, 160 n. in Crjtique of Pure Reason, 26 ff., 177,
Equality,235 181 fr.
Eudaemonism, 11] f., :116 n., 130. 265 enjoyment of, 114, Z19 f.
Evil, 1I7 n., 130, 136, 200 n., 127 f. and fact of pure reason, 166 n.
radical, 203 11. in Foundations of the Metaphysics of
cf. also Bad will Morals, 177
Ewing, A. C., vii and law, [79 f.
Examples, 110 n., 165. 195 n •• 228, 236 negative, I2Z f., ISo, 191, I¢, 103, 107
Exceptions, 83, 115 n .• 147, 151 neutral, :Z05 n.
Exposition, I 10, 126 positive, 113,168,172. lBo, 191, 197,203,
107
"Fact for pure reason," 167 if., 173 postulate of, 207 f., 248 n., 259, z61
"Fact of pure reason," ~6, 540 140 f., practical, 188 ff.
11ldexes 301
problematic judgment of, 16 f., 188 Hen. Marcus, 8,114

regulative and constitutive principle Heteronomy, 10% fl., 180n., 136n., 241,
of,99, 169 243 n.
as substitute for intuition, 171 f. Highest good; see SU1'I'lmUm bonum
transcendental, 130, 179, 189, 105 n., 111 Hobbes. Thomas, 101 f., 103 n., 104
"vicious circle" in argument for, 59, Hoffmann. Adolf Friedrich, 6 n.
73, 174 Holiness, 107, 167 fl.
"well-ordered," 114 Holy will, SO n., Jl4> 138, 199. 203, 216,
Friedrich, Carl Joachim, viii 119
Friedrich Wilhelm II, 15 Hope, 10,141,144> 248, 170
Horace, 58 n.
Garve, Christian, 3 n., 15 n., 57, 58 n., Hudson, Hoyt H., 91 n.
59 n., 60, 186 n. Hume, David. ix, 3 n., 6 n., 9, 21, 21 f.,
God, 10,14> 107 n., 114> 183 n., 199.102 n., 41, 60 f., 101 n., U4, 155, 170 f., 181 f.,
165, 279 11 1,130,255 n., 177 n.
arguments for existence of, 15, 17, Humility, 214 n., %8%
171 ff. Hutcheson, Frances, S, 6n., 94n., 104-
faith in, 1$4, 158 106,13 0
and freedom, 105 ff. Hypothesis, lSI, 161, 167
justice of, 169
Kingdom of, 141 n., 26c}, 273 n. Idea (Idee), 10, 24, 18,46, 4Bff., 59 n .,
postulate of existence of, 107 f., 161 ff. 117,119,134> 160 f., 174, -18 3,107,218,
as rewarder of virtue, 114
%39
and unity of reason, 49 n. as maxim, 164
will of, 103 f., 107, 241, 180 two theories of, in Critique of Pure
Goethe, Johann WoIlgang von, 4n., 258 Reason, 164
Good, concept of, 60, u8, 119, 136,143; Idea for a Universal History, 134
see also Bonitas; Gute, das; Wobl, Ideal, 141,165,168,171
das Idealism, 59 n., 60 f., 194, 263 f.
Good will, 11,41, 134> 173, 198 n., 203 Ignorance, 1SS ff.
Grabmann, Martin, 37 n. Imagination, 67
Gratitude, 161, 214 n. Immanent and transcendent, 48, 134
Green, Joseph, 15 n. Imperative, 54> 71, 84 fl.
Greene, Theodore M., 91 n. apodictic, 88, 117
Grundsatz, 77, 78 n., 97 assertoric, 8S, 88
Gute, das, 133, I¢ categorical, 71 tr., 77 n., 87, 117
Hatpann, Johann Georg, 255 n. heteronomous, 103 n., Ill, 123
Happiness, 10,71,81,97, 113, 135, 214f., hypothetical, 71 f .. 84ff., 96, J02, 112,
228 117 ff., 131, 196n.
a priori foem of, 114 and law, 77 n., 117,117
Idea of Universal, 161 modality of, 87
moral, 114> 116 pragmatic, 85
and pleasure, 97 and rule, 79
in Srmrmum bonum, 141 ff., 171, %74 ff. technical,85
worthiness for, 214f., 144> 274 n., 17S Inaugural Dissertation, 7 f., 15 n., 95
Hare, R. M., 154 n. Incentive, 68, 76, 90 f., 178, 211, ZJ6tr.,
Hedonism, 91 fl., 97 ff., 100 fl., 133 21 9. 1 35
Hedonistic calculus, 94 Inclination, 10 n., 90,119
Hegel, Georg Wilhelm Friedrich von, Individualism, 101
15 0,161, 219 Infield, Louis, xii
Heidemann, Ingeborg, 195 n. Inner sense, 93 n., 131 n.
Heimsoeth, Heinz, 195 n., 116 n. Intellectual intuition; see Intuition, in-
Helvetius, Claude Adrien, 104 tellectual
Heracleitus, 234 Intellecrus, 37
Herder, Johann Gottfried, 6, 18, 60 n., Intelligible world, 48, 171, 180, lU, 116,
115,133 216 f., 242 n., 147 n., 148, 178
302 Indexes
Interest, 35 f., 91 f .. 101 n., 216, 249 n. see also Moral law
moral, 92 n., lIS, 117, UI Lectures on Ethics, 54, 100 n., 103 n., 236
Interests of reason, 1740 185, 189, 119, Legality,79, 121,203,111,219,221,230
248 f .. Leges permissivae, 147
Intuition, 10, 2Z f., 67, 129, 140, ISS, 181, Leibniz, Gottfried Wilhelm, 190 n., 206,
151 214n.
intellectual. So, 143, 195,210, 269, 277, Lessing, Gotthold Ephraim, 231, 266 n.
27 8 Libertas phaenomenon and libertas nou-
moral, 128, 167 menon, 191 n., 205 n.
Life, definition of, 53, 218
Jachmann, Reinhold Bernhard, IS n., Locke, John, 90,181
282 n. Logic
Jacobi, Friedrich Heinrich, 279 n. in Critique of Practical Reason, 55, 112
Jakob, Ludwig Heinrich, 17 n. general, 65, 87, 239, 262 n.
James, William, 254 transcendental, 65, 67, 239, 262 n.
Jenisch, Daniel, 58 n. Love, 224 n., 2]2 f., 243 n.
Johnson, Samuel, 232 n. Lying, 8o, 160
Judgment (Urteil, Satz)
aesthetic, 114 MacBeath, Alexander, 115 n.
analytic and synthetic, 20, 52, 156 Mandeville, Bernard de, 102, 103 n., 104
practical, 70, 76 if., 96, 1440 156 Manthey-Zorn, Otto, viii
synthetic a priori, II, 20 f., 27, 47, 75, Marburg School of Neo-Kantianism,
121, 173,262 167 n.
Judgment (Urteilskraft) , 128, lS4ff., Martin, Gottfried, 251 n.
chap. ix Material and formal, 96, 134, 198
determinative and reflective, 154 n., Material principle, 72, 95 f., 102, 118, 120,
z79 122, 130, 133, %I 3
practical, 156 ff. Material of principle, 96, 118 n., 130
typic of, 154ff. Mathematics, 6, 20, 1I, 53, 60 f., 156 n.,
Jung-Stilling, Johann Heinrich, 144 n. 170 f., 25' f.
Justice, 147 Matson, W. I., 80n.
Juvenal, 236 n. Matter and form, 96, 134
Maupertuis, Pierre de,98 n.
Kiiubler, Bruno, 81 n., 91 n. Maxims, 70 ff., 80 ff., 91, 145 f., 178
Kahn, L. W., 200 n. evil, 2 04
Kemp Smith, Norman, vii, xi, 183 n. of lower faculty of desire, 95 If.
Khodoss, Claude, 177 n. mere, 82, 89, 95, 97, 116, 118, 146
Kiesewetter, Johann Gottfried Christian, order or levels of, 97, 118
38n., z80 n. as regulative, 83 n., 193
Kolenda, K., 202 n. and rules, 79 f.
Kraus, Christian Jacob, 58 n. Mechanism, 187, 190, 193,278
Krause, Albrecht, 154 n. Mellin, Georg Samuel Albert, 37 n., 144,
Kroner, Richard, 18 n. 15°,15' n.
Mendelssohn, Moses, 3 n., I I n., 105,
La Mettrie, Julien Offray, 190 n. 159 n., 206, 266 n.
Lambert, Johann Heinrich, 6, 8 n. Merian, Philip, 255 n.
Law, 145 ff. Messer, August, vii n., 248 n.
conception of, 34 if., 38,41 Metaphysical Foundations of Natural
empirical,82 Science, 13, 54, 144 n.
of nature (i.e., as a norm), 123, 158 n. Metaphysics, 45 n., 53
of nature (i.e .. scientific law), 83, critical and precritical views of, iff.,
12 3 f., 1590 2%% "8 ff., 11 ff., 214
prescriptive and descriptive, 40, 83 division of, 9
regulative and constitutive, 121 immanent and transcendent, 24, 226,
and rule, 97 25 2 , 26 5
Indexes 303
practical-dogmatic, 46 n., zz7, 145, Moral sense,S, 66,105 f., 128,212,230
163 n. Moralgesetz, das, origin of term, 280
Metaphysics of Morals, Ion., 14n., 17, Morgan, B. Q., 37 n.
51, 53 f., 99, l44 n ., 177, 209f., 112, Moritz, Manfred, 10] n.
124 f., 236, 276 Motive
Metaphysics of morals [division of meta- definition of, 91 n.
physics], 7, 9 f., 46 n., 53 f., 245, as object of mora~ judgme~~, 113 .
263 n. two theories of, m precnncal period,
"Metaphysics of Morals" [projected 214 fl.
book], 6 ff., 8 f., II fl., 17,46 and type, 158, 161
Metaphysics of nature [division of meta- Mysticism, 163 n., 23S
physics],9, 13 n., 23,43,54
Method Nahm, Milton C., l54n.
analytic, 6n., Il, 52,84,112 Natorp, Paul, xi, I l n., 116 n., uS n.
skeptical, 156 n., 240 f. Natura archetypa, 180
Socratic, 110, 235 f, Narure,23
synthetic, 12, 84 as ethical norm, 125
Methodology realm of, 158 f.
of Critique of Practical Reason, 210, see also Law, of nature
233 if. Need of reason, 252, 2540 26r
of Critique of Pure Reason, IS Neigung,9O
Mill, John Stuart, 158 n. Newton, Sir Isaac, 6 n., 159 n., 21]
Modality Noumena' see Causa noumenon; Phe-
categories of, 14 1 , 149 fl. nome~a and noumena; thing-in-it-
principles of, 152 self
Montaigne, Michel de, 104 f. Obedience, [07, [96,200,214 n.
Moral argument, 2SS, 260 fl., 266 n., 275, Object, 92 f.
277 of desire, 90 if., 118 fl.
Moral commonwealth, 99t 273 n. of Ideas, 263
Moral community, 147 immanent, 1]5 f.
Moral experience, as realm of discourse, of interest, 91 fl.
17 1 of postulates, 263
Moral feeling, 6?, 100, 103, 105, 107. of practical reason, 39 f.. 66, 129 if.,
213 fl., 122 fl., 229f. IS7 n., %l2, 243
Moral law, 116 fl. Objectivity, 22
as a priori synthetic, 27 Obligation, 73, II], Il3, n8, 197, 199; see
deduction of, chaps. viii and x also Duty
eifect on conduct, 21J, Z zz "Obliged to," meaning of term, I I]
formula of, 71, 73, 260 Observations on the Feeling of the Beau-
freedom and, 73 tiful and Sublime, 21], 214n., 211 n.
and imperative, 77, III f .. 165 n., 117 Observer; see Spectator
as incenti"e, 219, UI Only Possible Premise for Demonstra-
logical and social criteria of, 99 tion of the Ezistence of God, 27J
and maxim, 80 if., 89, 94 f., 98,116,146, Ontological argument, 271
178 Ontology, 56, U2
as postulate, 253 n. Orientation, 256
purity of, 41 n. Orientation in Thinking? What Is, 125
and rules, 79 f. "Ought," 49, 71., 117 n., 149, 153, 173, 189,
in SU'11Z'TItW1Z bonum, 243 f. 197,260
and unity of will, 138 and "can," 1]5, 189, 1.00n., 132
validated by Summum bonum, 244 f., and "is," 50, 1l4, J57, 163. 199. 117. 226
247, 2540 274 and "would." 117.226
Moral philosophy, relation of, to trans-
cendental philosophy, 7, 9 n., 46 Pain. 10 Q., 93,219
Moral satisfaction, 229 fl. Paradoxes in Kant's ethics, 123,201
304 11ldexes
Paralogisms, ¥I, 59, 195 n., z65 and concepts, 60, 66 f., 117, 131 fl., 145,
Passion, 940 Zl9 n. 173,111
"Patchwork theory," 3 n., 17 as incentives, 116
Paton, H. J., vii, viii, 3 n., 79 n., 80 n., and rules, 79 f., 145
86 n., lI8 n., 128 n., 147 n., 157 n., 159 tautological, loon.
Perfection, 94, 10 3 f., 106 f., IZ3, 199, 141, technical, 80
169 Prize Essay, 5, 17,113
Permitted and forbidden, the, 149, 150n. Progress, 134 ff., 167, 169 fl.
Perpetual Peace, 103 n. Prolegomena, 51, 51
Person, 142, 147, 148 n., 121, 226 Prudence, I l l , 147,103
Personalism, %16 counsel of, 85, 116
Personality, 148, 191, lZ1, 115 fl., 167 "Prussianism," 101
Phenomena and noumena, 17 n., 16 f., Publicity, criterion of, 103 n.
59 n., 186 fl., 188, 194, Ill, 117 Punishment, 114
Philosophy of history, 133 "Pure," 40 n .. 53
Physico-teleological argument, 171, 175 Pure ethics, Ion., 54
Picavet, Fran~ois, 91 n., 177 n. Purpose, 91, 130,113,177
Pietism, 165,135
Pistorius, Hermann Andreas, 16, 58, 59, Quality, categories of, 146 f., 149
60,127 n. Quantity, categories of, 144 n., 145 f.
Platner, Ernst, 166 n.
Plato, 37, 185,166 n. Rational psychology, 226
Platonism, 7, 166 n. Rationalism, 8 n., 25, 44,115,182,213,165
Pleasure, Ion., 53, 91 fl., 97,131 n., 131 f., Realm, 158, 161
117 fl. of ends, 99, 145 f., 160 fl., 127
contemplative, 92, 101 n. of grace, 206
intellectual, 215 of nature, 158 f.
negative, 220 Reason, 37, 38,80
pathological, 114 f. practical, 27, 29, 36, 37 fl., 44, 67 f.,
practical, 93, 124 f. 75 ff., 88, 137, 141
Policy, 35 f., 40,77 fl. categories of, 119, 136 fl., 153 ff.
Poseidonius, 132 n. cognition of, 39 f., 150
Possibility, 172 n. empirical, 40, 46, 76 ff., 90 ff., 95
category of, 137, 141 f. as higher faculty of desire, 118
moral, 133 f., 157 history of term, 37 f.
in postulates, 159 objects of, 118
practical, 137 postulates of, 114, 151 ff., chap. xiv
Postulate, 16, 28, ISO, lSI fl., chap. xiv primacy of, over theoretical reason,
of empirical thought, 114, 151, 171 149 ff., 263
and imperative, 153 n. pure, 40 f., 42 ff., ¢, 73, III, 110, 113,
law as, 153 n., 159 1J4o 14[, 144
in mathematics, 151 f. real and logical use of, 41, 52 f.,
truth of, Z61 128 n., 178, 196 n., 103
see also Freedom; God; Soul regulative and constitutive use of,
178n.
Practical knowledge, 39 and n., 40, 113
self-consciousness of, 168, ZZl n.
Pragmatism, 194,154
unity of, with theoretical reason,
Precept (VorscbTift), 78 n., 146
I I f., IS, 45 fl., 175, 245,178 f.
Predestination, 105 theoretical (or speculative), 13 f., 38 f.,
Presupposition, 151 n. 56,80
Pretense, permitted moral, 134 real and logical use of, 7, 38 n., 75,
Price, 116, 136 178
Price, Richard, 41 n., 106,130,131 n. unity of, with practical reason, I I f.,
Priestley, Joseph, 116 n. 15,45 ff., 175,245, 278 f.
Principles, 7°, 77 If., 111 in Wolff, 38 n., 95
Indexes 305
Regulative and constitutive, 10, 178, 19J Shaftesbury, Third Earl of,s. 6 n., 7, 105
Rehberg, A. W., 150 n. Sidgwick, Henry, 105 n.
Reid, Thomas, 37 n. Singer, Marcus G., 80 n.
Reiner, Hans, 131 n. Sittengesetz, 180
Reinhard, Walter, 174 n • Skepticism, 21, 27,44,61
Reinhold, Karl Leonard, 56 n., 57 n. Skill,l06
SocHres, ZZ2
Relation, categories of, 144 n., 147 f .. 149,
159, 160 Soul, 24. 183 n., zz6, 265
Religion, 133, 179 if. immonality of, 10, 25, 17, 107 f., 21 4,
Religion 'Within the Limits of Reason 217,148,261,265 ff.
Alone, 103 if., 1]1, 173 n., 176, 179 "Sour duty," 110 n., 135 n., 129 n.
Repentance, 116, 1]1 Space and time, 7, 17 n., 21, 25, 50, 67,
Respect, 54, 116, [41, 119 tr., lZ], %1.7, l66n.
136,14] n., 182 Spectator, 31 if., 193
Reward, 1I 4 Speculative philosophy, 140 n.
Romanticism, 115 Spiegelberg, Herben, 280 n.
Romundt, A., 225 n. Spinoza, Benedictus de, 206, 256 n.
Ross, Sir William David, vm, 120 n., Spontaneity, [77, [840 18<), 194 ff., 1 °3,
161 n. Z240 %26, 130
Rousseau, Jean-Jacques, 6 n., 135 n .. Stange, Carl, vii
159 n., 165, [95 n., 197, zoo, 101 n., Stephen, Leslie, 281 n.
ns n., :1.19n., 134n., 235 f., 158 Stilling, J., 154 n .
Rules, 70 f., 76 tr., 79 fr., 84 Stoicism, [040 1]1 n., 219n., 246, 281
and categories, 138 f. Strife of the Faculties, 207, 279
of commission and omission, 146 Strothmann, F. W., 37 n.
and laws, 97 Style, vii, 3 f.
and principles, 78 n., 98, 1]8, [45 "Subjective," 217,156
of skill, 85 Sublimity, 93, 220, 216, 282
"Sub ratione boni," 131 n.
Santayana, George, 201 n. Summum bonum, 16, 28, lS7 n., 208,
Scheler, Max, 118 n. 266 ff., 278
Schema, 68, 129, 138, 141 f .. ISS, 164 analysis of, 242 ff., 146
Schematism, 68 f., 117,141,155 if. antinomy of, 246 ff.,273
Schiller, Friedrich, 110 n., 2]1, 133 definitions of, 241 n., 268
Schopenhauer, Anhur, 4t 75 duty [0 promote, 244, 2B. 161. 26<).
Schrader, George A., 274 n. 175 f.
Schcecker, Paul, ]7 n. possibility and actuality of, 247, 159
Schutt, Christian Gattfried, 11, 13 n., 17, validates mor2l1aw, 144, 245, 247, 254,
46 n., 57 n., 60 n., 1440 150 174
Schultz, Johann, 58 Supersensuous nature, J92; see also Intel-
Schulz, Johann Heinrich, 59 n., 190 ligible world
Schwiirmerei, 115, 163 n. Superstition, 256, 183
Schweitzer, Alben, 20S n. Syllogism, 81, J z8, ~60, z64
Self, 188 f., 194 fr., :1.16 practical, 66, 81, 119, u8 f.
noumenal and phenomenal, 58 Symbol, 157 n., [62
Self-conceit, 219 Sympathy, 100, 113,23°, 133
Self-love, 72, 89, 97, JOO, 119, [SO, 1[6, Synthesis, zz, [0<), 130, 1]9, 181,111, 15 1
2[9,ZZ3 Synthetic unity, 136
Self-rewarding system of morality, 147 n. "System of Metaphysics" [projected
Selfishness, 100, 219 book], IJ
Seneca, 181 n. "System of Practical Philosophy" [pro-
Sensation, 93, 130 jected book], 13, 140 54
Sensibility, 22, IS, 56, 67, 93, 211 "System of Pure Moral Philosophy"
Sentiment, 219n. [projected book], 9
306 Indexes
Teleological argument, 166, 175 f. Universalization, III n., 159, 161
Teleology, J60, 191 f., 175, 178 Useful, the, 133
Territory, 16J
Tetens, Johann Nikolaus, 6on. Vaihinget, Hans, vii, 58 n., 139 n., 181 n.
Theology Value, intrinsic, 115
moral, 177, 181 n. Vaucanson, Jacques de, 190 n.
natural,6, [60, 151, 171, 177, :81 Virtue, 106, 134> 138, II7
rational, z77 Vorlander, Karl, 4 n., 131 n.
Thing-in-itself, 16, [81, [81, 186 ff., zoo,
z5 0 Weldon, T. D., vii, 106 n.
Thomasius, Christian, z80 Will, 15,39,67. chap. "j
Thought and action, chap. iii, passim. cognitive and dynamic components of,
H f., 40. 7 1, 76 ff., 80, 81, 85 f., 90,
Tindal, Mathew, 18[ n.
157, 21 5
Tittel, Gottlob August, 58 n., 165 definitions of, 36, 37, 38, 75
Tonelli, Giorgio, 221 n. and desire, 91; 178
Totality, Idea of, [86 f.
and passion, 94
Transcendent and transcendental, 48, 65, unity of, 138 ff., 148 n. .
163, 262; cf. also Immanent and see also Bad will; Freedom; Good will;
transcendent Holy will
Transcendental consciousness, 195 n. "Will to believe," 154
Transcendental philosophy, 7 ff., 10 n., Wille, der, 37, 75 n., 177 ff.
43,46, u8, 14On., 211, 261n. Willkiir, die, 75 n., 9[ n., 130, 139, 148 n.,
Triebfeder, 90, 216 f.
177 ff., 214
Truth Wizenmann, Thomas, 153 f., 156
coherence theory of, 163 f. Wlomer, Johann Heinrich, 4n.
as moral principle, 103 n. Wohl, dllS, 133, 146 .
"Two-world theory," 26, [91 Wolff, Christian, 6 n., 14 n., 37 n., 38 n.,
Type, 68, 99 n., 158 ff. 51, 90 n., 94, 95, 101 n., 106, 10'] n.,
Typic,68, 154 ff., 28z 12], U4> u8 n., 131 n., 16:, 177 n.,
213, :30n., 151 n., 162 n., 279n., :80
Ulrich, Johann August Heinrich, 58 n., Wolff, Heinrich, "
190, 195 Wundt, Max, 6 n.
Understanding, 18, 38, 56, 130
practical, 38, 118 n., [78 ZeIter, Karl Friedrich, 4 n. .
and reason, 95 Zum e'Wigen Frieden in dt'l Philosoph",
Wolff's conception of, 38 n., 94 Z07 n., 253 n.
Universality of moral judgment, 114 Zustand, 148
Philosophy
In this exegesis a leading Kantian scholar makes a full historical, literary,
and philosophical examination of the Critique of Practical Reason consid-
ering it in relation to the total eighteenth-century background.
"Indispensable for the proper understanding of Kant's moral phil os
ophy."-Marcus G. Singer, The Journal of Philosophy
Midway Reprints
Midway Reprints are limited-edition, paper-covered books. The purpose
of this series is to keep in print titles for which there is a steady but
small demand-out-of-print titles, as well as books which might oth-
erwise have been declared out of print. In order to keep the prices as
low as possible, Midway Reprints will not be issued in clothbound
editions.
The University of Chicago Press IS BN 0-226-0 40 76-3

A Commentary On Kant's Critique of Practical Reason (UCP 1960) - L.W. Beck

Uploaded by

Copyright:

Available Formats

A Commentary On Kant's Critique of Practical Reason (UCP 1960) - L.W. Beck

Uploaded by

Document Information

Original Title

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

A Commentary On Kant's Critique of Practical Reason (UCP 1960) - L.W. Beck

Uploaded by

Copyright:

Available Formats

Lewis White Beck

Midvvay Reprint '

LEWIS WHITE BECK

THE UNIVERSITY OF CHICAGO PRESS

© 1960 by The Cniversity of Chicago

Printed in the Cnited States of America

Odd though it may be, the Critique of Practical Reason is a neglected

opher. But it is a mistake to believe that what is of utmost importance

References to the Critique of Practical Reason are given as follows.

I THE WRITING OF THE "CRITIQUE OF PRACTICAL REASON" 3

V SURVEY OF THE ANALYTIC OF PRACTICAL REASON 6S

5 Are There Empirical Practical Laws? 82

X THE "TRANSCENDENTAL DEDUCTION" OF THE PlUNCIPLE OF PURE

XII THE "AESTHETIC" OF PORE PRACTICAL REASON; CoMMENTARY ON

3 Some Earlier Attempts at an Answer 21 3

XIII THE DIALECTIC OF PURE PRACTICAL REASON; CoMMENTARY ON

I TEXTS OF "KRITIK DER PRAKTISCHEN V ERNUNFT" 187

I INDEX OF PASSAGES CITED FROM "CRITIQUE OF PRACTICAL REASON" 197

The Writing of the

8 To Lamben, December 31, 1765 (X, 56).

work on "the first part of my ethics."23 We do not know whether he

2fS Reflexion 7202 ,

Notwithstanding this, the third section of the Foundations is entitled

21 To Schiitz, September 13, 1785 (X, 406).

The Limits of Theoretical Reason

ruents. He held that mathematical judgments, such as the theorems of

§ 5. THE THIRD ANTINOMY'

§ 6. TRANSITION FROM THE PROBLEMATIC TO THE ASSERTORIC

Thought, Action, and Practical Reason

§ I. TWO ASPECfS OF CONDUCT

§ 4. PURE PRACfICAL REASON

Name, Purpose, and Structure of the ("("Critique";

The Preface of the Critique of Practical Reason discusses the follow-

2 Critique of Judgment, Introduction, III (V, 179 [IS]).

In the preceding section, when evaluating Kant's reasons for having

§ 4. THE UNITY OF THEORETICAL AND PRACI'ICAL REASON

9 The Critical Philosophy of bmmmuel Kant, II, 164.

10 Cf. Critique of Pure Reason, A 548 = B 576; Critique of Practical Reason, 15

This question is different from that of task I (a), which is to show

17 Refiexion 7101 (XIX, z8z).

18 Foundtztiolls of the Metaphysics of Morals, 445 {64>.

§ 6. RELATION TO THE "METAPHYSICS OF MORALS"

21 Cf. above, p. 9, n. Zl.

22 Foundations of the Metaphysics of Morals, 41 %(%8).

Actual Structure of Remembered Structure of Structure of Second

Transcendental Aesthetic Transcendental Analytic Logic

§ 8. THE POLEMICS OF THE PREFACE

§ 2. THE STRUCTURE OF THE ANALYTIC

The Analytic of Empirical Practical Reason

§ I. EMPIRICAL PRACTICAL REASON

20 Critique of Prat;tical Reason, 26 (114); Critique of Pure Reason, A 159::::: B 198.

In this chapter we have studied the formal structure of practical rea-

The Ana ly tic of Empirical Practical Reason

1'1' Reflexion 72 51.

Kant equates the principle of hedonism with that of egoism. To seek

guide in satisfying them, a principle or a maxim whose content is the

principles, since it does preserve morality from all contamination with

45 Metaphysik der Sitten, Tugendlehre, Einleitung (VI, 399-400); Critique of

The t'~etaPhJsical Deduction" of the Moral